PROCEEDINGS

OF THE

ROYAL SOCIETY OF LONDON,

From April 20, 1890, to November 23, 1899.

VOL. LXV.

LONDON :

HARRISON AND SONS, ST. MARTIN'S LANE,
$rintcis in @rbinavjj ia ^'er Utajtstg.
J1DCCCC.

LONDON :

HAEBI80N AND SONS, PEINTEES IN OKDINAKY TO HEE MAJESTY,
ST. MAETLN'S LANE.

%

01

CONTENTS.

VOL. LXV.

Pat;e

No. 413.

Report of the Kew Observatory Committee for tlie Year ending
December 31, 1898 1

Croonian Lecture. — On the Eelation of Motion in Animals and
Plants to the Electrical Phenomena which are assoc iated with it. Bv
J. Burdon-Sanderson, M.A., M.D., F.R.S 37

No, 414.

Meeting of April 20, 1899, and List of Papers read 64

A Sugar Bacterium. Bv H. Marshall Ward, F.R.S., and J. Reynolds
Green, F.R.S 65

Experiments in Micro-metallurgy : — Effects of Strain. Preliminary
Notice. By J. A. Ewing, F.R.S., and Walter Rosenhain, 1851 Ex-
hibition Research Scholar, Melbourne University. (Plates 1 — 5) ... 85

The Physiological Action of Choline and Neurine. By F. W. Mott,
M.D., F.R.S., and W. D. Halliburton, M.D., F.R.S 91

On Intestinal Absorption, especially on the Absorption of Serum, Pep-
tone and Glucose. By E. Waymouth Reid, F.R.S 94

Studies in the Morphology of Spore-producing Members. IV. The
Leptosporangiate Ferns. By F. O. Bower, Sc. D., F.R.S 26

Note on the Fertility of different Breeds of Sheep, with Remarks on
the Prevalence of Abortion and Barrenness therein. By Walter
Heape, M.A., Trinity College, Cambridge. Communicated by W.
F. R. Weldon, F.R.S 99

Some further Remarks on Red-water or Texas Fever. By Alexander
Edington, M.B., F.R.S.E., Director of the Bacteriological Institute,
Cape Colony. Communicated by Dr. D. Gill, C.B., F.R.S Ill

No. 415.

Meeting of April 27, 1899, and List of Papers read 114

On the Luminosity of the Rare Earths when heated in Vacuo by means
of Cathode Rays. By A. A. Campbell Swinton. Communicated by
Lord Kelvin, F.R.S '„ ., 115

iv

Page

On the Electrical Conductivity of FJ-"~ containing Salt Vapours. By
Harold A. Wilson, B.Sc. (Lond. k Vic), 1851 Exhibition Scholar.
Communicated by Professor J. J. T» nnson, F.R.S. 120

On a Quartz-thread Gravity Balance. By Richard Threlfall, lately Pro-
fessor of Physics in the* University of Sydney, and James Arthur
Pollock, lately Demonstrator of Physics in the University of Sydney.
Communicated by Professor J. J. Thomson, F.E.S .. 123

Data for the Problem of Evolution in Man. I. A First Study of the
Variability and Correlation of the Hand. By Miss M. A. Whitelev,
B.Sc, and Karl Pearson, F.R.S 12G

Meeting of May 4, 1899, List of Candidates recommended for Election
and List of Papers read 152

Impact with a Liquid Surface, studied by the aid of Instantaneous
Photographv. Paper II. By A. M. Worthington, M.A., F.R.S.,
and R, S. Cole, M.A 153

An Observation on Inheritance in Parthenogenesis. By Ernest
Warren, D.Sc, University College, London. Communicated bv Pro-
fessor W. F. R. Weldon, F.R.S 154

Omjgena equina, Willd. : a Horn-destroying Fungus. By H. Marshall
Ward, D.Sc, F.R.S., Professor of Botany in the University of Cam-
bridge 158

The External Features in the Development of Lepidosiren paradosca,
Fitz. By Graham Kerr. Communicated by A. Sedgwick, F.R.S 160

The Thermal Expanson of Pure Nickel and Cobalt. By A. E. Tutton,
B.Sc Communicated by Professor Tilden, D.Sc, F.R.S 161

On the Presence of two Vermiform Nuclei in the Fertilised Embryo-
sac of Labium Martaqon. By Ethel Sargant. Communicated by Dr.
D. H. Scott, F.R.S. ' 163

No. 416.

Meeting of May 18, 1899, and List of Papers read 165

On a Self-recovering Coherer and the Study of the Cohering Action of
different Metals. By Jagadis. Chunder Bose, M.A., D.Sc, Professor
of Physical Science, Presidency College, Calcutta. Communicated
by Lord Rayleigh, F.R.S 1C6

Bakerian Lecture.— The Crystalline Structure of Metals. By J. A.
Ewing, F.R.S., Professor of Mechanism and Applied Mechanics in
the University of Cambridge, and W. Rosenhain, 1851 Exhibition
Research Scholar, Melbourne University 172

The Yellow Colouring Matters accompanying Chlorophyll, and their
Spectroscopic Relations." By C. A. Sehunck. Communicated by
Edward Schmick, F.R.S. (Plate 6) ; 177

On the Chemical ( Ratification of the Stars. By Sir Norman Lockyer,
K.C.B., F.RS. (Plate 7) 186

Page

The Diffusion of Ions into Gases. By John S. Townsend, M.A.
(Dublin), Clerk Maxwell Student, Cavendish Laboratory, Cambridge.
Communicated by Professor J. J. Thomson, F.R.S 192

On the Presence of Oxygen in the Atmospheres of certain Fixed Stars.
By David Gill, C.B., F.R.S., &c, Her Majesty's Astronomer at the
Cape of Good Hope. (Plate 8) 196

No. 417.

Annual Meeting for the Election of Fellows 206

Meeting of June 1, 1899, and List of Papers read 207

The Characteristic of Nerve. By Augustus D. Waller, M.D., F.R.S 207

The Parent-rock of the Diamond in South Africa. By T. G. Bonnev,
D.Sc, LL.D., V.P.R.S 223

Photographic Researches on Phosphorescent Spectra : on Victorium, a
new Element associated with Yttrium. By Sir William Crookes,
F.R.S. (Plate 9) '. 237

Experimental Contributions to the Theory of Heredity. A. Telegony.
By J. C. Ewart, M.D., F.R.S., University of Edinburgh 243

No. 418.

Meeting of June 8, 1899 (Discussion Meeting) 252

On Preventive Inoculation. By W. M. Haffkine, CLE. Communi-
cated by Lord Lister, P.R.S 252

Meeting of June 15, 1899, and List of Papers read 272

A Preliminary Note on the Morphology and Distribution of the
Organism found in the Tsetse Fly Disease. By H. G. Plimmer and
J. Rose Bradford, F.R.S., Professor Superintendent of the Brown
Institution 274

No. 419.

The Colour Sensations in Terms of Luminosity. By Captain W. de W.

282

Abney, C.B., D.C.L., F.R.S.

The Conductivity of Heat Insulators. By C. G. Lamb, M.A., B.Sc,
and W. G. Wilson, B.A. Communicated by Professor Ewing,
F.R.S 283

On the Orientation of Greek Temples, being the Results of some
Observations taken in Greece and Sicily, in May, 1898. By F. C.
Penrose, M.A., F.R.S =••<• 288

On the Comparative Efficiency as Condensation Nuclei of positively
I and negatively charged Ions. By G. T. R. Wilson, M.A. Communi-
cated by the Meteorological Council 289

vi

Page

Data for the Problem of Evolution in Man. II. A First Study of the
Inheritance of Longevity and the Selective Death-rate in Man. By
Miss Mary Beeton and Karl Pearson, F.R.S., University College,
London 290

Collimator Magnets and the Determination of the Earth's Horizontal
Magnetic Force. By C. Chree, Sc.D., LL.D., F.R.S., Superintendent
of the Kew Observatory. Communicated by the Kew Observatory
Committee of the Eoyal Society , 306

The Thermal Expansion of Pure Nickel and Cobalt. By A. E. Tutton,
B.Sc. Communicated by Professor Tilden, D.Sc, F.R.S 306

On the Waters of the Salt Lake of Urmi. By R. T. Gunther, M.A.,
and J. J. Manley, Daubeny Curator, Magdalen College. Communi-
cated by Sir John Murray, F.R.S , 312

On the Application of Fourier's Double Integrals to Optical Problems.
By Charles Godfrey, B.A., Scholar of Trinity College, Isaac Newton
Student in the University of Cambridge. Communicated by Pro-
fessor J. J. Thomson, F.R.S 318

On Diselectrification Produced by Magnetism. Preliminary Note. By
. C. E. S. Phillips. Communicated by Sir William Crookes, F.R.S. .... 320

On the Orbit of the Part of the Leonid Stream which the Earth
encountered on the Morning of 1898, November 15. By Arthur A.
Rambaut, M.A., D.Sc, Radcliffe Observer. Communicated by G.
Johnstone Stoney, M.A., D.Sc, F.R.S 321

A Comparison of Platinum and Gas Thermometers, including a Deter-
mination of the Boiling Point of Sulphur on the Nitrogen Scale : an
Account of Experiments made in the Laboratory, of the Bureau
International des Poids et Mesures, at Sevres. By Drs. J. A.
Harker and P. Chappuis. Communicated by the Kew Observatory
Committee 327

Agricultural, Botanical, and Chemical Results of Experiments on the
Mixed Herbage of Permanent Grasses, conducted for many years in
succession on the same Land. Part III. — The Chemical Results.
By Sir John Bennett Lawes, Bart., D.C.L., Sc.D., F.R.S., and Sir J.
Henry Gilbert, LL.D., Sc.D., F.R.S 329

No. 420.

On the Orientation of the Pyramids and Temples in the Sudan. By
E. A. Wallis Budge, M.A., Litt.D., D.Lit., F.S.A. Communicated by
Professor Sir Norman Lockyer, K.C.B., F.R.S 333

The Effect of Staleness of the Sexual Cells on the Development of
Echinoids. Bv H. M. Vernon, M.A., M.D., Fellow of Magdalen
College, Oxford. Communicated by W. F. R. Weldon, F.R.S 350

On the Influence of the Temperature of Liquid Hydrogen on the
Germinative Power of Seeds. By Sir William Thiselton-Dyer,

K.C.M.G., CLE., F.R.S., Director of the Royal Botanic Gardens,
Kew 361

Effects of Thyroid Feeding on Monkeys. By Walter Edmunds. Com-
municated by Professor J. Rose Bradford, F.R.S 368

vii

Page

On the Orientation of Greek Temples, being the results of some
Observations taken in Greece and Sicily in the month of May, 1898.
Bv F. C. Penrose, M.A., F.R.S., F.R.I.B.A., &c 370

No. 421.

Collimator Magnets and the Determination of the Earth's Horizontal
Magnetic Force. By C. Chree, Se.D., LL.D., F.R.S., Superintendent
of the Kew Observatory. Communicated by the Kew Observatory
Committee of the Royal Society 375

The Absorption of Rontgen's Rays by Aqueous Solutions of Metallic
Salts. By the Right Honourable Lord Blythswood, LL.D., and
E. W. Marchant, D.Sc. Communicated by Lord Kelvin, F.R.S 413

On the Resistance to Torsion of Certain Forms of Shafting, with
special Reference to the Effect of Keyways. By L. N. G. Filon,
M.A., Research Student of King's College, Cambridge, Fellow of
University College, London, 1851 Exhibition Science Research
Scholar 428

No. 422.

Meeting of November 16, and List of Papers read 432

Note on the Electromotive Force of the Organ Shock and the Elec-
trical Resistance of the Organ in Malapterurus electricus. Bv Francis
Gotch, M.A., F.R.S.. and G. J. Burch, M.A. Oxon 134

On the Formation of the Pelvic Plexus, w ith especial Reference to the
Nervus Collector, in the Genus Mustelus. By R. C. Punnett, B.A.,
Scholar of Gonville and Caius College, Cambridge. Communicated
by Hans Gadow, F.R.S 445

On the Least Potential Difference required to produce Discharg''
through various Gases. By the Hon. R. J. Strutt, B.A., Scholar of
Trinity College, Cambridge. Communicated by Lord Rayleigh,
F.R.S ! 446

Meeting of November 23, 1899, and List of Ollicers and Council
nominated for Election 448

List of Papers read 449

Note on the Spectrum of Silicium. By Sir Norman Lockyer, K.C.B.,
F.R.S 449

Preliminary Table of Wave-lengths of Enhanced Lines. By Sir
Norman Lockyer, K.C.B., F.R.S 452

The Colour-Physiology of Hippolyle varians. By F. W. Keeble, Caius
College, Cambridge, and F. W. Gamble, Owens College, Manchester.
Communicated b}^ Professor S. J. Hickson, F.R.S 461

An Experimental Research on some Standards of Light. By J. E.
PetaveL Communicated by Lord Rayleigh, F.R.S ........... 469

PROCEEDINGS

OF

THE ROYAL SOCIETY

Report of the Kew Observatory Committee for the Year
ending December 31, 1898.

The operations of The Kew Observatory, in the Old Deer Park,
Richmond, Surrey, are controlled by the Kew Observatory Committee,
which is constituted as follows : —

Mr. F. Galton, Chairman.
Captain W. de W. Abney, C.B., I Prof. A. W. Riicker.

R.E.

Prof. W. G. Adams.
Captain E. W. Creak, R.X.
Prof. G. C. Foster.
Prof. J. Perry.

Dr. R. H. Scott.
Mr. W. N. Shaw.
Lieut. -General Sir R. Strachey,
G.C.S.I.

Rear Admiral Sir W. J. L.

The Earl of Rosse, K.P. I Wharton, K.C.B.

The work at the Observatory may be considered under the fol-
lowing heads: —

I. Magnetic observations.
II. Meteorological observations.

III. Seismological observations.

IV. Experiments and Researches in connexion with any of the

departments.
V. Verification of instruments.
VI. Rating of Watches and Marine Chronometers,
VII. Miscellaneous.

VOL. LXV. B

Report of the Kew Observatory Committee.

I. Magnetic Observations.

The Magnetographs liave been in constant operation throughout
the year, and the usual determinations of the Scale Values were made
in January,

The ordinates of the various photographic curves representing
Declination, Horizontal Force, and Vertical Force were then found
to be as follows : —

Declinometer : 1 cm. = 0° 8'' 7.
Bifilar, January 11th, 1898, for 1 cm. SH = 000051 C.G.S. unit.
Balance, January 12th, 1898, for 1 cm. BV = 0-00050 C.G.S. unit.

Owing to the gradual secular change of declination, the distance
between the dots of light upon the cylinder of the magnefcograph
had become too small for satisfactory registration, and it Avas found
necessary to alter the position of the zero line. From a similar
cause it was also found necessary to re-adjust the balance of the
vertical force magnetometer.

During the past year two magnetic storms, or periods of con-
siderable disturbance of the needles, have been registered, the first on
March 14-15, the second on September 9-10.

The extreme amplitude of the March disturbance was : horizontal
force, 0-0050 C.G.S. unit; vertical force, 0'0057 C.G.S. unit, and
declination, 1° 26'. In eight minutes, from 10.40 to 10.48 p.m. on the
15th, the horizontal and vertical components exhibited falls of 0 002
and 0*003 C.G.S. unit respectively. The most rapid change of
declination occurred some thirty minutes later. Speaking generally,
the most salient features were fche large falls in both the horizontal
and vertical components and the movement of the declination needle
nearly 1° east of its normal position.

The second storm occurred on September 9 — 10. The principal
disturbance commenced somewhat gradually about noon on the 9th,
but one of its most striking features was an exceptionally rapid fall
occurring simultaneously at 3.5 p.m. in the horizontal and vertical
forces and in the westerly declination. The fall was so rapid as to
be shown somewhat indistinctly on the photographic traces, but it
amounted to at least 15' in the declination and 0'0023 C.G.S. unit
in the horizontal force. The recovery from this fall was also
rapid.

The decimation needle, on the same day, between 5.15 p.m. and
8.8 p.m. receded 54' to the east, then turned and in the course of the
next thirty-two minutes moved 59' to the west. The horizontal force
attained its extreme maximum and minimum at 2.42 p.m. and

Report of the Kew Observatory Committee,

3

8.30 p.m. respectively, the range amounting to [0*0050 C.G.S. unit
(or about 1/37 of the whole component). Between 7.30 and 8.30
p.m., this element fell 0*0036 C.G.S. unit. The vertical force
reached its maximum about 6 p.m., and its minimum about 8.30 p.m.,
but as the trace unfortunately got off the sheet near the minimum,
it can only be said that the range of vertical force exceeded 0*0036
C.G.S. unit.

Both storms were presumably associated with the aurora simul-
taneously seen in the British Isles. The March storm was the
largest recorded since August, 1894.

The hourly means and diurnal inequalities of the magnetic elements
for 1898, for the quiet days selected by the Astronomer "Royal, will
be found in Appendix I.

A correction has been applied for the diurnal variation of tempera-
ture, use being made of the records from a Richard thermograph as
well as of the eye observations of a thermometer placed under the
Vertical Force shade.

The mean values ut the noons preceding and succeeding the selected
quiet days are also given, but these of course are not employed in
calculating the daily means or inequalities.

The following are the mean results for the entire year : —

Mean Westerly Declination * 17° l'*4.

Mean Horizontal Force 0*18364 C.G.S. unit.

Mean Inclination 67° 17'*6.

Mean Vertical Force 0*43885 C.G.S. unit.

Observations of Absolute Declination, Horizontal Intensity, and
Inclination have been made weekly, as a rule.

A table of recent values of the magnetic elements at the Observa-
tories whose publications are received at Kew will be found in
Appendix Ia to the present report.

In September Professor Luigi Palazzo of the Ufficio Centrale
di Meteorologia, Rome, paid a visit to the Observatory for the
purpose of comparing the Kew magnetic instruments and his own.

Dr. van Rijckevorsel also spent some time in the summer in
making a further comparison between his magnetic instruments and
those at Kew.

Mr. Hough, Fellow of St. John's College, Cambridge, who has
recently been appointed chief assistant at the Royal Observatory,
Cape of Good Hope, visited the Observatory from August 18 to Sep-
tember 1, in order to gain a knowledge of the method of observing
with the Unifilar Magnetometer and Inclinometer.

At the request of Professor Moos, director of the Colaba Ob-
servatory, Bombay, copies of the horizontal force, the vertical force^

B 2

4

Report of the Keic Observatory Committee.

and the declination curves for certain selected days during the years
1892, 1893, and 1897 have been made and forwarded to him.

Information on matters relating to various magnetic data has been
supplied to Dr. von Bezold, Professor Milne, and Mr. Gray.

The Observatory has been visited by Dr. A. Schmidt, of Gotha,
Professor Eschenhagen, of Potsdam, and Professor Liznar, of Vienna,
members of the International Conference on Terrestrial Magnetism,
which was held at Bristol in September.

In spring the unifilar magnetometer and dip circle, previously lent
to the Jackson-Harmsworth Polar Expedition, were put in order and
lent to Mr. P. Baracchi, Acting Government Astronomer, Melbourne
Observatory, for observational use in Australia and New Zealand, or
in Antarctic exploration, as he might decide. Later in the year an
old dip circle was put in order at the cost of Sir George Newnes, and
lent for tlie use of the Antarctic Expedition, under Mr. Borchgrevink.
It was also agreed that if Mr. P. Baracchi should be willing to
transfer to Mr. Borchgrevink the unifilar magnetometer and dip
circle referred to above, the Committee would raise no objection,
provided Sir G. Newnes should become responsible for the safe
return of the instruments.

A course of magnetic instruction was given to the two magnetic
observers of Mr. Borchgrevink's expedition, Mr. Colbeck and Mr.
Bernacchi, the latter of whom had already practised the use of
magnetic instruments at Melbourne Observatory.

II. Meteorological Observations.

The several self-recording instruments for the continuous registra-
tion of Atmospheric Pressure, Temperature of Air and Wet-bulb,
Wind (direction and velocity), Bright Sunshine, and Rain, have been
maintained in regular operation throughout the year, and the
standard eye observations for the control of the automatic records
duly registered.

The tabulations of the meteorological traces have been regularly
made, and these, as well as copies of the eye observations, with
notes of weather, cloud, and sunshine, have been transmitted, as usual,
to the Meteorological Office.

With the sanction of the Meteorological Council, data have been
supplied to the Council of the Royal Meteorological Society, the
Institute of Mining Engineers, and the editor of ' Symons' Monthly
Meteorological Magazine.'

EledrograpJi. — This instrument worked in a satisfactory manner
till May, when the action markedly deteriorated. Tests of the
battery showed chat its E.M.P. had fallen off considerably ; this was
so far remedied by cleaning and recharging the top row of cells. At

Report of the Keiv Observatory Committee.

5

the same time a new silk suspension was fitted to the needle of the
electrometer, and the instrument generally overhauled, and a new
scale determination was carried out.

The records remained satisfactory until November, when the
battery potential again began to fall oft' rapidly. Between November
24 and 27 the whole sixty cells were cleaned and recharged with a
satisfactory result, and on the latter date one-third of the cells were
removed to contract the scale, in order to record high winter values,
as explained in last year's Report.

On several occasions it had been noted that the electrometer needle
had a tendency to " set " when the acid in the interior jar had been in
use for some time. This i; setting " largely interfered with the freedom
of the needle. It has, however, been considerably reduced, by substi-
tuting a single platinum wire connection for the double gridiron
form hitherto employed.

In May another portable electrometer, No. 80, was purchased from
White, of Glasgow-; it is furnished with some additions to the usual
pattern, which experience at the Observatory suggested as likely to
prove beneficial in reducing induction effects. This electrometer
has been used since, with the older instrument, White, No. 53,
in obtaining the scale value of the self-recording instruments, de-
terminations being made on February 7, April 1, May 26, June 16,
September 6, and November 28.

Inspections. — In compliance with the request of the Meteorological
Council, the following Observatories and Anemograph Stations have
been visited and inspected : — Stonyhurst, Yarmouth, North Shields,
Alnwick Castle, Fort William, Glasgow, Aberdeen, and Deerness
(Orkney), by Mr. Baker; and Radcliife Observatory (Oxford), Holy-
head, Fleetwood, Armagh, Dublin, Valencia, Falmouth, and St.
Mary's (Scilly Isles), by Mr. Constable.

III. SE1SMO LOGICAL OBSERVATIONS.

The seismograph referred to in last year's Report was delivered
by Mr. R. Munro in March. It is of Professor J. Milne's " unfelt
tremor" pattern, the motion recorded being that of a horizontal
pendulum or boom with along period of vibration (fifteen to eighteen
seconds from rest to rest). It is intended to measure tbe tilting of
the ground along an east- west line, the boom itself lying due north
and south.

At the suggestion of Professor Milne, who visited the Observa-
tory, the site selected for at least a temporary trial is in the base-
ment, inside the double-walled wooden room, originally designed for
pendulum observations, and sometimes used as a warm chamber for
chronometers. At first difficulties were encountered from wandering

6

Report of the Kew Observatory Committee.

of the boom, -winch, is still too liable to get off its pivot ; but the
record has been, on the whole, satisfactory for the latter half of the
year. The following table gives particulars respecting the time of
occurrence and amplitude in seconds of arc of the largest movements
actually recorded : —

Time (Q-.M.T.).

Amplitude.

Date.

h.

m.

//

7*

19-8 p.m.

£i • )

21-8

J5

3*4

26-7

JJ

3-0

■>■> •

. . . . . ,,

31*4

JJ

2-2

. . . 8

34-9

??

2-7

37-0

?)

1-5

37-8

J?

1-7

40-7

1-6

.... 1

443

55

0-5

46-4

5?

0-6

58-6

5>

06

The times deduced for the commencement of the above-mentioned
earthquakes were 6 h. 47*6 m., 8 h. 4*5 m., and 1 h. 37 m. re-
spectively.

Without special very careful experiments it would be difficult to
say what is the probable error in fixing the precise times. Inde-
pendent measurements of the photographic trace may agree to O'l or
0'2 of a minute, but there is room for a certain amount of doubt as to
the proper values to attribute to the time marks on the sheet.

In the case of the times of commencement of a disturbance the
uncertainty is greater, because the movement may be initially infini-
tesimal, and because a tiny movement arising from a different source
(such movements being not uncommon) might intervene.

IV. Experimental Work.

Fog and Mist. — The observations of a series of distant objects,
referred to in previous Reports, have been continued. A note is taken
of the most distant of the selected objects which is visible at each
observation hour.

Atmospheric Electricity. — The comparisons of the potential, at the
point where the jet from the water-dropper breaks up, and at a fixed
station on the Observatory lawn, referred to in last year's Report,
have been contir.ued, and the observations have been taken twice
every month.

During October some simultaneous observations were made with

Report of the Kew Observatory Committee.

7

the two portable electrometers, the ODe situated on the pillar in the
garden, the other at the same height on a tripod stand, at some
distance in the park. It is hoped that time will he found to repeat
the experiments on sufficiently numerous occasions to allow some
conclusions to be drawn.

Aneroid Barometers. — The experiments referred to in the last three
" Reports " were continued in the early part of the year. The results
have been discussed by the Superintendent in a paper recently pub-
lished in the Society's * Transactions.'

Platinum Thermometry. — The experimental work carried out at
the International Bureau of Weights and Measures at Sevres by
Dr. J. A. Harker .in co-operation with Dr. Chappuis lias only just
terminated. It has comprised a careful comparison of certain
platinum thermometers belonging to the Observatory with a gas
thermometer belonging to the Bureau, over the range —30° C, to
+ 600° C.

Dr.Harker brought back the platinum thermometers, vesistance box,
&c, to the Observatory late in December, and is about to be engaged
in preparing the results for publication. In view of this and other
special thermometric work in contemplation, the Committee have
temporarily secured the services of Dr. Harker in the capacity of
special assistant to the Superintendent.

Experiments have been continued at the Observatory itself on the
fixity of zero, and the general behaviour of platinum thermometers,
which have shown, amongst other things, the expediency of carefully
checking the behaviour of the " leads."

Experimental work on the comparison of platinum and mercury
thermometers has also been continued, and it is hoped that it will
shortly be possible, utilising the results of Dr. Harker's work at
Sevres, to issue certificates to high range mercury thermometers
embodying the results of direct comparison.

Mercury Thermometry. — The experiments on thermometers of
different kinds of glass made by Messrs. Powell and Sons, to which
reference was made last year, have been continued. Further ther-
mometers are being made by Messrs. Powell, of a pattern suggested
by the Superintendent, with which it is hoped to experiment at
higher temperatures.

V". Verification of Instruments.

The subjoined is a list of the instruments examined In the year
1898, with the corresponding results for 1897 : —

8

Report of the Kew Observatory Committee.

Number tested in the year

ending December 3 1 .

1897.

1898.

5

1

3

11

77

169

17

9

167

122

101

58

• 30

55

661

374

51

44

4

3

292

463

5

5

1U

-LO

2

2

J £

707

681

27

12

31

10

2

694

750

10

15

29

26

Thermometers, Avitreous, or Imnrisch's

5

10

17,270

17,962

119

79-

37

56

30

38

71

94

„ Meteorological

2,874

3,290

2

1 1 K

LI/

fid

DO

Tin iniQ

■i 4

6

4

3

23,457

24,434

Duplicate copies of corrections have been supplied in 84 cases.
The number of instruments rejected in 1897 and 189S on account
of excessive error, or for other reasons, was as follows : —

Report of the Kew Observatory Committee.

1897. 1898.

Thermometers, clinical 156 173

ordinary meteorological 38 92

Sextants 98 106

Telescopes 66 60

Binoculars 28 30

Various 56 26

Two Standard Thermometers have been constructed during the-
year.

There were at the end of the year in the Observatory, undergoing
verification, 7 Barometers, 550 Thermometers, 50 Sextants, 20
Telescopes, 59 Binoculars, 2 Hydrometers, 2 Sunsliine Recorders,.
5 Rain Measures, and 2 Rain Gauges.

VI. Ratixg of Watches axd Chroxometees.

The high standard of excellence to which attention has been drawn
in previous Reports has been maintained. Although the number of
watches sent for trial this year is less than last year, yet the general
average is as good, and 66 movements have obtained the highest
possible form of certificate (the class A, especially good), which
involves the attainment of 80 per cent, of the total marks.

The 483 Avatches received were entered for trial as below : —

For class A, 383 ; class B, 73 ; and 27 for the subsidiary trial.
Of these 17 passed the subsidiary test, 116 failed from various causes
to gain any certificate, 55 were awarded class B, and 295 class A.

In Appendix III will be found a table giving the results of trial
of the first 50 watches which gained the highest number of marks
during the year. The highest place was taken by Mr. S. Yeomans,
Coventry, with a keyless going-barrel, Karrusel lever-watch,
No. 76,152, which obtained 89*2 marks out of a maximum of 100.

Representations having been made to the Committee that some
changes were desirable in the system of marks and dates on certifi-
cates, a circular wras issued (as mentioned in last year's Report) to
ascertain the general opinion of manufacturers and others interested
in the matter, but the replies received showed no unanimity of
opinion in favour of any one specified change, whilst a considerable
number were quite satisfied with the existing conditions. Finally
some small alterations were made, mainly in matters of detail.

The objection to the certificates that sustained most support —
though even on this question opinions were fairly divided — was that
the date suggested to the customer, in the case of any but the most
recently tested watch, a line of criticism that would not naturally have
presented itself. In consequence it was urged that the possession of a

10 Report of the Kew Observatory Committee.

Kew certificate was a very doubtful advantage to any watch, remain-
ing unsold for several years in a retailer's hands. The Committee
could not see their way to alter the invariable practice of dating
Kew certificates, but they agreed, in order to minimise the source
of complaint, that a watch tested at the Observatory not less than
three years previously, should be admitted to a fresh trial at half the
usual fee. ,

Marine Chronometers. — During the year, 70 chronometers have been
entered for the Kew A and B trials ; of these 33 gained certificates,
21 failed, and there are 16 in hand.

The new cold-air chamber, to which a preliminary reference was
made in last year's Report, has been completed, and has proved
very convenient.

It consists of three separate divisions, each isolated from the
others, and separated by a 3-inch space packed with flake charcoal,
this same packing being continued on all sides of the divisions, the
size over all being ft. by 6^ ft. by 3 ft.

The centre chamber, 3 ft. by o ft. by 2 ft., is fitted with sliding
racks for the chronometers, and the division on either side is for the
ice. This is supplied in blocks, which rest on boards, and drain
away into a trap and gulley. The chronometer chamber is furnished
with trays to hold potassic chloride for drying purposes, and with
maximum and minimum thermometers.

The doors are packed with flake charcoal, and are so arranged that
the ice stores can be filled or emptied without any disturbance of the
chronometer chamber.

VII. Miscellaneous.

Paper. — Prepared photographic paper has been supplied to the
Observatories at Hong Kong, Mauritius, Oxford (Radcliffe), and
Stonyhurst, and through the Meteorological Office to Aberdeen,
Fort William, and Valencia.

Anemograph and Sunshine Sheets have also been sent to Hong Kong
and Mauritius.

Gas Thermometer. — Sir Andrew Noble, K.C.B., having generously
offered to present a gas thermometer to the Observatory, and to
defray the cost of sending an assistant to Berlin to study the method
of using a similar instrument at the Reich sanstalt, at Charlotten-
burg, the Committee gladly accepted the gift. The construction of
the instrument has not yet been completed.

Pendulum Observations. — In July Mr. F. Laurin and another officer
of the Royal Austrian Navy swung half second pendulums in the
sextant room on the spot where observations were made some years
ago by von Sterneck.

Report of the Kew Observatory Committee. It

Electric Tramways. — During the year a variety of schemes have been
promoted for applying electric traction on the trolley system to tram
lines in the neighbourhood of the Observatory, and one of these schemes,
promoted by the London United Tramway Company, for a new line
between Kew Bridge and Hounslow, passing within 1,300 yards of
the Observatory, has received the sanction of Parliament. The Com-
mittee, roused by the fate that has befallen the magnetic observa-
tories at Toronto and Washington, requested Professor Riicker and
Professor Perry to take the matter in hand. A series of experi-
ments made at various places in London and Leeds, under the
general supervision of Professor Riicker, showed that electric rail-
ways and tramways, satisfying presumably all the existing require-
ments of the Board of Trade, produced very sensible disturbances in
a declinometer at distances of two or three miles. This fact was
brought before the notice of the Royal Society, who in turn entered
into communication with the Board of Trade, with the result that
the following clauses were inserted in the London United Tramway
Company's Bill : —

1. The whole circuit used for the carrying of the current; to and
from the carriages in use on the railway shall consist of conductors,
which are insulated along the whole of their length to the satisfaction
in all respects of the Commissioners of Her Majesty's Works and
Public Buildings (in this section called the " Commissioners "), and
the said insulated conductors which convey the current to or from
any of such carriages shall not at any place be separated from
each other by a distance exceeding one-hundredth part of the
distance of either of the conductors at that place from Kew Obser-
vatory.

2. If, in the opinion of the Commissioners, there are at any time
reasonable grounds for assuming that, by reason of the insulation or
conductivity having ceased to be satisfactory, a sensible magnetic
field has been produced at the Observatory, the Commissioners shall
have the right of testing the insulation and conductivity upon giving
notice to the Company, who shall afford all necessary facilities to the
engineer or officers of the Commissioners, or other person appointed
by them for the purpose, and the Company shall forthwith take all
such steps, as shall in the opinion of the Commissioners be required
for preventing the production of such field.

3. The Company shall furnish to the Commissioners all necessary
particulars of the method of insulation proposed to be adopted,
and of the distances between the conductors which carry the current
to and from the carriages.

It is understood that the above clauses will be insisted on by the
Board of Trade in the case of any other tram line which can be
shown to be a probable source of danger to the Observatory.

12

Report of the Kew Observatory Committee.

The Committee are much indebted to Professor Riicker and Pro-
fessor Perry for the trouble they have taken in the matter, and they
are also glad to express their acknowledgment of the valuable
assistance rendered by the editors of scientific journals and various
eminent men of science in educating public opinion. The Committee-
even hope that ere long tramway companies themselves will recog-
nise the benefits accruing from improved insulation.

Whilst everything has been done, as far as can be foreseen, to pro-
tect the magnetographs, it is impossible to contemplate the future
without some misgivings.

National Physical Laboratory. — The Government Committee,
referred to in last year's Report, visited the Observatory on
January 18th. In the course of the summer, that Committee sub-
mitted to the Lords Commissioners of Her Majesty's Treasury a
report, embodying the following four recommendations : —

1. That a public institution should be founded for standardizing and
verifying instruments, for testing materials, and for the determination
of physical constants.

2. That the institution should be established by extending the Kew
Observatory in the Old Deer Park, Richmond, and that the scheme
should include the improvement of the existing buildings, and the
erection of new buildings at some distance from the present Obser-
vatory.

3. That the Royal Society should be invited to control the proposed
institution, and to nominate a Governing Body, on which commercial
interests should be represented, the choice of the members of sncli
Body not being confined to Fellows of the Society.

4. That the Permanent Secretary of the Board of Trade should
be an ex officio member of the Governing Body ; and that such Body
should be consulted by the Standards Office and the Electrical
Standardizing Department of the Board of Trade upon difficult
questions that may arise from time to time or as to proposed modi-
fications or developments.

In October, the Royal Society informed the Kew Observatory
Committee that the Government had adopted the report generally,
and were willing to provide funds for carrying it into effect ; conse-
quently the Royal Society asked for the concurrence of the Kew
Observatory Committee in their action.

In reply, the Committee expressed their willingness to facili-
tate the execution of the scheme, and to continue to administer
the Observatory pending the nomination of the new Governing-
Body. The arrangements were not completed before the close of
1898.

Library. — During the year the library has received publications
from

Report of the Kew Observatory Committee.

13

20 Scientific Societies and Institutions of Great Britain and
Ireland,

93 Foreign and Colonial Scientific Establishments, as well as
from several private individuals.

The card catalogue has been proceeded with.

Audit, fyc. — The accounts for 1898 have been audited by Mr. W. B.
Keen, Chartered Accountant, on behalf of the Royal Society, and
by Professor Carey Foster on behalf of the Committee.

The balance sheet, with a comparison of the expenditure for the
two years, 1897 and 1898, is appended.

Personal Establishment.

The staff employed is as follows : —

C. Chree, Sc.D., F.R.S., Superintendent.

T. W. Baker, Chief Assistant.

E. G. Constable, Observations and Rating.

W. Hugo, Verification Department.

J. Foster „ „

T. Gunter

W. J. Boxall „

G. E. Bailey, Accounts and Library.
E. Boxall, Observations and Rating-.

G. Badderly, Verification Department, and six other Assistants.
A Caretaker and a Housekeeper are also employed.

(Signed) FRANCIS G ALTON",

Chairman.

14

Report of the Kew Observatory Committee,

© US t~
© © t*i

113

E- C rt

2 ".S

J. gal
hi %.<

3 2

3"S 3
oi a

II

ill

ec oo >g

rj< 00 o

"3 o

Ill

Is s
If" -a

•IS
IS?!

.2 .so o

i5

C CO
OSS

^ O O i- o

^ o o «* o

OO S> GO

II .,

Eua

=+j 5£ £

o CI

© 35
© IN

© © ©
© 00 CO

Is

•3 o

§ 2 «o

tt cs fco

O s: £

o o s

J. S J 3
to <•* S3 i-l

5 a

,3 © © © r-t

• © oo co o

III*

ill!

t/2 £ W W

5 g

6

o
P3

J£j OS W2

owes

Report of the Kew Observatory Committee.

15

16 Report of the Kew Observatory Committee.

Comparison of Expenditure during the Years 1897 and 1898.

Expenditure.

1897.

1898.

Increase.

Decrease.

Administration : —

£

s.

d.

£ s.

d.

£

s.

d.

£ s.

d.

500

0

0

500 0

0

331

18

0

333 8

0

1

10

0

6

1

121 10

0

2

3

11

Kent, Fuel, Lighting, &c.

88

9

2

87 16

6

0 12

8

70

4

6

68 18

0

1 6

6

Incidental Expenses ....

113

2

3

137 12

1

24

9

10

1223

0

0

1249 4

7

28

3

9

1 19

2

-Normal Observatory: —

*

Salaries — Observations,

.

320

2

10

336 15

6

16

12

8

Incidental Expenses ....

48

1

4

41 1

7

6 19

9

Prop. Adm. Expenditure

244

12

0

187 10

0

57 2

0

Researches : —

110

0

0

158 S

0

48

8

0

Purchase of Apparatus,

209

11

1

64 9

2

145 1

11

Prop. Adm. Expenditure

366

18

0

375 0

0

8

2

0

Tests : —

898

11

6

918 6

0

19

14

6

Incidental Expenses ....

203

0

6

222 9

5

19

8

11

Prop. Adm. Expenditure

489

4

0

499 4

7

10

0

7

Commissions : —

Purchases for Colonial

398

18

2

529 3

1

130

4

11

Prop. Adm. Expenditure

122

6

0

187 10

0

65

4

0

55 15

0

55

15

0

G-ross Expenditure.. ..

3411

5

5

3575 12

4

373

10

7

209 3

8

(showing an increase of

£164 6*. lid.).

Extraordinary Expenditure.

Researches : —

110

0

0

158 8

0

18

8

0

Purchase of Apparatus,

206

0

7

61 15

10

144 4

9

Commissions : —

Purchases for Colonial

398

18

2

529 3

1

130

4

11

55 15

0

55

15

0

714

18

9

805 1

11

234

7

11

144 4

Leaving for Ordinary Nett

2696

6

8

2770 10

5

139

2

8

64 18 11

(showing an increase of

£74 3*. 9d.).

j

Report of the Kew Observatory Committee,

17

List of Instruments, Apparatus, &c, the Property of the Kew
Observatory Committee, at the present date out of the custody of
the Superintendent, on Loan.

lo wnom lent.

Articles.

Date
of loan.

G. J. Symons, F.R.S.

1869

The Science and Art

Articles specified in the list in the Annual

Department, South

1876

Kensington.

Professor W. G-rylls

Unifilar Magnetometer, by Jones, No. 101,

Adams, F.R.S.

1883

Pair 9-inch Dip Needles with Bar Magnets . . .

1887

Lord Rayleigh, F.R.S.

1885

Radcliffe Observa-

1897

tory, Oxford,

Mr. P. Baracchi

Unifilar Magnetometer, by Jones, marked

(Melbourne Ob-

1898

servatory) .

Dip Circle, by Barrow, with one pair of

1898

1898

The Borchgrevink-

Dip Circle, by Barrow, No. 24, with four

Newnes Antarctic

1898

Expedition.

VOL. LXV.

C

Report of the Kew Observatory Committee, 19

APPENDIX I.

Magnetioal Observations, 1898.

Made at the Kew Observatory, Old Deer Park, Bich-
mond, Lat. 51° 28' 6" N. and Long. 0h lm 15s'l W.

The results given in the following tables are deduced from the
magnetograph curves which have been standardised by observations
of deflection and vibration. These were made with the Collimator
Magnet K.C. I. and the Declinometer Magnet marked K.O. 90 in the
9-inch Unifilar Magnetometer by Jones.

The Inclination was observed with the Inclinometer by Barrow,
"No. 33, and needles 1 and 2, which are 3^ inches in length.

The Declination and Force values given in Tables I to VIII are
prepared in accordance with the suggestions made in the fifth report
of the Committee of the British Association on comparing and
reducing Magnetic Observations.

The following is a list of the days during the year 1898 which
were selected by the Astronomer Royal, as suitable for the deter-
mination of the magnetic diurnal inequalities, and which have been
employed in the preparation of the magnetic tables : —

January 3, 4, 7, 9, 23.

February 1, 3, 7, 26, 27.

March 1, 3, 4, 24, 31.

April s . 1, 9, 21, 22, 29.

May 7, 19, 21, 23, 25.

June 5, 13, 17, 20, 21.

July 2, 10, 15, 16, 18.

August 1, 8, 10, 15, 25.

September 6, 7, 12, 21, 26.

October 4, 8, 12, 16, 18.

November 5, 10. 14. 29, 30.

December 11, 12, 17, 23, 26.

20 Report of the Kew Observatory Committee.

Table I. — Hourly Means of the Declination, as determined from the

Hours

Preceding

Mid.

1.

2.

3.

4.

5.

6.

h

Q
O.

Q

in

XV.

1 1

XX.

noon.

(17° +) West

Winter.

1898.

Months.

/

/

Jan. . .

6-1

3'3

3-5

3-8

3-9

3-6

3-4

3-2

3-0

2-9

3-0

! 3-3

4*8

-C cU. . .

6-0

3-0

3-2

3-3

O O

3-3

2-9

2-8

2-7

2-4! 3 1

March .

5-4

1-3

1-3

1-3

1-2

0 9

1-0

1-0

0-8

o-i

-0-4! 0*5

2-9

Oct. ..

4-8

-1-7

-1-6

-1-5

-1-5

-1-5

-1-7

-1-7

-2-5

-3-3

-3-0

1-0-8

1-8

Nov. ..

2-2

-1-6

-1-7

-1-1

-0*8

-0-9

-1-0

-10

-0-9,-0-9

-0-8

0-3

1-5

Dec. ..

1-8

-1-5

-1-3

-0-8

-0-7

-0-8

-0-7

-0-8

-1-1

-1-3

-1-1-0-3

1

o-i

Means

4-4

0-5

0-6

0-8

0-9

0,

0-7

0'6

0*4

o-o

o-o

|,0

2-6

Summer.

/

i

April . .

6-2

0-6

0-8

0*6

0-5

0-4

o-i

0-3

-0-5

-1-0

-1 1

0-5

2-8

May ..

6-7

1-5

1-5

1-2

0-9

0-2

-0-8

-2-3

-3-4

-3 -1

-2-1

1-4

4-8

J une . .

5-7

1 -1

1-0

0-9

1-0

-0-3

-1-8

-2-8

-3-0

-2-8

-2-4

-0-3

2-9

July . .

5 3

0-9

0-3

0-2

-0-3

-l'O

-2 3

-2-9

-2-9

-2'6

-1 -6

0-5

2-7

Aug. . .

6-6

o-o

o-o

-0-3

-0-7

-1-0

-1-7

-1-9

-2-4

-2-4!

-0-9

1-3

3-4

Sept. . .

6 -4

-0-3

-0-5

-0-9

-0-8

-1-4

-1-8

-2-3

-2-4

-2-3

-1-6

0-8

3-2

Means

6'2

0-6

0-5

0-3

0,

-0*5

-1-4

-2-0

-2-4

-2-4

-1-6

0-7

3-3

Table II. — Diurnal Inequality of the

Hours

Mid.

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

Summer Means.

-0-7

-0-8

-i'o

/

-1-2

-1-8

-2-7

-3 3

-3-7

-3-7

-2-9

-0-6

/

+ 2-0

Winter Means.

-1-0

-0-9

-0-6

-0-6

-0-7

-0-7

-0-9

-1-1

-1-4

/

-1-4

-0-4

+ 1-2

Annual Means.

/

-0-8

-0-8

-0-8

-0-9

-1-3

-1-7

-2-1

-2-4

-2-6

-2-2

-05

+ 1-6

Note. — When the sign is + the magnet •

Report of the Kew Observatory Committee. 21

selected quiet Days in 1898. (The Mean for the Year = 17° l'-4 West.)

Noon.

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

Mid.

Succeeding

noon.

Winter.

/

/

/

.

,

5-6

5-7

5-0

4-4

4-2

4-0

3-7

3 4

3-1

2-8

2-9

3-0

3-3

5-6

6-0

6'4

6*4

5-9

4-9

4-5

4-0

3-6

3-6

3-2

3'0

2-9

2-6

5 6

5-5

6'6

6-6

5-5

4-3

3-5

3-3

2-9

2-6

2-4

1*8

1-8

1-6

5-1

3 3

3-7

3-1

1 -9

0*3

o-i

-0-1

-0-3

-0*8

-1-3

-1-6

-1-5

-1*8

4-5

2-6

2-9

2-0

1-2

0-8

0-6

-01

-0'4

-0-5

-0-8

-11

-1-1

-1-1

2-3

13

1-3

0-9

0 2

-o-i

-0-6

-0-6

-0'9

-ii

-1-4

-1-3

-1-2

-1-3

1*1

4-1

4-4

4-0

3-2

2-4

2-0

~1

1-4

i.

0-8

0-6

0-7

0-6

4-0

Summer.

/

5-6

7-3

7-3

5-8

4-5

3-4

2-4

1-8

1-9

1-8

1-4

1-1

0-8

6*6

7-7

8-4

7-8

5*8

3-8

2-0

1-2

1-3

1-7

1-7

1-6

1 -4

11

6-1

5*3

5'8

5 3

4-]

3'2

2-3

1-9

1-3

0-8

1-0

1 -5

1 -3

1*0

6-1

5-4

6-5

5-5

4-6

3'1

2'0

1-7

16

1-5

1-5

1 -3

1-0

0'9

6-2

5-8

7-2

6-8

5-9

4-0

2-4

1-6

1-2

1 -o

1 -o

0-9

0-6

0-4

6-9

5-5

6-4

5-3

3-2

1-3

0-2

-0-1

0-3

o-i

-0-7

-0*5

-0-4

-0-6

5-1

5-9

6*9

6-3

4-9

3'3

2-1

1-6

1-3

1*2

1-1

1-0

0-8

0-6

6-2

Declination as deduced from Table I.

Noon

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

Mid.

Summer Means.

f 4-6

+ 5-G

+ 5-0

+ 3-6

+ 2-0

+ 0-7

+ 0*1

-o-i

-o-i

-0-3

-0-3

-0-5

-0-7

Winter Means.

+ 2-6

+ 3-0

+ 2-5

+ 1-7

+ 0-9

+ 0-6

/

+ 0-2

/

-o-i

-0*3

-0-7

-0-8

-0-8

-09

Annual Means.

+ 3-6

+ 4-3

/

+ 3-8

+ 2-7

+ 1-5

+ 0-7

+ 0-2

-o-i

-0*2

-0-5

-0-6

-0-6

-0-8

points to the west of its mean position.

22

Report of the Keiv Observatory Committee.

Table III. — Hourly Means of the Horizontal Force in C.Gr.S. units (corrected

(The Mean for the

Hours

Preceding

Mid.

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

noon.

0-18000 +

Winter.

1898.

1

Months.

J an. . . .

349

351

352

351

352

353

355

358

357

355

351

347

348

Feb. ...

353

362

361

361

361

363

364

366

366

365

362

358

357

March . .

346

356

354

356

357

355

357

359

360

358

351

345

340

Oct. ...

355

369

370

369

366

368

369

368

366

361

353

348

348

Nor. .. .

366

369

369

368

370

371

374

377

376

372

365

359

361

Dec. . . .

378

381

382

382

383

384

384

384

385

383

384

385

382

Means. .

358

365

365

364

365

366

367

369

368

366

361

357

356

Summer.

April . . .

343

360

358

358

357

356

356

354

354

348

343

338

338

May . . .

362

373

372

369

369

369

367

362

352

345

342

340

341

June . . .

359

373

372

371

371

370

369

365

361

353

350

348

351

July ...

362

370

369

370

370

370

370

364

357

351

347

347

356

Aug. . . .

358

378

375

373

373

372

369

366

362

356

351

349

355

Sept

333

355

356

357

354

352

351

348

344

339

334,

328

331

Means. .

353

368

367

366

366

365

364

360

355

349

345

342

345

Table IV. — Diurnal Inequality of the

Hours

Mid.

2.

3.

4.

5.

6.

7.

8. 9.

10.

11.

Summer Means.

|+ -00005

+ -00004j+ -00004

+ -00003|+ -00002

+ -ooooi

- -00003

- -00008

- -00014

- -00018

- -00021

- -00017

Winter Means.

•00000

•00000

- -00001

•00000

+ -00001

+ -00002

+ -00004

+ -00003

+ -ooooi

- -00004

- -00008

- -00009

Annual Means.

+ -00003

+ -00002

+ -00001

+ -00001

+ -00001

+ -00002

•00000

- '00002

- -00007

- -00011

- -00015

- -00013

Note. — When the sign is + the

9

Report of the Kew Observatory Committee. 23

for Temperature) as determined from the selected quiet Days in 1898.
Year = 0-18364.)

Noon.

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

Mid.

Succeeding

noon.

Winter.

350

354

353

353

351

354

354

354

354

354

354

354

354

354

357

359

362

362

359

361

361

361

362

363

363

363

363

358

342

347

351

353

356

356

357

359

360

360

361

361

363

345

354

359

366

368

369

371

371

373

374

374

373

372

371

353

366

370

371

372

372

375

376

376

376

375

373

371

372

368

383

384

384

383

385

386

387

387

385

384

384

384

383

386

359

362

364

365

365

367

368

368

368

368

368

367

368

361

Summer.

343

350

353

354

356

360

366

366

365

363

360

361

361

342

347

353

360

364

369

374

376

380

381

380

378

377

375

361

359

365

371

371

373

376

378

380

379

376

375

373

372

355

363

365

369

375

375

377

379

380

378

380

379

376

375

360

361

361

363

366

370

374

380

382

383

384

382

382

380

364

340

349

351

352

354

357

360

363

365

365

362

362

362

350

352

357

361

364

366

370

373

375

375

375

373

372

371

355

Horizontal Force as deduced from Table III.

Noon

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

Mid.

Summer Means.

- -0001]

- -00006

- -00002

+ -ooooi

+ -00004

+ -00007J+ '00010

+ -00013

+ "00012

+ '00012

+ -00010

+ -00009

+ -00008

Winter Means.

- '00006

- -00003

- -00001

•ooooo

•ooooo

+ -00002

+ -00003

+ -00003

+ -00003

+ -00003

+ -00003 + -00002

1

+ -00003

Annual Means.

- -00008

- -00004

- -ooooi

+ -ooooi

+ '00002

+ -00005

+ -00006

+ '00008

+ '00008

+ -00008

+ '00006

+ -00006

+ '00005

reading is above the mean.

24 Report of the Kew Observatory Committee.

Table Y. — Hourly Means of the Vertical Force in C.G.S. units (corrected

(The Mean for the

Hours

Preceding

Mid.

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

J noon.

0

•43000 +

Winter.

1898.

Months.

J an. . . .

892

899

899

899

899

899

898

897

897

896

895

895

897

Feb. ...

897

902

902

901

901

901

901

901

900

900

900

898

896

March . .

891

908

908

908

906

905

905

904

904

904

902

897

891

Oct. ...

850

862

862

861

861

862

861

862

863

862

857

852

852

Nov. . . .

865

873

874

875

875

874

874

872

870

870

870

868

867

Dec. ...

868

863

863

862

862

863

864

864

864

863

863

863

862

Means

877

884

885

884

884

884

884

883

883

882

881

879

877

Summer.

April . . .

875

898

897

896

896

895

894

893

893

891

888

884

879

May . . .

878

898

897

896

896

898

898

899

897

892

885

878

873

June . . .

883

894

892

892

891

892

894

892

891

889

883

876

873

July ...

893

905

905

903

903

902

904

903

902

900

895

893

889

Aug. . . .

883

898

897

895

895

896

897

897

896

894

889

887

886

Sept. ...

830

853

852

851

850

850

850

851

851

849

846

840

837

Means

874

891

890

889

889

889

890

889

888

886

881

876

873

Table YI. — Diurnal Inequality of the

Hoursj Mid. 1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

Summer Means.

1

+ -0C003 + -00002! + -00001

i 1

+ -ooooi

+ -ooooi

+ -00002

+ -00002

+ '00001

- -00002 - -00007|- -00011

- -00015

Winter Means.

1 1

| + -00001 + -00001

1 1

+ -00001

+ -ooooi

+ -ooooi

•ooooo

•00000

•ooooo

- -ooooi

- -00002

- -00004

- -00006

Annual Means.

|+-00002j+ -00002

+ -ooooi

+ -ooooi

+ -ooooi

+ -ooooi

+ -ooooi

•ooooo

- -ooooi

- -00004

- -00008

- -00010

Note.— When the sign is + the

Report of the Kew Observatory Committee. 25

for Temperature), as determined from the selected quiet Days in 1898.
Year = 043885.)

JNoon.

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

Mid.

Succeeding

noon.

Winter.

897

898

902

901

901

901

900

900

900

900

899

899

899

897

897

899

901

904

905

905

906

905

905

904

902

902

902

892

889

892

896

898

901

902

903

904

904

905

904

903

903

899

853

855

859

863

864

863

863

862

862

862

862

862

861

850

869

873

876

877

877

878

877

877

877

875

874

874

874

869

862

864

867

866

865

866

865

865

864

863

863

864

864

859

879

880

883

885

885

886

886

885

885

885

884

884

884

878

Summer.

876

878

886

892

897

900

903

903

902

900

899

898

897

875

874

879

887

895

901

903

904

904

902

901

900

901

900

868

873

880

884

889

893

897

897

898

898

895

894

893

893

856

888

892

899

906

911

914

915

914

913

912

909

907

906

878

885

885

891

896

902

904

904

902

902

902

901

899

897

887

836

338

845

852

856

856

855

855

855

853

851

851

849

837

872

875

882

888

893

896

896

896

895

894

892

892

S90

867

Vertical Force as deduced from Table V.

I

Noon

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

Mid.

Summer Means.

- '00016

- -00012

- -00006| + -00001

+ -00006

+ '00008

+ -00009

+ -00008

+ -00008

+ -00006

-f -00005

+ -00004

+ -00003

Winter Means.

- -00006

- -00003 -00000

1

+ '00002

+ -00002

+ -00003

+ -00003

+ -00002

+ -00002

+ -00002

+ -ooooi

+ -ooooi

+ -ooooi

Annual Means.

- -00011

- -00008

- -00003

+ -00001

+ -00004

+ -00005

+ -00006

+ -00005

+ -00005

+ -00004

+ -00003

+ -00002

+ -00002

reading is above the mean.

26 Report of the Kew Observatory Committee,

Table VII. — Hourly Means of the Inclination, calculated from the Horizontal

-tiours

Preceding

Mid.

1.

Z.

Q
O.

4.

5.

6.

/ .

Q
O.

Q

lAJ.

1 1

-LA.

noon.

67° +

"Winter.

1898.

Months.

f

18-8

18-8

18-8

18-8

18-8

18-7

18-5

18 "3

18 -4

18-5

18-7

19 -0

19-0

18*6

182

18-2

18-2

18-2

18 -1

18 -0

17'9

17*9

17*9

18 *1

18*3

18 -3

March..

18-9

18 -8

18-9

18 '8

18 -6

18'7

18 -6

18-4

18-4

18-5

18-9

19*2

19-3

17-2

16-6

16 -5

16 -6

16-8

16-7

16-6

16-7

16-8

17-1

17-5

17-7

17*7

Nor

16-9

16'9

16-9

17'0

16-9

16-8

16 6

16 -4

16*4

16 -6

17-1

17-4

17-3

Dec

16-2

15-8

15-8

15-7

15-7

15-6

15-6

15-6

15-6

15 -7

156

15*6

15-7

Means . .

17-8

17'5

17 -5

17-5

17-5

17 *4

17 '3

17 2

17 -3

17-4

17 -7

17-9

17 -9

Summer.

t

/

April . . .

18-7

18'2

18-3

18'3

18-3

18-4

18 -4

18-5

185

18*8

19-0

19-3

19-1

May. . . .

17-5

17*3

17-4

17-5

17-5

17-6

17 -7

18-1

18 -7

19-0

19 -0

19 -0

18*8

June . . .

17 -8

17-2

172

17-3

17 -3

17*4

17-5

17-7

179

18'4

18 -4

18-4

18-1

July. . . .

17 -9

17-7

17 *8

17-7

17-7

17-6

17 -7

18 -1

18 '5

18 9

19 -0

18 -9

18 *2

Aug. ...

17 -9

17 0

17'2

17 '2

17 -3

17 3

17 '6

17 '8

18-0

18-3

18-5

18-6

18*2

Sept

18' 1

17*3

17 -2

17-1

17-3

17-4

17 -5

17*7

17-9

18-2

18-5

18-7

18 '4

Means . .

18 -0

17-5

17-5

17 -5

17-6

17*6

17-7

18-0

18-3

18-6

18-7

18-8

18'5

Table VIII.— Diurnal Inequality of the

Hours

Mid.

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

Summer Means.

-0-3

-0-2

-0-2

-02

-o-i

/

o-o

+ 0-2

+ 0-5

+ 0-9

/

+ 1-0

+ 1-1

+ 0-7

Winter Means.

/

K)*l

t

+ 0-1

t

+ 0'1

0 0

0 0

-o-i

/

-0'2

-0-2

-o-i

+ 0-2

/

+ 0*4

+ 0*4

Annual Means.

-o-i

-o-i

/

-o-i

-o-i

/

-o-i

-o-i

o-o

+ 0-2

/

+ 0*4

+0-6

+ 0-8

+ 0*6

Note. — When the sign is +

Report of the Kew Observatory Committee, 27

and Vertical Forces (Tables III and V). (The Mean for the Year = 67° 17' 6.)

1

JNoon.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

Mid.

Succeeding

noon.

Winter.

1

/

/

18-8

18-6

18-8

18-8

18 -9

18*7

18 "7

18 -7

18 -6

18*6

18 -7

18-7

18-7

18*6

18-4

18 3

18 -1

18-2

18-5

18 3

18 '4

18-3

18'3

18 -2

18-1

18-1

18*1

18-2

19-1

18-9

18-7

18 -7

18'6

18-6

18'5

18-4

18*4

18 -4

18-3

18-3

18-1

19-2

17 3

17-1

16-7

16-7

16-6

16-5

16-5

16-3

16 '3

16 3

16-3

16 -4

16-4

17*3

17 -0

16-8

16-8

16-8

16 -8

16 -6

16-5

16*5

16-5

16-6

16 7

16-8

16-7

16*9

15 -7

15-6

15 -7

15-8

15-6

15 6

15-5

15 5

15 "6

15-6

15 6

15 '6

15-6

15-4

17-7

17 6

17-5

17-5

17-5

17 '4

17 '4

17-3

17'3

17-3

17 "3

17*3

17-3

17 6

Summer.

/

18 7

18-3

18-3

18-4

18-4

18'2

17-9

17-9

18 -0

18-1

18-2

18-1

18-1

18'8

18-4

18-1

17-9

17-8

17*7

17 '4

17-3

17-0

16-9

17 -0

17-1

17-1

17-3

17'3

17 -6

17 '4

17-1

17 2

17-2

17-1

17 -0

16-9

16-9

17 1

17 -1

17 2

17-2

17 3

17-7

17-7

17-6

17 *4

17 6

17-5

17-4

17 -3

17-4

17-3

17'2

17-4

17*4

17-6

17-8

17 -8

17-8

177

17 -6

17 -4

17 -0

16-9

16-8

16-7

16-8

16-8

16-8

17-6

17-8

17-3

17 3

17 -4

17-4

17-2

17-0

16-8

16-7

16-6

16-7

16-7

16-7

17 -2

18-0

17'8

17-7

17 -7

17-6

17-5

17-3

17-1

17-1

17-1

17-2

17 *2

17-3

17-6

Inclination as deduced from Table VII.

Noon

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

Mid.

Summer Means.

+ 0*3

o-o

-o-i

-o-i

-o-i

/

-0-3

/

-0-5

-0-6

-0-6

-0-6

-06

-0'5

-0-5

Winter Means.

+ 0-3

+ 0-1

o-o

+ 0'1

+ 0-1

-o-i

-o-i

-0-2

-0-2

-0-2

-0-2

-o-i

-0'2

Annual Means.

+ 0-3

r

+ 0-1

o-o

o-o

0 0

-0-2

-0-3

-0-4

-0-4

-0-4

>

-0-4

-0-3

/

-0-3

the reading is above the mean.

28

Report of the Kew Observatory Committee.

APPENDIX Ik.

Mean Values, for the years specified, of the Magnetic Elements at Observatories
whose Publications are received at Kew Observatory.

Place.

Latitude.

Longitude.

Year.

Declination.

Inclination.

Hori-
zontal
Force.

Force
C.G-.S.

C.G-.S.
Units.

Units.

e

59

41

K

o

30

29 E.

1896

o

0

21 -3 E.

0

70

41

6

IS.

•16495

•47084

Katharinenburg

56

49

TS.

60

38 E.

1896

9

47 -5 E.

70

40

•o

TS.

•17811

'50765

Kasan

55

47

TS.

49

8E.

1892

7

30-8 E.

68

36

•2

TS.

•18551

•47345

•1895

10

35-3 W.

68

47

•o

TS.

•17400

•44821

Copenhagen . . .

55

41

N.

12

34 E.

1

1896

10

29 -5 W.

68

45

•6

TS.

•17422

'44824

1897

10

24 -4 W.

68

43

•8

TS.

'17450

•44826

Stony hurst . . • •

53

51

sr.

2

28 W.

1897

18

27 -6 W.

68

53

9

TS.

•17236

•44663

Hamburg

53

34

TS.

10

3 E.

1896

11

36 7 W.

67

38

8

TS.

•18061

• 43921

Wilhelmshaven

53

32

TS.

8

9E.

1897

12

41-6 W.

67

49

•o

TS.

"18028

•44213

Potsdam

52

23

TS.

13

4E.

1897

10

9-7 W.

66

36

•3

TS.

•18775

• 43398

52

16

TS.

104

16 E.

1896

2

5-2 E.

70

11

8

TS.

•20139

•55929

Utrecht

52

5

M".

5

HE.

1896

14

9-7 W.

67

4

5

TS.

•18448

'43618

Kew •

51

28

N.

0

19 W.

1898

17

1-4 W.

67

17

6

TS.

•18364

•43885

Greenwich*. . . .

51

28

N.

o

o

1897

16

50 '4 W.

167

7

6

1

•5

TS.}

•18387

( '43567
| -43546

TJccle (Brussels)

50

48

5T.

4

21 E.

1897

14

27*3 W.

66

19

5

TS.

•18917

'43145

Falmouth

50

9

1ST.

5

5W.

1897

18

42 -2 W.

•18595

Prague

50

5

TS.

14

25 E.

1897

9

21-1 W.

• 19884

St. Helier (Jer-

RPV \

49

12

TS.

2

5W.

1898

17

7-9 W.

65

52

•5

TS.

Pare St. Maur

(Paris) . . .

48

49

TS.

2

29 E.

1896

15

3-9 W.

65

1

•6

TS.

•19685

•42264

T1895

8

36 -0 W.

63

9

•o

TS.

'20731

•40951

48

15

TS.

16

21 E.

1

1896
1897

8
8

30 -5 W.
24 -8 W.

63

7

•1

TS.

•20756
•20785

•40944

1898

8

20 -8 W.

•20797

O'G-yal)a(Pesth)

47

53

TS.

18

12 E.

1896

7

46 -9 W.

•21105

46

26

IS.

30

46 E.

1897

4

47 -3 W.

62

30

•9

TS.

•22039

•42372

44

52

IS.

13

51 E.

1897

9

36-6 W.

60

28

•o

TS.

•22088

•38967

43

43

w.

7

16 E.

1897

12

18-8 W.

60

15

■4

TS.

•22318

•39059

43

40

N.

79

30 W.

1897

4

53 -0 W.

•16650

42

42

TS.

2

53 E.

1896

13

55-3 W.

60

5

•9

TS.

•22398

•38948

41

54

N.

12

27 E.

1891

10

45 -1 W.

58

4

•6

TS.

•2324

•3730

Tiflis

41

43

TS.

44.

48 E.

1896

1

53 -7 E.

55

48

1

TS.

•25670

•37775

'1894

9

41 -7 W.

Capodimonte

1895

9

37 -0W.

56

37

•9

TS.

•24007

•36454

(Naples) ....

40

52

IS.

14

15 E.

\

1896

9

32 -1W.

56

37

•1

TS.

•24040

•36484

,1897

9

26 -3W.

56

31

•4

TS.

•24075

•36406

* Of the two values of the Inclination and Vertical Force, the first is based on observations
with 3-inch dip needles only, the second on combined observations with needles of 3, 6, and
9 inches.

t Inclination and Vertical Force means from six summer months.

X Inclination and Vertical Force means from five months, January — May.

Report of the Kew Observatory Committee. 29

APPENDIX Ik— continued.

Place.

Latitude.

Longitude.

•

Year.

Declination.

Inclination.

Hori-
zontal
Force.
C.G-.S.
Units.

Vertical
Force.
C.G-.S.
Units.

- ■

o /

40 25 N,

o /

3 40 W.

1895

o /

16 6 *6 W.

o /

—

—

40 12 N.

8 25 W.

1896

i o/? • o TXT

17 35*8 W.

59 40 '2 IN .

•22620

•38662

Washington . .

38 55 N.

77 4W.

1894

3 39 '9 W.

70 34*3 1ST.

•19979

• 56646

T1896

17 35 *9 W.

58 11 *8 ~N.

•23346

'37648

38 43 N.

9 9W.

■I 1897

17 31 -6 W.

58 8*2 If.

•23385

•37624

[3898

17 27 "7 W.

58 7 -8 N.

•23413

•37660

m

Zi-ka*wei

31 12 N.

1 21 26 E

1895

2 15 *6 W.

45 55 *1 N.

•32679

•33743

Hong Kong. . . .

22 18 N.

114 10 E.

1897

0 23 3 E.'

31 36-9 N"!

•36547

•22497

19 24 N.

99 12 E.

1895

7 45 -6 E.

44 22 -2 K".

•33428

•32764

Colaba(Bombaj)

18 54 N.

72 49 E.

1896

0 33-8 E.

20 55 -6 3S».

•37463

•14326

14 35 N.

120 58 E.

1896

0 51 -0E.

16 39-7 N.

•37868

•11333

6 11 s:

106 49 E.

1896

1 22 -0E.

29 29-5 S.

•36768

•20795

20 6S.

57 33 E.

1896

9 48 -7 W.

54 32 -3 S.

•23913

•33572

37 50 S.

144 58 E.

1896

8 15'0 E.

67 18 -3 S.

•23392

•55936

30

Report of the Kew Obsewatory Committee.

o

c3
>

©

GO

O
©

3 s

•a 1

© OS

t
©

£h

©
Eh

©
i— t

•3

O

Pi
cc3
©

o .s

" 5

C5 CD-* 1> OOvOO CO WOON
A ^p OOi rfi 05000 <M IO
.5 C CM rH CM N CO CO tP CO CO CM CM

CO CD OCOK5C5

CO N CO W

tP CO CO *o J>» i> o

C5 05 OS OJ O Q CS Oi 05000005
CM <M CM CM CM CM CM CM CM CM CM CM

a ssasssSsig a sa-
fe -4 p; <5 pi g ^

r-j o th o cd oo oo o a 00 00 O -I

00 i— I 1> QO JO X> O i-l ©q CM 00 CM
N HH H H CO i-ICM

i-^—J ^P

rH CM 00 CO CO 05 rH O O O CM
.i rH 00 J>. J> OS 00 CO CM J> £* Tp
jh £>- CO CM CM COCMCO CO HH CO CO *D
06666'

OOO

OOOO

CO COCO CO CO CO CO CO CO CO CO CO

tP lOXO ION CD H3 N ID O X

CO CO 05 CM TP OS rH CM rH TP CD 00

CO OS 00 05 00 05 1— 1 O H CO 00 O

6 05 05050 6 6666

6 Ow ' — . z~. .. Z "V ~. r; _
CO CMCM CM CMCMCO CO CO CM <N CO

a b - a 2 a = a § a s a a

CM l> CD CO Tp CM Tp rH a CO CO rH CO

00

CO O CO X> CO rp r-l O HCf 05

a a s s

< fe

lO CO CO

o

O CM lO rH CM CO 1— I
rH

a a

< fe

rtf CO

rH rH 00 00 CO rH IO CM
CO rH CM CM rH ©3

00 CO CO TP

O CO Tp O N CO H ©5 CO rH O CO

00 it* tp thtpo 6 oo i>. 6 cb

10 IO lO CO i>l> 00 00 00 CO CD *0

O tP CO 1> NOJN rH O lO CO CM
Tp H CO o 10
CD CO tP Tp

OJ OO CO rPCOO CM NH 0«5

o oc i ii 6) w 6 05

rH J>» lO rH O "I 05 00 05 CO 05

•ui3aj\[

tP CM O CO HNN C5 00 tP 00 lO

rO h

S3 3 2

o »
fa hj

Report of the Kew Observatory Committee.

31

||-aT[«o

rH i—l

o

00

03

COTFrHCOCOfMCO^COCOr-ICvl

co

i

iO

a

OQ

o

CD

SB*

M0JWi0U5OC0OQ0t>W!0
i—l r-l iH i— 1

CD
05

*s

02

HNH1>M WCONTfCOiOCO

CO

umber

S.E.

iO .rtNH . .NHH^l .

iH

ft

•+7

<M .cq-^^r-icoco^coco .

CO

Wind

to

CO

to

iHOrH . CO l> CO CO iH . CO <N

1-1 .

o

w an

£ 2

Q 02

3 « W
- - -

H N lO . . ^ U5 00 N ^ lO

lCt>COt^C5COr-U>CiCS»OCD
<N N H N N

ainfall.'

Maxi-
mum.

.\OiOOOO»-OOiO»0 0 00
mOSNONN^HinmHOH

,h66666666666h

K

Total.

.oaioiooiaoooiflcio

'"OHHHNHOHOWNN

18-220

Mean
amount
of cloud

(0= clear,
10 = over-
cast).

^0M1>0 WNCOH1>I>0

cbi>j>^c©j>cbo^j>cbi>

i>

CO

o
EH

32

Report of the Kew Observatory Committee.

Horizontal movement
of the air.*

Date.

rH CM ^OCOlOOlQOCr.TflfM <M |>
CO (N CO NNHNHN cm

Grreatest
hourly
velocity.

m

V <M »0 CO CO IN 00 H* Tfl CO i> rH ©
•h M CO 00 CO CM CM CO CM CM CO lO

a

Average
hourly
velocity.

gt> co »p oo r-i Oi J> (M *> y1

9 r-Hr— Ir-lrHr-lrH rH rH

rH

rH

Minimum tempera-
ture on the ground.

<6

Is *-+
P

rH rH CO O rH CO rH O Oi CO CO O CO
»H CM rH rH CM rH CM CO CM

Lowest.

bflrH CD l> t~ 00 lO H^ C5 CO Oi Oi i>
CO CM rH rH rH CM CO CO CC CM CM rH rH

i
©

CuDtA r^t ■ \A z- *i rvn *ji <*4h

ZT^ HJ \JJ vJU rl s»J IXJ «'J UTJ xj'

CO CM CMC0-*^«?kO^ft CO CO

CO

Maximum tempera-
ture in sun's rays.
(Black bulb in vacuo.)

03

p

cm co co oo co o co ^ oo co w

CM rH CM rH i — i rH CM rH CM

Highest.

'_,•„© 1> Vfi C5 CO *C —1 HJ1 CO 05 rH TP
S*>Oi Oi rH CM CO Tf H^ CO O O X
r-w rH rH rH rH rH rH rH rH rH

Mean.

CO C^l CO ^H rH lft> CO Oi ""^ !>■
r2 CD 00 00 © rH CM CM CM rH 00 -t> CO
• rH rH >H rH rH rH

t>
Oi

Bright Sunshine.

Date.

J> CO XO CO I> rH HJI CM T? rH rH CO
CM H H rH CM rH CM

Grreatest
daily
record.

d 00 © ©©CMCDOOCMOO CO tP CM
HHJI CO CO rH CO tP rH TP CO CM rH

• U3 00 OSCMCO^TffCM© 00' CO CO
^ rH rH rH rH rH rH

Mean
percen-
tage of
possible
sunshine.

rH tP CO lO rH tP CM CO VO •— 1 CM "H

rH CM CM CO CO CO Tp tP lO CM CM <M

rH

CO

Total
number of

hours
recorded.

fZ CM CO tP tJI HH 00 CM CO CM 00 © CD
rH CO ICS 113 O H H CO H rH

• 1> 00 CM CO CO CD rH © 00 © rH
M CM CD O tP CD rH © rH CO CO W5
rH rH rH CM CM CM

1452 42

5 rQ £ &J b§J5»Sf."tt S

§ O k2 Rh^ § 3 J2 ST ■ «

Report of the Kew Observatory Committee. 33

03

©

©

be
a

• «H

02

o

©

d

GO
©

bo

-raoo aan-juiadmaj, j o

•uopisod jo sSuvvp { °<
li aiBJ jo 9Sireu.Q j o

jo uorvBUBA £\vecL j

CO
>J

<!

W

©

I

fa
O

to.
H

t3

Bel
«

X! g

123 *

£ s

< ©

^3

2

c3 !

*J ol

joj aj-BJt jo aSuuip uB9pi

Jf[rep JO UOITBUBA UBSJfl

•dn yeia

•dn }nepu8j

S5 *

fflCXQOOOOOOONt^t-.Nr^OOSfflrtO'*'*'^ CO CO CO

oooooooocooooooowoooooooooooooooooooooooo oo oo oo ao

o>co»'p^ip«irtNcoenco^^o^(»iococo>p «p a> cp <r>

t»l>t-0Clh.^l^00(ONtOt£>tO-#t-00l~l>C0Nt^ O CO 00 00

CO^O»O^OOOrtip<pi^^<piC^cpfflClCOOJ ip O CM 'p

eowKKicirannnMcoNMMnncoNmNM co eo so co

OOO^H^aNO^^OrtM^Olr(lh.Hipip © o — i co

cococococococococococococococococococorJco co co co co

Of-NNfJOO^OiOil5N00a)ONOClM«?lN t— © © ;!0

WCOCOCOu3^t-.'^TjH-^r*l^^lCOCOlOlO.O'*t^"*'<tl lO 00 O 00

n;cO'*'*C>J'*-*'*CV10'tlif5COCDOOCOrOCO'*<COT»<-* ift CM (M CM

Sooooooooooooooooooooo o ooo

w O O o o o o o o o o o o © © © o o o o o o © ©©©

0^co^coco:ocp^coco^^^co^-*-*f<Oiptp^ iO "* »p ^

So©©©©©©©©©©©©©©©©©©©© © ©oo

BiOfflfflMNfltDN^omofl'P-OOOi.ON'ilffl <M ^ 00

8!BONHH*OiO(MH(i)OOONNNH--0 CM OMh

m+i+1++I+I+++I I I I l + l 1+ + 1 + 1

S«0-^N'<i<NON"*-«0'NOOOrtHOO-«H — i O iO ?J

■" + I + I + + + + 1 i + l I I t I l + l 1+ + + + I

KrtNlO^- - ^TO^-t©OCOC>--'>J'-jt<05tOip>p'-H© ©NO

OOOJ(M<MON«MOO(NO-'Ot>J'*O^NNO) CO l£5 — ' CO

e0+l + !+ + + +|+ + +| + l l + l I I I + l + l

ol'OMNCltOtBClOClt^MOI-OOMOONiSOON CO >-H O CM

OfflNNNO-HOWO-'HOOOi'aOOHflO CO ^H — OO

M-hl + l+ + ++l+++| I I I |++| | + | + |

oJooi^cptp^aeftrocMoocot^cMoccoip©© — cm ip op co o

So — — <^H-HOJ©CO'MOCNI©--<-<CO'i'© — ,-i<Ni— I CM «IMM

co+l + l+ + 4+ ll++ l+l 11 + 11+ + l + l

: -

:use)

: f-i : u

:hronomet<

m., " Kan
ronograph

: 3 2 3
: a btc
: o o ?

L 1 1 -S -

- 3

- o

CO -

d o,
bib"2

6 ° o

.c ■ a> a> o : S> a>

-• ' oa « <n • S OT

g s 3 3 :2S

o " o ' -*

6 6 o 6 6 6 o * d

» H a! o! a tc ' '

s

d 6 d

> & °- d -d © _a c ^ ^2 d d d d d d d d d A A ^ -d
J ciiiotc^^^cicicibfibiiibiciciti bi bibJObi

i oo Q O

H t| fc| -! h u £ £ C c i\ u u £ u C *i
Q co od C Q O t/2 oo od co cb co co cc aj oo oo' co co' Q

oooti — o i>i — <

"i o o iO o ^
^ n n w m

is °

"er "3

S « o.

o->
© :

o O H
-««co(

s is « * •

OJ - CU t>»

gS^ a § I

i _ - - « g° S S3

~© -c>o . o
m - o o -oo1"©

SCO „ -C©=<!o

I t% s 1 s^s^

<u o o

© % 6
2©

_

goo
-o H 3

• c^

: CO

6 u i-s" t> £

VOL. LXV

34 Report of the Kew Observatory Committee,

•uoijesnad

uopisod jo aSaBip
i\yia. ajiu jo sSuBqo

I <=>

JO UOI1TBUBA .i\lU(J

£2222 £ 2£222£:°-2£2:222222:2:£

%%%%% 8 ££c33Sg£S8SSSS£§£8SS

•sajBi Saisoi ptre SuiureS

*J oT

JOJ JO Utjajtf

^l{Bp JO UOTJBUBA

•imop i^ja

•dn ibiq:

•dn auupua j

ill!!! ! iilltilililiiiillli

go«HHN iSOMONIOO^NHCOOOOHMNHJO

» + + ! 1+ + ++I 1+ + + + I I + + + + + + + + +

+ + + I f

-T++ 1 +

g f r °° ^ ^ '? r

»???T? + +?7T + + ++?°+?ci +++ + + +

: :

ffilii!

II

I

i p r -e s = s

II II]

i's sii f iii s f f s s f s s i f is if
iiiii i iMUiiimumi
mi i MiiiiiiMimiiM

llttilftitlPtPtflllPtl

Report of the Kew Observatory Committee. 35

J3 3

9

CO

CO

00 O rft HH

CO t> co

00

r-.t>.t>j>.

X X J> t>

00

CO CO

CO <M

^ O
O O CO

00 00 O LO

CO O ^ 00
i> 00 *>

3

-T" CO iO t>»

CJ CO X

~7<

X

IN

d

o
no"

5 S
QQ r

<u o

^ k>

O o o o

HP

. o
I § ' ^

O !3

IS -S
1114

£3 00 ^

t> C5

00 -T<
C5 CO
CO
<N CO

fH Tfl <M

CO (N CO CD

CM CO t> ©
rH O CO

<M X OS

H OH lO
^ N H
(MHH

S3 53

c «

9 *

0 o

c-> P-I

<a _,

m ™< £3
o

Ti S O

1 * s

Si o

© <D -

36

Report of the Kew Observatory Committee.

o a
& 5

00 N

lO rft r? CO

t-H C5 CO VO CO

! o

02

O

^ rf(

j> (?q n

lO <M Ol
<M O 00

1-1 to

CO

<N CO

Motion in Animals and Plants and Electrical Phenomena. 37

Csoonian Lectuee. — " On the Eelation of Motion in Animals and
Plants to the Electrical Phenomena which are associated with
it." By J. Burdon-Sandekson, M.A., M.D., F.E.S. Eeceived
March %— Eead March 16, 1899.

In a Croonian Lecture which I delivered to the Eoyal Society in
1867 — more than thirty years ago — I exhibited a number of diagrams
of graphic records, in evidence of the mechanical relations which I
then sought to establish between the movements of the heart and those
of respiration in the higher animals.

I have to-day to bring before you results which have also been
obtained by a graphic method, which however differs from the other in
that the records are written by light, and not by pen on paper ; that
the time taken in recording is measured in thousandths of seconds, not
tenths ; and finally, that the events recorded are not the movements of
the chest or heart, but the electrical changes which, as will be shown,
are found to associate themselves with all manifestations of functional
activity in living organisms, whenever these take place under condi-
tions which admit of their being investigated.

Our purpose is to consider the relation of two coincident and con-
current processes, with reference to which we make at the outset the
assumption that one is functional, the other concomitant, using the
word " function " in the biological sense to imply the doing by an
organ of the work for which it is adapted. In the observations which
I have made from time to time during the last twenty years relating
to the electrical phenomena of plants and animals, it has always been
my endeavour to regard them exclusively in relation to the functional
activity of the structures in which they manifest themselves. In in-
vestigating the function of a living organ or organism, you have to do
with a machine that you cannot take to pieces, and it is often the best
way to observe how, after its action has been arrested, it goes on again.
To do this under experimental conditions is one of the most frequently
used methods of the physiologist. The possibility of employing it
depends on the circumstance that all animal organisms, and certain
parts of plants, possess the faculty of being awakened from a state of
rest to normal activity.

Even under the most favourable conditions, the observation of this
transition is attended with difficulties which arise from the com-
plexity of the chemical and mechanical changes, and the shortness of
the time spent in their accomplishment. It is this crowding together
of chemical, thermal, and electrical phenomena into a very brief period
which determines the method for their elucidation, a method consisting
to a large extent of a determination of time-relations , i.e., of the order of

38 Prof. J. Burdon-Sanderson. Relation of Motion in Animals

succession of phenomena ; for it is evident that when you have to do
with a number of events which appear to be simultaneous, the most
effectual way to determine their causal relations is to ascertain the
order of their occurrence. For, inasmuch as one event which follows
another cannot be its cause, the proof of their sequence which accurate
time-measurement affords may be of infinite value in indicating where
the starting point in a complex series of changes is to be sought for.

The inquiry as to the relation between functional activity and the
electrical phenomena accompanying it, can only be entered upon by
finding instances in which both processes can be observed together.
Amongst these, those are to be preferred in which the question presents
itself in its simplest form, the experimental conditions can be most
easily controlled, and the observations can be made with the greatest
exactitude.

It might at first sight seem desirable to begin by describing the
electrical manifestations of functional activity in the simplest organisms
and organs. There are, however, important reasons for following the
reverse order. To do so is in conformity with the general rule that a
problem can be most easily solved when it presents itself in its simplest
form. In the lowest organisms the relation of function to structure,
so far from being simple, is necessarily very complex, for functions
of the most varied kind have to be discharged by one and the same
mechanism, and often in default of any mechanism at all that we can
discover; whereas in the higher plants and animals we find for the
most part that every kind of work has its instrument, every action its
agent. It is in the highest organisms therefore that elementary physio-
logical questions must be studied, and it is in them that they have been
most studied.

Of the elementary vital functions, motion was the one fixed upon as
the subject of this lecture by its founder. Its fitness for our purpose
is pre-eminent. Motion, in the physiological sense, is simple, control-
lable, measurable. It is, moreover, a function of paramount import-
ance as the means by which the animal organism maintains its relation
to the external world. In the higher animals, muscle is the instrument
of motion, and therefore claims our consideration. It has, in addition,
the advantage of being a structure of which the chemical, thermal,
and mechanical properties are better known than those of any other.
This advantage applies particularly to the muscles of the frog, which
on that account, as well as on the grounds which have been the occa-
sion of their being most studied, are to be preferred for our present
investigation. What we have first to do therefore this afternoon is to
determine the relation between the electrical concomitants of muscular
action and muscular action itself ; but before entering upon it, I must
occupy you for a few minutes in stating what is at present most cer-
tainly known as to the nature of that action.

and Plants to the Electrical Phenomena associated with it. 39

When a muscle is roused to activity by the presence of an exciting
cause, its mechanical properties suddenly change. It shortens and, if
the shortening is resisted, becomes tight. If the resistance is such as
cannot be overcome, it tightens without shortening. With reference
to this mechanical change, we know that it is dependent on chemical
change, and that that change is oxidation. Admitting these proposi-
tions, we must necessarily believe that the oxidation is sudden, i.e.,
explosive, because its mechanical effect (the tension or tautness I have
mentioned) attains its maximum at a very short period after the
moment at which the process begins.

At ordinary temperature we find that in a whole muscle the tension
which is induced by an excitation attains its maximum in about 3/100
second. But if we fix our attention on a single muscular element,
i.e., on one of the infinite number of molecular mechanisms by the
cooperation of which the mechanical change consequent on excitation
is brought about, it can be shown that in each taken separately a much
shorter duration than 3/100 second must be assigned to the process of
transition. According to Bernstein, less than 1/100 second must be
assigned to the chemical process which takes place in a muscular
element in response to a single stimulus.*

This chemical process of extreme suddenness is followed without
measurable loss of time by the conversion of the chemical energy of
the oxidisable material into mechanical energy, which immediately
manifests itself either in shortening or in the effort to shorten. The
way in which this transference takes place must for the present be
left an open question, for, as Professor Engelmann explained in the
Croonian Lecture for 1895, the transformation of chemical into
mechanical energy consequent on excitation of muscle, though by no
means an insoluble, is still an unsolved physical problem. We know
how much chemical energy is liberated, we know how much work is
done, and how much heat is wasted, but we cannot explain how it
happens ; it being difficult to suppose that the temperature required for
such transformation can exist in living muscle.

The absence of a sufficient physical theory of the origin of muscular
force does not, however, deprive the mechanical manifestations of the
process of their value as simple, measurable, and controllable indica-
tions of functional activity. Whether we take the case in which a

* See Pfluger's 'Archiv,' vol. 67, p. 211, 1897, where Bernstein describes his
method of measuring the period of latency. As in the method described by me in
the 'Journal of Physiology' in 1895, a magnified image of the moving surface of a
muscle excited directly is received on a slit, behind which a sensitive plate passes.
From the curve so obtained, Bernstein determines the moment at which the rate of
expansion increases most rapidly, and regards this as the moment at which the
moving force is at its maximum. This conclusion, of course, applies to the part of
the muscle immediately excited.

-40 Prof. J. Burdon-Sanderson. Relation of Motion in Animals

muscle strives against a resistance which it cannot overcome, or
shortens without resistance, or does both simultaneously, the change of
tension in the one case, of form in the other, or of both, are measurable
processes of which the time-relations can be ascertained with great
accuracy. We are on safe ground therefore in using either change of
tension or change of form as a means of estimating the vital activity of
muscle, and in fact both are required.

For every investigation in which muscular function is in question,
three points come prominently forward : (1) The moment at which
mechanical energy comes into play ; (2) the maximum energy dis-
played ; and (3) the time at which that display culminates. As
regards the first point, the time occupied before the mechanical
response begins was, for many years, believed to be 1/100 second.
This estimate was accepted as though on the authority of Helmholtz,
but was really based on a misunderstanding of his experimental data.
But we now know that the change of form resulting from the action of
a single instantaneous stimulus begins in the muscle element not later
than 4/1000 second after the moment of excitation, and I may be per
mittecl to show you how this result can be arrived at with absolute
certainty by the photographic method. (Photograph shown.)

As regards the second and third points, we find it better to measure
contractile activity by change of tension rather than by change of form,
firstly, because the method of measuring tension is less liable to error,
and, secondly, because the process of development of tension is more
rapid than the development of change of form. For, although with
the exquisite methods we now possess of getting rid of inertia in our
recording apparatus, it is possible to measure the shortening with great
exactitude, yet it is easier to guard against errors of observation when
the other (isometric) method is used.

Having seen how functional activity can be measured, we may
advert to the question that principally concerns us this afternoon
■ — the question, namely, whether the electrical phenomena may also
be regarded as expressions of functional activity. Assuming for the
moment the question to be answered in the affirmative, with what
part cf our tension curve should we expect the electrical concomitant
to coincide 1 Assuredly with the first and second hundredths of a
second after excitation, i.e., with the period of greatest activity of the
unknown process by which chemical is replaced by mechanical energy,
and, for a reason which I will at once explain, with the very begin-
ning of that process. For, as I have already indicated, the transition
does not occur at the same moment everywhere, and inasmuch as the
method which we use for the investigation of electrical change takes
cognisance only of what happens within an area of a couple of milli-
metres, we should expect it to occur not at a moment corresponding
to the maximum development of tension in the whole muscle, but at

and Plants to the Electrical Phenomena associated with it. 41

the moment at which the transition process is going on with the
greatest rapidity in the elements immediately concerned within that
limited area. Assuming for the moment that this rapidity is expres-
sible as a measurable electromotive force, we should expect the appear-
ance and disappearance of that electromotive force to be represented,
not by a curve resembling the tension curve, but by a curve of the
form indicated in the diagram. (Curve P' in Diagram 4.)

Before losing sight of the mechanical changes which have for the last
few minutes been occupying our attention, there are two other points
which must be shortly adverted to on account of their bearing on what
follows. The one is the terminableness and the cyclical character of the
mechanical process. The muscle returns to its status quo at a certain
time after it has been disturbed, a time strictly dependent on tempera-
ture and other well ascertained physiological conditions. We do not
know as yet how it relaxes, whether it is merely a physical reaction, or
whether it is by the intervention of a new chemical process. This is a
qua'stio vexata which for the present must remain open.

The second point is that although the mechanical process is limited
in time it is not limited in space. If it were possible to imagine a con-
tinuum of contractile protoplasm, an excitation once started would go on
for ever, i.e., it would be propagated from element to element — in every
direction if it were of the nature of cardiac muscle, in two directions
only if it were of the nature of skeletal muscle. For this process to
take place we suppose that each element excites its neighbour. In each
transmission the time lost is almost infinitesimal, yet by summation it
acquires a definite value, so that the relation between distance travelled
and time occupied can, when the temperature and other conditions are
known, be foretold. In so far as each element transmits its state of
change to its neighbour without loss it resembles the propagation of
light and sound, but the velocity of propagation is of so different an
order that the comparison must not- be carried too far.

Method of Observation.

We are now in a position to enter on the inquiry which more
immediately concerns us. Having the order of the mechanical changes
which constitute muscular action before us, it will be our purpose to
compare with this order that of its concomitant electrical phenomena.
Before I proceed with this comparison it is desirable to say that
it should be understood that no reference will be made to electrical
theory. We merely derive our modes of observation and of measure-
ment from the exact sciences, and aim at the utmost attainable
precision ; but the phenomena have their chief interest as outward
and visible signs of intimate vital processes, of which they afford us the
only knowledge that is within our reach.

42 Prof. J. Burdon-Sanderson. Relation of Motion in Animals

We choose as our subject of observation a muscle of nearly sym-
metrical form — a band of parallel fibres. AVe explore its electrical
state, by a conducting arch containing a galvanoscope, the ends of the
arch being in contact with its surface. If the muscle is no longer
living, the galvanoscope gives no evidence of current. If it is living,
there is again no current, provided that the two surfaces are in the
same physiological state. If one is less living than the other the fact is
indicated by a difference of potential between them, a current flowing
through the galvanoscope from the more living to the less living. Vitality is,
therefore, here indicated by difference of potential. By vitality we
mean nothing more than the capacity for discharging function. This
capacity diminishes by discharge, i.e., by activity. Accordingly we
find that when, for any reason, the muscular substance at one part
becomes more active than the muscular substance at another, the former
becomes negative to the latter.

Every observation of the electrical phenomena of muscle (or of any
other excitable structure) relates either to the state of capacity for
action (called in physiology "rest"), or to the state of action or dis-
charge. In either case it consists in comparing the states of two
contacts,* i.e., of two parts to which electrodes of a galvanoscope are
applied. It is obviously desirable for the investigation of the changes
at either, that those which take place at the other should be annulled
during the period of observation. On this consideration a rule is based,
to the mode of carrying out of which. I will advert presently.

Most of the results which I shall place before you were obtained
with the aid of the capillary electrometer, of the use of which as an
aid to electrophysiological investigations I brought before the Eoyal
Society some instances nearly twenty years ago. Its application has
since been studied with great completeness by Mr. Burch, to whose
skill I am indebted for the instruments which I have used for my work
during the last ten years, and more particularly for the one which has
enabled me to submit to you the photographic results I am now about
to exhibit. These photographs, I need scarcely explain, express the
excursions of the meniscus of the mercury column as a sensitive plate
moves rapidly past the slit on which it is projected, each upward move-
ment of the image indicating that the surface of contact connected
with the mercury has become at that moment positive to the other.

I do not propose to give this afternoon even the shortest description
of the instrument, and I should not occupy time in explaining why it
answers my purpose so perfectly, were it not that with the exception
of Professor Einthoven and Dr. R. du Bois-Reymond the leading au-
thorities on the other side of the Channel, and particularly Professor

* It may be well to note that the contacts referred to here and elsewhere are
made by means of non-polarisable electrodes of the kind originally devised by
du Bois-Reymond and always used in physiological work.

and Plants to the Electrical Phenomena associated vjith it. 43

Hermann, have condemned it as an instrument of which the defects
are essential and irremediable. As I have answered these criticisms
elsewhere, I need only say here that for the investigation of the order
and duration of a rapid succession of electrical changes, such as those
with which we are now concerned, the instrument surpasses all others ;
and that by means of it my colleague, Professor Gotch, has with Mr.
Burch's aid, successfully photographed phenomena in nerve, of which
the very existence could not be demonstrated a few years ago.*

The purposes to which we apply it are (1) for the measurement of
intervals of time between electrical changes which succeed each other
with great rapidity, and (2) the obtaining an estimate of their relative
intensities. The properties which make it so invaluable to us are
(1) that it responds to the action of a current promptly, beginning
when the current is closed, and indicating every change in its strength
or direction without measurable loss of time ; (2) that east. par., the rate
of ascent is proportional to the electromotive force of the current which
produces it ; and- (3) that the instrument can be graduated, and its
graduation verified by comparison with instruments of greater precision,
and thus used for the measurement of differences of potential of longer
duration.

The diagrams 1, 2, and 4 illustrate the bearing of these three pro-
perties on the cases we have to investigate. As we shall see, a muscle
can be brought into action either by an instantaneous stimulus, by a
series of stimuli, or by continuous stimulation. Each of these has its
mechanical and its electrical response. I will anticipate so far as to say
that the three forms of electrical response correspond to the three
forms of mechanical. They correspond to the changes indicated by the
black lines in the three diagrams. I will further premise that all known
excitatory responses — all electrical changes which are concomitants of action
— may he compared with one of these types.

Case 1. — Response to a continuous stimulation. A difference of
potential comes into existence at the contacts at the time t, and persists
long enough to produce its full effect on the column. (Diagram 1.)

Case 2. — Series of short continuous stimulations. The column,
moves in alternately opposite directions. (Diagram 2.)

Case 3. — Response to a single instantaneous stimulation. A differ- -

* Full information relating to the instrument will be found in Mr. Burch's work
'The Capillary Electrometer in Theory and Practice,' and his papers in the
'Proceedings' (rol. 48, 1890) and 'Transactions' (A, vol. 183, 1892) of the Eoyal
Society. A very perfect method of recording the excursions of the electrometer
photographically and of interpreting the curves was described by Prof. Einthoven
in Pfiuger's ' Archiv ' in 1894, and applied by him to the investigation of the •
electromotive phenomena of the human heart. It need scarcely be added that the
two methods are the same in principle. An important paper has also recently
been published by Dr. R. du Bois-Reymond in the ' Archiv f . An. u. Physiol.,' 1898,
p. 516.

44 Prof. J,

Burdon-Sanderson. Relation of Motion in Animals

•ence of potential comes into existence abruptly, and subsides abruptly
at first, afterwards less rapidly. (P in Diagram 4.)

Xow I have found that in the study of my experimental results it
is of great advantage to proceed a priori. Let us assume that there
are three types of stimulation, and that each has its form of response.
We can best begin by inquiring to which of these three forms the
•observed variation belongs, and then determine in what respects it
conforms with, or differs from, the type.

In the diagrams, I have shown the types of photographic curves
which correspond to the three forms of response to stimulation I have
indicated. The faint lines represent photographic curves ; the strong,
variations of potential-difference. In each diagram the strong and the
faint lines have been drawn in their true mathematical relation to each
other, i.e., so that the vertical distance apart of strong from faint is every-
where proportional to the gradient or slope of the photographic curve, the
proportion being such that if the E.M.F. of the current acting on the electro-

and Plants to the Electrical Phenomena associated with it. 45

meter varied according to the strong line, the movement of the head of the
mercury column would be expressed by the faint line. We shall see as
we proceed that one or other of the three forms of photographic curve,
which correspond to the three forms of electrical change, just desig-
nated as typical, presents itself in every excitatory response we have
to investigate, provided that, as I mentioned just now, the changes
under one contact only are recorded.

To ensure this, the exploring contacts must be so arranged, and the
muscle itself so prepared, as to enable us to separate the part of the
surface we desire to investigate from the rest, so far as concerns its
effect on the instrument we are using as indicator. It is obvious that
when we apply our leading-off electrodes to two parts of the surface,
both of which are at the same time undergoing change, there must
always be a difficulty in determining how far the effect is due to
changes at the one or at the other contact. It is therefore essential
for the correct observation of an electrical change at one of them, that
the other should be protected from disturbing influences.

The Fir.4 Fundamental Experiment.

An experiment will show how this may be accomplished. It will
also bring us face to face with a phenomenon which is, perhaps, the
most fundamental of those which at present concern us, the phenomenon
of the wave of excitation, or, to use the designation given to it by its
discoverer, the Reizwelle. The nature of the experiment is illustrated
by Diagram 3, in which the band of parallel fibres represents the
sartorius muscle. It is excited (instantaneously) at r. A change

Diagram 3.

occurs there which is propagated first to the proximal contact p, and
then onwards to the distal contact d, at a rate which in our preparation
may be 150 cm. per second; This change is essentially a vital one,
but it is attended by a mechanical change represented by the muscle

46 Prof. J. Burdon-Sanderson. Relation of Motion in Animals

curve, and an electrical change, which we record photographically.
Diagram 4 will serve to explain what (as will be immediately seen)

Diagram 4.

Explanation of Diagram 4. — The horizontal line
is that of equipotentiality of the two surfaces of
contact p and d. The curve P' expresses the rela-
tive negativity (negative difference of potential) of
the surface p ; the curve D' the corresponding rela-
tive negativity of the surface d. S is a curve of
which the ordinates are the algebraic sums of the
corresponding ordinates of P' and J)'. 8 is the
photographic curve which expresses S' ; P' is the
photographic curve which expresses P. The num-
bers under the horizontal line indicate hundredths
of a second. The distance 1 1' expresses the time
taken by the wave in its progress from p to d.

actually happens at the moment the wave passes under p. It means
that a current suddenly appears there, of which the direction is from
p to d. When the wave reaches d, a second effect of the same kind (D')
occurs, of which the direction is opposed to the first. What the gal-
vanoscopic effect of this must be is easily understood from Diagram 4,
in which the two curves P' and H are placed in a relative position to
each other which expresses their time-relation. The two effects sum
together. In the diagram the curve S' expresses the result of that sum-
mation, i.e., the actual variations of difference of potential between the
contacts which occurred while the wave was passing from p to d. It
will be seen at once why we call this effect the diphasic variation.

I explained before, that in accordance with the fundamental pro-
perties of our instrument the curve P would have as its photographic
expression the curve P. Similarly the combination-curve S', would

and Plants to the Electrical Phenomena associated with it. 47

have for its photographic counterpart the curve S. May I emphasize
the point that if you have the curve F of a parallel-fibred muscle, you
•can calculate from it S' and consequently S, but that from S alone you
cannot deduce the others. In other words, if you know the form
of P', you know everything as to the form of the electrical response —
the Reizwelle.

Let us now take the actual result. As before stated, the two contacts
are at p and d, and the muscle is excited at r. The wave affects the
muscle first abp then at d, and the consequent movement of the column
is photographed (Photo. 1).

Photograph 1.*

You recognise that it is the counterpart of the deduced curve S. In
other words it is the expression of the effects of two similar processes
having their seats at the two contacts. Our aim must now be, as I
have explained, to annul or suspend the effect of one of them, leaving
the other intact. The method is simple. After having obtained the
record I have shown you, I tie a fine thread round the muscle between
p and d. I tighten the ligature so as to constrict the muscle and again
record the variation. There is no change of effect, for the wave is still
able to pass the constriction. I tighten again : it still passes. I then
draw the ends of the ligature hard, and again photograph. I find
the photographic curve is no longer S but P, i.e., it has assumed the
characteristic form of the monophasic electrometer curve (Photo. 2).

"What has the ligature effected 1 It has exercised no influence on
either contact, but it has arrested the progress of the excitatory wave,
so that its effect at p only is manifested, and not that at d. The rela-
tion between the two curves (P and S) is obvious enough when they
are seen in succession. It will be still more obvious if I place them on
the screen together, in such a way that they are in synchronic rela-
tion to each other (Photo. 3).

* Photographs 1, 2, and 3. — jCurarised Sartorius kept for ahout twenty-four
hours iu 0*6 per cent, solution of chloride of sodium. Temperature during obser-
Tation 9° C. Contacts, &c, as in Diagram 5, but p much nearer to d.

48 Prof. J. Burdon-Sanderson. Relation of Motion in Animals

Photograph 2 *

o-7see.

The experiment may be further varied by altering the seat of excita-
tion from r to /. You thus obtain a photographic record which
represents what happened at d in the unligatured muscle. If the muscle
is in a normal state, this is an exact reversed counterpart of photo-
graph 2.

If instead of placing the ligature half way between p and d, we
place it close to the distal electrode d, the proximal may then be
placed in a succession of experiments at different distances from the
seat of excitation without altering the form of the recorded variation ;
the time at which it begins depends in each case on the distance of
the proximal contact from the seat of excitation.!

In all of these instances the ligature acts as a block. Without
interfering with the condition of any other parts it kills the part
which it grasps and makes it incapable of transmitting the excited

* Photographs 1, 2, and 3. — Curarised Sartorius kept for about twenty-four
hours in 0'6 per cent, solution of chloride of sodiuni. Temperature during obser-
vation 9° C. Contacts, &c, as in Diagram 5, but p much nearer to d.

t An experiment of this kind is by far the most exact method which we possess

of measuring the conduction -rate in muscle. This rate is most correctly expressed

, L. , . , Difference between the distances -, . ,

by the quotient = - as measured in two experiments

Difference between the times

in which the distances are different.

and Plants to the Electrical Phenomena associated with it. 49

state from the living structural elements on one side to those on the
other ; but if we compare the condition of the unexcited preparation
immediately before and after the application of the ligature, we find
evidence that breach of continuity of function is not the only effect
produced by it. If the one contact is placed on the ligatured part it
is found that, irrespectively of any excitation, there exists a large
difference of potential between the contacts, which may amount to
four or six hundredths of a volt.

The Muscle Current.

Now it is easy to prove that this difference is not due to breach of
continuity, for if you shove the electrode away from the ligature in
either direction it disappears. The phenomenon which is thus brought
to light is that to which the great founder of animal electricity, du
Bois-Eeymond, applied the term "muscle-current," and when the
method I have described is employed, it presents itself in its utmost
simplicity — for by the act of tightening the ligature previously applied
under an electrode, you at once bring into existence a state of things in
which the constricted part is negative to the living parts on either side.

What happens in this case 1 What is the difference between the
state of the surface of contact immediately before and immediately
after the tightening of the ligature ? Nothing more can be said than
that a certain process which was going on there and which provisionally
Are call " life," being ignorant of its nature, has been annulled. What
we actually observe may be represented diagrammatically thus : —

Diagram 5.

The divided line represents the graduated wire of a potentiometer ; at d is a ligature
as yet not tightened round a muscle ; p and d are equipotential. The galvano-
meter is at zero and the slider of the potentiometer is up to the block.
The ligature is tightened ; at once the needle indicates a current directed from
d top, but can be brought by the slider again to zero.
VOL. LXV. E

50 Prof. J. Burden-Sanderson. Relation of Motion in Animals

The contacts are as shown in the diagram. Before tightening the
ligature between them they are equipotential, because they both rest
on muscle in the same physiological state. I represent the electrical
concomitant of that state by an arrow, by which I mean nothing more
than that if it were possible to connect p with some other part of the
anuscle, without passing through another electromotive surface, there
would be a current in that circuit from p to the galvanometer. But
inasmuch as the actual circuit passes through d where the same condi-
tions exist as at p, but opposed in direction, there is no current. If by
tightening the ligature I annul the effect of d, the effect of p comes
into evidence. This statement is simple, and seems to arise naturally
from the observed facts ; but cannot be received without question, for
it suggests that what we call the " demarcation current" has its seat,
not at the surface of demarcation, but at the living surface, so that we
should have to consider the state of " Stromlosigkcit " not as a state of
electrical inaction, but as a state of balance.

A similar question would arise as regards the response to excitation.
For when, as we have seen, the Reizwelle passes under the proximal
contact (Exp. 1), what happens there (during the 100th of a second
that it is passing) is analogous to what I have just described as the
effect of suddenly tightening a ligature at that spot. The moment
before excitation a state of balance existed between p and d. As the
wave passes under p it upsets that balance by annulling the outgoing
current, then pursues its course until it is extinguished by the ligature.
From the moment that the tail of the wave has left the edge of the sur-
face of contact behind, it has no action whatever on the indicating
instrument. We have evidence of this in the curve of variation itself,
for the form of the curve is the same whether the wave is blocked by
the ligature at one centimetre from the point of observation, or at
three, which could not be the case if, as I once imagined, something
happened at the moment of extinction.

The complete proof that this is so, is however obtained by another
form of experiment in which the seat of excitation (r) is shifted from
the proximal side of p to the proximal side of d. The unligatured and
therefore equipotential muscle is excited in the two positions succes-
sively. The results show (1) that the excitation wave is propagated in
both directions, and (2) that the form of the curve varies according to
the order in which the electrodes are reached. This having been deter-
mined, the progress of the wave is stopped by a ligature under the
distal contact d, and the excitations in the two positions repeated. It
is now seen that the form of the wave is the same whatever the direc-
tion from which it approaches the point of observation p. When the
excitation is proximal to (/, it is not now anticipated by a variation at
</, and there is consequently a long delay (see Photograph 4) during
which the electrometer is unaffected. The experiment affords direct

and Plants to the Electrical Phenomena associated with it. 5.1

Photograph. 4.*

evidence that, although the whole muscle is in circuit, the presence of
the wave cannot reveal itself until it is under the electrode. As regards
the action-current therefore, the electromotive source is always the
surface of contact of the leading-off electrode with living substance,
not the surface of contact between dead and living.

We may now resume our consideration of the form of the propagated
monophasic variation, or excitatory wave. It would be easy to prove by
the exhibition of numerous photographs of the monophasic variation
relating to different muscles, that all have the same characteristic
features, indicating that in each muscular element the electrical change
culminates from two to six thousandths of a second after excitation,
according to the physiological state of the muscle and the time it has
been kept, subsiding at first abruptly, afterwards more gradually, so
that its whole duration (i.e., to the summit of the electrometer curve)
amounts to from two to six hundredths of a second.

The discoverer of the Pieizwelle, Professor Bernstein, assigned to it a;'
very different duration. " In every element of muscular structure^
the variation lasts between 1 250 and 1/300 second, and coincides with
the period of latent stimulation." At first sight this statement seems irre-
concileable with fact, but it is much less so than it appears to be. ' We
have only to assume that Bernstein's method of estimating a small and
transitory difference of potential between two surfaces, was not suffi-
ciently delicate to enable him to appreciate those which exist during
the period of decline, and that what he regarded as the duration of the
whole variation, was in reality the duration of its summit only. How-
ever this may be, it is clear that we may divide the period of variation
into two parts, which we may call respectively the initial rise and the
decline, of which the latter lasts eight to ten times as long as the for-
mer : and that we may regard the first as a period of upset, the second
as a period of restoration. Taking the period of upset as equivalent to

* Curarised Sartorius kept for twenty-four hours. Seat of excitation between
the leading off contacts, 4 mm. from d, 20 mm. from p.

52 Prof. J. Burdon-Sanderson. Relation of Motion in Animals

Bernstein's " negative Schwankung," we can accept all he says as to its
coincidence in time with the moment of greatest intensity of the pro-
cess by which chemical is transformed into mechanical energy — the
moment in the shortening of an unloaded muscle at which its rate of
change increases most rapidly. As regards the period of decline, it might
suggest itself that the return of each element to its previous state is in
every instance the expression of an anabolic process, not merely a
result of the cessation of the opposite process. The facts we are con-
sidering, however, lead us for the present to regard the whole varia-
tion as the concomitant of one and the same chemical process, and we
are confirmed in this view by the observation that, as Ave shall see im-
mediately, the modifications which the monophasic variation undergoes
under external or accidental conditions affect both stages equally.

Of these conditions one of the most important is temperature, par-
ticularly when muscles which have been kept for some time in physio-
logical salt solution are used.* We have hitherto had in view the Sar-
toriiis which has been kept for some twenty-four hours, and is at the
temperature of about 10° C. By placing it in a cooled chamber at
a temperature some 6° C. lower, and allowing it to remain there until
it has acquired the temperature of its environment, its mode of re-
sponding is not changed, but only in its relation to time. In shortening,
it takes a longer time to attain its minimum length, and if its con-

Photograph 5.

traction is resisted, its period of effort is of longer duration. Conse-
quently it is able to do more external work in a single effort than
before, although it is not able to support a heavier weight or maintain
a greater tension in a continuous effort. Now all these modifications
depend, so far as I have been able to ascertain, on diminution of the
rate of propagation of the excitatory wave. As has been already
stated, we are able to measure this rate with great facility and accu-
racy. By alternately cooling and warming our chamber we can deter-
mine in any number of instances the change of rate which a difference
of 2°, 4°, or 6° C. produces, and compare the data so obtained with the
effects of the same changes on the duration of the monophasic variation
and on that of the mechanical effort which it accompanies.

* ' Journ. of Physiol.,' vol. 23, p. 332.

and Plants to the Electrical Phenomena associated with it 53

Up to this point the phenomena we have had under consideration
have been associated with the response of a muscle to a single instan-
taneous excitation, i.e., the monophasic variation and the momentary
contraction which it ushers in. We must now pass on to the considera-
tion of the electrical concomitants of those forms of contraction which
more obviously resemble the natural action of muscles.

Physiologists have for half a cent my taught that natural muscular
action, whether reflex or voluntary, is made up of single contractions
of definite duration, such as those we have been considering, i.e., of a
rhythmical series of such contractions of definite frequency. This doc-
trine— that voluntary motion is a well organised system of twitches — is
now commonly expressed by calling it a tetanus, a word which was
some fifty years ago diverted from its medical signification to be
adopted as a technical term in physiology, but not precisely in its pre-
sent sense. What is now meant by it is that every contraction, how-
ever continuous it may appear to be, is in reality discontinuous. Tips
conclusion was arrived at by a method which, though sometimes of
great value to the physiologist, does not always lead to the discovery
of truth — the method which consists in first imitating a natural pro-
cess, and then mentally transferring the characteristics of the imitation
process to the natural process which it represents. In the present
instance the study of artificial tetanus has taught us a large proportion
of what we know as to the properties of muscle, but not much about
voluntary contraction. In assuming the identity of the latter with
experimental tetanus, physiologists have perhaps minimised certain
fundamental difficulties and assigned undue value to certain analogies.

Of the difficulties, the most obvious one is that discontinuity could
not, if it existed, be of any advantage. For if we regard the muscular
system as the mere instrument of the central nervous system, and every
muscular fibre as the instrument of the motor cell which governs it, it
is difficult to see how subjecting that muscular fibre to a rhythm of its
-own could have any other effect than to interfere with its efficiency.
Of the analogies the chief are, first, that just as when you listen to a
muscle in artificial tetanus you hear a musical sound of which the
frequency of vibration corresponds to that of the stimuli, so a muscle
when contracting voluntarily gives out the quasi-musical sound, which
Wollaston compared to the rumble of wheels over pavement. The
other analogy relates to the reflex spasm of strychnine,1 which is not
only rhythmical in itself, but is accompanied by a series of ' electrical
changes which are as rhythmical as if they were evoked by a series of
stimuli. The discussion of the muscle sound lies outside of our present
inquiry: the spasm of strychnine will be considered after we have
examined the electrical concomitants of artificial tetanus... "..

54 Prof. J. Burdon-Sanderson. Relation of Motion in Animals

Second Fundamental Experiment*

The point to which I have first to draw your attention is the form of
photographic curve, which is obtained when the Sartorius, injured under
one electrode by a ligature, is excited by a series of stimuli of which
the frequency is about 60 per second. The photograph shows that

Photograph 6 *

the .column rises at first abruptly, but afterwards in such a way
that; the rate at which it rises is at any moment proportional to its
distance from the point to which it will eventually arrive, i.e., to the
distance between the corresponding point of the curve and its asymp-
tote. The electrical state, therefore, which comes into existence when
a muscle is tetanised (i.e., subjected to a frequent series of excitations)
corresponds to diagram 1. In other words, the electrometer is acted
on by the same difference of potential between its terminals through-
out, with the exception that the effect of the first, or first couple, of
excitations is often greater than that of the succeeding ones. Although
this hardly needs proof, it can be easily verified by direct experiment.
With this view our eircuit is so arranged that we can, without altering-
the resistance, project on to a second photographic plate the effect of
allowing a constant difference of potential to act on the mercury column
just as the plate is passing behind the slit. On comparing this curve
with the tetanus curve they are found to be nearly identical.

Let us now take the case in which a muscle is tetanised in the same
way as in the last instance for a succession of periods of one-fifth second,
alternating with equal periods of rest (Photo. 7). The complete cor-
respondence of the photographic curve with that represented in

* Sartorius not cura^ised. Indirect excitation. The undulations on the line of
ascent indicate (he frequency of the stimuli — 60 per second. The radial line i
indicates here, not the moment of stimulation hut that at -which the short circuit
of the secondary coil was opened.

and Plants to the Electrical Phenomena associated with it. 55

diagram 2 indicates that the conditions correspond with those which
are there theoretically represented. During each period of excitation
(tetanus) the movement upwards of the meniscus is determined by the
difference of potential. During the intervals it follows in its fall the
similar curve of depolarisation.

Photograph 7.*

From this we may now proceed to other forms of experimenta
tetanus in which the excitations are less frequent. Provided that the
frequency is not much less than 40 per second, the general contour of
the curve resembles the other one, with the exception that the effect. of
each excitation is seen separately (Photo. 8). If the frequency is
diminished to 20 per second the undulations are more ample, while the

Photograph 8.f

curve rises to a lower level, the reason obviously being that the electro-
meter is acted on by a smaller number of excitations in a given time.

* Frequency of excitation as in Photo. 6. The original shows similar undulation*
in the ascents, which the copy by inadvertence does not show,
t The first four undulations have been imperfectly copied.

56 Prof. J. Burdon-Sanderson. Relation of Motion in Animals

Diminishing the frequency still further (to 14 per second) we obtain a
curve (Photo. 9) of which the character is that of a series of equal and
similar monophasic variations.

We now go on to compare the variation-curve of artificial tetanus
with the nearest approach to a normal contraction we can obtain for
investigation, viz., the reflex response of the motor apparatus of the
spinal cord to an instantaneous stimulation of the cutaneous surface.
A ligature is applied as before to the tibial end of the Sartorius under
the distal contact ; but inasmuch as the muscle must now be excited
through its nerve, the proximal leading-ofF contact is on the hilus.
The mode of excitation is the same as before, but in this case the effect
has first to be communicated to the motor cells of the spinal cord
through the sensory apparatus, a process which occupies a relatively
considerable length of time. The motor cells then deal with it auto-
matically, responding to it in their own way, and inducing in the
muscles under their control an action which is the faithful and exact
expression of the changes going on in themselves.

As is well known, it is not possible in a normal preparation to obtain
an unfailing response to an instantaneous stimulus applied to the
cutaneous surface, but the previous injection of a trace of a strychnine
salt (e.g., 1/30 milligram of the sulphate) is sufficient to give to the
motor apparatus of the cord the required degree of excitability. A
single induction current applied to the skin then evokes in the
Sartorius and other muscles, first a twitch which resembles the
response of the same muscle to a similar stimulus applied to its nerve ;
a little later, this twitch is replaced by a short, sometimes thrilling,
spasm resembling a short tetanus. What I have to show you is that,

Photograph 9.

The Reflex Electrical Response.

and Plants to the Electrical Phenomena associated tvith it. 57

although the reflex spasm resembles a short artificial tetanus as regards
the way in which the muscle contracts, the contractions are shown by
their electrical concomitants to be of a different nature. The strych-
nine spasm, as it is rightly called, is seen not to be a tetanus, i.e., not
to consist of a series of single twitches, but to be a succession of con-
tinuous contractions, the rhythm of which depends on the spinal cord,
not on the muscle.

Photograph 10.*

The grounds on which this conclusion is founded appear to me to be
unequivocal. The observation is a simple one. The automatic
mechanism, which carries the photographic plate, liberates as before,
at the beginning of the period of exposure, an induction current which
pricks the skin of the preparation. After an interval which may be
about a tenth of a second (during which a ^^^si-psychological process is
going on in the spinal cord) the muscle responds. A curve is drawn
simultaneously by the writing lever to which the end of the muscle is
attached, which indicates that it is in spasm ;f but it is the photo-
graphic curve which tells us the nature of that spasm. Each ascent of
the meniscus is seen to be the response, not to a single instantaneous,
but to a short continuous, stimulation, of which the duration can be
easily deduced by measuring the time interval between the beginning
and the culmination of an excursion. By subjecting the muscle arti-
ficially to series of excitations of similar duration with corresponding
intervals of inactivity, one can produce an imitation of the strychnine
spasm which, both in its mechanical and electrical characters, resembles
the natural one (see Photo. 7).

* Freshly prepared Sartorius attached to pelvis and connected to spinal cord by
its nerve. Leading off electrodes on hilus and tibial end. Exciting electrodes
applied close together to skin of flank of decapitated preparation.

f The curve is often toothed, the teeth corresponding in frequency with the
electrical undulations.

58 Prof. J. Burdon-Sanderson. Relation of Motion in Animals

Before leaving the subject of the strychnine reflex, I must refer very .
briefly to such previous observations as bear on our present inquiry.
The phenomenon is of interest as being one which could not have been
discovered had we not possessed the capillary electrometer. Its dis-
covery was, indeed, the outcome of the first attempt made by Professor
C. Loven to use that instrument for the investigation of the electrical
properties of muscle just twenty years ago. He was good enough to
make for me the electrometer which was used in some of my own
earliest experiments. Shortly afterwards, Mr. Page devised the
method of obtaining photographic records of our own results and,
amongst others, of those of Loven relating to the strychnine spasm.
Loven's observation has served ever since as a support for the doctrine
of discontinuity. No one would be tnore willing than he would, if he
were with us this afternoon, to recognise its true meaning.

The conclusion to which all the facts we have had before us up to
this moment lead, is that normal muscular action is the manifestation
of what happens in the motor nervous system. If this motor impulse
is so short that we are obliged to call it instantaneous, the response is
correspondingly brief ; if it lasts longer, we call it continuous, recognis-
ing that the difference between the two is merely one of duration.
In either case it is of the essence of the response that it is terminable.
There is no difficulty in understanding on teleological grounds why a
muscle must relax ; but of the mechanism by which it is brought about
we know little, excepting that it is localised in the muscular structure.
Each element — each tagma — returns to its status quo in the same way
in a curarised muscle as in a normal one ; but whether this power of
recovery is a process by itself, as some physiologists hold, is a question
which is at this moment much debated, but by no means settled. It is ,
only in so far as it relates to the electrical concomitants that it here
concerns us. Without prejudice to the question whether, as Fick and
Gad maintain, the relaxation of a muscle is dependent on a special
chemical process or not, it falls within our present scope to inquire
whether by comparing with a normal muscle, one which not only does
not relax but has been deprived of the faculty of relaxing, we can
arrive at any electrical indication of such a process. Fortunately we
have within reach a means by which this experiment can be made.

The Continuous Response of a Veratrinised Muscle.

The alkaloid veratrine* is an agent by which a muscle excited by an
instantaneous stimulus is deprived of its power of recovering itself .
The quantity of the alkaloid required to produce the effect is extremely
small. The addition of one part in a million of veratrine to the

* The reratrine used was kindly prepared by my friend Professor Dunstan,
F.R.S.

and Plants to the Elect rical Phenomena associated with it. 59

physiological salt solution in which a muscle has been kept for several
hours, is sufficient to give it this property or, as it may be expressed,
to " veratrinise " it thoroughly. The alteration of the properties of a
muscle by veratrine in such a way that it must continue an effort once
begun, has been long known. It is an example of perfectly continuous
contraction. Normal muscular contraction being regarded, as I have
said, as discontinuous, the relation between it and the continuous con-
traction of veratrinised muscle has not been sufficiently considered.
AVhen therefore we set to work to measure the m aximum contractile
effect of a " veratrine spasm," I was both surprised and gratified to
discover that the tension of a veratrinised muscle* when excited by a
single instantaneous stimulus, was as great as that of a similar but
un veratrinised muscle when subjected to a succession of stimuli, i.e.,
when artificially tetanised. It can also lift as great a load and hold it
up for several (10 — 20) seconds at as great a height. (Tracings shown.)

We then proceeded to investigate the electrical concomitant of the
veratrine "tetanus," if I may so call it (Photo. 11) and found it to
be identical with that of an artificial tetanus produced by a succession
of stimuli of sufficient frequency. Its true character can be best
judged of by comparing it with Photo. 1'2, which was obtained by

introducing into the unchanged circuit a constant difference of potential
in the way before explained (p. 54).

The fact that the veratrine spasm has the mechanical and electrical
character of a continuous contraction is of value, not from its bearing

* Electrical response of curarised and veratrinised Sartorius to an instantaneous
stimulation. Leading off contacts at middle and tibial end, exciting electrodes
near pelvic end. The initial rise of the curve is steeper than that of the comparison
curve (Photo. 12).

€0 Prof. J. Burdon-Sauderson. Relation of Motion in Animals

Photograph 12.*

on the mode of action of a particular chemical substance, but from the
evidence it affords that discontinuity is not essential to energetic
display of contractile force. In this respect it would be wholly
irrelevant to object that the data derived from experiments on a
poisoned muscle cannot be applied to a normal one. All that it is
required to prove is that it is possible for a spasm which is not dis-
continuous to be as effectual for the doing of external work as a normal
contraction. It can hardly be disputed that the contraction of a vera-
trinised muscle is continuous. It is, therefore, no longer possible to
assert that discontinuity is essential to functional capacity.

That our results differ from those of other observers is to be
attributed to the mode of using the alkaloid, and to the homoeopathic
minuteness of the dose. We estimate the quantity of veratrine which
actually enters the muscle not to exceed 1/10,000 milligram.

The Heart

We now turn from the skeletal muscles to the organ by the
rhythmical contractions of which the circulation is maintained. The
mechanical response of cardiac, like that of skeletal, muscle can be
evoked either directly or indirectly, but the heart has this peculiarity
that each part of it has attributes which we are accustomed to regard
as nervous rather than muscular. It has above all the property which
belongs, as we have seen from our experiments with strychnine, to the
motor cells of the spinal cord — that of discharging itself rhythmically
when in a state of continuous excitation. It is characteristic of heart-
muscle that it exhibits alternating periods of rest and activity, and we
have now the clearest evidence that it is not in virtue of its possessing

* Comparison curre obtained by leading off from the compensator a current of
E.M.F. equal to that of the " action current " ; leaving the unexcited muscle in
circuit.

and Plants to the Electrical Phenomena associated with it. 61

an intrinsic nervous system that it has this property. In another
important respect it resembles the motor apparatus of the cord, namely,
that its relations to stimuli are governed by what has been called the
" all or not at all " principle. It either does not respond or, if at all,
responds completely. In these respects, therefore, the action of the
heart is comparable neither with that of muscle acting independently,
nor even with that of the muscle nerve preparation, but rather with
that of muscle acting under the direction of the motor neuron which
governs it.

I began the investigation of the electrical phenomena of the heart's
beat in 1881 with Mr. Page. We made out two new facts, namely,
that the electrical change which is evoked by excitation of the surface
is propagated, at a rate dependent on temperature, not in one direction
only but in all, as Engelmann had already shown to be the case with
regard to the wave of contraction ; and secondly, that the monophasic
variation is not, as had been supposed by previous observers, an instan-
taneous change, but lasts during the whole period of energetic systole.
But neither Mr. Page nor I understood then the nature of the initial
" spike," which is so striking a feature in the photographic record of
the variation in the uninjured heart. For its explanation I am indebted
to Mr. Burch, whose investigations on the use of the capillary electro-
meter for measuring the electromotive force of currents of short
duration have been of so much value to physiologists. The moment it
was understood that the spike indicated a diphasic variation analogous
to that of the muscle, I felt that I had the key to the complete under-
standing of my own previous observations. I was, moreover, able to
bring these into "omplete harmony with those of Professor Engelmann
made about the same time with the rheotome and galvanometer.

Let me ask your attention to the photographic curves of the diphasic
and monophasic variations which I have placed one above the other in
synchronic relation to each other. It is to be noticed that the move-
ment of the recording surface is very slow, about a centimetre a second
only. To obtain the monophasic curve you have to place the distal
electrode on a spot which has been devitalised by scorching, and which
is consequently physiologically inactive, the proximal electrode on the
living surface near the junction between auricle and ventricle. The
instantaneous stimulation is applied to the auricle some couple of
millimetres distant from the proximal leading-off electrode. The Reiz-
welle is propagated from the auricle to the base of the ventricle and
then on to the devitalised spot, so that before it arrives at the contact
it is extinguished.* Consequently the change which is expressed by
the electrometer-curve takes place exclusively at the proximal contact-
surface. It differs only from the monophasic variation of skeletal

* This mode of observation corresponds to the first fundamental experiment in
muscle (see. p. 45).

62 Prof. J. Burdon-Sanderson. Relation of Motion in Animals

Photographs 13 and 14.*

muscle in the longer duration of the period which intervenes between
culmination and decline, and consequently bears a greater resemblance
to the effect of a short continuous excitation of muscle than to that of
an instantaneous one.

Turning to the diphasic variation obtained when the surface under-
lying the distal contact is not devitalised, we see that during the whole
intervening period just referred to, the two contact surfaces are ap-
proximately equipotential. This of course does not mean that both
are physiologically inactive, but simply that the influence of the one
exactly balances that of the other. This meaning of the diphasic
variation is (with the exception of the initial spike) that which was
assigned to it in 1882. It results from the mutual interference of two
monophasic variations, the dip of the curve at the end indicating that
the effect of the distal contact overlasts that at the proximal.

The general result of these observations is that, just as from the
mechanical point of view the systole of the ventricle has lately been
shown to be entirely analogous to the response of a muscle to an in-
stantaneous stimulus, provided that we substitute volume for length
and lateral pressure for tension,! so as regards the electrical phenomena

* Ventricle of hea"t of R, esculenta arrested by Staminas' ligature. Exciting
electrodes on auricle. Leading off contacts at base and apex. In 13, apex
surface devitalised by heat ; in 14, both surfaces uninjured.

f O. Frank, " Zur Dynamik des Herzniuskels," 'Zeits. f. Biol.,' vol. 32, p. 370.

and Plants to the Electrical Phenomena associated ivith it. 63

there is a complete analogy between the monophasic and diphasic
variation of the heart and of muscle, respectively, provided that we
hear in mind that the one is a response to a short continuous stimula-
tion, the other to an instantaneous one.

Dioncea.

My last example of motion and its accompanying electrical phe-
nomena I will take from the plant. As everyone knows, there are
certain parts of some of the higher plants which respond to stimulation
like the motor organs of animals. These instances have been regarded
as indications of the close relationship which exists between plants and
animals as regards their elementary physiology. The subject attracted
the attention of Mr. Darwin in relation to certain insectivorous plants,
and it was at his suggestion that the observations to which I am now
about briefly to refer, were made. The electrical changes can be most
•easily studied arid appear in the most striking way in the leaf of
Dionaea. The leading-off contacts are applied to the opposite surfaces
•of one lobe of the leaf. In the resting state the one surface is found
to be positive to the other. At a certain moment, a hair on one lobe
some 10 or 12 millimetres away from the place under investigation, is
touched by a camel-hair pencil or excited by an induction current.
The surface which was before positive becomes less so, and the curve
described resembles, as you see, the monophasic heart curve.

It is not necessary on the present occasion to do more than refer to
this typical experiment, by which it was shown for the first time that
the migration of liquid, and consequent sudden closure of the lobes on
excitation, is accompanied by an electrical change analogous to that in
contracting muscle, and that in the leaf this is propagated at a rate
varying with temperature. Although the experiment is one of extreme
simplicity the method of investigation has not, so far as I know, been
pursued by any plant physiologist. The criticisms which were bestowed
on it by animal physiologists I was able to answer in my second com-
munication to the Eoyal Society, and have now the satisfaction to find
that the experimental data set forth in that paper are given in full in
Biedermann's important treatise on 1 Electro-Physiology.'

I have now, though in a very incomplete way, described the pheno-
mena bearing on my subject so far as I have been able to observe
them. -May I be permitted to submit to you the indications which
they seem to me to afford 1

In striated muscle the primary effect of every excitation is a process
oi oxidation having its seat at the excited part. It may be surmised
that this consists of two stages, namely, liberation of previously intra-
molecular oxygen, and actual oxidation. In a single element of

VOL. LXV. F

64

Proceedings and List of Papers read.

muscular structure the duration of this process, when induced by an
instantaneous stimulus, must be exceedingly short, and corresponds
with that of the excitatory variation ; but in the whole organ may
hist until the development of tension has reached its maximum.

We have further learned that the monophasic variation is a pheno-
menon of great regularity, and may be taken as the type from which
all other forms of response to stimulation may be derived, either by
repetition, prolongation, or interference.

Although no attempt has been made to settle the question whether
the natural contraction of muscle is discontinuous, it has been shown
that the electrical phenomena of reflex contraction afford no ground
for supposing that it is so. The efficiency of the veratrine spasm
seems, at least, to justify us in doubting whether discontinuity is an
essential quality of muscular contraction.

Finally, reasons have been given for thinking that the phenomena
known as the "muscle current "and the "demarcation current "are
manifestations of processes which have their seat at the surface of
contact between electrode and living ninscle.

April 20, 1899.

The LOED LISTEE, F.E.C.S., D.C.L., President, in the Chair.

A List of the Presents received was laid on the table, and thanks
ordered for them.

The following Papers were read : —

I. " The Physiological Action of Choline and Xeivrme." By Dr.
Mott, F.E.S., and Dr. Halliburton, F.E.S.

II. " On Intestinal Absorption, especially on the Absorption of Serum,

Peptone, and Glucose." By Professor E. \Vaymouth Eeid,
F.E.S.

III. " Studies in the Morphology of Spore-producing Members. IV.

The Leptosporangiate Ferns." Bv Professor F. 0. Bower,
F.E.S.

IV. " Note on the Fertility of different Breeds of Sheep, with Remarks

on the Prevalence of Abortion and Barrenness therein." By
W. Heape, M.A. Communicated by Professor Weldon, F.B.S.

V. " Some further Eemarks on Eed-water or Texas Fever." By
A. Edingtox, M.B. Communicated by Dr. Gill, F.E.S.

A Sugar Bacterium.

65

" A Sugar Bacterium." By H. Marshall Ward, F.B.S., and
J. Keynolds Green, F.B.S. Eeceivecl March 2, — Bead
March 9, 1899.

In the 'Annals of Botany ' for 1897,* one of us published a short
note on a curious organism — or rather association of organisms —
obtained in Paris, and said to have come from Madagascar, where it
occurs as " an excrescence on the sugar-cane."

It consists of a bacterium associated with at least one yeast, and
grows in saccharine solutions, producing clumps so like the ginger-beer
plant f that the assumption seemed warranted that we had here a
symbiosis of the same kind as that proved to occur there.

Moreover, the general course of events in the use of this body, which
is employed to make a fermented effervescing drink from common
brown sugar in water, points to the same conclusion.

In moderately strong solutions containing 15 to 20 per cent, of
common sugar in water, the clumps referred to induce a powerful fer-
mentation, resulting in the liberation of relatively enormous quantities
of carbon dioxide and some acid, the saccharine liquid being thus
converted into a not unpleasant acid drink, with some resemblance
to lemonade or ginger-beer.

From the fact that this fermentation occurs rapidly when the corked
flask is entirely filled with the recently boiled sugar solution, infected
with a few clumps of the organism, it is clear that oxygen is not
necessary in any quantity.

This conclusion is also confirmed by the observation that if a bottle
of soda-water is opened, and a handful of sugar added with a few of
the clumps, and at once corked and wired, the pressure of the carbon
dioxide liberated during the active fermentation which at once ensues
becomes so great in three or four days at 22° C, as to cause danger of
bursting the flask ; and if a manometer tube with mercury is attached,
as described in the paper above referred to,{ the bubbles of gas pressed
out come off steadily for many days, or even weeks, at ordinary
temperatures, until no more sugar is left.

The general resemblances to the well known kephir, also referred to
in the previous paper, led one of us to repeat the above experiments
with sugar and milk, instead of soda-water, with the result that carbon
dioxide came off as before until all the sugar had disappeared, the milk
meanwhile undergoing coagulation into clots, but since these clots
remained unaltered for weeks or months, this experiment suggests

* Marshall Ward, " On the Gingpr-beev Plant," 1 Annals of Botany,' 1897, vol. ii,
p. 341.

f ' Phil. Trans.,' B, 1892, pp. 125—197.
X Loc. ext., p. 137.

F 2

66 Profs. H. Marshall Ward and J. Eeynolds Green.

that the organism — unlike kephir — does not ferment the milk itself,
but only the added sugar, and that the clots are simply due to the acids
liberated as the sugar is destroyed, a conclusion fully borne out by
subsequent investigations.

A preliminary experiment with the ginger-beer plant showed that it
also behaves in the same way as regards milk ; the fermentation only
occurs if sugar be added, and lasts only so long as any sugar remains.

In the cases both of the present organism and of the ginger-beer
plant, if a clump be placed in sterilised beer wort, the resulting fermen-
tation gives a frothing liquid with a beer-like taste and smell, and a
rapid deposit of yeast occurs. This primary fermentation is very soon
finished, and it is abundantly proved by the experiments that this
medium favours the yeast — or one of the yeasts — in the clumps, so
that we may regard this fermentation as merely a particular case of
an impure alcoholic fermentation, such as is got in ordinary brewing.

During these preliminary trials, and with the object of testing
whether acidification of the liquid from the first would affect the matter,
it happened to one of us to select lemon-juice as the medium in one
case, partly to acidify the medium, and partly to add vegetable matter
other than sugar. The result was somewhat astonishing. The mixture
of soda-water, sugar, and lemon passed into violent fermentation in
three days, and carbon dioxide came off abundantly, under a pressure
of about 12 inches of mercury, and continued to do for many days.

Such a flask started on April 9, 1897, was evolving gas actively on
the 12th ; this went on without any apparent diminution until May 24,
and even on June 24 gas was still coming off, though now under less
pressure. Xo sugar had been added in the interval, and the flask,
still unopened, remained on a side bench in the laboratory until
January 18, 1898. On that date it was opened and examined. It
still stood at a pressure of 6 inches of mercury, and gave off gas as
soon as the pressure was reduced. The microscopic examination
showed that here, again, as in the beer wort, the medium had favoured
one of the yeasts, and although the bacterium was discoverable, it was in
abeyance, and the compound organism as a whole was not increasing.

In another similar case the flask was started on May 28, 1897, and
the pressure of the gas evolved was still supporting nearly 12 inches of
mercury on April 23, 1898, and again only yeast predominated in the
deposit, and examination showed this to be fully alive; it at once
renewed its activity when placed in sugar solutions.

These preliminary experiments will suffice to show that we have in
this compound of associated organisms an agent or agents capable of
setting up very active fermentation in various saccharine liquids, such
as ordinary sugar and water, or soda-water, beer wort, milk, and sugar,
or an infusion of vegetable substance, such as lemon pulp, and it is
clear that the fermentation, though differing in details in each case,

A Sugar Bacterium.

67

always results in the destruction of the sugar, and the production of
enormous quantities of carbon dioxide. Obviously, also, these fermen-
tations are anaerobic.

In order to put this last point beyond all cavil, however, we placed
a clump of the compound organism into a mixture of a 15 per cent,
solution of sugar to which 10 per cent, gelatine had been added, and
kept the tubes at 30 3 C, while the air was pumped out from the fluid
mass and pure carbon dioxide allowed to filter in, and this process was
repeated four or five times, and the de-oxygenated gelatine, still in an
atmosphere of carbon dioxide, was then allowed to set. In a fortnight
the solid gelatine in all these tubes had visible submerged colonies
throughout the mass, and examination showed these to consist of the
bacterium and yeast found in the original clumps.

Similarly, streak cultures on the same sugar-gelatine medium, grew
normally on the sloped surface in tubes filled with carbon dioxide, and
in these again were observed the same bacterium and the same yeast as
had been foimd predominating in the original clumps.

These cultures — still preliminary in nature and only dealing with
the composite organisms of the clumps as a whole — suggested an
obvious method for separating at least this prominent bacterium and
yeast from the clumps.

Sugar-gelatine, made as before, was infected with a small piece of a
clump, rubbed up by a platinum loop in sterile water, and plates made
in the ordinary way in Petri dishes ; the dishes were then placed trader
a receiver, attached to the pump and exhausted, and then filled with
carbon dioxide, with proper precautions as to the purity of the gas,
filtration through cotton-wool plugs, &c., and the cultures put aside in
an atmosphere of carbon dioxide at the ordinary temperature. In a
week the solid gelatine showed two kinds of colonies, one consisting
entirely of a yeast, the other of a bacterium, and closer investigation
showed them to be identical with the prevailing yeast and bacterium
in the original clumps.

Repeated plate-ciiltures made in this way gave consistently the same
results, and there was no room for doubt that these are the two essential
organisms of the clumps, though they are not the only species found
in the original material, there being at least one other yeast-like
organism, so common that for some time it was thought it must play
an essential part. Since this latter — and certain much rarer forms
occasionally found — will not grow in the atmosphere of carbon dioxide,
however, there can be little doubt that the two anaerobic microbes
isolated by the above process are the essential constituents in the
fermentations referred to.

Their further separation by means of repeated plate-cultures, as
above, was comparatively easy, the yeast especially being readily
picked out and further cultivated in sugar-gelatine tubes.

68 Profs. H. Marshall Ward and. J. Reynolds Green.

As it is not at present proposed to deal with the yeast, which ap-
pears to be a mere variety of S. cerevisice, we pass on to the cultures of
the bacterium only.

On repeating the plate cultures exactly as before, except the exhaus-
tion and filling with carbon dioxide, it was found that mixtures of the
yeast and bacillus grew as well in air as in carbon dioxide. At first it
seemed possible that this was because the yeast rapidly consumed the
oxygen and so prepared an oxygen-free atmosphere for the bacterium,
but further experiments proved that this is not necessary, and that
both yeast and bacterium can be grown in air as well as in carbon
dioxide or in hydrogen.

The appearance of the separated bacterium on the sugar-gelatine
plates is that of circular, raised, dome-shaped, watery-looking colonies,
stiff, like a firm jelly, and lifting as a whole on the needle. Each
colony is, in fact, a firm zooglcea composed of short rodlets in pairs or
chains, the cell-walls of which are so swollen as to furnish the zoogloea
jelly. The average size of these rodlets is 2 — 3 fi long by 1 p thick,
though much longer rods and filaments occur in other media.

Having once obtained the organism in pure culture, it was, of
course, easy to test its behaviour on various media. It is unnecessary
to enumerate all the media tried, or to give details of the cultures,
which amount to several hundreds ; enough that all ordinary media
employed by bacteriologists were tried, as well as a long series of
special ones devised to meet the suggestions which arose during the
course of the investigation.

A striking fact comes out on surveying these cultures, namely, this
Schizomycete practically refuses to grow in or on any pabulum devoid
of sugar, and, further, only certain sugars are capable of supplying it
with its necessary food. No growth at any temperature could be
obtained in normal gelatine-peptone media, or in broth, milk, or other
animal extracts, e.g., serum-agar, such as is used by the animal
pathologists.

Gelatine Cultures.

Gelatine, 10 per cent., added to "black sugar" solution* 15 per
cent, is a capital medium for cultures at 18° C, or thereabouts, and in
air, in hydrogen, or in carbon dioxide, the bacterium formed prominent
domed colonies looking like drops of stiff gum or gelatine.

This black-sugar gelatine was also used for cultures in vacuo. Plates
kept under a receiver permanently attached to the pump, going day
and night continuously for a week, showed traces of colonies in eight
days, and in fourteen days the colonies in the exhausted receiver were

* This " black sugar " is a very dark coarse Demerara sugar and probably contains
considerable quantities of mineral and other matter.

A Sugar Bacterium.

69

as well developed as those in control plates in air at the same (low)
temperature.

Streak cultures on black-sugar gelatine with yeast extract developed
very rapidly at and near 18° C. Instead of a spreading slimy mass
as on agar, these formed dense gelatinous, almost brittle, tear-like
drops and streaks standing a millimetre or more high, and curling up
oft' the surface of the gelatine in a most remarkable and characteristic
manner.

On saccharonal-yeast-extract-gelatine similar raised streaks were
obtained, but less luxuriant than on brown sugar; possibly because
the temperature was lower, or only 10 per cent, saccharose was em-
ployed, or from lack of minerals.

Beer-wort gelatine gave very slight indications of growth, soon
stopping, and never attaining anything like these dimensions, the mere
traces observed being probably due to saccharose in the wort.

Certain mineral solutions — e.g., Kleb's solution — appeared to inhibit
the growth in the black sugar and gelatine.

We have here gathered together the principal facts of its behaviour
on gelatine media, because they seem to bring out clearly that the
gelatine itself has little or no part to play in the nourishment of the
bacillus : mere traces of superficial liquefaction occur, and several
experiments show that it is of no slight importance what is added to
the gelatine, e.g., the surprising results of beer-wort gelatine.

Agar.

Agar (2 per cent.) made up with black sugar and yeast water rapidly
formed large slimy blister-like colonies in four days, both at 25° C. and
at 19° C. The contrast between these large, slimy, flat, and extended
colonies and the small, raised, dome-like, stiff ones, on gelatine was very
marked.

Streak cultures on black sugar agar, with yeast extract, grew rapidly
as a dull, honey-like slime, spreading all over the surface in two days
at 16—18° and at 23—25° C, as well as at 36—37° C.

Agar made up with peptone, &c, was of no use whatever, nor was
potato-agar. Even peptone-agar with black sugar proved unsuitable.

Similarly unsuitable was agar made up with yeast extract and sac-
charose, which had been filtered through porcelain, though a pale
dotted streak formed at first at 25, 27, and 31° C.

Agar (2 per cent.) made up with 10 per cent, saccharose and yeast
extract not filtered through porcelain, on the other hand, was an
excellent medium. In twenty-four hours at 30 — 31° C, a rapidly
growing slimy streak had formed, consisting of filaments up to 60 //.
long and more, breaking up into segments of all lengths down to 1*5 \x
by 1 /x. These showed no sheaths — they had turned slimy— stained

70

Profs. H. Marshall Ward and J. Eeynolds Green.

well in gentian violet by Gram's method, and were non-motile. At
23° and 20° C. the growth was similar bnt slower.

Wort-agar, however, was unsuitable, cultures parallel with the above,
showing very slight growths at any temperature ; results quite conform-
able with previous experience.

Here, again, we must conclude that the agar is of little or no import-
ance, except as a support. The significance of the failure on agar, to
which porcelain-filtered yeast extract and saccharose was added, appears
significant, and we shall return to this in order to discuss what happens
to the yeast extract and sugar when this mode of sterilisation alone is
relied on.

Beer Wort.

In beer wort (unhopped) the bacterium grows fairly well at first,
rapid turbidity following the infection, and a dirty yellowish deposit
soon falls, consisting of the flocculent bacterial masses carrying down
colouring matter ; but the liquid is not viscous, and the deposit scarcely
slimy, and growth soon ceases.

When we reflect that beer wort — i.e., malt extract — is usually an
excellent medium for the growth of fungi, it is somewhat surprising
that this sugar-loving bacterium should do so badly in it. Thinking
that the failure might be owing to the kind of sugar in beer wort being
unsuitable, we tried adding saccharose, but no obvious improvement
was effected. We have seen (p. 69, above) that beer wort and gelatine
gave poor results, and we concluded provisionally that some unfavour-
able substance occurs in wort.

Struck by the success of the preliminary trials with cane sugar, and
remembering the alleged original habitat of the organism, it was deter-
mined to try the effect of beet : this was done not only because beet
is a well-known source of cane sugar, but also because one of us had
previously observed in a certain beet disease, that a bacterium very
similar to the one under investigation spreads in the tissues of the
sugar-beet, and is connected in some way with the disease itself.

Beet Extract.

Cold water extract of crushed sugar-beet was found at an early stage
of the work to be a favourable medium. In tubes at 25° C, cultures
in carbon dioxide rapidly become- turbid, and on the third day, a dense
slimy zooglcea-like deposit had fallen to the bottom carrying with it the
colouring matters, and containing embedded bacteria. The same at
15° C. in carbon dioxide, the growth being simply somewhat slower.

Parallel cultures — from the same tubes — in broth, gelatine, or agar,
devoid of sugars, showed no growth either in air or in carbon dioxide,
at 15° or at 25° C.

A Sugar Bacterium.

71

Tubes of raw beetroot, infected with the gelatinous bacterial clumps,
showed evident sinking into the tissues in twenty-four to forty-eight
hours at 18°, 25°, and 31°, and in six days the depressed area showed
collapsed and browned cells under the microscope. A curious white
zone surrounded the discoloured patches. The impression gained was
that the bacterium draws out the sugary sap and thrives on it : no
proof of direct entry into the cell walls could be obtained.*

Cooked, i.e., sterilised, beet gave a mere wet patch at first, but in a
week raised gelatinous lumps of the typical kind were formed.

Here, then, we appeared to have proof that it is really the cane sugar
which decides the success or otherwise of the cultiues of this gelatinous
bacterium, and we entered on what proved to be a very long series of
trials with various kinds of sugars to decide this.

Trials with Sugars.

A series of preliminary experiments in which cane-sugar, glucose,
and milk-sugar were employed, made up in various ways, soon showed
that this bacterium grows far better in cane-sugar than in the others,
and better in solutions made up with yeast-water than in such containing
mineral salts, asparagin, tartrates, &c.

We, therefore, made series of parallel cultures at various tempera-
tures, in air and in carbon dioxide, and all infected from the same tube,
using the following sugars and solutions : — Levulose, pure glucose,
cane-sugar, saccharon, lactose, maltose, dextrin, and a body known in
Grubler's catalogue as " dextrin-zucker-losung." The sugars were all
made up in 10 per cent, solutions, and definite quantities — equal in
each case — of the other ingredients added.

Summing up the results of numerous experiments, it was found that
no growth occurred in any medium at temperatures of 35° C. and
upwards, except in the case of certain agar cultures, where rapid
growth occurred at and near 37° C. for a few days only.

Cane-Sugar.

The most striking results were obtained in all cases with cane-sugar,,
especially in a very pure re-crystallised form labelled " Saccharon,"
though we also employed ordinary lump-sugar, and a coarse dark brown
moist sugar known locally as " Black sugar."

Made up as Mayer's solution, the brown sugar rapidly became turbid
and viscous, and a dense gelatinous deposit of bacteria formed below

* Attempts at infection were made in view of the alleged connection between
bacteria and certain diseases of beet and sugar-cane. See, for instance, ' Kew
Bulletin,' No. 85, January, 1894, p. 1, and ' Zeitschr. f . Pflanzenkrankh.,' 1897,
No. 7, p. 65.

72 Profs. H. Marshall Ward and J. Reynolds Green.

the surface. This occurred even in tubes to which a few drops of abso-
lute alcohol were added — a result not obtainable with dextrin-Mayer.

Far better growth was got with cane-sugar and yeast-extract how-
ever.

The best growth was got with saccharon (10 per cent.) and yeast
extract, where at all temperatures from 16 — 27*5° to 31° C, the liquid
became opalescent and viscous in two or three days, and deposited the
typical gelatinous zooglcea at the bottom of the tubes.

The results so far show that only the various forms of saccharose
and beet extract (containing this sugar) afford any pronounced growth
of this bacterium, the best results being got with pure saccharon and
yeast extract, as indicated by the rapid turbidity and viscosity and the
clump of gelatinous deposit. Since other sugars seem quite unsuit-
able, or only induced slight growths expressed as temporary turbidity
and flocculent deposits, it may probably be assumed that in " dextrin-
zucker "* and in " glucose " solutions, where indication of the viscosity
and gelatinous deposit occur, traces of saccharose were contained.

Glucose.

Pure glucose and yeast-water encouraged an excellent growth at
first, the liquid becoming turbid and a flocculent deposit settling down
in a few days. No signs of viscosity appeared, however, and the
deposit was quite loose and easily shaken up. The fairly abundant
flocculent growth at first led to the expectation that prolonged cul-
tures, or cultivation at different temperatures, might result in the
development of the typical sliminess and viscosity, but repeated
attempts show that such is not the case. This sugar did not appear to
injure the bacterium, for the deposit was alive after three weeks.

Nor would it grow on or in gelatine made up with glucose. The
slight growth obtained in certain cases where commercial glucose was
used may have been due to admixtures. The total results show that
glucose is not a favourable medium.

Levulose (Fructose).

Solutions of levulose prepared as the other solutions became slightly
turbid in three or four days, and then the bacteria deposited as a very
thin layer, no trace of slime or gelatinous matrix being formed, and
the impression resulted that either traces of some other sugar must
have sufficed for what activity was evinced in ordinary glucose solu-
tions, because no growth worth mentioning compared with the pre-
ceding had occurred, or levulose is to a slight extent a food for the
•organism.

* This proved to be true of this medium ; it consists of dextrin and cane sugar
nvith a little alcohol.

A Sugar Bacterium.

— o

j O

Similar tubes placed in a vacuum also showed no growth beyond the
formation of the flocculent deposit, hardly slimy, and easily shaken
up into the clear supernatant liquid — a point of constrast of some
importance, for in successful cultures in cane-sugar the viscosity of the
liquid above is so great that it is very difficult to shake up the deposit,
which is also dense and gelatinous.

Repetitions of these cultures gave the same results — a rapid develop-
ment of a very slight turbidity and formation of a small non-gelatin-
ous deposit which falls and leaves the liquid clear.

Some attempts were made to see if varying the strengths and
composition of the levulose solutions would affect the matter — e.g.,
Mayer's solution made with levulose gave little or no signs of growth
at all with the bacterium, though it proved a splendid medium for the
yeast.

Curiously enough, a mixture of beet extract and this levulose-
Mayer's solution, though it encouraged abundant growth of the
bacillus, rapidly becoming turbid and then clearing as the deposit
fell, showed no viscosity nor was the deposit slimy, as occurs in beet
extract alone.

Even made up with yeast-extract, no viscosity could be obtained :
nothing beyond the slight flocculence, and it was clear that levulose is
at best a poor food material for this bacillus.

Mixed Sugar*.

Proceeding from the observation that the bacterium undoubtedly
inverts saccharose, cultures were made as follows : — Dextrose-yeast-
extract and levulose-yeast-extract were mixed, and infected with the
bacterium : an excellent growth of the organism resulted, at both
32° and 24° C, but the deposit which resulted, consisted entirely of
the non-sheathed bacterium in flocculent masses, and it remains a
puzzle why the sheaths are formed in saccharose and not in the
two sugars resulting from its inversion. The only conclusion seems
to be that proportion of constituents has something to do with the
matter.

' 1 Dextrin-Zi t eke i --Losu ng. ' '

Under the above . name, Grubler supplies a syrupy sugar in which
the microscope shows delicate needle crystals, and — merely in the
spirit of trying all experiments — this was tried with the following
results : —

Made up with yeast water as a 10 per cent, or 15 per cent, solution,
the bacterium slowly formed a large slimy deposit in from five to ten
or twelve days at 15° and 23°, but not at 35°, in which rods and chains
existed in the colourless matrix. The rods measured 1'5/x x 1/x to 3/x x 1/z,

74 Profs. H. Marshall Ward and J. Eeynolds Green.

but filaments up to 20 /x long were found. The growth resulted in a.
curious opalescent change in the liquid above the whitish dense deposit,,
and it became markedly viscous so as to draw out into strings-
Made up in Mayer's solution also, the same viscidity and gelatinous
growth occurred, and this both in air and in hydrogen. We have since
learnt that this syrup consists of dextrin, cane-sugar, and a little alcohol..

Dextrin.

The difficulty as to the nature of " Dextrin-Zucker " referred to
above, led us to try dextrin itself, and owing to the kindness of Dr..
Ruhremann some very pure material was to hand. With Mayer's
solution several experiments demonstrated that no obvious growth,
occurred, and it was clear that this dextrin is not the same for
nutritive purposes as the " Dextrin-Zucker " used previously..

Nor was dextrin made up with yeast-water of any use, though
occasionally very slight indications of growth occurred during the
first twenty-four hours, but no viscosity or sheathed bacteria formed.

Malto-dextrin,

Owing to the kindness of Mr. Ling, we were furnished with some
prepared malto-dextrin, but neither as Mayer's solution, nor made up
with yeast extract, was this sugar of any use as a medium, for the
growth of this bacterium.

Maltose.

This sugar, made up as Mayer's solution, was found unsuited for the
bacterium, either at high or low temperatures. Nor was it more suc-
cessful made up with yeast extract ; a faint turbidity and flocculence
occurred, but no trace of viscosity at 25° occurred in four days.

Milk-Sugar.

Milk-sugar, whether made up with yeast-water, Mayer's solution,
asparagin, or peptone, proved quite useless as a food for the bacterium :
no signs of growth appeared in the solutions at all, at high or low tem-
peratures, in air, or in carbon dioxide. The organism was- dormant
only, however.

Soluble Starch.

It seemed worth while, in view of the gelatinous nature of vigorous
growths, to see how the bacterium would behave towards soluble
starch, but in no case could any evidence of growth be obtained in
solutions containing 2 per cent, of soluble starch made up with Mayer's

A Sugar Bacterium.

75

solution, or with yeast-water. The bacteria lay dormant only at the
bottom, as experiments showed.

It is evident from the foregoing that of all the sugars tried, saccha-
rose is the one which favours the growth of the bacterium, but even in
saccharon the growth is distinctly favoured by the addition of yeast
extract. Some experiments were consequently started to test the
effect of the yeast extract.

Raw Yeast-Water.

Fresh yeast, squeezed and drained, and then ground up with kiesel-
guhr and extracted with water, was allowed to stand all night and
"filtered through porcelain. Employed alone it was of no use as a
medium for the growth of the bacterium ; nor would the latter develop
in this raw yeast-water, to which 5 per cent, of cane-sugar was added.
The latter fact was thought to be possibly due to the raw yeast-water
liaving inverted the saccharose, and we have seen that glucose and
levulose are not suitable media ; and we explained similarly the failure
of saccharon-Mayer's solution, to which this raw yeast-water was
•added, as well as failures with brown sugar, and with beet extract
similarly made up. That failure should follow with dextrin, with
maltose, with levulose, milk-sugar, and with beer wort similarly made
Tup, was only to be expected from experience with these media con-
cocted with boiled yeast-water.

Further experiments, however, led to the conviction that matters
were more complicated than would be implied by this explanation.

At one stage in the investigation, being impressed by the stimulus
to growth afforded by the addition of (boiled) yeast- water to the saccha-
rose solutions, we tried the effect of adding sugar to the raw yeast
extract and sterilising by filtration through porcelain only, and were
surprised to find that no growth whatever occurred at any temperature,
e.g., 18°, 23°, and 30° C. This was afterwards explained as above —
the yeast extract inverts 'the sugar before it has time to filter.

We then tried a series of experiments as follows : — Cultures at
17 — 20°, 25°, and 31° C, were made in 10 per cent, saccharon +
10 per cent, yeast-water mixed raw and sterilised by filtration only;
no traces of growth occurred in 10 days. The same failure was
realised with 10 per cent, saccharon alone sterilised by filtration only ;
with yeast extract only sterilised by filtration only ; and with yeast
extract sterilised by filtration and then boiled before adding to the
filtered sterile 10 per cent, saccharon solution. In no case did any sign
of growth occur.

It is therefore clear that the filter either holds back some body
necessary for the nutrition of the bacterium, or destroys it in its passage
through the pores. Also that raw yeast extract in some way spoils the

76 Profs. H. Marshall Ward and J. Eeynolcls Green.

sugar (saccharose) as a food material, probably by inverting- it. And
nevertheless the bacterium nourishes in conjunction with the living
yeast in saccharose solutions. Here is a puzzle which we have not
succeeded iu explaining.

It may be merely noted that numerous trials were made in other
media than those mentioned, among which glycerine and yeast extract,
alcohol with saccharon and yeast extract, starch treated with diastase,
also potato, carrot, and milk are the most important. No growth of
significance Avas obtained in any case, and the results may be neg-
lected.

The Acidity of the Cultures.

Several tests showed that the cultures of the bacterium are acid,
and the following experiments were made. Sterilised blocks of marble
were 'placed in the culture tubes before they were steamed, and then
the infections made as before. In saccharose yeast extract, the active
growth which resulted was accompanied by a more diffused viscosity
than before, and gas bubbles (C02) ascended for days. Tubes of mixed
dextrose and levulose with yeast extract, treated exactly similarly,
became turbid, and gave off bubbles, but no trace of viscosity resulted ;
the abundant flocculent deposit consisted entirely of the non-sheathed
form of the bacterium.

As will appear later, the acid which causes this liberation of COo is
mainly acetic acid.

On the Nature of the Gelatinous Matrix or Slime.

A number of experiments were made to determine the nature of the
viscous slime and the jelly-like matrix holding the bacteria. Although
reasons have already been given for concluding that this is really
nothing more than the swollen cell-walls or sheaths investing the
bacteria, the possibility of the opalescence and viscosity of the saccha-
rose media being due to the direct action on the sugar of some enzyme
or other body excreted by the organism requires investigation, for it is
conceivable that such slimes might arise in any of three ways.

(1) As products of metabolism from the interior of the cells, such as
certainly occur in the glands of higher plants, e.g., the mucilage hairs
of ferns.*

(2) As products of the action outside the cells of some enzyme-like
body which escapes from the organism and acts directly on the sugar. f

(3) As products of conversion of the cell-walls of the organism,
these swelling up and becoming diffluent as in the case of the ginger-
beer plant. |

* See Gardiner and Ito, ' Annals of Botany,' vol. 1, p. 27.
f See Kitsert, ' Cent. f. Bakt.,' vol. 11. p. 830.
X ' Phil. Trans.,' B, 1892, loc. cit.

A Sugar Bacterium.

11

It seemed impossible to test the first suggestion on such minute cells,
but attempts were made to test the second one by the following
experiment : —

The bacterium was grown in saccharose yeast extract inside a porce-
lain filter plunged into the same solution ; the gelatinous matrix was
formed in abundance inside the filter, but none was developed outside
although the liquids communicated freely through the pores of the
filter. Of course the reply may be made that an enzyme of this
nature may be unable to traverse the fine pores, and the question
must be regarded as still open.

These gelatinous sheaths or " capsules " are now known in many
Schizomycetes, among the best examples being that of B. vermiforme*
and Leuconostoc, j and it is now pretty generally agreed that these
sheaths are composed of dextran,j: and Liesenberg and Zopf§ made a
curious observation with reference to its formation • they found that
the addition of. calcium chloride favoured the development of the
sheath.

Calcium Chloride.

Zopf's paper suggested that we should test the effect of CaCl2, and
accordingly solutions of Liebig's extract, peptone-saccharon, and CaClo
were tried.

In the slightly alkaline liquid at 32° a mere shimmering turbidity
was observed in twenty-four hours, and in four days a dense gelatinous
clot formed below the turbid liquid.

At 23° the alkaline liquid showed similar turbidity on the second
day, and had formed very little gelatinous deposit in four days. In a
week, however, it resembled that at 32° C.

The same solution slightly acidulated with a drop of HC1 was
slightly more turbid in twenty-four hours, but very little of the jelly
formed even in a week.

At the end of the week both set of tubes were distinctly acid, and
evidently the formation of the jelly is favoured by the slight alkalinity
of CaCl2.

This seemed to strengthen the supposition that our bacteiium may
be the same as Van Tieghem's Leuconostoc, but a close comparison does
not bear out this view.

It is nevertheless interesting to observe that Leuconostoc inverts
saccharose, and only forms sheaths in presence of that sugar or of
grape-sugar ; that these sheaths are explained as the swollen cell-walls,
and many other features exist in common with our form.

* Marshall Ward, ' Phil. Trans.,' B, 1892, loc. cit.

f Tan Tieghem, 'Ann. des Sci. Nat.,' 6th Series, Bot., vol. 4, 1878.

X Scheibler, ' Yereinzeitsch. f. Kubenzncker Ind.,' (1874), p. 24.

§ 1 Beitr. z. Phys. u. Morph. niedcrer Organismen,' Heft 1, 1892, p. 1.

78

Profs. H. Marshall Ward and J. Eeynolds Green.

The differences may he important in various degrees. Leuconostoc
appears smaller and shorter, and withstands high temperatures ; it
succeeds well in grape-sugar ; its characters on gelatine media appear
to he different ; it can he cultivated in milk, and it forms lactic acid
in sugar solutions.

How far the differences can he insisted upon cannot be determined
until hoth organisms have been tested side by side.

There is one further observation to be made in support of our con-
tention that the viscosity depends on the deliquescence of the swollen
cell-walls. On agar media, which exude water, the growths are slimy
rather than gelatinous, and the longer the gelatine cultures are kept,
provided they are not allowed to evaporate, the more diffluent the
gelatinous lumps become. In liquid cultures, moreover, the gelatinous
clot does not spread evenly through the liquid, but remains around
the motionless organism.

Mixed Cultures.

Several attempts were made to obtain the typical clumps of the
compound organism by infecting tubes of yeast extract and saccharon
and other media with both the bacterium and the yeast separately
cultivated pme. The success was only partial, however, though it was
not difficult to obtain a viscous clot at the bottom of the tubes in
which both bacteria and yeast were embedded, the typical stiff jelly
clumps floating in the liquid were not formed, and here again we
must conclude that some definite proportion of each is necessary.

On the Chemical Changes incident to the Fermentations.

The jelly-like masses of which the organism consisted set up a very
vigorous fermentation in beer wort and in solutions of various sugars.
Careful cultures showed, as already stated, that the jelly was composed
essentially of a bacterium associated with a yeast, and a long series of
our experiments has been directed towards ascertaining what part
each played in the fermentation, and how far they assisted or impeded
each other.

In the greater number of these experiments five flasks were used,
each of about 250 c.c. capacity. 150 c.c. of the culture fluid were
placed in each, and four of them were sown with pme cultures : (a) of
the yeast alone ; (/3) of the bacterium alone ; (7) of the yeast and the
bacterium separately; (8) of the two in their ordinary condition of
association, forming a lump of the jelly. The fourth flask contained
only the culture-fluid with no organism of either land.

The culture-fluids were usually a 10 per cent, solution of some par-
ticular sugar to which 10 per cent, of its volume of yeast-water had

A Sugar Bacterium.

7£

been added to furnish the necessary combined nitrogen for the growth
of the organism. The yeast-water was prepared by boiling yeast in
water for several minutes, and then filtering and sterilising. In one
experiment an ordinary beer wort was used without addition of more
nitrogenous matter.

The liquids were carefully and repeatedly sterilised in the flasks
before the organisms were added.

The fermentations were conducted at a temperature of about 20° C
The marked feature of the fermentation set up by the conjoint organ-
ism was the production of a considerable acidity, the liquid after a
few days having the appearance and flavour of lemonade. The acidity
proved to be due to acetic and succinic acids.

As considerable differences of behaviour were soon manifested, we-
subjoin the results of, typical fermentations.

Acetic Succinic
Alcohol. acid. acid.
Per cent. Per cent. Per cent.

I. Beer wort, 150 c.c. :• —

The yeast produced 3-5 0*0124 0-068

The bacterium produced 0 0-1734 0-3

II. Cane sugar, 150 c.c. of 10 per cent,
solution containing 15 c.c. yeast
water : —

Yeast produced

Bacterium produced

Yeast + bacterium (in the form
of the conjoint organism) pro-
duced

III. Dark brown sugar (mixture of cane

sugar and levulose), proportions as
in II :—

Yeast produced

Bacterium produced

Yeast + bacterium separately
sown produced

Conjoint organism produced ...

IV. Grape sugar, proportions as in II : —

Yeast produced

Bacterium produced

Yeast + bacterium produced . . .
Conjoint organism produced ...

5-0

o-oi

0-057

0

0-7

0-57

5-0

0-048

0-078

4-75

0-026

0-137

0

0-596

0-416

4-0

0-124

o-i

3-7

0-306

0-168

2*2

0-013

0-049

0

0-012

0-018

2-0

0-02

0-097

2-0

0-15

0-046

The formation of the alcohol was thus shown to be due exclusively
to the yeast, and in its production the influence of the bacterium was
VOL. LXV. G

80 Profs. H. Marshall Ward and J. Keynolds Green.

not manifested. The yeast was found to be capable of fermenting
both glucose and fructose (levulose), but to be more active in the pre-
sence than the absence of the latter. Only half the amount of alcohol
was formed when the fructose was excluded. The latter sugar was
more favourable also to the acid fermentation by the yeast. If we
compare Experiments II and III, we find that while the same quantity
of alcohol was formed in both cases, the proportion of both acetic and
succinic acids was about doubled in the presence of fructose.

The bacterium, however, was responsible for the greater amount of
the acid formation. While it caused the production of both acetic and
succinic acids in greater proportion than the yeast, it yielded far more
relatively of the former. In the experiment with beer wort it produced
fifteen times as much acetic acid as the yeast ; it only gave rise to four
to five times as much succinic. With cane-sugar solution it formed
seventy times as much acetic and ten times as much succinic acid as
the yeast. The same result was obtained with cane-sugar and
fructose.

With grape-sugar the bacterium, like the yeast, could do but little.
It produced about the same amount of acetic acid, but scarcely more
than one-third as much succinic.

The experiments on the fermentations confirm the view expressed in
the early portion of the paper, that cane-sugar is the most favourable
medium for the bacterium, but before it is of use to it, it undergoes
inversion.

The association of the. two organisms together introduced a some-
what curious feature of the fermentation. The yeast continued to
produce alcohol in the same proportion as when alone, but the acid
fermentations were modified considerably. Comparative experiments
were made with cane-sugar, dark brown sugar containing fructose, and
grape-sugar. In the first two cases the amount of both acetic and
succinic acids produced by the conjoint organism was distinctly less
than that which was formed by the bacterium alone. In the grape-
sugar fermentation the contrary was the case.

The association of the two organisms into the jelly-like clumps was
not without its effect upon the progress of the fermentation, and this
effect again appeared in connection with the process of the acidification.

With the dark brown sugar the conjoint organism produced twice as
much acetic acid as the two separate constituents working together, but
only one and a half times as much succinic. With grape-sugar there
was a diminution of the succinic acid, which fell to one-half the
quantity produced by the yeast and the bacterium, while not in such
close association. The acetic acid, on the other hand, increased. In the
first two cases again, the presence of the yeast was very considerably
inhibitory of the activity of the bacterium, the latter producing far
more acid of both kinds when it was alone in the fermenting liquid.

A Sugar Bacterium.

81

One or two features of the fermentations call for comment. The
progress of the fermentation in an acid liquid was so great as to sug-
gest that an enzyme was excreted by the organism, but careful
search proved that this was not the case. Both the conjoint organ-
ism and its two constituents are normally aerobic, but in all cases the
fermentations will proceed, though with slightly less vigour, in an
atmosphere of C02. The decomposition effected by the bacterium is
accompanied by only a very slow evolution of this gas, not greater
indeed than would be due to its respiration. In one experiment which
was carried on for two months, there was an output of 0-005 gram of
C02 per day, the gas being absorbed by caustic potash as it was given
off, and the fermentation being consequently aerobic.

The intimate association of the two organisms in the lumps of jelly
suggested a symbiotic relationship. Experiments made to ascertain
how the two affected each other failed however to bear out this view.
The influence of the bacterium we have seen to be largely shown in
the formation of relatively considerable quantities of both acetic and
succinic acids. Cultivations of the yeast were consequently made in
the presence of different proportions of acid, with the expectation of
finding that an acid medium would be advantageous to it.

Three flasks were prepared, each containing 90 c.c. of 10 per cent,
solution of cane-sugar to which 10 c.c. of yeast- water were added.
These were then carefully sterilised. To flask A, acetic acid was added
till O'l per cent, was present ; B contained 0*25 per cent., and C 0*5 per
cent, of the same acid. Each was then infected with 1 c.c. of a pure
culture of the yeast, and they were allowed to ferment at the labora-
tory temperature (about 10° C). A control was prepared in a fourth
flask, no acid being added. The fermentation which resulted in A
was about equal to that in the control ; that in B was less vigorous,
and that in C was very feeble. After two days, while the fermentation
was still active, they were all filtered on to tared filters, and the
quantity of yeast weighed. The control was then found to contain
0'09 gram of yeast, while the other flasks contained respectively — A,
0-106, B, 0-079, and C, 0-027 gram. So far as the growth of the yeast
was concerned, 0-1 per cent, of acetic acid was favourable ; but 0*25
per cent, was slightly inhibitory, while 0*5 per cent, was markedly so.

The alcohol was then distilled off' and estimated. A contained 2 per
cent., B 1*33 per cent., and C 1 per cent., while the control contained
1*66 per cent, of the spirit. These figures agree with the conclusion
based upon the weight of the yeast. The same results were obtained
when lactic acid was substituted for acetic.

As in the original experiments made with cane-sugar, the bacterium
produced more acetic acid than the highest proportion used in these
fermentations, it is clear that the bacterium does not conduce in this
way to any increase of growth or activity of the yeast.

82

Profs. H. Marshall Ward and J. Beynolds Green.

It seemed possible that the bacterium might assist the latter by
fixing nitrogen from the air, as bacteria have been found to do in other
cases of symbiosis. Careful experiments showed, however, that neither
the conjoint organism nor the bacterium alone could grow in a culture
fluid which did not contain combined nitrogen. A quantitative esti-
mation of nitrogen was made of two fermentations of cane-sugar, one
by yeast alone, the other by the conjoint organism, a control of the
culture fluid alone being examined simultaneously. The nitrogen was
determined by Kjeldahl's process. The initial amount of nitrogen
present was 0*013 gram. Neither flask showed any variation from
this quantity at the end of the fermentation.

The presence of the bacterium was seen consequently to be of no
service to the yeast, but on the other hand to be disadvantageous to
its growth.

The yeast was, on the other hand, found to be of some value to the
bacterium. During the fermentations the former was found to excrete
a certain amount of various extractives into the liquid, which had a
distinctly nutritive value to the latter. The bacterium grew very
much better at the expense of these extractives than it did when sup-
plied with combined nitrogen in the form of ammonium tartrate, or
of asparagin. Comparative experiments made with the extractives
prepared from a fermentation showed that neither asparagin nor an
ammonium salt could minister to its development so readily as they
did. This view was also supported by the fact that yeast-water had
already been found to be the best form of supplying combined nitrogen
artificially to cultures of the isolated bacterium.

The alcohol was of no nutritive value to the latter. When it was
cultivated in the presence of varying quantities of the spirit, it made
no use of it, and the original quantity of alcohol was found to be pre-
sent in the liquid at the conclusion of the experiment.

The relationship between the two is not therefore one of symbiosis.
The bacterium appears to be a saprophyte, thriving at the expense of
the nitrogenous excreta of the yeast. It is not at all parasitic on its
neighbour, the yeast never being injured by its presence until sufficient
acid has been produced to cause a secondary inhibition.

In the formation of the acetic acid this bacterium is peculiar. It
cannot be referred to the ordinary group of acetifying organisms,* as
it has not the power of affecting alcohol as they do. Its action on the
sugar seems to be a direct one, causing a formation of acid at the
expense of the latter, without setting up any preliminary fermenta-
tion. Bacterium aceti (Brown) has also this power in the presence of
oxygen, but it can oxidise alcohol in addition.

The immediate antecedent of the acid appears to be fructose.

* See Beijerinck, ' Cent. f. Bakt.,' vol. 4, 1898, for a summary of the kno$ra
acetic organisms.

A Sugar Bacterium.

83

Though the bacterium flourishes in solutions of cane-sugar, it does not
convert the latter immediately into the acids of the fermentation, but
hydrolysis takes place first. In one experiment 150 c.c. of a 10 per
cent, solution of cane-sugar, containing 15 c.c. of yeast water, were
infected in a sterile flask with a pure culture of the bacterium, and
fermentation was allowed to proceed for twelve days at the tempera-
ture of the laboratory. At the end of that time the acetic acid was
distilled off, and the residue poured into alcohol to precipitate another
constituent, which will be referred to below. The remainder of the
sugar was left in solution in the alcohol. After filtration the alcoholic
liquid was evaporated to a syrupy consistency, and the residue taken
up with water. • It was then divided into two equal portions, and one
of them was acidified with sulphuric acid till the concentration of the
latter was 2 per cent. It was then boiled on a water-bath for two
hours and carefully neutralised. An aliquot part of each was then
titrated with Fehling's solution, and the cupric oxide formed was
weighed. The weights of the two were almost identical, differing by
only 0*002 gram. The inversion of the cane-sugar by the organism
had consequently been practically complete.

A few bacteria only have been found to secrete invertase. Fermi
and Montesano found that Bacillus megatherium, B. fluorescens lique-
faciens, the red Kiel bacillus, and Proteus vulgaris were capable of
producing it in bouillon to which cane-sugar had been added. Van*
Tieghem showed that it was secreted by Leuconostoc mesentenoides. To
these the bacterium under observation must now be added. The
following experiment shows that the inversion noticed in the last case
was due to an exoreteel enzyme.

A good culture was made in cane-sugar solution, and, after a few
days, was filtered under pressure through a Chamberlandt's porcelain
tube. Three flasks were taken, and 15 c.c. of cane-sugar solution of ■
2 per cent, concentration placed in each. They were then carefully
sterilised. To A and B, 2 c.c. of the filtrate from the culture were
added, the filtrate having been found free from organisms by careful
microscopic examination; B was then boiled and allowed to cool
slowly. They were all kept for a few days at the temperature of
22° C. in an incubator. On titration with Fehling's solution, C, the
control, gave no reduction, B a slight one, and A threw down a con-
siderable precipitate of cuprous oxide. The small reduction in B was
due no doubt to a trace of invert-sugar present in the culture fluid
added to the flask. The much greater reduction in A was due to
invertase which the bacteria had excreted.

The experiments quoted in the early part of this paper show, how-
ever, that the inversion must take place gradually and the fructose
be supplied almost as it is wanted. In presence of fructose only, growth
is almost impossible.

84

A Sugar Bacterium.

Attention has been called in the first part of this paper to the
amount of viscous material which the bacterium produces in solutions
of cane sugar. There is a great similarity between its action in this
respect and that of Van Tieghem's Leuconostoc.

An examination was made of the viscous material, a special culture
being made in a large flask for the purpose. The slimy material was
found to be slowly soluble in water, yielding an opalescent solution.
When poured into an excess of alcohol, it gave a bulky flocculent pre-
cipitate. This was allowed to settle and the alcohol was decanted, and the
wet precipitate thrown on to a filter. When all the spirit had drained
away, it was taken from the filter and stirred with water in a beaker.
A considerable quantity dissolved, but a good deal of residue was left
in suspension. The watery solution was filtered off and examined.

It gave a reddish-purple colour on the addition of iodine, but had no
action on Fehling's fluid. Heated with 2 per cent, of sulphuric acid
on a water-bath for two hours and then neutralised, it gave evidence
of the presence of sugar. It reduced Fehling's fluid on boiling, and
yielded an osazone when treated with phenylhydrazine acetate. It
deflected a ray of polarised light, and had a specific rotatory power of
[a]D = +130.

The residue, which was insoluble in water, was coloured violet on
the addition of iodine. It had no action on Fehling's solution, but
was converted into a reducing sugar by boiling it with 2 per cent, of
sulphuric acid. It was readily soluble in a 10 per cent, solution of
caustic soda, and less freely in a 1 per cent, solution. Neutralisation
or dilution did not cause it to be reprecipitatecl. The solution had no
action on polarised light.

There were thus found to be present in the viscous liquids two dis-
tinct carbohydrates, which possessed much in common with Scheibler's
dextran, but which were not quite identical with the latter. They both
appeared to be members of the hemi-celluloses.

The exact relation of these bodies to the bacterium has not been
determined. While there are some grounds for thinking that they are
not so much products of fermentation in the strict sense as of its ordi-
nary biological processes, being perhaps only the substance of the
different sheaths to which allusion has been made in the first part of
this paper, it is not certain that they may not be regarded as products
produced altogether outside the organism, in consequence of an altera-
tion of the fructose produced by the hydrolysis. It is altogether
unlikely that they resulted from the glucose moiety of the inversion,
as this sugar is not a favourable medium for the growth of the organ-
ism, and such cultivations as are possible in solutions of glucose do not
show the presence of any viscous material.

Experiments in Micro-metallurgy.

85

"Experiments in Micro-metallurgy: — Effects of Strain. Pre-
liminary Notice." By James A. Ewing-, F.E.S., and Walter
Bosenhain, 1851 Exhibition Besearch Scholar, Melbourne
University. Eeceived and Bead March 16, 1899.

[Plates 1 — 5.]

Much information has been obtained regarding the structure of
metals by the methods of microscopic examination initiated by Sorby and
successfully pursued by Andrews, Arnold, Charpy, Martens, Osmond,
Boberts- Austen, Stead, and others. When a highly polished surface of
metal is lightly etched and examined under the microscope, it reveals
a structure which shows that the metal is made up in general of irre-
gularly shaped grains with well defined bounding surfaces. The
exposed face of each grain has been found to consist of a multitude of
crystal facets with a definite orientation. Seen under oblique illumi-
nation, these facets exhibit themselves by reflecting the light in a
uniform manner over each single grain, but in very various manners
over different grains, and, by changing the angle of incidence of the
light, one or another grain is made to flash out comparatively brightly
over its whole exposed surface, while others become dark.

It is also well known that the grains are deformed when the metal
is subjected to such processes as cold hammering, or cold rolling, or
wire-drawing. On polishing and etching a piece strained in any such
way, the grains are found to be on the whole longer in the direction in
which the metal is extended than in other directions. But on heating
the metal sufficiently a re-formation of structure occurs, and the grains
are found to have again assumed forms in which there is no direction
of predominating length. In iron this recrystallisation occurs at a red
heat. It is also known that prolonged exposure of iron to a tempera-
ture of about 700° C. tends to produce a larger granular structure than
is found if the metal is somewhat quickly cooled from a higher tem-
perature.

The grains appear to be produced by crystallisation proceeding, more
or less simultaneously, from as many centres or nuclei as there are
grains, and the irregular more or less polygonal boundaries which are
seen on a polished and etched surface result from the meeting of these
crystal growths. The grains are, in fact, crystals, except that each of
their bounding surfaces is casually determined by the meeting of one
growth with another. This is, we believe, the view usually accepted
by metallurgists •* but there is considerable difference of opinion as to
the part played by foreign matter in possibly contributing to form a
cement at the intergranular junctions.

* See especially two papers by Mr. J". E. Stead ('Jour. Iron and Steel Inst.,'
1898).

86

Mr. J. A. Ewing and Mr. W. Bosenhain.

The experiments^ of which this is a preliminary account, have been
directed to examine the behaviour of the crystalline grains when the
metal is subjected to strain.

For this purpose we have watched a polished surface under the micro-
scope while the metal was .gradually extended until it broke. By
arranging a small straining machine on the stage of the microscope, we
have been able to keep under continuous observation a particular
group of crystalline grains while the piece was being stretched, and
have obtained series of photographs showing the same group at various
stages in the process. Strips of annealed sheet iron, sheet copper, and
other metals have been examined in this way. We have also observed
the effects of strain on the polished surfaces of bars in a 50-ton testing
machine by means of a microscope hung from the bar itself, and have
further observed the effects of compression and of torsion.

When a piece of iron or other metal exhibiting the usual granular
structure is stretched beyond its elastic limit a remarkable change
occurs in the appearance of the polished and etched surface, as seen
by the usual method of " vertical " illumination. A number of sharp
black lines appear on the faces of the crystalline grains : at first they
appear on a few grains only, and as the straining is continued they
appear on more and more grains. On each grain they are more or
less straight and parallel, but their directions are different on different
grains. At first, just as the yield-point of the material is passed, the
few lines which can be seen are for the most part transverse to the
direction of the pull. As the stretch becomes greater oblique systems
of lines on other grains come into view.

The photograph, fig. 1 (Plate 1), taken from a strip of transformer
plate (rolled from Swedish iron and annealed after rolling), gives a
characteristic view of these lines as they appear after a moderate
amount of permanent stretching, but long before the iron has reached
its breaking limit.

The appearance of each grain is so like that of a crevassed glacier,
that these dark lines might readily be taken for cracks. Against this,
however, was the consideration that an over-strained piece of iron
recovers its original elasticity after a period of rest, though, as we
found, the dark lines did not disappear when recovery took place, and
further that sharp lines of the same nature were not seen on the surface
of metal which was polished and etched after straining.

The real character of the lines is apparent when the crystalline
constitution of each grain is considered. They are not cracks, but
slips along planes of cleavage or gliding planes.

Fig. 2 is intended to represent a section through the upper part of
two contiguous surface grains, having cleavage or gliding planes as
indicated by the cross-hatching, AB being a portion of the polished
surface. When the metal is pulled beyond its elastic limit, in the

Experiments in Micro-metallurgy.

87

direction of the arrows, yielding takes place by finite amounts of slips
occurring at a limited number of places in the manner shown at
a, b, c, d, e (fig. 3). This slip exposes short portions of inclined
surfaces, and when viewed under normally incident light, these surfaces
appear black because they return no light to the microscope. They are
consequently seen as dark lines or narrow bands, extending over the
polished surface in directions which depend on the intersection of the
polished surface with the surfaces of slip.

We have proved the correctness of this view by examining these
bands under oblique light. When the light is incident at only a small
angle to the polished surface, the surface appears for the most part
dark; but here and there a system of the parallel bands shines out
brilliantly, in consequence of the short cleavage or gliding surfaces which
constitute the bands having the proper inclination for reflecting the
light into the microscope. Fig. 4 is the photograph of a strained piece
of Swedish iron illuminated in this way. The magnification is 280
diameters. The- groups of parallel bright bands which appear in the
photograph may readily be observed under the microscope to be
exactly coincident with the black bands seen under vertical illumina-
tion ; and by changing the angle of incidence of the oblique light, the
same bands may be made to appear dark on a faintly luminous ground.
Eotation of the stage to which the strained specimen is fixed makes
the bands on one or another of the grains flash out successively, with
kaleidoscopic effect. In what follows we shall speak of these lines as
slip bands. Fig. 1, through a mixed illumination, shows some of the
slip bands bright and some dark.

Incidentally fig. 4 illustrates the fact that oblique lighting picks out
the boundaries of the crystalline grains, showing that these boundaries
are marked by inclined surfaces connecting grains whose faces are at
different levels. This is observed also in the etched surface of the
metal before straining. The boundaries, which appear dark under
vertical light, are bright on one side of each crystalline grain, when
the light falls with grazing incidence from one side. But the sloping
surfaces which mark the boundaries between the grains have by no
means the sharply definite inclination which characterises the surfaces
which form the slip bands. One or more groups of slip bands will
shine out very brightly when the light has a particular angle of inci-
dence, and will vanish when the incidence is slightly changed. The
boundaries are not in general so bright, but they remain fairly bright
while the incidence is changed through wide limits.

When the metal is much strained a second system of bands appears
on some of the grains, crossing the first system at an angle, and in
some cases showing little steps where the lines Cross. These bands are
clearly due to slips occurring in a second set of cleavage or gliding
surfaces. An example of the crossed systems of bands will be seen in

88

Mr. J. A. Ewing and Mr. W. Eosenhain.

fig. 8. The crystals in metals are generally cubical, but the angle at
which the intersecting systems of bands cross depends on the inclina-
tion of the polished surface to the planes of cleavage. Occasionally a
third system of bands may be seen.

As straining proceeds the originally smooth surface of the specimen
becomes roughened by the surface grains changing in their relative
levels and also becoming more or less inclined, as well as more or less
stepped. All this happens in consequence of the slips which they and
their neighbours undergo. To this is due the dull appearance which
an originally bright surface assumes when the metal is overstrained.
Under the microscope the strained surface is seen to be full of ups and
downs, and a continuous alteration in focus • is required to trace the
system of bands, even over the face of a single grain.

When the experiment is made with a polished but unetched specimen
the slip bands appear equally well. The boundaries of the grains are
invisible before straining ; but they can be distinguished as the strain
proceeds, for the slip bands form a cross-hatching which serves to mark
out the surface of each grain. To strain a polished but unetched
specimen reveals in a striking way the granular character of the
structure.*

Figs. 5, 6, and 7 are selected from a series of photographs showing
under a magnification of 140 diameters, the same group of crystalline
grains in a specimen of soft wrought iron at various stages of straining
by pull. The arrows show the direction in which the pull was applied.

Fig. 5 shows the group before straining began. Fig. 6 is the same
group after the strain had been carried some way past the yield point.
Fig. 7 is the same group after the piece had suffered considerable
further extension in the same direction. Comparison of the three will
show how the grains change their shape in consequence of the slips
which occur in them, and also how the faces of the grains become tilted
and altered in relative level.

Fig 8 is another sample of iron strained by pull. The specimen in
this case was a bar of Swedish iron, in which a comparatively large
crystalline structure had been developed by annealing for some hours at
700° C. The photograph was taken after the bar had been broken in
the testing machine, and shows with a magnification of 400 diameters
a portion of the surface not far from the place of fracture. The large
grain which appears to the left in fig. 8 measured 0*16 mm. in the
direction of the arrows before straining, and was extended by the
strain in that direction to about 0*2 mm. The slip bands upon it are
on the average 1/400 mm. apart. This applies to both of the two
systems of bands which appear in the photograph. The apparent

* The fact that the crystalline structure is revealed when a specimen with a
polished unetched surface is strained has already been pointed out by Charpy
(' Comptes Eendus,' vol. 123, 1896, p. 225).

Experiments in Micro -metallurgy \

89

width of the slip bands themselves is too small to be measured
with any accuracy ; it does not appear at any place here to exceed
1/2000 mm.

The slip-bands are developed by compression as well as by extension.
Fig. 9 is a photograph (at 400 diameters) from the polished and
etched side of a block of Low Moor iron compressed in the testing
machine sufficiently to give it a considerable amount of permanent set.
The bands developed by compression have apparently all the character-
istics which they present in stretched pieces, and we could not, by
microscopic examination of the surface, distinguish, in this respect
between the effects of compression and extension. The irregular dark
patches in fig.- 9 are streaks of slag.

By twisting an iron bar well beyond the elastic limit the slip-bands
are made to appear, for the most part in directions parallel and per-
pendicular to the axis of twist.

A strip of sheet metal, such as iron or copper, in the soft state, when
bent and unbent in the fingers shows them well developed by the
extension and compression of the surface.

We have developed the slip-bands in iron, steel, copper, silver, lead,
bismuth, tin, gun-metal, and brass. In silver they show particularly
well, the crystalline structure being large and the lines straight. In
copper also the lines are straighter and more regularly spaced than is
general in iron. Most of these metals have been tested in the form
of blocks under compression. A beautiful development of slip-bands
may readily be produced by pinching a button of polished silver or
copper in a vice.

In carbon steels we have found the slip-bands considerably more
difficult to observe than in wrought iron. The smaller granular struc-
ture of steel apparently makes the slip-bands correspondingly minute.
In mild steel they are seen readily enough, but in a rather high carbon
steel we succeeded in seeing them only with difficulty in the " ferrite "
areas under a magnification of 1,000 diameters. A cast piece of the
nearly pure iron used for dynamo magnets showed a relatively very
large granular structure and well marked slip-bands.

These experiments throw what appears to us to be new light on the
character of plastic strain in metals and other irregular crystalline
aggregates. Plasticity is due to slip on the part of the crystals along
cleavage or gliding surfaces. Each crystalline grain is deformed by
numerous internal slips occurring at intervals throughout its mass. In
general these slips no doubt occur in three planes, or possibly more,
and the combination of the three allows the grain to accommodate
itself to its envelope of neighbouring grains as the strain proceeds.
The action is discontinuous : it is not a homogeneous shear but a series
of finite slips, the portion of the crystal between one slip and the next
behaving like a rigid solid. The process of slipping is one which takes

90

Experiments in Micro-metallurgy.

time, and in this respect the aggregate effect is not easily distinguish-
able from the deformation of a viscous liquid.

We infer from the experiments that " flow " or non-elastic deforma-
tion in metals occurs through slip within each crystalline grain of
portions of the crystal on one another along surfaces of cleavage
or gliding surfaces. There is no need to suppose the portions which
slip to be other than perfectly elastic. The slip, when it occurs, in-
volves the expenditure of work in an irreversible manner.

It is because the metal is an aggregate of irregular crystals that it is
plastic as a whole, and is able to be deformed in any manner as a result
of the slips occurring in individual crystals. Plasticity requires that
each portion should be able to change its shape and its position. Each
crystalline grain changes its shape through slips occurring within itself,
and its position through slips occurring in other grains.*

From what is known about the break-down in elasticity which occurs
as the immediate effect of overstrain and the subsequent recovery of
elasticity after a period of rest, it would seem that the surfaces over
which slip has occurred are at first weak, but heal with the lapse of
time. To discuss these points, however, lies beyond the scope of a
preliminary notice.

The experiments were made in the engineering laboratory at Cam-
bridge, and are being continued. We wish to take this opportunity of
thanking Sir W. Eoberts- Austen and Mr. T. Andrews for the great
kindness with which, at the outset of our work, they gave us the
benefit of their experience in preparing and observing microscopic
specimens of metals.

[Note added April 14, 1899. — In a specimen of cast nickel, which
showed after etching a crystalline structure much resembling that of iron,
but on a considerably smaller scale, straining developed minute slip-
bands, which are clearly apparent under a power of 1,000 diameters.

We have also examined a specimen of pure gold by compressing a
cast button with a polished face, not etched. The straining reveals
crystalline structure on a large scale, and in each of the crystalline
grains there is a superb development of slip-bands. They are long,
nearly straight, exceedingly numerous, and very closely spaced. A
power of 1,000 diameters is required to see them well. Two inter-
secting systems are common, and three are well seen in some of the
grains, forming a regular geometrical network. The intergranular
boundaries are sharply denned by the meeting of the slip-bands on
each grain with those on its neighbours. The slip-bands in adjacent
crystals meet in a way which demonstrates the absence of any appre-
ciable quantity of foreign matter in the intergranular junctions.]

* Attention should be called in this connection to the experiments of Messrs.
MeConnel and Kidd on the plasticity of glacier ice (' Roy. Soc. Proc.,' vol. 44,
p. 331). They found that bars cut from glacier ice, which is an aggregate of
irregular crystals, are plastic.

Ewing and Rosenhain.

Roy. Soc. Proc.y Vol. 65, Plate t.

■Jt*. .^....1 ' ^^IbtK^— _

FiO. 1. — Soft Sheet Iron strained by tension. 400 diameters.

A B

Fig. a. Before straining.

Fig. J. >4#er straining

Ewing and Rosenhain. Roy. Soc. Proc, Vol. 65, Plate 2.

Ewing and Roienhain.

Roy. Soc. Proc. , Vol. 65, Plate 3.

Fig. 6. — After moderate straining. 140 diameters.

I

Ewing and Rosenhain.

Roy. Soc. Proc, Vol. 65, Plate 4.

Fig. 8. — Swedish. Iron, much strained. 400 diameters.

Ewing and Rosenhain.

Roy. Sof. Proc, Vol. 65, Plate 5.

The Physiological Action of Choline and Neurine.

91

" The Physiological Action of Choline and Neurine." By F. W.
Mott, MIX, F.E.S., and W. 1). Halliburton, M.D., F.RS.
Eeceived March 13 — Bead April 20, 1899.

(Abstract.)

The cerebro-spinal fluid removed from cases of brain atrophy, particu-
larly from cases of General Paralysis of the Insane, produces when
injected into the circulation of anaesthetised animals (dogs, cats,
rabbits), a fall of arterial blood pressure, with little or no effect on
respiration. This pathological fluid is richer in proteid matter than
the normal fluid, and among the proteids, nucleo-proteid is present.
The fall of blood pressure, is, however, due not to proteid, nor to inor-
ganic constituents, but to an organic substance, which is soluble in
alcohol. This substance is precipitable by phospho-tungstic acid, and
by chemical methods was identified as choline. The crystals of the
platinum double salt, which, when crystallised from 15 percent, alcohol,
are characteristic octahedra, form the most convenient test for the
separation and identification of this base.

The nucleo-proteid and choline doubtless originate from the disinte-
gration of the brain tissue, and their presence indicates that possibly
some of the symptoms of General Paralysis may be due to auto-intoxi-
cation ; these substances pass into the blood, for the cerebro-spinal fluid
functions as the lymph of the central nervous system. We have iden-
tified choline in the blood removed by venesection from these patients
during the convulsive seizures which form a prominent symptom in the
disease.

Normal cerebro-spinal fluid does not contain nucleo-proteid or choline,
or if these substances are present, their amount is so small that they
cannot be identified. Normal cerebro-spinal fluid produces no effect on
arterial pressure ; neither does the alcoholic extract of normal blood or
of ordinary dropsical effusions.

The presence of choline in the pathological fluids will not explain
the symptoms of General Paralysis ; for instance, it will not account
for the fits just referred to. Its presence, however, is an indication
that an acute disintegration of the cerebral tissues has occurred. If
other poisonous substances are also present, they have still to be dis-
covered.

Our proof that the toxic material we have specially worked with is
choline, rests not only on chemical tests, but also on the evidence
afforded by physiological experiments ; the action of the cerebro-spinal
substance exactly resembles that of choline. Neurine, an alkaloid
closely related to choline, is not present in the fluid ; its toxic action is
much more powerful, and its effects differ considerably from those of
choline.

92

Drs. P. W. Mott and W. D. Halliburton.

Physiological Action of Choline.

The doses employed were from 1 to 10 c.c. of a 0*2 per cent, solution,
either of choline or of its hydrochloride. These were injected intra-
venously.

The fall of blood pressure is in some measure due to its action on
the heart, but is mainly produced by dilatation of the peripheral vessels,
especially in the intestinal area. This was dmonstrated by the use of
an intestinal oncometer. The limbs and kidneys are somewhat lessened
in volume ; this appears to be a passive effect, secondary to the fall in
general blood pressure. The drug causes a marked contraction of the
spleen, followed by an exaggeration, of the normal curves, due to the
alternate systole and diastole of that organ.

The action on the splanchnic vessels is due to the direct action of the
base on the neuro-muscular mechanism of the blood vessels themselves ;
for after the influence of the central nervous system has been removed
by section of the spinal cord, or of the splanchnic nerves, choline still
causes the typical fall of blood pressure. The action of peripheral
ganglia was in other experiments excluded by poisoning the animal
previously with nicotine.

Section of the vagi produces no effect on the results of injecting
choline.

We have obtained no evidence of any direct action of the base on
the cerebral vessels.

Choline has little or no action on nerve trunks, as tested by their
electrical response to stimulation. This aspect of the subject has been
taken up by Dr. Waller and Miss Sowton, who will publish their
results fully in a separate paper.

Choline has no effect on respiration.

The effect of choline soon passes off, and the blood pressure returns
to its previous level. This is due partly to the great dilution of the
substance injected by the whole volume of the blood, and may be
partly due to the excretion of the alkaloid, or to its being broken up
into simpler substances by metabolic processes. We could not find it
in the urine.

If the animal has been previously anaesthetised with a mixture of
morphine and atropine, the effect produced by choline is a rise of
arterial pressure, accompanied by a rise of the lever of the intestinal
oncometer. Other anaesthetics cause no change in the usual results.
We consider this observation of some importance, for it shows how the
action of one poison may be modified by the presence of another.
This has some bearing on General Paralysis, for the arterial tension in
that disease is usually high, not low, as it would be if choline were the
only toxic agent at work.

The Physiological Action of Choline and Neurine.

93

Physiological Action of Neurine.

The doses employed varied from 1 to 5 c.c. of a 0*1 per cent, solu-
tion. These were injected intravenously.

Neurine produces a fall of arterial pressure, followed by a marked
rise, and a subsequent fall to the normal level. Sometimes, especially
with small doses, the preliminary fall may be absent. Sometimes,
especially with large doses, by which presumably the heart is more
profoundly affected, the rise is absent.

The effect of neurine on the heart of both frog and mammal is much
more marked than is the case with choline ; in the case of both choline
and neurine, the action on the frog's heart is antagonised by atropine.

The slowing and weakening of the heart appear to account for the
preliminary fall of blood pressure ; in some cases this is apparently
combined with a direct dilating influence on the peripheral vessels.

The rise of blood pressure which occurs after the fall, is due to the
constriction of the peripheral vessels, evidence of which we have
obtained by the use of oncometers for intestine, spleen, and kidney.

After the influence of the central nervous system has been removed
by section of the spinal cord, or of the splanchnic nerves, neurine still
produces its typical effects.

After, however, the action of peripheral ganglia has been cut off by
the use of nicotine, neurine produces only a fall of blood pressure.
It therefore appears that the constriction of the vessels is due to the
action of the drug on the ganglia; in this, it would agree with nicotine,
coniine, and piperidine.

Section of the -<Tagi produces no influence on the results of injecting
neurine.

In animals anaesthetised with morphine and atropine, injection of
neurine causes only a rise of blood pressure, which is accompanied with
constriction of peripheral vessels.

Xeurine produces no direct results, so far as we could ascertain, on
the cerebral blood vessels.

Neurine is intensely toxic to nerve-trunks (Dr. Waller and Miss
Sowton).

It produces a marked effect on the respiration. This is first greatly
increased, but with each successive dose the effect is less, and ultimately
the respiration becomes weaker, and ceases altogether. The animal
can still be kept alive by artificial respiration.

The exacerbation of respiratory movements will not account for the
rise of arterial pressure ; the two events are usually not synchronous,
and an intense rise of arterial pressure (due, as previously stated, to
contraction of peripheral blood vessels) may occur when there is little
or no increase of respiratory activity or during artificial respiration.

As confirmatory of Cervello's statement that neurine acts like

94

Prof. E. Waymouth Eeicl.

curare on the nerve endings of voluntary muscle, and to which he
attributes the cessation of respiration, we may mention that after an
animal has been poisoned with neurine, asphyxiation causes little or
none of the usual convulsions.

The full paper contains references to previous work on the subject,
and complete details of the methods used, and the cases investigated ;
it is illustrated by reproductions of numerous tracings.

[Note added April 20, 1899. — It should be mentioned that in the
cases of brain atrophy referred to, the cerebro-spinal fluid was removed
soon after death. Since the foregoing abstract was written, we have,
however, had the opportunity of examining two specimens removed
during life by lumbar puncture, and the results of our experiments
with these corroborate the conclusions previously arrived at.]

" On Intestinal Absorption, especially on the Absorption of
Serum, Peptone, and Glucose." By E. Waymouth Eeid,
F.B.S., Professor of Physiology in University College, Dundee,
St. Andrew's University, 1ST.B. Eeceived March 30, — Eead
April 20, 1899.

(Abstract.)

The experiments detailed in the full paper deal with the absorption
from the intestine of the animal's own serum, and of solutions of
glucose and peptone. The method employed has been that introduced
by Leubuscher, in which two loops of intestine are simultaneously
employed, the one the experimental, and the other the control, loop.

The conclusions arrived at are as follows : —

1. A physiological activity of the intestinal epithelium in the act of
absorption is demonstrated by —

(a) The absorption by an animal of its own serum (or even plasma)

under conditions in which filtration into blood capillaries or
lacteals, osmosis, and adsorption are excluded.

(b) By the cessation or diminution of the absorption of serum when

the epithelium is removed, injured, or poisoned, in spite of the
fact that removal, at any rate, must increase the facilities for
osmosis and filtration.

2. The activity of the cells is characterised by a slower uptake of
the organic solids of the serum than of the water, and a rather quicker
uptake of the salts than of the water. The relations to one another
of the absorptions of these various constituents is variable in different
regions of the intestinal canal (upper ileum, lower ileum, and colon).

On Intestinal Absorption.

95

3. No evidence can be obtained of specific absorptive fibres in the
mesenteric nerves.

4. The state of nutrition of the cells is the main factor in their
activity, and this is intimately associated with the blood supply.

5. In reduction of the rate of absorption, without detachment of
epithelium, the absorption of the various constituents of serum is
reduced in the proportion in which they exist in the original fluid.

6. The activity of the cells may be raised by stimulation with weak
alcohol, without evidence of concomitant increase of blood supply.

7. The bile has no stimulant action on the cells.

8. The cells exhibit an orienting action upon salts in solution (sodic
chloride especially). In a loop of gut with injured cells sodic chloride
enters the lumen from the blood, at a time when it is being actively
absorbed from a normal control loop in the same animal. (This fact
was first noted by 0. Cohnheim.)

9. The absorption of water from solutions introduced into the gut
is dependent upon -two factors : —

(a) The physical relation of the osmotic pressure of the solution in

the gut to the osmotic pressure of the blood plasma.

(b) The physiological regulation of the difference of osmotic pres-

sure by the orienting mechanism of the cells.

10. The chief factor in the absorption of peptone is an assimilation
(or adsorption) by the cells, while in the absorption of glucose, diffu-
sion, variable by the permeability of the cells (and so, probably, related
to their physiological condition) is the main factor.

11. By removal of the epithelium, the normal ratio of peptone to-
glucose absorption is upset, and the value tends to approach that of
diffusion of these substances through parchment paper into serum.

12. Absorption in the lower ileum is greater for the organic solids
of serum, and less for peptone and glucose than in the upper ileum.
The relative absorption of water in the upper and lower ileum is.
variable.

13. The relative impermeability of the lower ileum to glucose dis-
appears with removal of the epithelium.

14. Absorption in the colon is for all constituents of serum, and for
peptone and glucose far less per unit of measured surface than in the
middle region of the ileum.

15. The normal relative excess of salt absorption from serum over
water absorption, observed throughout the intestine, is most marked in
the colon, and more marked in the lower than in the upper ileum.

16. Finally it is suggested that the cell activity which causes serum
to pass over to the blood is of the same nature as that involved in the
orienting action of the cells upon salts in solution.

VOL. LXY.

H

96

Prof. F. 0. Bower.

" Studies in the Morphology of Spore-producing Members. IV.
The Leptosporangiate Ferns." By F. 0. Bower, Sc.D., F.E.S.,
Regius Professor of Botany in the University of Glasgow.
Received April 6 —Read April 20, 1899.

(Abstract.)

The characters used in current classifications of Ferns need streng-
thening. In recent years the more detailed knowledge of the prothallus
has been used for this purpose, but while not denying its value in
certain specific cases, the author holds that the vegetative development
of the prothallus is an uncertain guide to a general classification. On
the other hand the archegonium is so uniform in its character that it
gives little help ; the comparison of the antheridium is, however, a
useful aid, though not sufficiently varied to serve in detail.* Accord-
ingly the sporophyte must be the main basis. Its vegetative organs
have lately been largely used for systematic purposes by Christ ;f but
the same objection holds here as in Phanerogams to the use of these
as characters of first rank for comparison. An attempt has therefore
been made in this memoir to strengthen the characters derived from
the sorus by a fresh examination of its details, and certain of its
features will now be used for purposes of general comparison, which
have hitherto received too little attention ; they are : —

1. The relative time of appearance of sporangia of the same sorus.

2. Certain details of structure of the sporangium, including its stalk.

3. The orientation of the sporangia relatively to the whole sorus.

4. The potential productiveness of the sporangium as estimated by

its spore-mother cells, and the actual spore-output.

Observations of these features extending over all the more important
living genera, coupled with data of habit and the characters of the
Gametophyte as collateral evidence, have led the author to divide the
homosporous Ferns thus : —

Marattiacese Eusporangiate
Osmundacege
Schizaeacese
Gleicheniacese
Matonineae
Loxsomaceae

Hymenophyllaceae J- Leptosporangiate
Gradatae i Cyatheaceae
Dicksonieaa
Dennstaedtiinae
The bulk of the
Polypocliaceae

* Heim, < Flora/ 1896, p. 355, &c.
f 1 Die Farrnkriiuter der Erde,' Jena, 1897.

Simplices

Mixtae

Studies in the Morphology of Spore-producing Members. 97

These divisions are primarily based on the order of appearance of the
sporangia in the sorus, the Simplices having all the sporangia of the
sorus formed simultaneously, the Gradatae having them disposed in
basipetal succession, and the Mixtae having the sporangia of different
ages intermixed. But it is found that other important characters run
parallel with these : thus the Simplices and Gradatae have an oblique
annulus (where definitely present), the Mixtse (with very few excep-
tions) have a vertical annulus. None of the Mixtae have been f ound to
have a higher spore-output per sporangium than sixty-four, but this
number is exceeded by some of the Gradatae, and large numbers are the
rule in the Simplices. The Simplices and Gradatae have relatively short
thick stalks, the Mixtae usually have long and thin stalks. The
orientation of the sporangia in the Simplices and Gradatae is usually
definite, in the Mixtae it is indefinite. The receptacle is often elon-
gated in the Gradatae, but not in the Simplices or Mixtae. The sum of
these characters, which for the most part run parallel to one another,
appears to give a substantial basis to the classification.

Evidence as to the transition from type to type has been collected.
In the case of the transition from a simultaneous to a successive sorus
it does not amount to a demonstration : but it is specially pointed out
how slight a step it is from such a sorus as that of Gleichenia dichotoma
to that of an Alsophila : that given a basal indusium and marginal
position, the similarity of sporangial structure and dehiscence between
Gleichenia and Loxsonia is suggestive ; as also the sporangial structure
and high spore-output in Hyrnenophyllum. Though we may recognise
these lines of similarity, they do not focus upon any one genus as the
actual transitional link from the simultaneous to the basipetal. But
the transition from the basipetal to the mixed sorus can be followed in
detail ; intermediate steps are seen in the Dennstaecltiinae, while the
fully mixed type is seen in the closely allied Damllia. Probably this
is only one of several such lines of transition from the basipetal to the
mixed type.

It is shown that a biological advantage would be gained by the
suggested transitions. In the Simplices the few sporangia are large,
and, arising simultaneously, make a demand all at once on the nutri-
tive resources of the part. In the Gradatae the smaller sporangia are
produced in succession upon an elongating receptacle, and the drain on
the part is spread over a longer period. But with the assumption of
the mixed character the drain may be spread over an equally long time,
while, as the elongated receptacle disappears, the surface from which
nourishment can be derived is enlarged, and the distance through
which it has to be transferred is shortened. Thus it appears biologi-
cally reasonable that the succession should be as suggested.

It is shown how the various types of dehiscence, and the action of
the annulus stand in close relation to the orientation of the sporangia,

H 2

98 Studies in the Morphology of Spore-producing Members.

and to their arrangement in the sorus. Thus the position of the
annulus, which has played so important a part in classification, has
been placed upon a footing of adaptation.

Estimates of numerical output of spores per sporangium have been
made with a view to illustrating the relation of the Eusporangiate and
Leptosporangiate ferns in this respect. The estimated output in the
Marattiaceae has been shown to be high ;* that of the Polypodiaceae
is sixty-four or less. The result of numerous countings is to show
that, of all Leptosporangiate ferns, Gleichenia approaches most nearly
to the Marattiaceae (Gl. flabellata may produce over 800 per sporan-
gium) ; Osmunda may have over 500, and Lyg odium 256. The most
interesting results were derived from the Hymenophyllaceae, in which
Hym. tunbridgense may have over 400, while species of Trichomanes
may produce as few as thirty-two per sporangium. These results,
when taken with those derived from the filmy Todeas, make it seem
probable that the filmy habit is a condition leading to reduction of
output per sporangium, and indicate that the Hymenophyllacea3 are
a derivative series of reduction.

A most important commentary upon the classification proposed is
derived from comparison of the antheridia, which Heimf found to be
the most dependable part of the Gametophyte for comparative pur-
poses. He recognises two types according to their dehiscence : the one
type includes, with the exception of two genera of Schizaeaceae, our
Simplices and Gradatae, while the other includes the Mixtae. I can
only regard this correspondence of parts, so aloof from one another as
the antheridium and the sporangium, as establishing the relations of
the Simplices and Gradatge upon a firmer footing ; the facts also give
substantial support to the distinction of the Gradatae and Mixtae.

The effect of the observations and comparisons in this memoir is
rather confirmatory of the current classifications than disturbing. The
divisions suggested would supersede those of Eusporangiatae and Lep-
tosporangiatae, though these terms would still be retained in a
descriptive sense. If the sub-orders Osmundaceae, Schizaeaceae, and
Marattiaceae be transferred from the end of the Synopsis Filicum to
the beginning, and grouped with Gleichenia and Matonia, we have the
" Simplices " before us. The Gradatae include the Cyatheaceae, Dick-
sonieae (Excl. Dennstaedtia), Hymenophyllaceae, and Loxsomaceae,
sequences probably of distinct descent, and, in my view, derivative
from some prior forms such as the Simplices ; and in the arrangement
of Sir Win. Hooker they hold a position following on the Gleichenia-
ceae. The family of Dennstaedtiinae, founded by Prantl to include
Dennstaedtia and Microlepia, also has its place here, but it leads on by
intermediate steps to undoubtedly mixed forms such as Davallia,

* 1 Studies,' No. 3, p. 60.
t ' Flora/ 1896, p. 355, &c.

On the Fertility of different Breeds of Sheep, tyc.

99

Cystopteris, Lindsaya, and the Pteridese. But this sequence is already
laid out in this order in the Synopsis, and it illustrates one at least of
the lines along which mixed forms are believed to have been derived
from the Gradatse. No attempt has been made to follow the natural
grouping of the Mixtas into detail, or to test the arrangement of them
in the Synopsis. Sufficient has, however, been said to show that the
systematic divisions of the ferns now proposed fall in readily with the
system of Sir William Hooker, notwithstanding that they are based
upon details of which he cannot have been aware.

" Note on the Fertility of different Breeds of Sheep, with Bemarks
on the Prevalence of Abortion and Barrenness therein." By
Walter Heape, M.A., Trinity College, Cambridge. Communi-
cated by W. F. B. Weldon, F.B.S. Beceived March 9,—
Bead April 20, 1899.

The importance of fertility as a factor in the survival of a species is
admirably demonstrated by Hafkine,* whilst Professor Karl Pearsonf
shows that fertility when correlated with other characteristics works a
progressive change, and that not only is fertility a race characteristic,
but may be a class characteristic in the human species.

Among domesticated animals, although fertility ma}' be a racial
characteristic, its importance may be much reduced from a variety of
circumstances.

Among sheep there is undoubted evidence of the racial character of
fertility, but the quality of the wool or the value of fat sheep of a par-
ticular breed may render that breed worth keeping in spite of a low
rate of fertility as compared with other breeds. Then the rate of fer-
tility may be artificially increased, as when certain rams of undoubted
value as progenitors, but useless as breeders if left to themselves,
become valuable sires by the help of the shepherd. In the same way a
certain breed of sheep, kept in one district and managed in a particular
manner, may be more liable to abortion, or to barrenness, or to mor-
tality among the lambs, than the same breed in another district
managed in another way, and yet the former may, on account of the
supply of food which it is possible to grow there per acre, prove the
most remunerative.

From these and many other similar reasons the survival of a particular
breed or its retention in, or importation to, a particular district is not
necessarily due to natural fitness or adaptability. At the same time a

* " Recherches sur l'adaptation au milieu chez les Inf usoires et les Bacteries ;
contribution a 1' etude de l'immunite," ' Annales de l'lnstitut Pasteur,' vol. 4, 1890.

f " Contributions to the Mathematical Theory of Evolution. Note on Repro-
ductive Selection," ' Roy. Soc. Proc.,' vol. 59, 1896.

100

Mr. W. Heape.

wider knowledge of the racial character of fertility among sheep might
exercise considerable influence on the survival of certain breeds in the
future. Just as the number of Long-horn cattle is rapidly becoming
reduced in this country on account of the length of time they require
to come to maturity and fatten, so the low fertility of Southdowns if
it cannot be increased may lead to the retention of that breed in the
hands of only a few special breeders.

The following account is a brief abstract of information obtained
from 397 sheep breeders who have supplied me with records of flocks
containing 122,673 ewes for the breeding season of 1896-97. Table IV
records certain particulars of eight pure breeds separately, and of ten
pure breeds jointly (" Various Pure Breeds " ; details of only a small
number of flocks were supplied for each of these ten breeds). These
are totalled, particulars of fifty-nine flocks of various cross-bred sheep
subjoined, and the totals for all breeds finally arrived at.

Besides the figures contained in Table IV, other statistics were sup-
plied regarding the age of rams and ewes, the size of breeding flocks,
the proportions of rams and ewes therein, the usual and the highest
percentages of barrenness experienced, &c. To this were added remarks
on the food given to rams and ewes at specified times, on their condi-
tion during the breeding season, on various methods of management of
breeding flocks, and on numerous views regarding matters which are
supposed to influence fertility, abortion, and barrenness amongst ewes ;
finally particulars of districts, subsoils, and weather were noted, and the
subject considered statistically from all these various points of view for
each breed.

The variation in the numbers of flocks and of ewes concerned in the
various calculations for each breed (see Table IV) is due to the inability
of all flock-masters to supply the whole of the information asked for in
the schedule, which, by the kindness of the Eoyal Agricultural Society,
was distributed to them, and is not due to my selection of flocks.

Fertility.

It is to be noted the number of lambs returned does not always repre-
sent the number born ; that information, although asked for, was not
always forthcoming and sometimes the number of lambs alive when
the schedule was filled up was given instead. The percentage of twins
given is, however, a check on this element of error, and in the follow-
ing account the columns under " lambs " and " twins," Table IV, are
considered together. The error is most apparent in the Hampshire
Down, Oxford Down, Dorset Horn and Lincoln breeds ; in the other
four breeds, in whiah the records have been most carefully kept, the
Suffolk, Shropshire, Kent and Southdown breeds, their fertility is
demonstrated to be a racial character.

On the Fertility of different Breeds of Sheep, §c. 101

The Suffolk breed is by far the most fertile, while the Southdowns
are at the bottom of the list : the value of the former as a prolific
breed is incontrovertible, while the record of the latter, as shown both
by tie percentage of twins and lambs, is so low as to suggest cause for
some anxiety and to show urgent need for close attention on the part
of breeders.

The Hampshire breed also stands low both with regard to twins and
lambs, especially when its low percentage of loss from abortion and
barrenness is considered ; whereas the Lincoln breed, although recording
only 111-1 per cent, of lambs, shows 29*09 per cent, of ewes bearing
twins, and the fault, with this breed, is obviously not so much low
fertility as heavy mortality among lambs and a high percentage of
loss from abortion and barrenness.

The most fertile of all pure breeds of which I have records, is the
Wensleydale breed, in which six flocks with a total of 319 ewes pro-
duce 177'43 per cent, of lambs (included in "Various Pure Breeds").
It is to be noted that both Wensleydale and Suffolk ewes when covered
by rams of other breeds, do not appear to produce a larger percentage
of lambs than they produce when covered by rams of their own breed ;
if anything, they seem to produce a somewhat smaller percentage of
lambs. On the other hand, Dorset Horn ewes are more usually fertile
and are more prolific with Down rams than with Dorset Horn rams.

It is a very usual practice with Dorset Horn flock-masters to keep a
Down ram to cover those ewes which fail to become pregnant to a Dorset
Horn ram, and, when pure-bred ewes are discarded from their breeding
flocks, they generally sell them " in young " to Down rams, when they
frequently produce as many as 170 per cent, of lambs.

It appears, therefore, that whereas the fertility of Suffolk and Wens-
leydale ewes is at its maximum when they breed to rams of their own
kind, Dorset Horn ewes require to be covered by rams of another
breed, in order that they may be stimulated to the greatest generative
activity.

Apart from the percentage of ewes which abort or are barren from
accident or constitutional defects, the fertility of ewes is chiefly affected
by their condition and by the condition of the ram at tupping time. I
have overwelming evidence that flocks in strong condition at tupping
time produce a higher percentage of lambs than flocks in poor condi-
tion at that period (Suffolk, Kent, Hampshire, Dorset Horn, and
Lincoln breeds) ; on the other hand, fat rams and ewes are associated
with a high percentage of barrenness, and not with a high percentage
of lambs.

In close connection with this subject is the production of twins.
Fifty-five per cent, of the flock-masters who send information on this
head, report that twins are usually born early in the lambing season,
and many of them add that otherwise the crop of lambs is small. That

102

Mr. W. Heape.

is to say, that most twins are produced by the ewes which first come in
season, and I interpret that to mean that it is the ewes with the most
active and most vigorous generative system, those which are in strong
breeding condition at that time, which bear the most twins. It is of
interest to note that among Southdown and Hampshire flocks, a com-
paratively small proportion of twins are born early in the lambing
season, and these breeds produce the smallest percentage of twins.

The effect of locality upon the fertility of a breed is worthy of con-
sideration, and statistics are given in Table V to illustrate this point.*
In considering these figures, numerous other important influences
must be borne in mind, which the information at my disposal makes it
impossible to dissociate from locality ; and in all cases, except those of
the Shropshire and Lincoln breeds, where a wide difference in fertility
is shown in different districts, it is possible that the variation in
fertility may be due to one or other or all of these various influences,
apart altogether from locality.

The fertility of a flock is greatly influenced by its management and
by the conditions under which it lives ; the condition, kind, and
amount of food available before tupping time will affect the condition
of the ewes and the percentage of twins subsequently born ; the season
may be more favourable for tupping or for lambing in one district
than in another ; cold, rain, or want of rain, will affect the feed, the
ground, and the ewes themselves ; while, owing to mortality among
ewes during lambing or at other times in the preceding year, the
flocks in one district may consist of a larger proportion of shearling
ewes than the flocks in another district, and this may affect the birth
rate of a flock.

These and numerous other such influences, combined with the undue
proportionate value of excessive loss or fertility in one flock, where
only a few are kept in a district, is sufficient to account for a much
greater variation than is shown for most breeds in Table Y. At the
same time the difference in fertility of Shropshire flocks kept in
Staffordshire, as compared with other flocks of the same breed kept
elsewhere, and of Lincoln flocks kept in Yorkshire, as compared with
those kept in the home county, is very remarkable. They certainly
suggest that Yorkshire is a more satisfactory habitat for the Lincoln
breed than is the home county, and although it is quite possible the
method of farming and other influences, in Lincolnshire, are responsible
for some of the difference, they can hardly be responsible for all of it.
I do not believe that the high percentage of loss in the Lincolnshire
wold flocks is due altogether to mismanagement, but if 50 per cent, of
the difference between the losses of the wold flocks and the Yorkshire
flocks was added to the percentage of twins of the former, there would
still be a difference of 17 per cent, of twins in favour of the Yorkshire
* See " Percentage of Twins."

On the Fertility of different Breeds of Sheep, §c. 103

flocks. So also with the Shropshire breed in Staffordshire and the
home county, if 50 per cent, of the difference between the losses was
added to the percentage of twins in the home county, they would still
have 12 per cent, less than the Staffordshire flocks; and it is not
probable that a different method of farming in these two neighbouring
districts can account for such difference in fertility.

The returns of the Suffolk flocks in Essex are misleading, inasmuch
as some of the flocks with the highest percentage of lambs do not show
the percentage of twins ; at the same time I am disposed to think that
Essex is not so favourable a county as Suffolk for this breed of sheep.

The difference in the return of lambs for Southdowns in East
Anglia and in the South, is probably due to the more careful records
which the smaller size of the flocks and the method of farming in
East Anglia allows, and I suspect the difference in the percentage of
twins may be similarly accounted for.

In neither of the cases where different breeds are represented in the
same locality, does the percentage of twins of the foreign flocks
approach that of the home breed. Southdowns do not approach the
fertility of Suffolks in East Anglia, nor do Hampshires become as
fertile as Dorset Horns in the West country.

The Suffolks, Hampshires, and Dorset Horns are most fertile in their
home districts ; in the case of the Shropshire breed, Staffordshire may
be considered a part of the home district ; and only in the case of the
Lincolns is it demonstrated that any breed thrives better in a foreign
district than at home. This was to be expected, for it can hardly be
doubted that natural selection, as well as artificial selection, has been
at work on the different breeds of sheep in this country.

In the case of the Lincoln breed it is of interest to note, in this con-
nection, that the method of farming in some parts of Lincolnshire has
greatly altered in modern times ; for instance, the facility which the
soil on the wolds affords to grow especially fine crops of roots, enables
the flock-masters in that district to keep more sheep per acre than is
possible on the low-lying farms ; this leads to crowding and to other
even more artificial conditions which, as the statistics indicate, are not
favourable to the fertility of the breed ; on the other hand several of
the more successful flocks m Yorkshire are run on grass chiefly and
under more natural conditions for sheep.

The variation in fertility in different districts does not, however,
affect the racial character of the fertility of the different breeds (com-
pare Tables IY and Y — " Per cent, of Twins "), except in the case of
the Lincolns, where the returns of the Yorkshire flocks suggest that
this breed should be placed in the first rank with the Suffolks and
Shropshires, instead of in the second rank with Dorset Horns, Oxford
Downs, and Kents. The position of the Southdowns at the bottom of
the list remains unaltered.

104

Mr. W. Heape.

While the percentage of lambs and of twins in cross-bred flocks is
greater than the same percentage for the total pure-bred flocks, the
ewes of certain pure breeds are undoubtedly more fertile than the
average cross-bred ewe.

The flock percentage of lambs in 306 pure-bred flocks, ranges from
203*8 to 59*09 per cent., the percentage for 89,370 ewes being 120-4
per cent. The most frequent percentage in these 306 flocks is between
110 and 120 per cent. ; the following Table (I) shows this, and as the
column for "under 110 per cent." includes all failures, the excess in
the 110 per cent, column is all the more marked.

There are more flocks of between 100 and 200 ewes than of any
other number, and it is in these flocks the highest percentage of lambs
occur ; broadly speaking, the frequency of a high percentage of lambs,
and the height of that percentage, vary in proportion to the number
of flocks and to their size.

Abortion and Barrenness.

There is an element of error in these statistics, due to the fact that
some ewes abort at an early stage of gestation, when the foetus is small
and the circumstance liable to be overlooked by the shepherd ; some of
these ewes come " in use" again and are again served by the ram, but
some do not again come " in use," or, owing to the fact that they have
already been drafted into a flock without a ram, they are not again
covered ; in either of these cases they are put down as barren. Thus,
although I do not believe the error is a great one, the percentage of
abortion in ewes may be higher and the percentage of barrenness lower
than is represented in Table IY.

The total loss from both these causes amounts to 7-1 per cent, for
all pure-bred ewes ; of these the Lincoln sheep suffer the most (12 per
cent.) and the Hampshires the least (4*01 per cent.), a very startling
difference and of specially grave significance to Lincolnshire flock-
masters on the wolds (see Table V).

Abortion.

In 300 pure-bred flocks the percentage of abortion varies from
23-75 per cent, to 0, while the percentage for 85,878 ewes is 2-39.
The Dorset Horn (4*11 per cent.) and the Lincoln (4 per cent.) breeds
surfer most, while all other breeds except Southdowns have less than
2 per cent. The causes which induce a high percentage of abortion
are little understood, and severe losses are from time to time experi-
enced from no knc vvn cause (Lincolns 20 to 30 per cent, and Dorset
Horns).

Statistics supplied indicate, that an undue proportion of shearling

On the Fertility of different Breeds of Sheep, SfC

'I'otals.

rH

00

co

§

<M

o

00

CO XO

rH

8

CO

200 p.c.
and over.

1

rH

190 p.c.

i

CM

CO

180 p.c.

CM

CO

100 ewe

6

HJ<

CM

n

i>

bs — per

160 p.c.

x>

00

rH

s. Lam

150 p.c.

CO

CM

rH

00

rH

CM

Pure-bred Flock

140 p.c.

00

O
rH

o

CM

TjH
CO

130 p.c.

CO

"3

i

CO

co

rH

CM

CM r-

H

rH

Q

—Total

120 p.c.

o

rH

CM

t>

CO

lO

CM

. i—

rH

CO

kffl

Table I.

110 p.c.

i>

J>

CO
rH

cm

Ci

CO

rH

r-

CO

co

Under
110 p.c.

r—

i

Ci

CO
rH

Ci

\

CO

CO CO

CO
lO

Flocks of

c
c

«

5

100—199

200—299

a

cr.

CO

I

CO

a
cr.

4

500—599

600—699

Ci
a

800—899

900—999

Total I

106

Mr. W. Heape.

ewes in a flock is associated with a high percentage of abortion
(Lin coins), and that ewes of a particular breed run on certain subsoils
or in certain districts are more liable to abortion than the average for
that breed, as, for instance, Lincolns run on the wolds and Hampshires
on oolite formation.

In some parts of Lincolnshire abortion sometimes approaches, if it
does not actually assume, an epidemic form; at such times several
neighbouring flocks may experience between 30 and 40 per cent, of
aborted ewes ; I am unaware of a similar form of abortion in any other
district.

Unsuitable food, causing indigestion and intestinal irritation, and
poor food, resulting in poor nutrition, are probably responsible for the
greatest proportion of abortion in ewes. It is not the kind of food, as
is frequently supposed, but the condition of that food which is at fault,
and, as a result, my schedules show that poor condition of ewes during
gestation is undoubtedly associated with a relatively high percentage
of abortion.

The highly artificial conditions under which sheep are kept in many
districts in this country, renders the question of the most suitable food
for breeding ewes a very important question, and it is one regarding
which but little attention has been paid.

The most frequent percentage of abortion experienced in 300 pure-
bred flocks is shown in Table II to be under 1 per cent. The highest
percentage is relatively more frequent in large than in small flocks, and
this table shows that much more irregularity is experienced in abortion
than Table I shows is the case for fertility.

Barrenness.

In 327 flocks of pure-bred ewes the percentage of barrenness varies
from 51-42 per cent, to 0, while the percentage for 96,520 ewes is 4*71.
The Lincoln (8 per cent.) and the Shropshire (6*06 per cent.) breeds
suffer the most, while, with the exception of Hampshire Downs, Dorset
Horns, and Suffolks, no pure breeds record less than 5 per cent, loss
from this cause.

The district or subsoil on which ewes are run is associated, in certain
breeds, with the proportion of barrenness experienced : thus, Lincoln
sheep run on the wolds, Shropshire sheep on new red sandstone subsoil,
and Hampshire sheep elsewhere than on chalk downs, are associated
statistically with a relatively high percentage of barrenness ; an exces-
sive proportion of shearling ewes in a flock is also frequently found
associated with a high percentage of barrenness \ but the quality of
the food given and the condition of the rams and ewes at tupping time
is no doubt the chief factor which influences the barrenness percentage.

Fat is well described as an enemy to fruitfulness, while the want of

On the Fertility of different Breeds of Sheep, fyc. 1

Total.

lO 00
^ 00

Hjl c
CO t

CO rH lO HH
rH

rH
rH

s

o

3 p. c.
id over.

r-i

CQ rH 03

rH

rH
rH

r-1 ,*

04

PH

Oi OS
C5 CI
00 OS

108

Mr. W. Heape.

OS

rH co o

rH | <M
CO

eo rH co eg

CO I

rH W3 <N <N

CO <N

CO O ^ !M

o

05

i>

rH

CO

CO

<M

rH

i>

co

10

CO CO ID <M 'M

p.c.

1 VD

00

00

CO

rH

CO

M

CO

6

1 CO

CO
rH

■H

rH

rH

rH

.

CO

rH CO X> (M rH

CO CO CD OS CO rH

OS

OS

OS

C2

OS

OS

OS

OS

cs

B

©s

OS

OS

os

OS

(M

co

1

1(5

CO

JO

00

1

OS

1

1

O

o

1

O

1

1

o

1

o

o

o

o

o

o

8

8

o

!N

00

LO

s

as

On the Fertility of different Breeds of Sheep, §c.

109

•uajjeq pire
paj.IOql? ssoi i^ox

CM © -* — i O — i © OD
tO CR CR O Tfi O >0 CR

to t-» -jf © t— t~ CM CO

S £ i
- * 1

©
to

Barrenness.

•n9.T.ix?q
•juao J9j

CO CM 00 >C tO tO 00
CM WJ O — O CR © WE

4< co |

00

■saM9 na.i
-jreq jo -O.M

iO CO

■*"

ic"

•S9AV9 JO "0^

coeocoo>?©ooi^2 I

CT. rH r~ iff 30 »« CO 1

b» ©*jO? a w a" N o t«* |

CM § 1

ws o
to" 00*

CT. rH

©

oo

W3 |

•sjpou jo -ojs;

SDiOCOr-IO OSCOOip
CO i-H . CM iO CM CM iC © CO

Ol O

co

•*

CO

CO
CM

<N

Abortion.

•pg'jjoq'B

•{^190 .19 J

■* TO «5 © -* — O

CO CO 00 »fl CO rH -*

rH— .CMrHi-|-rlr-(-*l —

CR CM
CO lO

CM rH

•S9.W9 p9J

-aoq« jo •ojij

C-l <* <C rH tO r-i <M OC CO
CR ITS l<3 r~ CO 1— — < t© CR

cm co co i-i ee

CM CO
iO fc-

CM

CM
CM

cm"

•S9A9 JO '0^[

rH i-H ~H 'O 00 O SO 1^ SO

^©d'OcOOICMOCO
Ct.' OS CO l~ tO © -r © to

to co co" co* of of t~ to" to"

CM -H

00 rH

t- to

CO

WJ —"
00 rH

CO
CM

CR

•sjpou' JO "OM

CO R * H R "18 IO M

© ©

o m

CO

©

I %
1 £

EH

•^1190 J9J

8 (8 N a "«i ift. ft *
cm co tp © © o ri., © ©

CM -h oo ,o r— © CR CO
irieOrHCMCOCO-rXCMTl

CM

© ©

co co

C^

Suuvaq

S9.VV9 JO 'OX;

1,490
2,104
1,229
5,093
Ml
3,225
1,932
3,326
1,268

K CR
•O 00

3

to

rfi

co

•S9.W9 JO 'O^J

;C©co-*OOOCMCOt-(

CM* to" tD rH* N » 1< -J ^<"

oo" of

!D

©
wo

00

i—

'^^TDOrr to *o vt

© rH 00 -H -f (— tO tO lO
Hr^HilrtNnfC-l

CO CO
CM

iO

ol

Lambs.

•S9.V\9 00 T
J9d SqUXBrJ

t- W5 CR OS © CO CR ©

1 - O 00 tO rH -O 1 - —

r-l — CR CR -0 tO rH tO
T1<CM© — — CMOOrHOI

© 1-

O CR

CO
CM

•SqUU}[ JO *o^[

IQ rH cr. M O "M -ri M 00
tO CM © rH © CR © iO
rH 1Q © lO OP © © If} CO
© O* oo" no* CO ©" r-H* 1- tC
rH rH Ol rH rH —

CO rH
© Mt

r-T ir:"

© rH

-3<

CO

CM

•S9AV9 JO 'O^

7,170
8,481
7,834

21,860
3,189
8,163
8,014

15,789
5,840

of cf

CO rH

>o

co

I

-inboBp-ok S S a g 'S S 8 ■ J5- 'tS

tO CM

© >o

CO

CO
CO

*S9A\9 JO "0^1

tOrH-fOiOirjCM©©

© co co o mi ■< cr. co r-i

iff OS r- -r* IO CM "* TO ©
O* CR C31 tO CO ©" CO* I- ©"

CM r-< rH rH

CO © i CO

2 S8 | 5

co* of 1 cm"

O rH CM
rH " r-

•SUTEJ JO "0^

rH -rJ ©CO CO O CO «-iO
tOlOOOl^-OOt^CRtOCR"
rH CM rH rH rH CO i-H

W5 W5

CO rH
© TH

cm"

©
o

iO
CM

•S5[0OU JO -OM

COIOCOCOOrH©(M-o
COrHCMOCMCO©tOeO

CO CR

OS
CO

Breed.

"3 4—

as p

S a
m t*

rH ex

3. Southdown ...

4. Hampshire ...

5. Oxford Down

6. Dorset Horn...

7. Shropshire ...

9 to 18. Various
pure breeds

I ®

rO £

£ to
S to

3. 2
3 °

o

Eh cr

1

rC

0)
43

"3

o
EH

110 On the Fertility of different Breeds of Sheep, Sfc.

-L3 ©

s u

bO o

^ tn '£

2 ^
O*o
O 1—1 ,

"Th S)_i 4i

. o c
«M 35

a © £
o

S » S 2

© O -ts § 2

tp cq
^ to

c3 35

si

I "tfl CO CO (M I © CO lO

1*1

c3

©

§ O £

CO 00
' OS O

ao to
co"i-T

00 CD I ©
CO CO j

t» OS

10

co x>

OS O

i>

r-i tO

CO

CO

rj< ^

to CO

CO CO

CO

co i>

to 1— 1

CO

00 CO

to 0

CD

rH O

CO

10 cq

tH CD

OS 05

I 00 CO

CO 0

"if- CO |

rH

10

O rH

oq co

3

to CO j

O

rH CXI

to —1 © r>. 00

co oq 00 cs cs

tt* CD 'ID ' 01

»o OS CO
CD CO

« O 1 GO

r-T-T I co"

rH OS

os caj

O CD

« 02 ,

3£

I

OS OS I CO

© r^H

. e.sp

tg 0

© ©

£r3

53 oq W

©

^3 ©

8.2

m o
© £

o .2

©

© no
a 5?

S3 .«

I §

O 2

la

O o

S3 o

e3 0
O
© 33

On Bed-water or Texas Fever.

Ill

sufficiently nutritious food also results in poor returns of lambs.
Among Suffolk and Shropshire ewes, which are highly fed as a rule, a
high percentage of barrenness occurs in cases where they are exces-
sively highly fed ; on the other hand, among Dorset Horn, Lincoln,
and Kent ewes, which are certainly not too highly fed as a rule at tup-
ping time, the highest percentage of barrenness occurs among the
poorest kept flocks.

Jhe most frequent percentage of barrenness experienced in 327
flocks is 1 to 2 per cent., but, as Table III shows, the returns are
much more irregular than was the case for abortion, and there is a
much larger proportion of flocks in the "10 per cent, and over "
column. The slightly excessive proportion of flocks in the 5 per cent,
column suggests generalised results rather than accurate returns, but
the number of flock-masters who are responsible for this is obviously
very small.

In conclusion : —

1. AVhereas the -total loss from abortion and barrenness is fairly con-
stant for most pure breeds of sheep, the SufFolks and Hampshires are
markedly free from, and the Lincolns markedly liable to, heavy loss
from these causes.

2. Although the loss from the above causes does exert an influence
on the returns of fertility of the various breeds, it does not account for
the wide variation which exists in this respect.

3. The ewes of certain pure breeds are conspicuously more fertile
than the average cross-bred ewe ; and

4. The fertility of certain pure breeds is sufficiently marked to con-
stitute a racial characteristic.

" Some further Remarks on Red-water or Texas Fever." By
Alexander Edington, M.B., F.R.S.E., Director of the Bacterio-
logical Institute, Cape Colony. Communicated by Dr. D.
Gill, C.B., F.R.S. Received March 13 —Read April 20, 1899.

Since my communication* to the Royal Society of London, by
Professor Thomas R. Frazer, I have been able to obtain valuable addi-

* The conclusions arrived at in that communication (received June 6, 1898)
were as follows : —

1. The blood of animals, themselves healthy, from a red-water area is dangerous
if inoculated into an animal which suffers coincidentally from another disease.

2. That the blood of animals suffering from mild or modified red- water may be
safely used to inoculate a healthy animal subcutaneously, but is dangerous when
injected into a vein.

3. That the subcutaneous inoculation of mild or modified red-water blood
conveys a mild form of the disease, and since the blood of such an animal is viru-

VOL. LXV. I

112

Mr. A. Edington.

tional evidence as to the communicability of the disease by the use of
blood derived from animals which have been either recovered from the
sickness for a very considerable time or which have been inoculated
many months previously to the date on which their blood has been used.

On the 8th December, 1898, I withdrew some blood from animal
No. 18, which has been continuously under observation since it was
inoculated on the 22nd December, 1897. After defibrinating the
blood, 20 c.c. was used to inoculate a young ox (No. 54) by intravenous
injection. On the following day a sharp rise of temperature occurred,
which reached to 106-6 F. On the following morning it was observed to
have fallen to 99*8° F. Three days later the temperature was again over
104° F., but fell previous to the next morning. From this time onward
an erratic course of temperature was observed, and on the twenty-fifth
day, subsequent to inoculation, it was seen to be ill, refused food, but had
no definite symptoms of " red-water." Three days later it died. The
blood on examination was seen to contain the spherical forms of the
parasite.

On post-mortem examination, the bladder and urine were quite
normal. The liver was not enlarged, but was somewhat discoloured
in patches, and the biliary ducts were distended with bile. The bile
was much altered, being stringy and of a greenish-yellow colour. The
spleen was normal in size and consistence. The kidneys were enlarged
and the pelves were filled up by a yellowish gelatinous exudation.
The cortex was somewhat congested, but there was no evidence of any
true inflammatory change. The general muscles were pale in colour,
and there was slight evidence of jaundice. This experiment serves to
show that an animal which has been inoculated with infected blood,
while it may not develop much illness as a result of it, is really infected
and, moreover, its blood, if drawn as late as a year subsequently, is
yet so infective that an intravenous injection of it, into susceptible

lent when injected into a vein in another animal, it is safely to be inferred that
the animal suffering from the mild form becomes more or less immunised or
" salted."

On these grounds I would suggest a method of protective inoculation against red-
water in the following manner. Haying procured a healthy animal from a red-
water area, or one which is known to have been " salted," inoculate it by injecting
5 c.c. of red-water blood into the jugular vein and 5 c.c. subcutaneously. In cases
where the operator is unable to attempt the vein inoculation I would recommend
the subcutaneous inoculation of 5 c.c. in four different sites.

Allow at least twenty-eight days to elapse, and if any degree of illness is recog-
nised, the blood of this animal may be used, after being defibrinated, to inoculate
healthy cattle. For such inoculation only 5 c.c. should be injected into small
animals and not more than 10 c.c. into larger.

Seeing, however, that the presence of other maladies renders such a proceeding
unsafe, I would recommend that it should only be practised during the autumn or
winter, when the veld diseases are, as a rule, in abeyance, and in no case when any
epidemic disease is in the near neighbourhood.

On Red-ivater or Texas Fever.

113

animals, will certainly infect, and may even kill, although after a some-
what extended period of time.

Very important corroboration of this is furnished by the experience
of inoculation for red-water, which has lately been adopted in the Cape
Colony. Four animals which were immune to red-water (three by reason
of having had the disease and recovered, and one by being born and
reared on permanently infected veld) were sent from Fort Beaufort to
Queenstown to be used by the veterinary surgeon there for inoculation
purposes. The animals to be used for inoculation had been " fortified,"
i.e., re-inoculated with virulent blood, seven weeks previously.

Twenty animals were inoculated with defibrinated blood from one
animal, the doses used being 10 to 20 c.c, according to age. All had
a febrile reaction and some slight symptoms of the sickness, but easily
recovered. From one of the other of the four animals blood was taken
and used to inoculate seven head, giving closes of 10 to 15 c.c. These
also all had a reaction, but made good recovery.

On November 1st the four animals were re-inoculated with virulent
red-water blood, and in each case 5 c.c. was injected intravenously and
10 c.c. subcutaneously. Twenty-nine days later they were bled. With
this blood two lots of cattle were inoculated.

One lot consisted of 107 animals which had not ever been exposed to
red-water infection. The doses used were increased beyond those
which I had recommended, namely, 10 to 25 c.c. were used, according
to age. Of these animals no less than seventeen died of characteristic
red-water. The remainder made a good recovery.

The second lot consisted of fifty-three head of cattle, all of which,
with one exception (an imported animal) had been born and reared on
red-water veld. The imported animal was the only one which showed
any signs of reaction, but it made a good recovery.

This experience has sufficed to show that it is not always safe to
exceed the doses which I have recommended, unless the animals which
have been used for withdrawing blood have been untouched for at
least a considerable number of months.

I have been able, with the co-operation of several farmers, to carry
out experiments by which inoculated cattle have been fully exposed to
infection at later dates. In May, 1898, I inoculated ten head of old
cattle with blood from an animal which had been inoculated, six months
previously, with virulent blood. These cattle were immediately removed
from the Institute, and later sent to an infected area in company with
ten head of young animals which were uninoculated, but, as is com-
monly known in this colony, are not so liable to death from this dis-
ease as are older animals. Of the young stock all have been infected
by exposure in the veld, and , three have died. Of the older, more
susceptible, animals not one has shown the slightest signs of illness,
and the cows have given birth to healthy calves.

VOL. LXV. . K

114

Proceedings and List of Papers read.

Mr. J. H. Webber had twenty-eight head of Fish River cattle inocu-
lated on the 7th November, 1898, and subsequently had them removed
to his farm, which is well known to be one of the worst infested
areas in the eastern province. Previous experience has shown that if
clean cattle are placed there they become very quickly affected with
the disease. On the 5th December one died from gall-sickness, but,
with this exception, all have done very well, and are at this date
in perfect health.

This method of inoculation has proved so satisfactory to the
farmers themselves that it is being very generally adopted, and the
farmers have petitioned the Government to arrange for an inoculating
station being placed at Graham's Town, so that clean cattle coming
from clean Karroo areas for transmission to the coast may be inocu-
lated previous to entering the infested belt.

April 27, 1899.

The LOED LISTER, F.R.C.S., D.C.L., President, in the Chair.

A List of the Presents received was laid on the table, and thanks
ordered for them.

The following Papers were read : —

I. " Data for the Problem of Evolution in Man. I. A First Study of
the Variability and Correlation of the Hand." By Miss M. A.
Whiteley, B.Sc, and Karl Pearson, F.R.S.

II. " On the Luminosity of the Rare Earths when heated in Vacuo by
means of Cathode Rays." By A. A. Campbell Swinton.
Communicated by Lord Kelvin, F.R.S.

III. " On a Quartz-thread Gravity Balance." By Richard Threlfall
and J. A. Pollock. Communicated by Professor J. J.
Thomson, F.R.S.

IY. "On the Electrical Conductivity of Flames containing Salt
Vapours." By Harold A. Wilson, B.Sc. Communicated by
Professor J. J. Thomson, F.R.S.

V. "On a Self -recovering Coherer and the Study of the Cohering
Action of different Metals." By Professor Jagadis Chunder
Bose, M.A., D.Sc. Communicated by Lord Rayleigh, F.R.S.

VI. " On the Presence of Oxygen in the Atmospheres of certain Fixed
Stars." By David Gill, C.B., F.R.S.

Lu minosity of Bare Earths when heated by Cathode Bays. 115

" On the Luminosity of the Eare Earths when heated in Vacuo by
means of Cathode Bays." By A. A. Campbell Swinton.
Communicated by Lord Kelvin, RE.S. Eeceived March 20,
— Eead April 27, 1899.

For incandescent gas mantles, it is found that certain definite mix-
tures of the rare earths are necessary, in order to obtain the maximum
luminosity. For instance, in the ordinary Bunsen gas flame, a mantle
consisting of pure thorium oxide, or of pure cerium oxide, will only
give about one-eleventh of the light that is given by a mantle com-
posed of 99 per cent, of thorium oxide, and 1 per cent, of cerium
oxide, which is the mixture at present used by the Welsbach Com-
pany.

In order to explain this remarkable fact, several different and some-
what contradictory theories have been propounded, one of which implies
catalytic or other - chemical action between the oxides and the consti-
tuents of the Bunsen flame.

In order to investigate this question, it is obviously important to
note the behaviour of the rare earths at high temperatures without
contact with any flame, and endeavours have already been made to
effect this by heating the oxides in specially constructed furnaces.
Under these conditions, only very minute differences could be detected
in the amount of light given by different oxides and mixtures, but it
appears doubtful whether the very high temperature of the Bunsen
flame was really attained.

It has occurred to the writer, that very high incandescence could be
produced by enclosing mantles in a vacuum tube, and subjecting them
to bombardment by means of cathode rays, when the mantles would
not be in contact with anything except the cathode rays themselves,
and the comparatively small amount of residual gas that remains in
the tube at the requisite high degree of exhaustion.

Since the date of Sir William Crookes's early researches, it has been
known that a very high temperature could be produced in a body
placed at the focus of the convergent rays from a concave cathode. In
this manner Crookes melted platinum and glass, and brought carbon
wool to bright incandescence. The writer has made many experi-
ments on this subject, using instead of the interrupted continuous
currents employed by previous investigators, alternating electric-
currents, which appear to. have many advantages for this purpose.
At the Eoyal Institution, in February, 1898, the writer showed that
very brilliant incandescence could be obtained for a short time in a
small block of lime, placed in a suitably exhausted tube midway
between two concave electrodes, connected to an alternating electric-
supply at about 6000 volts pressure, and in June, 1898, at the Boyal

K 2

116 Mr. A. A. Campbell Swinton. On the Luminosity

Society, he exhibited in action a similar arrangement, but with the
block of lime replaced by a flat plate of thorium oxide. In this case
the concave electrodes were of such a curvature and were placed so far
apart, that the two sides of the thorium plate in each case intersected
the diverging cone of cathode rays. Under these conditions nearly a
square inch of the thoria surface on each side of the plate became
highly incandescent, and a very powerful light was obtained for some
minutes at a time, but only at a critical and highly unstable degree of
vacuum.

The writer has now applied this method to the investigation of the
comparative luminosity of different mixtures of the rare earths.

One form of the tube employed was constructed as shown in the above
figure, where C, C are two spherically concave discs of aluminium 1-125
inches diameter, and 6 inches radius of curvature. These electrodes 4
are placed about 7 inches apart, and were connected to the secondary
terminals of a 10-inch Ruhmkorff coil, the primary of which was
supplied through a variable resistance, with alternating electric current
at 100 volts pressure from the main. The tube was connected
through a drying tube, containing phosphorus pentoxide, to a pair of
Toepler pumps, and also to a McLeod gauge.

The mantle M M to be experimented upon, was mounted on a
platinum wire frame and placed between the two electrodes, so that
as the electric current alternated, and each electrode became in turn
cathode, the mantb was subjected on alternate sides to cathode ray
bombardment. The curvature of the electrodes was such as to give
almost parallel beams of cathode rays, so that a considerable ring

of the Bare Earths when heated in Vacuo by Cathode Rays. 117

shaped, and slightly hollow, area on each side of the mantle was
subjected to the rays, and could be brought to high incandescence.

A preliminary experiment was made with a mantle of asbestos,
powdered over in patches with pure thorium oxide. With this it was
found that at a suitable degree of exhaustion, the patches of thoria
became brilliantly incandescent, with an intensity of cathode rays that
made the asbestos barely red hot.

Experiments were next made with mantles consisting entirely of
thoria and ceria, both separately, and mixed in different proportions.
These mantles were prepared in a similar manner to the Welsbach
incandescent gas mantles, by saturating a carefully purified cotton
fabric with ammonium nitrate of thorium and cerium, and then
binning out the cotton. Very thick and closely woven cotton lamp
wick, freed from foreign matter by treatment with caustic soda,
hydrochloric acid, and ammonia, was employed in place of the thin
fabric usually used, so that the resulting mantle after burning out the
cotton, was very close in texture, and fully 0*2 inch thick. This was
found necessary, as otherwise some of the cathode rays passed through
the mantle and melted the opposite aluminium electrode.

In order to obtain accurate comparisons between pure oxides and
different mixtures, the mantles were made in patchwork, each complete
mantle being made up of two or four sections, separately impregnated
with different solutions, and then sewn together with impregnated
cotton before being burnt.

The mantles were so mounted in the vacuum tube that the cathode
rays impinged equally upon the portions that consisted of different
mixtures, so that an equal amount of energy was imparted to each
sample.

With a compound mantle prepared in this way, composed one-half
of pure thorium oxide, and the other half of a mixture of 99 per cent,
thorium oxide plus 1 per cent, of cerium oxide, it was found after
exhaustion that on starting the cathode discharge the thoria plus ceria
heated up to incandescence more rapidly, and, on stopping the dis-
charge, cooled more rapidly than the pure thoria. Further, when at
full incandescence and observed through a dark glass, the thoria plus
ceria was slightly more luminous than the pure thoria, though the
difference was very small, probably not more than 5 per cent. Owing
to the difficulty of maintaining a constant vacuum, accurate photo-
metrical measurements were not possible, but the amount of light
under favourable conditions was roughly estimated at, at least, 150-
candle power per square inch of incandescent surface, this being
obtained with an expenditure of electrical energy in the secondary
circuit at about 8,000 volts pressure of approximately 1 watt per
candle. The amount of exhaustion suited to give the best results
varied with the dimensions of the tube and the conditions mentioned

118 Mr. A. A. Campbell Swinton. On the Luminosity

below, but was approximately about 0*00005 atmosphere, the maxi-
mum luminosity being obtained when the dark spaces of the two
cathodes just crossed at the centre of the bulb. Owing to the large
amount of gas occluded by the mantle, a proper degree of permanent
exhaustion was very difficult to arrive at, and required continuous
pumping for many hours with the cathode rays turned on at intervals.
Even then the conditions of maximum luminosity were exceedingly
unstable, owing to the further liberation of occluded gas on the one
hand, and on the other to the rapid increase in the degree of exhaus-
tion owing to absorption of the residual gas by the electrodes. That
such absorption probably took place in the aluminium electrodes, and
not in the mantle, was demonstrated by other experiments with a tube
in which there was no mantle, but only two electrodes of aluminium
wire.

After the cathode rays had been allowed to bombard the mantle for
a short time, the latter was found to have become discoloured where
bombarded. That portion which was composed of pure thoria became
dark blue, while the thoria plus ceria became brown. This effect,
which appears to be analogous to those observed by Goldstein with
lithium chloride and sodium chloride,* seems to be due to a partial
reduction of the oxides by the cathode rays. The admission of a very
minute quantity of air to the tube while the cathode rays are acting on
the mantle, and the latter is in parts incandescent, causes the dis-
coloration to disappear instantaneously on the incandescent, but not
upon the cool portions, probably by re-oxidation of the partially
reduced oxides, while the discoloration also slowly vanishes in a day
or two with the mantle cold if air at ordinary atmospheric pressure is
admitted to the tube. By continuing to bombard the mantle with
cathode rays, and alternately allowing the vacuum to increase and
letting in small quantities of air, the discoloration can be made to
appear and to disappear over and over again as often as desired. At
the moment of admitting the air, the amount of light was found
momentarily to increase, this being probably due to the increased
temperature due to the re-oxidation of the partially reduced oxides.

After repeating this process of letting in small quantities of air, and
allowing them to be absorbed, several times, it was found that the
degree of exhaustion which gave the maximum incandescence had
altered from 0-000047 to 0*000112 atmosphere, as measured by the
McLeod gauge. Similar effects were obtained with a tube containing
no mantle, but only aluminium wire electrodes, the inference being
that some change takes place in the residual gas which renders it less
conducting.

At a higher degree of exhaustion than that which produced incan-
descence of the mantle, the pure thoria was found to fluoresce blue,
* ' Wied. Ann.,' 1895, No. 54, p. 371.

of the Bare Earths when heated in Vacuo by Cathode Bays. 119

and the thoria plus ceria with a yellowish light. The fluorescence in
each case was much less bright when the oxides were white than when
they had become discoloured by previous bombardment. With very
higlj exhaustions the thoria plus ceria fluoresced the more brightly ; at
lower exhaustions the pure thoria gave the brighter fluorescence.

On the suggestion of Mr. W. Mackean, the tube was pumped up to
a very high vacuum and oxygen admitted. A similar experiment was
made with hydrogen, the tube being completely filled with the gas,
and then pumped to the proper degree of exhaustion. Though at low
exhaustions these gases gave distinctive appearances to the discharge
in the tube, no difference in the behaviour of the mantles with them
and with air . could be detected when once the vacuum reached the
degree required for producing incandescence of the mantle.

Further experiments were made with a similar tube containing a
compound mantle made up of four sections, composed as follows : —
(1) pure ceria, (2) pure thoria, (3) 50 per cent, thoria 50 per cent, ceria,
(4) 99 per cent.. thoria 1 per cent, ceria.

With an intensity of cathode rays that gave a brilliant light with
Nos. 2 and 4, Nos. 1 and 3 were found to give practically no light,
becoming barely red hot ; while, as before, No. 4 was found to give
slightly more light than No. 2, and to heat up more rapidly and cool
more rapidly than the latter.

These experiments show that thoria and ceria, both alone and mixed,
behave quite differently when heated by cathode ray bombardment
than when heated in a Bunsen flame. In the latter, 99 per cent,
thoria plus 1 per cent, ceria gives many times as much light as pure
thoria alone, while, when incandesced by cathode rays of equal inten-
sity, the difference, though in a similar direction, is exceedingly small.
Again, in the flame pure ceria gives just about the same amount of
light as pure thoria, while with a given intensity of cathode ray bom-
bardment thoria gives a brilliant light, while ceria gives practically
none.

In arriving at any finally satisfactory theory of the luminescent pro-
perties of the rare earths, these results with cathode rays, which differ
materially from those obtained by other methods of heating, will
require to be taken into account.

I am indebted to the courtesy of the Welsbach Incandescent Gas
Light Company for the samples of the rare earths with which the above
investigations were made ; also to the assistance of Mr. J. C. M.
Stanton and Mr. H. Tyson Wolff in carrying out the experiments.

120

Mr. H. A. Wilson. On the Electrical

" On the Electrical Conductivity of Flames containing Salt
Vapours." By Harold A. Wilson, B.Sc. (Lond. and Vic),
1851 Exhibition Scholar. Communicated by Professor J. J.
Thomson, F.K.S. Eeceived March 10— Bead April 27, 1899.

(Abstract.)

The experiments described in this paper were undertaken with the
object of following up the analogy between the conductivity of salt
vapours and that of Bontgenised gases, and especially of getting some
information about the velocities of the ions in the flame itself.

They are to some extent a continuation of the research of which
an abstract has already been published in the 'Proceedings of the
Boyal Society.'*

The paper is divided into the following sections : —

(1) Description of the apparatus for producing the flame.

(2) The relation between the current and E.M.F. in the flame.

(3) The fall of potential between the electrodes.

(4) The ionisation of the salt vapour.

(5) The relative velocities of the ions in the flame.

(6) The relative velocities of the ions in hot air.

(7) Conclusion.

The apparatus used for producing the flame was similar in principle
to that used in the investigation referred to above. Carefully regu-
lated supplies of coal gas and air were mixed together along with spray
of a salt solution, and the mixture burnt from a brass tube 0*7 cm. in
diameter.

The flame thus obtained was steady, and measurements of its con-
ductivity, when a particular salt solution was sprayed, did not differ
more than 1 or 2 per cent, on different days. The height of the inner
sharply defined green cone was 1*5 cm., and that of the outer cone
7-5 cm.

The current between two gauzes of platinum wire, each 14 cm. in
diameter, and placed horizontally one above the other in the flame, was
measured for E.M.F.'s up to 800 volts, and with various distances
between the gauzes.

The current with a large E.M.F. was found to be independent of the
distance between the electrodes when the upper electrode or gauze
was positively charged, provided that the distance between the elec-
trodes was not so great that the upper one was in the cooler parts of
the flame near its point. When the upper gauze was comparatively

* " The Electrical Conductivity and Luminosity of Flames containing Vaporised
Salts," by Arthur Smithells, H. M. Dawson, and H. A. Wilson, * Eoy. Soc. Proc.,'
toI. 64, p. 142.

Conductivity of Flames containing Salt Vapours. 121

cool the current was much smaller, but if the upper gauze was kept
hot by passing a current through it, then the current with a large
E.M.F. was independent of the distance between the electrodes, even
when the upper electrode was above the point of the flame.

If both of the electrodes were hot, then the current, as the E.M.F.
was increased, attained a nearly constant value. Cooling the positive
electrode by raising it in the flame caused the current to increase
towards this saturation value much more slowly than before, while
cooling the negative electrode, the positive one being hot, caused the
current to show no sign of arriving at a maximum value. The current
was much greater when the negative electrode was hot, and the positive
electrode cool, than when the negative electrode was cool, and the
positive one hot.

The fall of potential in the flame between the gauzes was examined
by putting in an insulated platinum wire, and finding the potential it
took up. When both the electrodes were hot, the fall of potential
closely resembled that observed in gases at low pressures- That is to say,
near each electrode there was a comparatively sudden fall of potential
much greater near the negative electrode than near the positive, with
a small and nearly uniform gradient in between. If either of the
electrodes was cooled, then the fall of potential near that electrode
became much greater, and often was nearly equal to the total drop of
potential between the electrodes. This effect was usually much more
marked in the case of cooling the negative electrode than with the
positive electrode.

If the positive electrode was uppermost and somewhat cool, then
with small E.M.F.'s practically all the potential fall occurred near to
the positive electrode; but if the E.M.F. was sufficiently increased,
then a drop of potential appeared at the negative electrode, and with
a still greater E.M.F. this became greater than that at the positive
electrode, as it is in gases at low pressures.

Some of the results obtained pointed to the conclusion that nearly
all the ionisation of the salt vapour takes place at the surfaces of the
glowing electrodes, and not throughout the volume of the flame. A
variety of experiments were tried to test this view, all of which con-
firmed its correctness. Thus, with two platinum foil electrodes opposite
one another in the flame, no increase in the current between them
occurred when a bead of salt was put between them, so that the salt
vapour from it passed between them without touching either electrode.
If the vapour came in contact with the negative electrode, there was
a great increase in the current, and a considerable but smaller increase
when it came in contact with the positive electrode.

The relative velocities of the ions of alkali metal salts in the flame
were estimated by finding the potential gradient necessary to make the
ions travel down the flame against the upward current of gases..

122 Electrical Conductivity of Flames containing Salt Vapours.

This was done by putting a bead of salt between the two gauze
electrodes, and finding what E.M.F. was necessary to produce an
increase in the current between the electrodes when the bead was
put in.

The potential gradient corresponding to this least E.M.F. was then
determined. In this way it was found that the positive ions of salts
of Li, Na, K, Eb, and Cs, all have nearly the same velocity in the
flame, whilst the negative ions of various salts of these metals also
have equal velocities which are about seventeen times as great as the
velocities of the positive ions.

The velocity of the positive ions was estimated to be about 60 cm.
per second for one volt a cm., and that of the negative ions was about
1000 cm. per second.

The relative velocities of the ions of various salts was also deter-
mined in a current of air at about 1000° C, which was obtained by
passing the air through a platinum tube 1*3 cm. in diameter, and 50
cm. long, heated in a gas-tube furnace. The method used was exactly
analogous to that used in the flame. The ions could be divided into
three classes, in each of which all the ions had equal velocities, viz. : —

Telocity.

1. Negative ions of salts of Li, Na, K, Kb, Cs, Ca, Sr,

and Ba 26'0 cm.-sec.

2. Positive ions of salts of Li, Na, K, Rb, and Cs 7*2 „

3. Positive ions of salts of Ca, Sr, and Ba 3-8 „

It thus appears that those ions which in solutions carry equal
charges have equal velocities in the gaseous state. This points to the
conclusion that the velocity of a gaseous ion in a given medium de-
pends only on its charge. The velocities are less than those calculated
for ions consisting of one atom, so that each ion appears to be a cluster
of atoms. If we regard this cluster as held together by the charge on
it, then it is reasonable to suppose that the size of the cluster will be
determined by the charge. Hence those ions having equal charges
will be of equal sizes, and consequently of equal masses, since the
atoms forming the cluster probably come from the medium rather
than from the small quantity of salt present. Consequently they all
have the same velocity under similar conditions.

The two main results arrived at in this paper, viz., that the ionisa-
tion of the salt vapour in the flame takes place only at the surfaces of
the glowing electrodes, and that the velocity of the negative ions in
the flame is very much greater than the corresponding velocity of the
positive ions, enable the phenomena of unipolar conduction to be very
easily explained. For example, if one electrode is much hotter than
the other, then if the hot electrode is negative, it will give off negative
ions very freely, and there will be a large current; but if the hot

On a Quartz-thread Gravity Balance.

123

electrode is positive, then the small velocity of the positive ions is not
favourable to their being dragged away from the electrode before they
can recombine, so that the current is very small unless a very great
E.M.F. is applied.

* On a Quartz-thread Gravity Balance." By Kichard Threlfall,
lately Professor of Physics in the University of Sydney, and
James Arthur Pollock, lately Demonstrator of Physics in
the University of Sydney. Communicated by Professor J. J.
Thomson, F.K.S. Eeceived April 11— Eead April 27, 1899.

(Abstract.)

The first part of the paper contains an account of the instrument in
its present form, an account of the investigations leading up to the
form adopted being relegated to an appendix.

The principle of construction is as follows : — A quartz thread
(which requires to be prepared with much care) is stretched horizon-
tally between two supports, to which it is soldered. At one end the
point of attachment is the centre of a spring of peculiar construction,
designed so as to be capable of displacement in the direction of the
thread, but incapable of any transverse motion or vibration.

At the other end the thread is attached to the axle of a vernier arm
moving over a sextant arc ; by turning the axle the thread may be
more or less twisted, the amount of twist being ascertained in terms of
the divisions of the sextant arc.

Midway between the two supports the thread is soldered to a short
length of fine brass wire, which is adjusted so that the centre of
gravity of the wire does not lie immediately above or below the thread,
but at some distance from it. The wire forming the " lever " is then
rotated about the thread as axis in such a manner that the two halves
of the thread are twisted through about three whole turns, and the
torsion of the thread is then of such a value that the lever assumes a
horizontal position. This adjustment is made by weighting the lever
with a small speck of fusible metal. The " balance," which determines
the position of the lever with respect to the horizontal plane through
the thread, is composed of the earth's gravitational force on the one
hand, and the forces of resilience of the twisted thread on the other.
Were gravitational force to increase, the centre of gravity of the lever
would fall, the end of the lever would move out of its sighted posi-
tion, and the thread would have to be slightly twisted by the vernier
axle in order to bring the lever back to its sighted position.

Differences in the gravitational intensity at different stations are
expressed in terms of the amount by which one end of the thread

124

Messrs. E. Threlfall and J. A. Pollock.

has to be twisted or untwisted to bring the lever to its sighted
position.

In carrying out the construction of the instrument on the principle
thus explained the following conditions have to be fulfilled : —

The instrument must be portable, and must be able to withstand
the rough usage inseparable from travelling, without being put out of
adjustment. It must have a sensitiveness of at least one part in
100,000 of the value of " g," i.e., a change in the value of " g "
amounting to one part in 100,000 must be shown by the balance.

These conditions have been satisfied in the following manner : — The
thread supports form part of a girder mechanism which is itself con-
tained in a thermally insulated tube. During transport the lever is
arrested by a mechanism which clamps it with a definite pressure in a
definite position. The end of the lever is observed by a microscope
which is always brought into the same relative position with respect to
the horizontal plane through the thread by means of sensitive striding
levels. It is shown as a consequence of the mechanical conditions
that the lever will be in unstable equilibrium when its centre of
gravity rises above the horizontal plane through the thread by about
3°. Consequently the accuracy with which the lever can be brought
to its sighted position is very great, for the position selected as the
sighted position is within a small fraction of a degree of the position of
instability.

As it is necessary either to keep the balance in an atmosphere of con-
stant density or to correct the observations for changes in the barome-
trical pressure, the former course was decided upon, and consequently
the instrument is contained in an air-tight space. This involves working
the vernier axle through a stuffing box which must be practically Mo-
tionless, a condition satisfied by a sort of mercury sealing.

The difficulty in making the apparatus arises from the fact that
quartz fibres, though infinitely better than any other material, are not
really sufficiently perfect in their elastic properties for the present pur-
pose, and it is only by a judicious balancing of advantages that it is
possible to arrive at the necessary sensitiveness. Even after two years'
twisting the thread of the instrument still undergoes a continual,
though small, viscous deformation; this, however, becomes sensibly
constant, and can be allowed for.

A further complication arises from the fact that as the temperature
rises the quartz becomes stiffer, so that at a given station the circle
readings are a function of the temperature. We have found that the
relation between the circle reading and the platinum temperature is a
linear one at ordinary temperatures.

An essential feature of the apparatus, therefore, is a platinum wire
thermometer placed alongside the thread.

The following statement shows concisely the effect on a determi-

On a Quartz-thread Gravity Balance.

125

nation of gravity of the various observational errors which are
possible. The instrument of course only refers differences of gravity
to a known difference. The results are expressed in round numbers,
gravity being taken at its value in the latitude of Sydney.

Our temperature observations may be inconsistent by at most one-
hundredth of a centigrade degree ; this would correspond to an uncer-
tainty of one part in 700,000 in the value of " g."

The accuracy with which the microscope can be set on the lever is
much greater than the accuracy of reading the sextant arc. If our
estimate of the latter is wrong by 5" the resulting value of "g " is
affected to the extent of one part in 1,300,000.

The errors of levelling may amount to one part in 700,000 in the
value of " g." This gives a possible maximum uncertainty of one part
in 300,000 in the value of " g." The daily rate of the instrument does
not introduce an uncertainty of anything like the amounts mentioned
above, and can in any case be eliminated by observing alternately at
two stations, the difference in the value of gravity between them being
the subject of observation.

Observations. — Two observers are required, one for the balance and
one at the thermometer resistance box.

It is only possible to observe with sufficient accuracy when the tem-
perature is nearly steady ; we always observe therefore at a time when
the temperature passes through a maximum or a minimum value.
With the instrument as constructed of various metals it is also neces-
sary to avoid observing too soon after any great and sudden variation
of temperature.

Journeys. — We have travelled with the balance from Sydney to Mel-
bourne by train, from Melbourne to Hobart by steamer, from Hobart
to Launceston (in Tasmania) by train, back to Melbourne by steamer,
and to Sydney by train. We have also made many less extended
journeys in Xew South Wales, having travelled over more than 6000
miles with the instrument.

Most of these journeys led to our making improvements in the
instrument, and therefore are not to be regarded as forming surveys.

If, however, a consistency of one part in 50,000 in the value of " g "
be considered satisfactory, then the Tasmanian stations may be consi-
dered as surveyed, and the values assigned to gravity at these stations
to be referred to the Melbourne-Sydney difference. Since this journey
was undertaken the instrument has been so much improved in detail
that we do not discuss its results from a gravitational point of view.

We have, however, made three test journeys between . Sydney and
Hornsby in New South Wales under proper conditions, and the result
of these observations shows that at Hornsby the thread has to be
untwisted at one end by the following amounts as referred to the
reading at Sydney : —

126

Miss M. A. Whiteley and Prof. Karl Pearson.

Journey 1. Mean of Sydney — Hornsby and Hornsby — Sydney.

Difference 18*5 sextant minutes.

Journey 2. Mean of Sydney — Hornsby and Hornsby — Sydney.

Difference 18'1 sextant minutes.

Journey 3. — Mean of Sydney — Hornsby and Hornsby — Sydney.

Difference 18'1 sextant minutes.

The maximum difference is thus 0'4 sextant minute, and corresponds
to an uncertainty in the value to be assigned to the acceleration of
gravity at Hornsby as compared with that at Sydney taken as known
of one part in 500,000. This we believe to fairly represent the accu-
racy attainable by the instrument in actual field work. It is about
double of the outside accuracy attainable by invariable pendulums, not
connected by telegraph, and the observation takes about half an hour,
but the time depends on the time required for the temperature to
become steady. The observations quoted took about three hours each.
Packing and unpacking takes about an hour and a half, and the actual
observing about five minutes, but the temperature must be watched to
the maximum or minimum before the observations begin.

The weight of the instrument and of appliances taken directly from
the laboratory and packed in strong boxes is 226 lbs. ; by making
special appliances with a view to lightness this weight might be reduced
to one-half.

The paper is illustrated by working drawings, &c.

" Data for the Problem of Evolution in Man. I. A First Study
of the Variability and Correlation of the Hand." By Miss
M. A. Whiteley, B.Sc., and Karl Pearson, F.K.S. Eeceived
April 6 — Bead April 27, 1899.

1. In a more purely theoretical discussion of the influence of natural
selection on the variability and correlation of species, which one of the.
present writers hopes shortly to publish, a number of theorems are
proved which it is desirable to illustrate numerically. But the quanti-
tative measures of the variability and correlation hitherto published
are comparatively few in number, especially when, as in the present
case, we desire to have their values for a number of local races of the
same species. When we have once realised that neither variability nor
correlation are constant for local races but are modified in a determi-
nate manner by natural selection, and further that their differences are
the sure key to the problem of how selection has differentiated local
races, then the nnportance of putting on record all the quantitative
measures we can possibly ascertain of variability and correlation
becomes apparent. For some five years past various members of the

Data for the Problem of Evolution in Man.

127

Department of Applied Mathematics in University College, London,
have, so far as their other work allowed, been collecting and reducing
data concerning the variability and correlation of different organs and
characters in man. So far as variability is concerned, 160 cases of
organs in divers races of man were worked out by Miss Alice Lee, Mr.
G. U. Yule, and one of the present writers some years ago,* and since
then the more laborious task of measuring the correlation of characters
and organs in man has been going steadily forward, until at the present
time a considerable mass of material is reduced and ready for publica-
tion. The present series of short papers is intended to cover this
ground. It will simply state the numerical results reached and any
obvious conclusions to be drawn from them, leaving to a later date the
consideration of the material as a whole, and in particular its bearing
on the general problem of evolution and the relationship of local races
of man.

2. This first study deals only with one character of the hand in one
sex and one race. A wider range of material on the skeleton of the
hand in another local race is already being dealt with. But while the
correlation of the anatomically simple parts of the hand is of very
great importance, it does not follow that the complex members of the
living hand may not be equally, or even more, significant when we have
to deal with fitness for the struggle for existence. So far as we have
been able to ascertain, although much has been written as to the fitness
of the hand for its tasks, no attempt has ever been made to ascertain
quantitatively the degree of correlation of its parts, f Hence our
first object was to get some idea of the correlation of the parts of the
hand from an easily measured and in practice important part. Is the
hand as highly correlated as the long bones, or as loosely correlated as
the parts of the skull, or does it occupy some intermediate position like
that of strength to stature 1 We accordingly selected as an easily
measured but still important character the first joint of the fingers.
The measurement therefore covers, besides the fleshy parts, the head of
the metacarpal bone together with the proximal phalange. It is thus
not anatomically simple, but it probably has much importance for the
fitness of the hand, and is a measurement which with a little care can
be made with considerable accuracy. Our measurements were taken
with a small boxwood spanner graduated to 1/10 inch, and provided

* A diagram was exhibited at a soiree of the Koyal Society three years ago, and
we shall be glad to send a photograph of that diagram to any one working at the
problem of variation. The data without the diagram are published in a paper on
" Variation in Man and Woman," ' The Chances of Death,' vol. 1, pp. 256 — 277.

f Here, as in other cases, both zoologists and anatomists have since the days of
Cuvier, talked a good deal about correlation, but would even to-day be unable to
reconstruct, with anything like quantitative accuracy, a skeleton from a long bone,
a hand from a finger- joint, or a skull from a fragment.

128

Miss M. A. Whiteley and Prof. Karl Pearson.

with a vernier, so that the readings conld be nominally made to
1/100 inch. Both the hands of 551 women were measured. At first it
was proposed to include only those of more than 20 years of age, but
no sensible difference was found for the means of those between 18
and 20, and accordingly some sixty or more between these years
were included in the final results. While more than a moiety of the
measurements and nearly all the laborious arithmetical reductions were
made by one of us, Miss M. A. Whiteley, we owe measurements on
the students of University, Girton, Newnham, and Westfield Colleges
to the energetic assistance of Miss Dorothy Marshall, B.Sc, and a
further ninety sets, principally from the students of Bedford College,
to Miss Edith Humphrey, B.Sc, to both of whom we wish to acknow-
ledge our great indebtedness.

In the tabulation of results the grouping was done to 1/20 inch, and
the means, standard deviations, coefficients of variation, and coefficients
of correlation, together with their probable errors, calculated by the
processes and formulae already fully described in papers of the series
" Mathematical Contributions to the Theory of Evolution," by one of
the present writers. Pianists were specially noted on the data cards,
but their numbers did not seem sufficiently large to justify any con-
clusions as to the effect of use on variability and correlation — a subject
which deserves very careful and special investigation.*

The following notation is used : —

Ei =

first joint of right-hand index finger.

Eii =

„ „ middle „

Eiii =

ring

Eiv =

„ „ little „

Li =

„ left-hand index „

Lii =

„ „ middle „

Liii =

ring

Liv =

little

3. Relative Size of the Hands. — Turning first to the absolute dimen- *
sions of these joints we have, the measurements being in inches : —

Table I. — Lengths of First Joints of Fingers.

R. L.

i. 2-2482 ± 0-0030 2*2252 ± 0*0031

ii. 2-3879 ± 0-0033 2*3667 ± 0-0033

iii. 2-2108 ± 0-0031 2-1878 ± 0*0031

iv. 1-8427 ± 0-0028 1-8197 ± 0-0028

* What effect ma; particular trades or forms of exercise have in modifying the
variability of the limbs used and their correlation to other limbs ? The relative
importance of use and of selection in determining the current values of variability
and correlation will one day require very careful investigation.

Data for the Problem of Evolution in Man.

129

We conclude at once that these joints in the right hand are very
sensibly larger than in the left. In every case there is a difference of
about 0*02, and this is many times larger than the probable error of
the difference, i.e., s/2 x 0*003 about.

We might, therefore, conclude that the right hand is larger than the
left. This conclusion is directly opposed to that of W. Pfitzner :* he
asserts that there is no quantitative difference between right and left
for the simple anatomical parts of the hand skeleton. His own
measurements, however, really do show such a sensible difference for
the first phalange. All then we assert at present is that the first joint
and the first phalange are larger in the right than in the left hand of
women. We prefer to state no more sweeping view at present as to
other parts of the hand, however strong our private opinion may be.

4. Variability of the Hand. — The following are the numerical results
reached : —

Table II.

Standard deviation. Coefficient of variation.

Ri 0-1055 ±0-0021 4*6945 ± 0*0954

Rii 0-1133 ±0-0023 4-7432 ± 0*0964

Eiii 0-1091 ±0-0022 4*9345 ± 0-0100

Riv 0-0986 ±0-0020 5-3537 ±0*0109

Li 0*1088 ±0*0022 4*8917 ± 0*0994

Lii 0*1137 ±0*0023 4*8033 ±0*0976

Liii 0*1082 ±0*0022 4*9481 ±0*0101

Liv 0*0975 ±0*0020 5*3614 ±0*0109

If we were to judge by absolute variations the index and middle
fingers of the right hand are less, the ring and little fingers more
variable than those of the left hand. But if we use the more
reasonable coefficient of variation, we see that all the first joints for
the left hand are more variable than the corresponding joints for the
right hand, and this is precisely what we might expect if there be
greater adaptation by selection, or by use of the right hand. The
greater the selection, the less the variability.

In the left hand the relative order of variability (as measured by
the coefficient of variation) is that of the relative size of the fingers :
in the right hand this is slightly modified.! The work has been care-

* Dr. G-ustav Schwalbe's ' Morphologische Arbeiten'; W. Pfitzner. 'Das
Menschliche Extrernitatenskelet,' Bd. I, pp. 21—35, and Bd. II, pp. 99—106,
1892 and 1893.

t The divergence is not one on which real stress can be laid considering the
probable error of the coefficient of variation. The hand confirms what we have
already learnt from the long bones ("Roy. Soc. Proc.,' vol. 61, pp. 347 — 348),
that 5 per cent, closely measures the variability of the chief parts of the human
body.

VOL. LXV. L

130

Miss M. A. Whiteley and Prof. Karl Pearson.

fully re-done but no error has been discovered. It would thus appear
that in the right hand the index finger is less variable than the middle
finger. The general order of utility of the fingers would appear to be
middle finger, index finger, ring finger, little finger, and this exactly
agrees with the order of increasing variability in the left hand. The
only doubt about this order appears in the relative efficiency and utility
of the middle and index fingers, which have a different order of varia-
bility in the right hand.

As all our subjects belonged to the educated classes, it is just possible
that the great use of the right hand index finger in writing has some-
thing to do with this diversity.

5. Correlation of the First Finger Joints : —

Table IV. — Correlation Coefficients.
(a) Eight Hand.

!

Ri.

Eii.

Eiii.

Eiv.

K r ,.
Eii..
R iii .
E iv .

1

0-8994 ± 0-0055
0 -8753 ± 0 -0067
0-8173 ± 0-0095

0-8994 ± 0-0055
1

0 -9031 ± 0 *0053
0 -8243 ± 0 -0092

0-8753 ± 0-0067
0 -9031 ± 0 -0053
1

0 -8629 db 0 -0073

0 -8173 ± 0 -0095
0 -8243 ± 0 -0092
0-8629 ± 0-0073
1

(b) Left Hand.

Li.

Lii.

L iii.

!

L iv.

Li ..
Lii..

! Liii .
L iy. .

1

0-9097 ± 0-0050
0-8798 db 0-0065
0-8204 ± 0-0094

0 -9097 ± 0 -0050
1

0-9141 db 0-0047
0-8227 d= 0-0093

0-8798 i 0-0065
0 -9141 ± 0 -0047
1

0-8710 ± 0-0069

0 -8204 ± 0 -0094
0 -8227 ± 0 -0093
0 -8710 ± 0 -0069

1

(c) Eight and Left Hands.*

Ei.

Eii.

Eiii.

Eiv.

Li...
Lii..

L iii .
L iy .

0 -9249 ± 0 -0042

0 -9341 ± 0 -0037

0-9287 ± 0*0039

0 -9039 ± 0 -0053

* The great labour involved in forming and reducing the seventeen correlation
tables of this paper precluded the determination of further right and left-hand

correlation coefficients for the present.

Data for the Problem of Evolution in Man.

131

Now these tables indicate very important conclusions : —

(i) The hand is a very highly correlated organ, far more highly
correlated than the skull and even somewhat more so than the long
hones.* We are accustomed to give man precedence in life on
account of his brain power, and it might, perhaps, be thought that
the brain case would be highly correlated in its parts. Yet what we
find is that the skull is extremely individual, its correlations are low
and a man could be readily identified by head measurements, whereas
hand measurements would be immensely less safe. In other words
the hand so far as its dimensions go (we put aside markings) is far
closer to a type than the skull.

(ii) The parts of the left hand are distinctly more closely corre-
lated than those of the right. The only exception is the correlation
of R ii and R iv, which is greater than that of L ii and L iv, but the
difference here is considerably less than the probable error of the differ-
ence, and the general rule appears to be quite certain. Now this is a
most remarkable- result, but again how is it to be interpreted 1 Is it a
result of selection or a use effect 1 For the same organ it is a rule
that the greater the selection the less the variability and the less the
correlation. Exceptions there can be, which will be discussed else-
where, but this appears the general rule. Is the less variability and
correlation of the right hand a result of greater selection, or is it after
all a result of use 1 If the latter we see how hopeless it is to associate
constancy of correlation, or even of regression coefficients with the idea
of local races. Indeed the further we enter into the quantitative side
of the problem of evolution the more important appears the de-
termination of the influence of growth and use on both variability and
correlation. Why is the right hand less variable and less highly
correlated than the left 1 Is the answer the same as to the question :
Why is civilised man less variable and less highly correlated than
civilised woman 1

(iii) The order of correlation of the first finger joints is identical for
both hands. This order is as follows : —

(a) The external fingers have the least correlation and the little

finger always less than the index.

(b) A finger has always more correlation with a second than with

any other finger from which it is separated by the second.

Table IV(c) exhibits the correlation of corresponding members on
both sides. It will be observed that again the extreme pairs show

* Compare the table on p. 181 of the memoir " On the Beconstruction of the
Stature of Prehistoric Races" ('Phil. Trans.,' A, vol. 192). The index and
middle finger first joints are more highly correlated than femur and tibia ; the
middle and ring finger first joints than humerus and radius, the index and ring-
finger first joints than femur and humerus.

L 2

132

Miss M. A. Whiteley and Prof. Karl Pearson.

least correlation, and the pair of middle fingers higher correlation than
the pair of ring fingers.

Dr. Warren* has been the first to consider the correlation of cor-

Hence we are compelled to conclude that the correlation between cor-
responding long bones (with the possible exception of that of the
radii, which is within the probable error of the value for the middle
fingers) is greater than that between corresponding parts of the two
hands.

6. Index Correlations. — One of the present writers has previously
expressed doubts of the validity of using index correlations as a
measure of organic correlation.! At the same time it may not be
without value to put on record the correlations between the finger
joints expressed in terms of the first little finger joint as unit.

There are two methods of obtaining index correlations, either
directly by forming the actual ratios and then grouping them in cor-
relation tables, or indirectly from the variations and correlations of the
absolute quantities by means of the formulae given in the memoir cited
in the footnote. The latter is by far the easier process, but it neglects
what are usually small quantities of the third order. In order to
justify the use of the latter method, the values of the constants for
i14 = Ei/Eiv and i24 = Eii/Eiv were found by both methods. They
gave the following results, 214, 224 being the standard deviations of
the indices, Vu, V04 the coefficients of variation, and p the coefficient
of correlation.

responding right and left parts. He gives for $ series of Naqada
bones : —

E and L femur 0-9618 ± 0*0045

R and L tibia 0*9505 ± 0*0047

E and L humerus 0*9643 ± 0*0047

E and L radius 0*9322 ± 0*0124

Table V.

Directly.
By correlation table.

Indirectly.
By formulae.

Difference.

— 14

y

1 -2216
1 -2970
0 -0368
0 -0389
3 -0097
3 -0034
0-7388

1 -2210
1 -2968
0 -0379
0 -0395
3 -1013
3 -0487
0 7631

+ 0 -0006
+ 0-0002
-0-0011
-0-0006
-0 0916
-0-0453
-0-0243

* ' Phil. Trans.,' B, vol. 189, p. 178.

t ' Boy. Soc. Proc.,' vol. 60, pp. 489—498.

Data for the Problem of Evolution in Man. 133

It will be seen at once that the means and standard deviations
obtained by the two methods are very close, but that in the coefficients
of variation and correlation there may be a difference of some 3 per
cent. Sensible as this is, its amount did not seem to justify the
immense additional labour of index correlation tables — until at any
rate the biologists have shown what possible use can be made of index
correlations for organic relationship.

The following results were obtained : —

Table VI.

Index.

Mean value.

Standard deviation.

R i/R ir
R ii/R iv
R iii/R iv

1 -2210
1 -2968
1 -2004

0 -03787
0 -03954
0 -03270

L i/L iv
L ii/L iv
L iii/L iv

1 -2238
1 -3016
1 -2030

0 -03799
0 -04001
0 -03186

It would thus appear that the indices for the left hand are all
larger than for the right, or the index, middle and ring fingers
relatively larger with respect to the little finger in the left than the
right hand. On the whole the variability of the right hand still
appears less than that of the left, i.e., two cases against one.

Turning to correlation, the following values were found : —

Table VII. — Total Correlations of Indices.

1

Ri/Riv
Rii/Riv
R iii/R iv

Ri/Riv.

Rii/Riv.

Riii/Riv.

Li/Liv.

Lii/Liv.

Li ii L iv.

Li/Liv
Lii/Liv
Liii/Liv

1

0 -7631
0-6632

0-7631
1

0 -7310

0 -6632
0 -7310
1

1

0 -7774
0-6587

0 -7774
1

0 -7590

0 -6587
0 -7590
1

Here, but not so decisively as in the case of absolute magnitudes,
the left hand exhibits higher correlation. This higher correlation
becomes absolutely decisive, however, if we consider the spurious cor-
relations given below.

Table VIII. — Spurious Correlations of Indices.

Ri/Riv
Rii/Riv
Riii/Riv

Ri/Riv.

Rii/Riv.

Riii/Riv.

Li/Liv.

Lii/Liv.

L iii/L iv.

Li/L iv
Lii/Liv
Liii/Liv

1

0 -5628
0-5529

0-5628
1

0 -5504

0-5529
0-5504
1

1

0-5502
0-5429

0 -5502
1

0 -5473

0-5429
0 -5473
1

134

Miss M. A. Whiteley and Prof. Karl Pearson.

In every case the right hand exhibits more spumous correlation than
the left, and our previous conclusion is thus thoroughly confirmed :
the left hand exhibits higher organic correlation of its parts than the
right. How is this to be explained ? It is all important that further
researches shoidd determine whether it is selection or use which
differentiates the two hands. It would be hardly possible to find a
sufficiently large group of left-handed persons to mark how far varia-
tion and correlation were modified ; but measurements on the hands
of children, of the educated and uneducated, and of workmen following
particular trades might possibly throw light on the extent to which use
modifies correlation.

"We append the correlation tables giving the data upon which our
numerical values are based.

Data for the Problem, of Evolution in Man.

135

P3

eg

* e

^> bJD

-.5 • i— i
R fi

* s

S3

Totals.

l

| to to O lOWifllOWW »0

i onnowot?h«:5dhomt}ih

rlCOOO^OOiOlMH

LO

tO

2-55

to
2 -60.

H

o to
tO p >p

to to

d o

to ©
• o >o

TP ^ •

tQ tO *Q
rH <M rH

to

OS

Q kO
O

<N <N

to lO
«5 N N LO O

H H 05 CO N O

ta

rH

CO

to d

M Of
IM ©3

Iffl to
tp 1> N

rH Ol

»0

2-30

to
2*35.

to to

© SO CO (M SO ©
M CO H

§3

2-25

to
2 *30

to to

J> tO to <N VO
Ol © GO © rH ©

to

©
©

O to
N p N

<M *~ <M

to to

tO to t~ <N <N

H O H O t» N
CO CO rH

GO
©

ua d

H ON

• ■+=> •

to to

to lO tO

H N 00 SO H N O
<M CO rH

to

5-1

GO

2-10

to
2 -15.

to to to to
© eo «a <m <M ©

HHH

GO
'H

to d

O OH
OJ <N

ip >p to

N N lO N SD H
rH

iO

<m

2-00

to
2 *05.

LO to to to

(N CD CO ©

CO

1-95

to
2 '00.

•

to to to
OONHO

o

I
P3

to©to©to©to©»o©to©to©to

©rHrH(MCNC0C0TTi^t0tO«OSOl>l>
1 1 1 1 1 | 1 I I 1 1

1 1 1 1 1 1 f 1 • 1 < V ' 1 1

©to©to©to©to©to©to©to©

OOHHNOqCOCO-rJHlOlOSOSON

1

136

Miss M. A. Whiteley and Prof. Karl Pearson.

c5

be

1 1

o
EH

tb'ua >.o«5lo »p

O CO CO OS Hf< CO CO CO CO l> CM CO l> rH rH 1
NTfJ>OOMOMH |
rH rH

1

LO
LO

to d

CM CM

o »g

tp O V5

oq cm

IB *S

o ©

rH

to d

CM CM

LO LO

CM CM LO »p
CM CO CO O

LO

O lo'
o ^

CM CM '

tO iO

LO tC CM CM

O CM CO O O CM
rH rH

to

CO

tffl d

05 Of
CM CM

tO J> lO »0 CM

CM GO LO © rH rH
CM rH

LO

•O K3
CO p CO

CM CM

WVO 1C LO tO tO

CM CM CM CM CM

OO^NffiONH

CM CM CM

CM
X

o d

CM O CO
CM CM

lO LO

CM CM LO LO

i-JO^eqasoO'^irH

CM CO CM

LO

© ■

l—>

o o

CM O CM
CM CM

mo in m m

LO CM CM CM

rH rH CO CM Oi l>
rH tjI CM

CO
OS

lo d

rH p (M
CM CM

CO LO

lo lo cm »p lo cm lo m

© CM OS tO CO i> CO
CO CM

LO

CM
CO

2-10

to

^5 J-t).

in m m lo

CM J> <M LO J> LO

O LO OS 4j< CO rH
rH rH

TJt

o d

O OH

CM CM

lo m

fc* i> LO lo

HHOOWHHO
rH rH

LO

OS

CM

O LO

O O o

• -4i •

N CM

tO LO
O O rH 00 CM

S

to d
OS p ©

rH CM

to to

CM I-
rH CO O

LO

t

5

Oiooioomowoaotflowo

C50500HH(MNWM^TiiiO»0<0
rHrHCMCMCMCMCMCMOqCMCMCMCMCMCM

1 1 1 1 1 I 1 1 1 1 1 1 1 I!

»O©LO©iO©*O©L0©LO©L0©L0
QOOSOJOOHHNNCOCO'frJtirjlfS

rHrHrHCMCMCMCMCMCMCMCMCMCMCMCM

or
6

■*=

c

Data for the Problem of Evolution in Mom.

137

OB
1

m in m in in o

HOCNOt»H005NMMMCOH
HNOOOOOCCON
r-t rH

m
m

2-55
to
2 -60

r-(

rH

o m

CM CM

•

i— i

iH

in O

CM 43 CM

m «3 cm J> o

m
©

' O

O

cm cm

m m m m
j> t> ip t> -i>

CM t> t> OS CO

in

io o •
cm cm

o ip m

iH O 00 rfl CO CM
HHH

in

to

2 35

m m
cm m in in m cm in

OHMHHHN
i-H CM CO i— 1

00

m ©
CM o «
CM CM

in m o io
in i> m t> in x> cm m

OMHOlOrf
H CO CO H

m

©
o

o in
cm 0 cm

CM CM

ic m m m o !
i> cm i> in

© rf< in W ^ «D iH | CO '
CM CO CM ©

m o
h o?q

cm cq

m in in m

CM CM i>» CM

■*<DO00H(M
HCOHH

kp

CM
00

O m
ih o T1

CM CM

in m m o

CM CM CM CM

<M 00 CO CO OS r-t rH
iH rH

00

in o
© o T*

CM CM

m m in in

CM CM .t- J>» in

00 1> J> rH

m
©

CM

2-00

to
2 -05

m m
cm m in

CO

m o
o Oo

■ *

iH CM

rH CM rH i— I

in

t

5

inoinoinomOinomomom
^inmcDOi>i>ooooo50iOOi— i h

1 M 1 1 1 1 1 1 1 II 1 1 1
©momcioomomowoin©

rfi^iomcocoj>i>oooooso5©©>-i

i— IrHiHfHrHi— IfHr—lTHrHrHrHCMCMCM

m

3

o
Eh

138

Miss M. A. Whiteley and Prof. Karl Pearson.

P3
5

w

■+3

rG

'QJO

a

O

H

to in to 1(3 to xo

OeOCOOi'*COCOCOC01>CM^5l>r-lrH
CM TiNOOMOWH

rH

in

2-70

to
2-75

w >n

CM CM

>p

m O

CM CM

HHH rH

H3

o o
© o?

<M ^ CM

in m

m cm in

HHOM

CO
■H

2-55

to
2 "60

m in
j>. in cm

H CC O LO H

in
o

CM

o m
o p o

<N CM

m in
in m i>.

© CM J> CO CO CO rH
CM rH

in

—- 1

m

m o

o 1(5
CM CM

in in in in in
O co oo h cb m

CM CM

in

m
co

O m
o

CM <M

.

m m
m in x.^- cm in

rH m CM CM O
rH CO

»
$

CO o ^

* -43 *

CM CM

in in

j ^ cm m

CD OS 00 1>

CO CO

m

rH

o o

CO o CO
• -*3 •
CM CM

»n m

cm cm in in

© © 00 to 00 CM rH

in

H3 O
CM o 03

CM CM

m m in in

cm cm cm j>- in m in

(~\ \f\ \fs rt\ /— \

LJ IIJ VJU T— ^

rH rH

rH

o o

CM o CM
CM CM

m m

CM

O rH ^ t> rH

m

03

«3 o

H 0(N
CM 43 CM

o in in in

CM CM in CM 1>
CD CM J^*

to

rH

O W3
r-i O rH
• -»s •

CM CM

m m m m

CM CM in J> fc-

<H> r— 1 w>

in

M3 O
9 OH

CM CM

in

in

CM

O to

? SP

CM CM

in
c

in
o

t

n

omoinomomoinoinomo

OJaOOHHNNcOCO^^iOiOCD
rHrHCMCMCMCMCMCMCMCMCMCMCMCMCM

II 1 1 1 ! 1 1 1 1 II 1 1 i

voomoinomomoinomom

OOCSQOOHHNIMCOCO^^WSIO
i— IrHrHCMCMCMCMCMCMCMCMCMCMCMCM

CO

is

O •
rH

Data for the Problem of Evolution in Man,

139

Totals.

tO O lOW «5 O
rHOO50©.l>rHO0St*e9J>C0 00rH
i-l rH

r-l

tO

to

© LO

co <m

lo >o

CO CO
O O r-l

»p

rH

2 65

to
2-70

lO LO
rH O r-l O

HH

2-60

to
2-65

»o IO
rH rj» ttJI co O

co

rH

to o

»p p CO
CO. CO

1

to to

to 10 tO J> CO
rH O to to to CO O

to

k

2-50
to

to »o

CO CO

CO rH CO rH 00 CO
H H H

to

rH

tO

tO ©

o ia

CO "CO

to to
to to i ^ to co

O © LO Oi CO CO rH

to

to

CO

O iO

Tfl O

CO CO

to to to to

tO to <N1>. t> i>

OHt-NHNN

N CO N

tp

CO
Ci

2-35

to
2 '40

to >o

t» to CO tO tO

NOMMNOH
CO CO CO

iO

rH
OS

LO
«*
05

2-30

to
2 '35

to to to to

I.ONNONN

rH JS tO rH
rH CO CM rH

ua o

N OM
CM CO

to »p

H CO 00 W O 05 H
rH rH

»o

2-20

to
2 *25

to tO tO lO

CM SO 00 00 CO
rH

tO
CO

2-15
to
2-20

to to

CO CO >p
HH CO tO O

CO
-H

2-10

to
2-15

' to lO
CM CO

rH rH CO CO O

tp
i>

2*05

to
2-10

to lo

t O J> CO

© O O rH

to

CO

o to
9 ©9

CO CO

tO

o

to
o

t

3

tootootootootootootooto

^»fllOCOCDJ>l>CO(30a>0500HH
rH rH rH rH rHrHrHrHrHrHrHCOCOCOCO
1 1 1 1 1 1 I II 1 1 I 1 I I

o to o to o to o o o to o »o o to o

rJ(^tOtOCOC01>l>00 00 05CSOOrH
l — IrHiHrHr-trHrHrHrHrHrHrHSOCOCO

CO

Is
"o
■h

140 Miss M. A. Whiteley and Prof. Karl Pearson.

Totals.

xp »p xo xp xp xo

HOOMONH005bMJ>(NIX)H
HNlOQOOOOCCCSI

i . MH

rH

XO
XO

XO O
xp _0 cp

<M cq

rH

rH

(

o xo
xp o xo

oq eq

rH

rH

xa ©

O VO

• -+S •

cm cm

XO XO

t>. j> xa

O CM CM O

O xa

O

CM CM

XO XO
1> (M

©
rH

XO O

WO'*

CM CM

1

xa CM t» <M i>

(MrlOCCOrl
rH rH

xp

CM
CO

O xa

co o co

i xo xo

j>j> xa

© O CD 00 rH
rH CM rH

xo

Hand

XO O
NOW

CM CM

xO XO
<M xa CM

rH CM OS CO O rH
rH CM CM

CO

i>

, Right

o xa

WON

CM CM

xo xo xa xa

t>- 1> CM <M XO
rH cs xa 00 00 o

CO CO rH

xa

co
©

I

SH

a>
bJD
fl

xa ©

HON

CM CM

XO XO
XO i> J>» XO

rH CO xa rH -"J
CO CO CM

XO

CO

©

Little

"2-10

to
2 *15

XO us

XO CM

* XO 00 rH Tf
rH CM CM

XO

CO

x>

iiig and

xo o
9071
cm ~ cm

XO XO
XO CM XO 1^

NHCMO
rH rH rH

H/t

P

>

O XO

© o ©
oq cm

XO XO
XO CM l> XO

CO © CO rH rH ©
rH

1

©
(M

xa o
9 3 9

rH CM

XO XO
rH rH CM © rH ©

©

o xa
<T. o cs

tH i— 1

XO

r- > CM ©

XO
CO

XO o

X OOJ

iH r-i

XO

©

XO

©

t

XO©XO©XO©XO©XO©XO©XO©XO
*»OlOOCDl>t^(»000)0)OOHH

1 1 I 1 I Ml 1 1 1 1 I 1 I

©xo©xo©xo©xo©xoOxo©xo©

^T^XOxOCDC0t>.i>0000OS©©©rH

CO

o

Data for the Problem of Evolution in Man.

141

1

■Jl

i

in in lO lO W >n lO lO lO W

Tot

©mC5CDCOHHr^CO©— lrHCii>CMrH
OJ CO t» C5 OJ 00 00 ^1 H

rH

in
m

2 55

to
2-60

©

©

© in

>p

cq ~ <m

rH ©

rH

2 45

to
2-50

10 J> J>
NOHH

CD

2-40

to
2-45

its m m io in

rH CO 00 rH ©

to

CM
CM

lO ©
CO o ^
• <M CM

in HO IQ
i> CM 1^ (N

i— 1 CM CM W5 ICS rH
rH rH

00

co

© W

co -o co

CM <M

lO lO
CM J>

C5 00 O 0 CO
rH CO rH

CO

i>

2-25

to
2 -30

>o in
x o in

OWHI>HO
rH CM CO

in

CO
Jt>

2-20

to
2 25

in m in

CO J> 00 in
CO CO (M

»o
©

rH

m ©

H ON

CM CM

H M N CD O N
CM CO CM

©

OS

2-10

to
2 -15

in m

CM ip (M »p

rH CO HH CM CO
rH CO rH

lO

CO

2-05

to
2 *10

vo in

I> lO in

CM © HH GO rH ©
rH rH

lO
CO

2-00

to
2 *05

v.° ^

CM CD rH CO ©
rH

I

i>

CM

1 -95
to

2 '00

rH rH rH

CO

© VO
i-l rH

CM rH

CO

in©m©in©in©in©m©w©in

©r-trHCMCMCOCOTflHHinmCOCOi>^

t

CMCMCMCMCMCMCMCMCMCMCMCMCMIMCM

Li

1 1 1 1 1 ! 1 1 1 1 1 1 I I I

©m©m©w©m©m©m©m©

OOHHNNCOW^-*W»CCOC01>

otals

(MCMCMCMCMCMCMCMCMCMCMCMCMCMCM

H

142

Miss M. A. Wliiteley and Prof. Karl Pearson.

.

Vf5 »p O U3V3 VO

rH

Mi

O

l!3 j'J
• •

eg eg

o

vfi
O

O «3

U0

CsJ " eg

O rH

rH

2 '45

to
2 *50

r-i rH r-l <M O

CO

o o
eg eg

Mi «5

«a i>. eg

H«5t»)>H

o

eg
eg

2-35

to
2*40

Mi lO U5«5

eg i> j> eg

rH i— 1

60
co

and (L

2-30

to
2 -35

xa Mi

o j> eg

4? j> co i-h co ©
rn eg eg

CO

t>

Left H

M3 O

eg o co
eg ^ eg

Mi Mi lO «5«5«5

t> j>. eg vo j>- eg eg
eg eg rH

Mi

CO

Ring Fingers,

2 '20

to
2*25

eg j> ip ua

eg t» OS 05 lO rH
H CO M

©

2-15

to
2 -20

Mi Mi
ip *p *«■ eg

O rH 00 00 CO 00

co co

o

©

CD

c o
i— i o 7*

eg *~ eg

OlflvpiO
O Oi CO CO

eg eg

•X)

I

2-05

to
2 -10

Mi iq io ia ia

rH

co eg co to rH

rH rH

Mi
CO

IIA

eg eg

i> eg i> eg
eg \a Mi co C

rH

eg

*a o
O o o

o io
Mi Mi t>. eg

i-i eg

OriOO

co

O io
9 o?

|H rH

r-i r-^ rH

CO

o >o O O O O O VO O lO o »o O
QOOOHHNNCOCO^'i'lO

t
rH*

rHrHegegegcgegegegegcgegeg
1 1 1 1 1 II 1 1 1 II 1

O O O O * O ifl O W O «5 O W
OOQOOOHHNiMMM'*-*

.HrHiHegeiegegegegcgegcgeg

Totals . .

Dot a for tTie Problem of E 'volution in Man.

IT.

1

LO LO LO MS MS MS

a

C0T*X>HjHrHt^l>COe0©COLO©
HCiCO'MCOONH

rH

L5
IO

ms o

CO CO

lO

©

MS
©

1

o lo

O o 10

V5 j

LO

• ^) •
CO CO

rH

LO ©

lO ip

CO ~ CO

CO

o lo

CO ~ CO

LO LO LO LO LO
CO LO CO CO rH O

LO

CO
CO

LO ©
CO o ^
• -ra •
CO CO

<0 LO
CO LO LO LO ^ MS

CO CO CO rH ©
•— 1 r— <

00
CO

Q lO
CO o CO

LO MS
MS J>

CO "~ CO

f-l O CO CO CO CO CO
rH CO CO

g

MS ©
CO o CO

MS MS
CO LO LO CM

LO

CO *" CO

CO O X> O CO
rH CO CO I— 1

g

O LO

LO MS

CO MS 1> MS

CO ^ CO

CO Tfl rH

\(S
©
rH

MS ©

MS MS
lO MS fc- CO MS MS

CO *" CO

o co ^ w a co o

rH CO CO

o

OS

O MS

1

LO MS
CO t> MS

LO

CO ^ CO

CO CO

*#

CO.

m o
9 o t1

lO lO
LO LO CO ^» LO

>p

CM ^ CO

rHOCOOOlCOCOO

CO

o o
9 o 9

MS MS
MS MS A"» CO LO LO

1

CO ~ CO

rH rH LO © LO CO rH

CO

1-95
to
2 '00

MS LO
rH rH O O

!

CO

0 >o

01 © 9

rH *" rH

LO \0
rH O rH O

CO

iflOWOtOOWOOOiOOiO
O W CC N l> 00 00OOOOHH

'.

t

Hi

HHHHHHHHHIMlMlNN

II 1 1 1 1 M M 1 1 I

©LO©LO©\O©L0©iO©LO©
OW5C0C0J>J>Q0XC5 0!OOH

an

~o

^^rHrHrHrHrHrHrHrHCOCOCO

EH

144

Miss 11 A. Whiteley and Prof. Karl Pearson.

Totals.

VO VO vp lO VQ

©cor-icscoi-irH'Mao.i>.oivo

HCCLO05HQU5^NH

VO

w

VO

o
i>

■>i

VO

o \

iO

1 \°

rH

vo
i— I

2-65

O

■43 •

VO \Q
O iH

O

©

j

vo

cq;

©

vo
»-> co

VO

oi

O CO N H

vo

o

r\ CO

3 •

O lO VO VO

C5

©

113
P V3

43 CM

i-l 6 X 30 00 CO

i-l

■^1

vo

o
o *^

4° i> \o vo

VO

■43 *

rH ^ iH OS O

cq i-i

i-l
VO

o

vo
o y

vo vo »o vo

X> cq Cq <N

vp

N

43 cq

H Tf H

O
00

\o

CO

o§

vo vo

cq vo vo

lO

<N

-43 •
cq

rH CO

CD
OS

O

co

o

vo vo
vo sq

VO

sq

-43 '

CO 79 C5 CO CO

•"t1 co

l>

C5

VO

<N

o

o M

vo

<M

•43 *

<M

co cq i— < ©

N CO H

i>

O

(M

VO

O <N

W H5 W lO

VO

EQ

rH <N CO TjH
rH iH

CO

co

lO
rH

O

o <N

VO VO
lO 1> vo J>-

vo

(M

O VO VO CO rH
i-l

CO
<M

O
sq

lO

o y

VO VO
CO -H ^

Ci

VO

9

<M

o

O -H

-43 "

vo VO » p vo
Ohhh

»o

t

OvoOvoOvoOvOOvoOvoO

:

rHrHCqcqcqcqcqcqcqsqcqNCq

CO

1111111111111

vo © »o O vo O vo O vo O vo O vra
OOOJCJOOHHWNMCO'l'll

rHrHiHCqcqcqcqcqcqcqcqcqeq

o
H

Data for the Problem of Evolution in Man.

145

►J

o
hp

Totals.

in in in «a 0 m

1 M^N^HNNCOCMOJCOinO
1 rH CO CO O CM « CO CM rH
rH rH

L

1 m

1 0

© uo

J> ON

CM CM

1

in

O rH

1-

lO O
O ON
• -4-a •
CM CM

1

in

O CM

1

l m ■

|

O m

CM CM

10 vp m

rH CO CM O O

2 55

to
2 '60

rH CM CO CO ^ <M

1

C5
rH

0 m

W O K5
• •

CM CM

»o lO CM CM in

CM 00 H3 rH CO rH
rH rH

rH

HH

2-45

to
2 *50

10 0 10 »o 0 m

NNNNNN

O00WNNH
rH rH

515

2-40

to
2*45

in in in in in
cqo»OMooo

rH CM CM rH

in

O

00

2-35

to
2 -40

m in

t>. s>. to 0

rH CO CM CO CO "?
CM Hi CM

98-5

2 30

to
2-35

»p i> >p in in n

H^OlffiOH
rH CO CM iH

in

N
OS

2-25

to
2 30

ujiqo in

t> NN W3 N

C5 N CO CM CM rH
rH CM rH

UO
N

2-20

to
2 -25

m m m m

rH VO r— 1 CM CO CM

in

CO
CO

2-15

to
2-20

LOO IQ
H O lO CO 00 ^? H

in

CO

CM

2-10

to
2-15

rH rH CM CM rH rH

O

in 0

O OH
• -» *

CM CM

V5 *P

r— 1 rH O CM

I !

* 1
1

loowooooottowoio

lCSCOONNrOOOOCiOOOHH

L f 1 1 1 i r mi 1 11

VOXOCOCONNOOQOCSCiOOrH

Totals

VOL. LXV.

M

Miss M. A. Whiteley and Prof. Karl Pearson.

a

{

H
O

1

O t.O lO MS MS MS

HMCOONOOONH
rH rH

rH

to
to

© !

CM CM

IS >oo I lO
OHH 1 CO

o to

^ Of
CM !M

tO MS MS MS I
(M W«CNO 1 "5
1 r_l

to o

CM CM

tO LO

O N (N N ^ H CM
rH CM

O to

tO MS

CO .05 — rH CM J>
N H -<3<

to O

<tt CM

»0 to
W5ifll> t, | ip

rH X (M ,_| 1 CO
CM CM 1 MS

2-20

to
2-25

MS MS

rH rH CM © O «
rH Tfl CM rH 1 O

lO ©
rH O *3

CI IM

MS tO
MS CM CM

OC5NCOQW rH
W ^ H | rH

O id
CM CM

>o to

MS MS MS i>

ONOnHOOH
CM 'T* CM

•H

©

US o

O C rH

• -+3 *

CM CM

>o to

tO tO CM 5-1

CO CM O CM tO O
rH CM rH

LO
CO

to

o *s
o o ©

• .4J •
CM CM

HJ>T?O^CJ

es

CO

MS ©

as o o

. ^3 .

rH CM

I 1

tp MS

rH "? H rH

rH

1 -90

to

J. VO

! to J >o

| rH CO CM 1 CO

I 1

MS ©
CO p O

rH rH

1 *

1 o

1-

t

to o to o ms o*oo>oo to o to

lOCOtO^NtOOCOiOOOr-iH

I I I I II M I I I M

o»oo»ootootoo*ootoo

vCKOOON^opoOflSOJpOH
rHrHrHrHrHrHrHrHiHrHCMCMCM

Data for the Problem of Evolution in Man.

j3
©
EH

vp VO vo vo vo vo

NMW050J>J>MIM
rH

rH

VO
VO

255
to
2 60

VO lO

o o

rH
rH

O vO

Vs o *P

<N (M

vo vp

o ©

iO o

tJI 0 lO
• -w *
<N <N

VO vo
eq vo i>

O CO rH rH

vp
Ci

2-40

to
2-45

VO VO
N VO J>

C0 CO <M <N
i— 1 rH

vp

w-i

CO

»o o

co © y

<N *~ <N

VO VO
l* M Lt 1>» VO

© <M i> CO rH
N rH

vo

k

2-30

to
2 35

VO VO

VO (N VO !N VO

rH VO 00 r-4 VO

00

2 25
to

1 '30

vo vO
t« N O

h a io o t» h

co

vp

o
©

© vO
<N <N

>o VO
VO <N (N

rH rH 1> VO VO 00
CO rf<

00

o

»0 O
? o «
<N 43 <M

vo vO

vp i> *a

CO <M 00 i> rH
CO CO

vp

<M
00

O »0
<N <N

VO 10

rH CM

00

VO O

Per1

<M <N

LO

rH CO Ol CO
rH rH

lO

©

N

2-00
to

2 '05

VO VO

CO
rH

l© O
OS o?
iH ** <N

vo VO
rH rH rH rH

VO

t

5

voOvOOvOOvoOvoOvOOvOO
QOOHHNNncO'f^lOLOO

rH <M <M <N <N NNNNNNNNIM

I I 1 1 1 i I 1 I I I ! i I

OVOOvoOvOOvoOvoOvoOvO

OOSOOrHrHOq<MCOCOTflTffi^)iO
HHNNNiNNNNNNNNN

O

M 2

Miss M. A. Whiteley and Prof. Karl Pearson.

C

<5

Totals.

vp »p >p ip \p »p U3 lO »p to

lO©CD«D^NQOOHHOH>NH
03 CO 1> OS OS 00 O tJI rH

i— (
iD
«3

O vo

> oJ>

<M 03

pi cq
Oh

»p
rH

2 65
to

2-70

rIOONO

HJ<

2-60

to
2 -65

»p »p pi £>.
OOHCOt?

CO
iH

2 55

to
2 "60

03 ID \Q 03
rH i> OS 03

»D

©

03

O ID

ID o

• •

03 03

lO m lO m
O »p .!> 03 03 03

O 03 A* CO CO O
00 03

»p
rH

ID

2-45

to
2-50

\a »p ip

O CO O lO CO X
03 03

\p
\o

CO

H* O h3
• -w •
03 03

-

ip i> J> »p

<=> ^ ^ ^ ^

»p

CO
OS

2-35

to
2-40

»p 03 X>

ID CO i> lO

^jl CO

ip
rH

o ws

M OM
03 CM

ID lO
»D ID J> 1>

H CO CO H
CO "tf tH

)p

H3
Oi

id o

03 o
03 03

ip »p »p »p

i— 1 CO 03 03 rH O
iH CO

rH

>o

2*20

to
2 25

\Q ID

i— 1 CO CO
rH tH

CO

2-15
to
2*25

\D \D
03 03 »p

03 ^ 00 i-i

CO

2 -10

to
2-15

»p |> 03
tH CO 03

>D
I>

2-05

to
2-10

»p

—1 iH

ip

03

8 Oo°
oi oq

6

up
o

t

5

0"50«30»00>OOW50»OOiO

HHoqwcow^^wiotocot^^

O303O3 03 03O3O3 03 O3O3O303O3O3

1 1 II 1 1 1 I 1 1 1 1 1 1
\D©iOO*D©vD©iD©iD©iD©
OHHCAKMCOCO-^^tfllOCOiXlN

O3O3O3O3O3O303 03 03 03 03 O3O303

Totals. . . .

Data for the Problem of Evolution in Man.

149

w

la
-g

H

«jw wo wo wo wo

O«DH0>WHHN00J>NO05
1— 1

rH

wo

wo

wo o
w cto

• -f-2 •
CM CM

wo wo
0 b

rH

o 10
wo _o wo

CM (M

1— 1

rH

2-45

to
2-50

IO ICO

fc> CM

o wo

y* 3

OJ CM

WO lO

a lo n x>

CM 00 O

2-35

to
2 "40

lO wo wo lO
CM 1> W0 J> CM

hoconhm

1— 1

■

wo

CM
CO

O WO

cc on
• •

CM <M

no 1010 O
(M (M 1> CM W0

O CM CO t> CM iH

!M CM

t--

iO

WO O
CM O CO

!M CM

ICO W0
i> M

O O H © Vfl
r-i CO CM

co
t>

o wo

<M © <M

cm <M

wo wo ico wo

N MO IN f)
W0 CO 00 XO O

wo

§

;

iC o

H ON

<m cm

lO lO
>p I> J> WO

O rH W0 00 00 O
CO *J<

W0
CO

0

2*10
to

W0 WO WO
CM rft

wo

CO

WO o
O O rH

CM CM

wo wo

CM CM W0

rH (Mi> 00 ^ r-i
-H 1— 1

O WO

9 o 9

CM CM

wo WO
WO CM «> W0

rH

C5
CM

wo O
o> o o

r-l CM

wo WO WO wo

O H H N O O

O WO

009

1-1 rH

wo
CO 0

WO

co

>~ O

00 OQ

1-1 i-l

wo
O

W0

0

t

s

owoowoowoowoowoowoo

OOOOHHMMWM'fTjda
r-irHCMCMCMCMCMCMCMCMCMCMCM

hi 1 J I -l 1 1 M 1 I I

W0OW0OW0OW0OW0OW0OW0
I QOOiQOOHHNNCOCC<#i(

rHrHrHCMCM(MCMCMCMCMCMCM(M

m

Is

-13
O

En

150

Miss M. A. Whiteley and Prof. Karl Pearson.

S

I?

3

I

(—5

!>

Totals.

\p \p \p on in

HCOCOONQOCONH
rH rH

rH

U3

0 ia

H 0H

cq cq

r-l

10

rH

10 0
O 0

• -w "
cq Cq

O IO (N lO
HNHNO

lp
00

2-00

to
2 05

tfi >o cq cq

OO HfflOH
1— 1

22-5

1*95

to
2 *00

l;0 -t> IO cq 1^

OHOOO^H
rH iH

xa

t>

co

1-90

to
1 *95

»o vo i> cq

OHG5MU5N
CO CO

CO

10 0

00 O Oi

• -+a •

i-l r-l

NO 10
\o cq cq

0 CO t- i> CO
CO

0

O »o

00 O 00
rH rH

ip cq cq vo

Ht>0 00 (N
Hjl rH

109-5

1-75

to
1'80

j> i>i>- cq

r-l Cq 00 rH

cq uo

10
b

OS

1-70

to
1 75

j>. j>. 10

CO t> rH CO rH rH

tH cq

rH

IO

1-65

to

1 70

t~ cq

CJ CO CO iH
rH

cq

1-60
to

1 -65

cq 10 cq

rH Cq Hji rH iH

©

1-55

to

1 -60

rH r-l

cq

1 -50

to

1 -55

O

1-45
to
1 -50

O

1-40
to
1 -45

rH

t

s

lflOlflO>OOWOlOOlOO«5
>(5?0^1>MjCX050500HH

1 1 1 1 I 1 I 1 1 1 1 1 1
OiOOWOiOOWOvOOOO
in«5C0C0NJ>»0Ci05OlOOH

Totals. . . J

Data for the Problem of Evolution in Man.

151

u
o

g

r-l

o
C

s

^3

Pi

X

o

>
X

>

5

£
5

e8

s

Totals.

lO W «5 K5

HHHOJN^lONNlH^OqOOH
O CO QO <M
r— i rH

1

»o

vo

1 375

to
1 -400

rH

_ ]

rH

O vO

vo _
<x> s V>

r-i i— l

1

©

1 '325

to
1 350

VO vO

rH rH © O

1-300

to
1 325

vo »o

VO cq <M

rH cq cq o

1-275

to
1 -300

vo tio vo

~ t/j cn w

vo
cq

] -250

to
1 -275

vO lO

VO VO J> X>

, 1 \f> f/*) A1 ofN MT| i
^-1 G>l ^1

cq co

VO

t>

1-225

to
1 -250

vp vo vo vo
O © © qo cq co

CO vO CO

CO
rH

1-200

to
1 -225

VO vO

j> cq

— u ff\ ff\

w»

cq cc co

rH

1-175

to
1 "200

VO VO

vo vo cq io vo

/vi *r-H I*** fVl _J 1
v^J U\| l-IJ '«N H"^ r~n

cq co cq rH

00

©

1-150

to
1 "175

vo vo vo vo
cq cq j> cq

U«J ^ SL-' UN P"^
rH rH

VO
H^

1125

to
1 "150

VO VO VO VO

cq cq j> cq

(O r^ *0

VO

1-100
to

o cq

vo

1-075
to

1 1UU

VO

©

VO

o

t

s

©vO©vO©vO©vo<DvO©vo©vO©
«5J>ONinNONiOJ>ONiOJ>0

iHrHcqcqcqcqcocococoTpT?Hj(Tf(i_o
1 1 II i 1 1 1 1 1 1 1 1 1 1

vO©vO©vo©vO©vo©vOOv<0©vO

(Mvot>ocqvot>ocqioi>ocqtO-t>
r-rHrHcqcqcqcqcococccoHjiT^HHH^

Totals .. ..

152

List of Candidates recommended for Election.

May 4, 1899.

The LOED LISTEK, F.E.C.S., D.C.L., President, in the Chair.

A List of the Presents received was laid on the table, and thanks
ordered for them.

In pursuance of the Statutes, the names of the Candidates recom-
mended for election into the Society were read, as follows : —

Barrett, Professor William.
Booth, Charles, D.Sc.
Bruce, David, Major E.A.M.C.
Fenton, Henry John Horstman,
M.A.

Gamble, James Sykes, M.A.
Haddon, Professor Alfred Cort,

M.A.
Head, Henry, M.D.
Hele-Shaw, Professor Henry Selby,

M.Inst.C.E.

The following Papers were read :-

Morgan, Professor Conwy Lloyd,

F.G.S.
Eeid, Clement, F.G.S.
Starling, Ernest Henry, M.D.
Tanner, Professor Henry William

Lloyd, M.A.
Threlfall, Eichard, M.A.
Tutton, Alfred E., B.Sc.
Windle, Professor Bertram Coghill

Allen, M.D.

I. " Photographic Eesearches on Phosphorescent Spectra." By Sir
W. Crookes, F.E.S.

II. " On the Chemical Classification of the Stars." By Sir Norman
Lockyer, K.C.B., F.E.S.

III. "On the Presence of two Vermiform Nuclei in the Fertilised

Embryo-sac of Lilium Martagon" By Miss E. Sargant.
Communicated by Dr. D. H. Scott, F.E.S.

IV. " Onygena equina (Willd.) : a Horn-destroying Fungus." By

Professor H. Marshall Ward, F.E.S.

V. " Impact with a Liquid Surface, studied by the aid of Instan-
: taneous Photography. Paper II." By Professor Worth-
ington, F.E.S., and E. C. Cole.

VI. " The external Features in the Development of Lepidosireu para-
doxa (Fitz.)." By J. G. Kerr. Communicated by A.
Sedgwick, F.E.S.

VII. " An Observation on Inheritance in Parthenogenesis." By Dr. E.
Warren. Communicated by Professor Weldon, F.E.S.

Impact with a Liquid Surface.

153

VIII. "The Thermal Expansion of Pure Nickel and Cobalt." By
A. E. Tuttox, B.Sc. Communicated by Professor TlLDEN,
F.R.S.

The Society adjourned over Ascension Day to Thursday, May 18.

" Impact with a Liquid Surface, studied by the aid of Instanta-
neous Photography. Paper II." By A. M. Worthtngton,
M.A., F.R.S., and E. S. Cole, M A. Received March 21 —
Read May 4, 1899.

(Abstract.)

This paper is a continuation of a paper under a similar title, published
in the ' Philosophical Transactions,' A, vol. 189, 1897.

It was there shown that between the splash of a rough and that of a
polished sphere falling the same distance into water, there is a remark-
able difference from the first moment of contact. The causes of this
difference are now investigated.

The configuration of the water surface below the general level, when
a rough sphere enters, is first studied by instantaneous photography,
and the origin is traced of the bubble that follows in the wake of the
sphere and of the emergent jet which follows its disappearance. The
depression or crater formed round the entering sphere is surprisingly
deep. This cavity segments, the lower part following as a bubble in
the wake of the sphere, while the upper part fills up by the influx of
surrounding water, which gathers velocity as it converges towards the
axis of the disturbance, and so produces the upward spurt of the jet.

Experiments are described in which some idea of the actual dis-
placements in the liquid has been obtained by letting the sphere
descend between two vertical slowly ascending streams of minute
bubbles liberated by electrolysis from two pointed electrodes.

It is found that with a gradual increase in the height of fall of a
well-polished sphere, the splash changes in character, and that the
sphere soon begins to take down air. But the height at which this is
first noticeable is largely dependent on minute differences in the con-
dition of the surface, and even on its temperature. It was further
found that dropping a smooth sphere through a flame, under certain
conditions, invariably alters entirely the course of the splash. This
action of the flame is proved to be no action of electrical discharge,
and reasons are given for attributing it to the burning off of fine dust
which has collected on the surface during the fall.

The influence of dust was proved by dusting one side only of a
polished sphere, a proceeding which always results in completely
changing the character of the splash on the dusted side.

154

Dr. E. Warren.

A satisfactory general explanation of all the phenomena is found in
the view that with a smooth sphere, cohesion is operative in guiding
the advancing edge of the liquid sheath which rises over and closely
envelops the sphere. If the surface is not rigid (e.g., is dusty), or is
rough, then the momentum of the sheath carries it, once for all, away
from the surface of the sphere, and the subsequent motion is quite
different. The persistence of the remarkable radial ribs or flutings
observable in the film that ensheathes a smooth entering sphere is
completely explained by the assumption of a viscous drag spreading
from the surface of the sphere outwards, and these flutings are always
absent from any part of the sheath that has left the sphere. Their
presence is an indication that there is no finite slip at the solid
surface.

Experiments made with water mixed with glycerine show that, up
to a certain point, the character of the disturbance is but slightly
affected by large changes in viscosity. With pure glycerine, however,
a thin film of water absorbed from the atmosphere equivalent to a
layer ^ mm. thick, was found completely to change the course of a
splash, a striking proof of the importance of the initial motion in
determining that which is to follow.

Experiments conducted in vacuo prove that the presence of the air
has no noticeable influence on the early course of a splash, but that its
» pressure subsequently prevents cavitation of the liquid under what
would otherwise be negative pressures.

The paper concludes with a reference to the remarkable similarity
between the splash at the surface of a liquid and that caused at the
surface of a hard-steel armour-plate by the impact of a projectile, and
with the suggestion that the explanation may be found in the argument
of Poynting,* which demands an increase of molecular mobility with
increase of pressure.

" An Observation on Inheritance in Parthenogenesis." By Eenest <
Waeken, D.Sc, University College, London. Communicated
by Professor W. F. E. Weldon, E.E.S. Eeceived March 22,
— Eead May 4, 1899.

On certain theoretical grounds it has been supposed by Weismann
that offspring produced by parthenogenesis exhibit little or no vari-
ability. To determine how far this conclusion was warranted by fact,
some measurements were made on Daphnia magna (Straus). f

* Poynting, "Change of State, Solid-Liquid," 'Phil. Mag.,' July, 1881; see
also two very important papers by Tresca on the " Flow of Solids," ( Proceedings-
of Institution of Mechanical Engineers,' June, 1867, and June, 1878.

f The measurements were made under the microscope with Zeiss's screw-
micrometer.

An Observation on Parthenogenesis.

155

The dimensions taken were : —

(1) The total length of the body measured along a line passing ventrally
from the base of the spine and cutting the convex surface of the head
opposite the middle of the compound eye (AB, see figure).

(2) The length of Uie protopodite of the 2nd antenna of the right side.
The measurement was made on the posterior surface of the proto-
podite along a line parallel to the dorsal edge. At the articulation
with the head the exo-skeleton of the protopodite possesses a well
defined point, which forms a good inner limit to the measurement
(CD).

Since, under favourable conditions, these animals continue to grow
throughout life, the second dimension was expressed in terms of the
first, thus Length of B protopodite x W0Q
lotal length of body

The mean of the relative length of the protopodite sinks as the
animal grows, but between a body length of 2'4 mm. and 3*6 mm. (the
total range of size) the change would not be large. I find that at the
time of measurement the offspring were constantly somewhat smaller
(0'4 to 0*5 mm.) than the parents, but as this applies to all the broods
which were measured, the rise in the mean of the offspring would not
affect the correlation surface.

From twenty-three Daphnia (themselves originating by partheno-
genesis) broods were produced consisting of three to six individuals.
The parents were measured, and the offspring were allowed to grow up.
On measuring the offspring it was at once obvious that the children of
the same brood exhibited very considerable variability.

In the following table (p. 156) the results of the measurements are
displayed in a correlation surface.

The table illustrates the variability of children of the same partheno-
genetic family, and we can further see, for example, that offspring
with a parentage of 169*5 thousandths exhibited a range of variation
159-5— 181-5 thousandths.

The following constants were calculated : —

6

Dr. E. Warren.

Actual
No. of

mothers

CO
<M

96

Weighted
mothers.

rH <N tH <M rH

CD
OS

Correlation surface.

I

6ST — 09T

z

n ! h w ! ! ! ! ' !

T9T— S9T

rH • 03 03 » rH . Oi • •

8

89T— ^9T

8

03 Z . CD CO r— I I 03 • rH

f

S9T — 99T

ST

. 05 CO N H . . .03

S

Z9T — 89T

II

* ! lO N H • • • I *

9

69T— 02,1

8

I I rH CO CO 03 I " * rH

ui—za

Oi

. • I CvJ ■ H H H • •

8

211— m

g

! 1 OI I ^ 03 03 CO I

6

81

I I !rH ! rjl rH 03 03 I

01

in — SIX

01

IT

62,1— 08T

9

• ! ! ! <h i'Ih •

ST

T8T — 381

z

8T

881—^81

0

I '. '. 1 . I - '. h I S5

n

S8T — 98T

T

rH<MCOTfliOCDt>OOOiO

•Suudego

No. of Offspring.

93 X

bJD

| «^

g 2 o °

©O3rJICO00©O3H3tO00
COCDCOCDCO.t^i>t>*-.t>

1 1 1 II 1 1 i 1 1
C5rHC0lOl>C5r-IC0lOJ>
uatOCDCDCOCDJ>i>l-»l>»
iHrHi— 1 H H rl rl H ri H

An Observation on Parthenogenesis.

157

1. The standard deviation (S. D.) of the mothers

weighted according to the number of offspring

produced . = 2*2208

2. The standard deviation of the offspring = 2*9503

3. The standard deviation of array of offspring = 2-6104

4. The coefficient of correlation - 0*466 ±0*0539

5. The coefficient of regression of offspring on

mothers.... - 0*619 ±0*0809

According to Mr. Galton's theory of ancestral heredity, a child, on
the average, inherits l/4th of any inherited character from either of its
parents, 1/1 6th from any one of its grandparents, l/64th from any one
of its eight great. grandparents, and so on.

From a mathematical standpoint Professor Pearson* has examined
Mr. Galton's theory, and he finds that if it be expressed in the form

/ 1 y S.D. of offspring

S.D. of individual parent of the nth generation

the coefficients of correlation and regression between offspring and any
generation of ancestors flow directly from it. Professor Pearson shows
that the total regression of the progeny on the mid-parent of any genera-
tion is constant and is equal to 0*6, while the correlation and regression of
an individual parent of the nth generation (supposing equal variability
for all generations) = 0*6(J)n and the correlation of the mid-parent

of the nth generation = 0*6/ -i-

\ si* I

Hence the coefficients of correlation and regression of an individual
parent of the 1st generation {i.e., father or mother) = O^J)1 = 0*3, and

the coefficient of regression, as we have just seen above, = 0*6.

Now, on comparing observation with theory, we see that the par-
thenogenetic mother appears to act like a mid-parent ; the coefficients
of correlation and regression being respectively 0*466 and 0*619.

Further, we know — P' °^ Pare— = -i- = 0*71, and in the present
S.D. of progeny ^2 r

S.D. of parthenogenetic mothers 2*22 A *K

case — - — ~ = ttxt — yj'iD.

b.D. oi progeny 2*95

Among my notes there are recorded the measurements of twenty-
six grandchildren, the offspring of seven grandparents. With these
the coefficients of correlation and regression were calculated. On
account of the altogether insufficient number of individuals, the results
were bound to be very uncertain, but they appear to favour the view
that inheritance in parthenogenetic generations resembles that from

* ' Eoy. Soc. Proc.,' vol. 62, pp. 386—412.

the coefficient of correlation of the mid-parent = 0*6( — — j = 0*424, and

158

Prof. H. Marshall Ward.

mid-grandparent to grandchildren. The coefficient of correlation was
0*272 ±0-12, and the coefficient of regression = 0*5 ±0*2, while, accord-
ing to theory, they should be 0*3 and 0*6 respectively.

The evidence of these measurements cannot be said to be conclusive,
and I am about to test the theory on some other parthenogenetic
animal. If, however, this kind of inheritance be found to hold at all
generally in parthenogenesis, it would be a fact of very considerable
significance, and might conceivably give some insight into the physio-
logical causes of heredity and variation.

" Onygena equina (Willd.) : a Horn-destroying Fungus." By H.
Marshall Ward, D.Sc, F.K.S., Professor of Botany in the
University of Cambridge. Eeceived April 6, — Eead May 4,
1899.

(Abstract.)

The genus Onygena comprises half a dozen species of fungi, all very
imperfectly known, remarkable for their growth on feathers, hair,
horn, hoofs, &c, on which their sporocarps appear as drum-stick
shaped bodies 5 — 10 mm. high. A cow's horn, thoroughly infested
with the mycelium of the present species, yielded material for the
investigation, and the author has not only verified what little was
known, but has been able to cultivate the fungus and trace its life-
history, neither of which had been done before, and to supply some
details of its action on the horn.

The principal new points concern the development of the sporo
phores, which arise as domed or club-shaped masses of hyphse and
stand up into the air covered with a glistening white powder. Closer
investigation shows this to consist of chlamydospores, formed at the
free ends of the up-growing hyphse. Their details of structure and
development are fully described, and their spore nature proved by
culture in hanging drops. The germination, growth into mycelia,
and peculiar biology of these hitherto unknown spores were followed
in detail, and in some cases new crops of chlamydospores obtained
direct in the cultures.

When the crop of chlamydospores on the outside of the young
sporophore is exhausted, the hyphse which bore the spores fuse to
form the peridium clothing the head of the sporocarp, and peculiar
changes begin in the internal hyphse below.

Minute tufts or knots of claw-like filaments spring from the hypha?
forming the main mass of the fungus, push their way in between the
latter, and so find room in the mesh-like cavities. Here the closely
segmented claws form asci — they are the ascogenous hyphse — and the

Onygena equina : a Horn- destroying Fungus. 159

details of development of the asci, their nucleated contents, and the
spores are determined. As the spores ripen, the asci, which are
extremely evanescent, disappear, and in the ripe sporocarp only spores
can be seen lying loose in the meshes of the gleba. The ascomycetous
character of the fungus is thus put beyond question, though the
peculiar behaviour of the developing ascogenous tufts at one time
rendered it questionable whether the older views as to the relationships
were not more probable.

No one had hitherto been able to trace the germination of these
ascospores — the only spores known previously — and De Bary ex-
pressly stated his failure to do it. The author finds that they require
digesting in gastric juice, and so in nature they have to pass through
the stomach of- the animal. By using artificial gastric juice, and
employing glue and other products of hydrolysis of horn, the details of
germination and growth into mycelia, capable of infecting horn, were
traced step by step under the microscope and fully described.

No trace of any morphological structure comparable to sexual organs
could be discovered, though many points suggest the alliance of this
fungus with Erysiphese and Truffles.

The author also found that similar digestion promotes the germina-
tion of the chlamydospores, and in both cases has not only traced the
germination step by step, but has made measurements of the growth
of the mycelium, induced the formation of chlamydospores on the
mycelium again, and by transferring vigorous young mycelia to thin
shavings of horn has observed the infection of the latter.

It thus becomes evident that the spores of Onygena pass through the
body of an animal in nature, and, as might be expected from this,
extract of the animal's dung affords a suitable food medinm to re-start
the growth on horn. Probably the cattle lick the Onygena spores from
their own or each other's hides, hoofs, horns, &c, and this may explain
why the fungus is so rarely observed on the living animal : it is recorded
from such in at least one case however.

Very little is known as to the constitution of horn, and some experi-
ments have been made to try to answer the question — what changes the
fungus brings about. The research has also obvious bearings on the
question of the decomposition of hair, horn, feathers, hoofs, &c, used
as manure in agriculture. Although a bacterial decomposition of hoof
substance is known to the author, special investigation of the question
showed that in the present case no symbiosis between bacteria and the
Onygena exists.

For the details as to the literature, the discussion as to the syste-
matic position of Onygena, the experimental cultures, growth measure-
ments, and the histology, the reader is referred to the full paper, which
is illustrated by plates and numerous drawings.

160 Development of Lepidosiren paradoxa, Fitz.

" The External Features in the Development of Lepidosiren
paradoxa, Fitz. By J. Graham Kerr. Communicated by
A. Sedgwick, F.E.S. Eeceived April 11, — Eead May 4,
1899.

(Abstract.)

The paper opens with a short account of the habits of Lepidosiren as
observed in the Gran Chaco. A description is then given of the
"external features in the development. The more important points in
this may be summarised as follows.

The egg is very large, 6 -5 — 7 mm. in diameter. It is surrounded
by a special capsule at first thick and almost jelly-like in appearance,
later on (after fertilisation) thin and horny. Outside this was found in
rare cases a thick jelly resembling that of the common frog's egg.
The egg is without a trace of dark pigment. Segmentation is com-
plete, resembling most nearly that of the egg of Amia, and leads to a
condition with an upper hemisphere of small cells with large segmenta-
tion cavity, and a lower of large yolk cells. Gastrulation begins with
the appearance of a row of depressions, or a continuous groove along
about one-third of the whole extent of the margin between small and
large cells. During its progress the small-celled portion spreads over
the lower yolk cells by the addition to its margin of small cells split
off from the yolk cells. As the groove referred to deepens into a slit
to form the archenteron, it becomes gradually shorter, and the
eventual complete blastopore is a crescentic slit only about a quarter of
the length of the original groove. The medullary folds soon appear
running forwards from the blastopore. There is no trace externally of
a blastoporic or protostomcd seam running along the back between the
medullary folds. The folds are low and inconspicuous, and they are
continued into one another behind the blastopore, which becomes the
anus. There are only slight traces of overarching of the medullary
folds to enclose a neural canal. During the later stages of intraoval
development, the posterior end of the body becomes much more con-
spicuously folded off the yolk than the head end. The Lepidosiren
hatches out as a tadpole-shaped larva, still completely devoid of dark
pigment. Just about the time of hatching the cloacal opening closes
temporarily. As the larva develops it becomes extraordinarily
amphibian-like. It possesses large pinnate external or somatic gills,
four on each side, corresponding to branchial arches I, II, III, and IV.
A large cement organ is also present, which during its early stages is
of the characteristic crescent shape so usual in the embryos of Anura.
Pigment begins to appear about ten days after hatching — first in the
retina, then ove^ the dorsal surface, especially anteriorly. The larval
condition lasts during the first six weeks after hatching. Towards the i
end of this period the cement organ undergoes atrophy. The somatic

Thermal Expansion of Pure Nickel and Cobalt. 161

gills atrophy later. During the process of their doing so, the Lepido-
siren passes through a condition in which the stumps persist evidently
corresponding to that well known in the young Protopterus, the group
of external gills with their common stalk having come by differential
development to be situated immediately above the fore limb. After
the close of the larval period the Lepidosirens become much darker in
colour and more lively in their movements. Young were obtained from
the nest up to a length of 60 mm. About this time the cornea begins
to assume the white unhealthy appearance that it has in the adult. In
the young of this size, small yellow spots appear, and in the young of
90 mm. these are conspicuous. Occasional yellow blotches persist in the
young Lepiclosiren of eighteen months, but in the adult they disappear.

The paper concludes with general remarks on the phenomena de-
scribed. The segmentation approaches most closely that of Ganoids.
The shortening up of the invaginating groove is considered to illus-
trate a process which has taken place in phylogeny in the passage from
the primitive holoblastic egg to the meroblastic condition. The con-
tinuity of the medullary folds behind the anus is adduced, together
with the evidence accumulating of the prolongation of the blastopore
along the floor of the medullary groove in other forms (Amphibia,
Ceratodus, e.g.) as affording potent evidence in favour of the hypothesis
which derives the Vertebrata from ancestral forms as primitive as the
Ccelenterata, and possessing a nelorigated mouth traversing the neural
surface. The occurrence of external gills in the young of three so
comparatively primitive groups of . Vertebrata as Crossopterygians,
Dipnoans, and Amphibians ; their occurrence on four branchial arches
in Lepidosiren, and on at least the hyoicl arch in Crossopterygians, and
the occurrence of a probable homologue on the mandibular arch in
Urodela, are taken as suggesting that these structures are organs of
great antiquity in the Vertebrate stem, and that there was formerly one
present on each visceral arch. It is pointed out that were this so, it
would afford a theory of the origin of the vertebrate limb, which
would be supported by much of the evidence brought forward by the
supporters of the Gegenbaur view, and which at the same time would
avoid the most important difficulties in the way of this view.

" The Thermal Expansion of Pure Mckel and Cobalt." By A. E.
Tutton, B.Sc. Communicated by Professor Tilden, D.Sc,
F.E.S. Eeceived April 18,— Bead May 4, 1899.

(Abstract.)

The author has carried out a series of re-determinations of the co-
efficients of thermal expansion of these two metals with the aid of the
interference clilatometer described in a former communication to the

VOL. LXV. N

162 Thermal Expansion of Pure Nickel and Cobalt.

Society.* Since the determinations made by Fizeau in the year 1869,
a large amount of additional knowledge has been accumulated with
reference to nickel and cobalt, including the discovery of the liquid
nickel carbonyl, which places processes of puiification in the hands of
the chemist of a character so superior to the older methods, as to
render it highly desirable that re-determinations of the physical con-
stants of these interesting elements should be carried out with speci-
mens of the metals thus purified. By the kindness of Professor
Tilden, who has prepared such specimens with infinite care for the pur-
poses of the investigation of other physical and chemical characters,
the author has been enabled to carry out determinations of the thermal
expansion with rectangular blocks varying in thickness from 8 to
13 mm. The blocks were furnished with parallel and truly plane
surfaces by the makers of the dilatometer, Messrs. Troughton and
Simms. The range of temperature of the observations was from
6° to 121°.

The results of the determinations of the coefficients of linear expan-
sion a are as follows : —

a = a + 2bt.

For nickel cc = O000 012 48 + 0-000 000 014 St.

For cobalt a = 0-000 012 08 + 0-000 000 012 8L

Nine different determinations were carried out for each metal, three
in each of the three rectangular directions, in order to eliminate any
slight error due to directional strain in the metallic blocks. As the
metals crystallise in the regular system, the expansion should be the
same in all directions. The metal in each case had solidified after
fusion in an oxy-hydrogen flame in presence at the last of excess of
oxygen. The individual results are highly concordant, the highest
result for cobalt being still lower than the lowest of the nine values
obtained for nickel. Hence there can be no doubt that the above
coefficients represent the true relationships.

The main result of the investigation may be summarised as fol-
lows : — The coefficients of linear expansion a of pure nickel and cobalt
exhibit a slight but real difference, the coefficient of nickel being
distinctly greater than that of cobalt. This is true with respect to
both the constant a, the coefficient for 0°, and the increment per
degree, 2b, of the general expression for the coefficient at any tempera-
ture t, cl = a + 2bt. The difference is consequently one which augments
with the temperature ; at 0° it amounts to 3*2 per cent., while at 120°,
the upper limit of the temperatures of the observations, it attains 4"5
per cent. Similar rules apply naturally to the cubical coefficients.
The metal possessing the slightly lower atomic weight, nickel, is thus
found to expand to a greater extent than the metal, cobalt, which is
endowed with the higher atomic weight.

* 'Phil. Trans.,' A, vol. 191, p. 313; 'Koy. Soc. Proc.,' vol. 63, p. 208.

On Nuclei in the Fertilised Embryo-sac of Lilium Martagon. 163

" On the Presence of two Vermiform Nuclei in the Fertilised
Embryo-sac of Lilium Martagon" By Ethel Saegant. Com-
municated by Dr. D. H. Scott, F.B.S. Beceived April 28, —
Eead May 4, 1899.

In a communication to the Russian Scientific Congress, which met
at Kieff, last summer, Professor S. Nawaschin summarised the brilliant
results of his recent work on the fertilised embryo-sac of Lilium Mar-
tagon and Fritillaria tenella (August 30, 1898). The report of this
paper, published in the 'Botanisches Centralblatt ' for January 4, 1899,
led Professor Leon Guignard to contribute a short account of his hitherto
unpublished researches on similar stages in the life-history of some
species of Lilium (L. martagon, L. pyrenaicum, and others) to the Aca-
demie des Sciences of Paris (April 4, 1899).

The results thus obtained independently by two distinguished
botanists are in perfect accord, and present the greatest theoretical
interest. They find that both the male generative nuclei on emerging
from the pollen tube are elongated in shape, and that each is more or
less twisted on its own axis. The nuclei, in fact, appear to have been
killed by the fixative in the act of spontaneous movement within the
embryo-sac. M. Guignard compares this motion to that of a non-
ciliated antherozoid.* The " vermiform " shape can be traced in the
male nucleus for some time after it has joined the nucleus of the
ovum.f

The most startling discovery, however, is that the second generative
nucleus unites with the upper polar nucleus of the embryo-sac, and
that both then fuse with the lower polar nucleus. Thus the definitive
nucleus of the embryo-sac, which later on gives rise by repeated divi-
sion to the endosperm nuclei, is formed by the coalescence of three
nuclei of very different origin. One is the sister-nucleus of the male
element in the fertilised ovum ; another, the sister-nucleus of the
female element ; and the third has all the characters of a vegetative
nucleus. Professors Nawaschin and Guignard are in complete agree-
ment as to these facts. M. Guignard adds that occasionally the polar
nuclei have united before the arrival of the " antherozoid," and gives
a number of figures in which the triple fusion is perfectly clear.

I am fortunate enough to possess a few preparations from the fer-
tilised embryo-sac of Lilium Martagon, which, so far as they go, com-
pletely confirm the results of Professors Nawaschin and Guignard.
The material was fixed in absolute alcohol for researches which were
never even begun, but I cut a few hand sections from it immediately

* Guignard, ' Comptes Rendus,' April 4, 1899, p. 3 of the separate copy,
f Guignard, loc. cit., p. 6 and figs. 3 — 5, 7 — 11.

164 Nuclei in the Fertilised Embryo-sac of Lilium Martagon.

after fixing, to make sure that it really contained fertilised embryo-sacs.
Fourteen sections were kept, all of them stained with methyl green
and acid fuchsin. As this, though a brilliant, is rather a diffuse stain,
I have lately re-stained eight of the preparations with Renault's hema-
toxylin and. eosin, which gives more precise results.

None of these preparations show the vermiform nuclei free in the
embryo-sac. In every case conjugation has already taken place ; the
male nucleus is applied to the female nucleus in the micropylar end of
the embryo-sac, and the second generative nucleus is applied to both
polar nuclei. In one case only the two polar nuclei are not in contact.
The much elongated " antherozoid " unites them like a bridge, one end
in contact with the lower, the other end coiled round the upper
nucleus (fig. 1, v.ri).

nvcropyle.

Excluding all doubtful cases, eight embryo-sacs show the male and
female nuclei not } et fused but in contact. In six of these the male
nucleus is more or less elongated. It may be distinctly coiled (fig. 1,-g.ri),
or merely horse-shoe or kidney-shaped, and commonly lies on the upper

Proceedings and List of Papers read.

165

or lower side of the much larger female nucleus. (See Guignard's
figs. 3, 4, 8, and 10.) In two cases the male nucleus is rounded, or
but slightly elongated.

Eight embryo-sacs show the polar nuclei near the centre. In five
cases the mass is clearly made up of three nuclei, and the generative
nucleus is distinguished from the other two by its irregular shape, the
differentiation of a slender chromatic ribbon, and by the absence of a
nucleolus. In three embryo-sacs two resting nuclei are applied to
each other near the centre.

The pollen tube is very clear in several preparations, and it com-
monly contains two small nuclei, stained green, and of irregular shape.
Since both generative nuclei are accounted for, these are probably due
to division of the vegetative nucleus.

May 18, 1899.

The LORD LISTER, F.R.C.S., D.C.L., President, in the Chair.

A List of the Presents received was laid on the table, and thanks
ordered for them.

The Bakerian Lecture, on " The Crystalline Structure of Metals,"
was delivered by Professor Ewing, F.R.S., and Mr. W. Rosenhain.

The following Papers were read : —

I. " The Yellow Colouring Matters accompanying Chlorophyll, and
their Spectroscopic Relations." By C. A. Schunck. Com-
municated by Dr. Schunck, F.R.S.

II. "The Diffusion of Ions into Gases." By J. S. Townsend. Com-
municated by Professor J. J. Thomson, F.R.S.

III. " The Diurnal Range of Rain at the seven Observatories in connec-
tion with the Meteorological Office, 1871—1890." By Dr.
R. H. Scott, F.R.S.

The Society adjourned over the Whitsuntide Recess to Thursday.
June 1.

VOT. LXV.

o

166 Prof. J. C. Bose. On a Self-recovering Coherer and the

" On a Self -recovering Coherer and the Study of the Cohering
Action of different Metals." By Jagadis Chunder Bose,
M.A., D.Sc., Professor of Physical Science, Presidency College,
Calcutta. Communicated by Lord Eayleigh, F.B.S. Be-
ceived March 6,— Eead April 27, 1899.

In working with coherers, made of iron or steel, some special diffi-
culties are encountered in the warm and damp climate of Bengal. The
surface of the metals soon gets oxidised, and this is attended with
variation of sensitiveness of coherer. The sensitiveness, it is true,
does not altogether disappear, but it undergoes a considerable diminu-
tion. The presence of excessive moisture in the atmosphere introduces
another difficulty. Substances to be experimented on become more or
less opaque by absorption of water vapour. As fairly dry weather
lasts in Bengal only for a few weeks in winter, the difficulties alluded
to above are for the greater part of the year serious drawbacks in
carrying out delicate experiments. To avoid as far as possible the
partial loss of sensibility of the receiver due to oxidation, I tried to
use metals less oxidisable than iron for the construction of the coherer.
In my earlier experiments I derived considerable advantage by coating
the steel spirals with deposits of various metals. Finding that the
sensitiveness depends on the coating metal and not on the substratum,
I used in my later experiments fine silver threads wound in narrow
spirals. They were then coated with cobalt in an electrolytic bath.
The coating of cobalt was at first apt to strip off, but with a suitable
modification of the electrolyte and a proper adjustment of the current,
a deposit was obtained which was very coherent. The contact surface
of cobalt was found to be highly sensitive to electric radiation, and the
surface is not liable to such chemical changes as are experienced in the
case of steel.

I next proceeded to make a systematic study of the action of dif-
ferent metals as regards their cohering properties. In a previous
paper* I enumerated the conditions which are favourable for making
the coherer sensitive to electric radiation. These are the proper
adjustment of the E.M.F. and pressure of contact suitable for each
particular receiver. The E.M.F. is adjusted by a potentiometer slide.
For very delicate adjustments of pressure I used in some of the fol-
lowing experiments an U-tube filled with mercury, with a plunger in
one of the limbs ; various substances were adjusted to touch barely the
mercury in the other limb. A thin rod, acting as a plunger, was made
to dip to a more or less extent in the mercury by a slide arrangement.
In this way the mercury displaced was made to make contact with the

# "On Polarisation of Electric Ray," 'Journal of Asiatic Society of Bengal,'
May, 1895.

Study of the Cohering Action of different Metals. 167

given metal with gradually increasing pressure, this increase of pres-
sure being capable of the finest adjustments. The circuit was com-
pleted through the metal and mercury. Sometimes the variation of
pressure was produced by a pressure bulb. In the arrangement
described above the contact is between different metals and mercury —
metals which were even amalgamated by mercury still exhibited
sensitiveness to electric radiation when the amalgamation did not pro-
ceed too far. In this way I was able to detect the cohering action of
many conductors, including carbon. For studying the contact-sensitive-
ness of similar metals I made an iron float on which was soldered a
split-tube in which the given metal could be fixed, a similar piece of
metal being adjusted above the float, so that by working the plunger
or the pressure bulb the two metals could be brought into contact with
graduated pressure. The other arrangements adopted were the contact
of spirals compressed by micrometer screw, and filings similarly com-
pressed between two electrodes.

With the arrangement described above the action of radiation on
metallic contacts was studied, a brief account of which will be given
under their respective groupings. It may here be mentioned that
certain metals which do not usually show any contact-sensitiveness can
be made to exhibit it by very careful manipulation. The nature of the
response of a coherer is to a certain extent modified by its condition
and particular adjustment. A coherer freshly made is more difficult to
adjust, but at the same time far more sensitive. The action is more
easily under control and more consistent after a few days' rest, but the
sensitiveness is not so abnormally great. The contacts of bright and
clear surfaces are difficult to adjust, but such contacts are more sensi-
tive than those made by tarnished surfaces. Pressure and E.M.F., as
previously stated, also modify the reaction. For example, a freshly
made and very delicately adjusted coherer subjected to slight pressure
and small E.M.F. showed an increase of resistance by the action of
radiation. The galvanometer spot, after a short interval, resumed its
former position, exhibiting a recovery from the effect of radiation.
The coherer continued to exhibit this effect for some time, then it
relapsed into the more stable condition in which a diminution of
resistance is produced by the action of radiation. Another coherer
was found apparently irresponsive to radiation, there being the merest
throb (sometimes even this was wanting) in the galvanometer spot,
when a flash of radiation fell on the receiver. Thinking that this
apparent immobility of the galvanometer spot may be due to response,
followed by instantaneous recovery, the galvanometer needle being
subjected to opposite impulses in rapid succession, I interposed a tele-
phone in the circuit ; each time a flash of radiation fell on the receiver
the telephone sounded, no tapping being necessary to restore the sensi-
tiveness. The recovery was here automatic and rapid. After twenty

O 2

168 Prof. J. C. Bose. On a Self-recovering Coherer and the

or thirty flashes, however, the receiver lost its power of automatic
recovery, and the sensitiveness had then to be restored by tapping.
An interesting observation was made to the effect that on the last occa-
sion the receiver responded without previous tapping, a rumbling noise
was heard in the telephone which lasted for a short time, evidently due
to the re-arrangement of the surface molecules to a more stable condi-
tion, after which the power of self-recovery was lost.

The state of sensibility described above is more or less transitory,
and is induced, generally speaking, by a somewhat unstable contact
and low E.M.F. acting in the circuit. In the majority of metals, the-
normal tendency is towards a diminution of contact resistance by the-
action of electric waves. The occasional increase of resistance, in
general, disappears when the pressure and E.M.F. are increased. But
in the case to be presently described we have an interesting exception,,
where the normal state of things is just the reverse of what prevails in.
the majority of metals.

Alkali Metals.

In the following investigations the radiator is a platinum sphere
9*7 mm. in diameter. The coherer was placed at a short distance, so
that the intensity of incident radiation was fairly strong.

Potassium. — In working with this metal, the exceptional nature of
the reaction became at once evident. The effect of radiation was to
produce an increase of resistance. The pressure of contact was adjusted
till a current flowed through the galvanometer, the galvanometer spot
of light being at one end of the scale. On subjecting the receiver to*
radiation the spot of light was deflected to the opposite end, exhibiting
a great increase of resistance. When the pressure and E.M.F. were
suitably adjusted a condition was soon attained, when a flash of
radiation made the spot of light swing energetically in one direction,
indicating an increase of resistance : the receiver, however, recovered
instantaneously with the cessation of radiation, and the spot violently
swung back to the opposite end, indicating the normal current that ^
flows in the circuit. This condition was found to persist, the receiver
uniformly responding with an increase of resistance followed by auto-
matic and instantaneous recovery. To prevent oxidation, the receiver
was kept immersed in kerosene. When the receiver was lifted from
the protecting bath, it still continued to respond with an increase of
resistance, but with a gradual loss of power of automatic recovery*
This power was again restored on again immersing the coherer in
kerosene. The receiver in vacuo, or under reduced hydrogen pressure,,
would have been preferred, had the necessary appliances been available.

Sodium. — As we pass from potassium to the neighbouring metals,
there is a gradual transition of property as regards the nature of
response to electric waves. With sodium the adjustment is a little

Study of the Cohering Action of different Metals.

169

more difficult than with potassium, but the response is somewhat
similar to that of potassium. Though in general there is an increase
of resistance produced by electric radiation, there are occasional excep-
tions when a diminution of resistance is produced. With some trouble
the adjustment could be made so that the recovery is also automatic,
but it is not so energetic as in the case of potassium.

Lithium. — Specimens of this metal not being available, I obtained a
deposit of it on iron electrodes by electrolysis of the fused chloride.
The action produced by electric radiation was sometimes an increase
and sometimes a diminution of resistance, the increase of resistance
being the more frequent. With some difficulty it was possible to
.adjust the sensitiveness so that the recovery was automatic, but it was
not energetic nor did this power persist for a long time.

Metals of the Alkaline Earth.

Pure metals of this group being not available, I had to rely on the
deposit obtained by electrolysis. Chloride of calcium was fused in a
crucible, and deposits were produced on iron cathodes, the anode being
a carbon rod. The deposit was not very even. One of the iron rods
with the deposit was tested by immersion under water, when hydrogen
was evolved. I did not succeed in getting deposits of either barium or
strontium, the temperature available not being sufficiently high.

On making a coherer with calcium, and keeping it immersed in
kerosene, an action similar to that produced by sodium was observed.
The tendency of self-recovery was, however, very slight.

Magnesium, Zinc, and Cadmium.

In these metals and in the succeeding groups there is a pronounced
tendency towards a diminution of resistance by the action of electric
radiation. Magnesium being easily oxidisable, there is a thin coating
of oxide on the surface. When this is scraped, the metal makes a very
highly sensitive receiver. The adjustment is not difficult, the metal
allowing a considerable latitude of pressure and E.M.F. It has
already been stated that the metals which are slightly tarnished can
be more easily adjusted.

Though there is in this metal a decided tendency towards a reduction
of contact resistance, yet it is possible by careful adjustment to obtain
an increase of resistance. Indeed it is sometimes possible to so adjust
matters that one flash of radiation produces a diminution of resistance,
and the very next flash an increase of resistance. Thus a series of
flashes may be made to produce alternate throws of the galvanometer
needle. The more stable adjustment, however, gives a diminution of
resistance, and receivers made with this metal could be made extremely
.sensitive. The tendency towards recovery is almost wanting.

170 Prof. J. C. Bose. On a Self-recovering Coherer and the

Zinc. — This metal also exhibits moderate sensitiveness j it, however,
requires a more careful adjustment.

Cadmium. — The action of this metal is somewhat similar to that of
zinc, but the sensitiveness is very much less.

Bismuth and Antimony.

Both bismuth and antimony make very sensitive receivers. Mode-
rately small E.M.F. with slight pressure is best suited for these metals.

Iron and the Allied Metals.

Iron. — The action of this metal is well known. In one of my experi-
ments I used it in connection with mercury. When the contact is
very lightly made, there is a tendency towards an increase of resistance
by the action of radiation. But after a time the action became normal^
that is to say, there was a diminution of resistance.

Nickel and Cobalt. — These are also very sensitive. The surface being
bright, the E.M.F. and pressure are to be adjusted with some care.

Manganese and Chromium. — These were obtained in the form of
powder. Their action is similar to the other metals of this group.

Aluminium. — This also makes a sensitive receiver.

Tin, Lead, and Thallium.

It is somewhat difficult to adjust tin, but when this is done the metal
exhibits fair sensitiveness. Lead is also sensitive. The sensitiveness,
of thallium is only moderate.

Molybdenum and Uranium.

The specimen obtained was in the form of powder, and very tar-
nished in appearance. The sensitiveness exhibited was slight.

Metals of the Platinum Group.

Platinum exhibited a moderate amount of sensitiveness. Spongy
platinum also showed the same action. The absorption of hydrogen
made the action slightly better, but the improvement was not very
marked.

Palladium. — This made a more sensitive coherer than platinum..
The adjustment is, however, more troublesome.

Osmium. — The specimen was in the form of powder. It requires a
higher E.M.F. to bring it to a sensitive condition. The sensitiveness
was moderate.

Rhodium was found to be more sensitive than osmium.

Study of the Cohering Action of different Metals. 171

Copper, Gold, and Silver.

Copper required a much smaller E.M.F. The sensitiveness was only
moderate.

Gold was more difficult to adjust, but the action is a little stronger.
Silver. — The receiver was extremely unstable. It exhibited some-
times a diminution and at other times an increase of resistance.

It will be seen from the above that all metals exhibit contact sensi-
tiveness to electric radiation, the general tendency being towards a
diminution of resistance. '

The most interesting and typically exceptional case, however, is the
receiver made with potassium, which not only exhibits an increase of
resistance by the action of radiation, but also a remarkable power of
self-recovery. In the accidental instances of increase of resistance
exhibited by other metals, an increase of pressure or E.M.F. generally
brought the coherer to the normal condition, which showed a diminu-
tion of contact resistance by the action of electric waves. With potas-
sium I gradually increased the pressure till the receiver grew insensi-
tive. All along it indicated an increase of resistance, even when one
piece was partially flattened against the other. I increased the E.M.F.
many times the normal value • this increase (till the limit of sensitive-
ness was reached) rather augmented the sensibility and power of auto-
matic recovery. I allowed the receiver a period of rest, the nature
of response remaining the same. As far as I have tried, potassium
receivers always gave an increase of resistance, a property which seems
to be characteristic of this metal, and to a less extent, of the allied
metals.

It will thus be seen that the action of potassium receiver is not,
strictly speaking, a cohering one. For it is difficult to see how a
cohering action and consequent better contact could produce an in-
crease of resistance. It may be thought that the sudden increase of
current may, by something like a Trevelyan rocker action, produce an
interruption of contact. But such a supposition does not explain the
instantaneous action, and the equally instantaneous recovery.

In arranging the metals according to their property of change of
contact resistance, I was struck by the similarity of action of electric
radiation on potassium in increasing the contact resistance, and the
checking action of visible radiation on the spark discharge. In the
latter case too potasium. is also photo-electrically the most sensitive.
But the action is confined to visible radiation, and is most efficient in
the ultra-violet region. I was indeed apprehensive that the action on
potassium receiver which I observed might be in some way due to the
ultra-violet radiation of the oscillatory spark. But this misgiving was
put to rest from the consideration that the receiver was placed in a

172

Prof. J. A. Ewing and Mr. W. Bosenhain.

glass vessel filled with kerosene, through which no ultra-violet light
could have been transmitted. To put the matter to final test, I lighted a
magnesium wire in close proximity to the receiver without producing
any effect. Thick blocks of wood of ebonite and of pitch were inter-
posed without checking the action. I then used polarised electric
radiation, and interposed a book analyzer, 6 cm. in thickness; when
the analyzer was held parallel, there was a vigorous action, but when
it was held in a crossed position all action was stopped. No visible or
heat radiation could have been transmitted through such a structure,
and there can be no doubt that the action was entirely due to electric
radiation.

It would be interesting to investigate whether the observed action
of electric radiation on a potassium receiver is in any way analogous
to the photo-electric action of visible light. I have commenced an in-
vestigation on this subject, the results of which I hope to communicate
on another occasion.

Bakerian Lecture. — " The Crystalline Structure of Metals." By
J. A. Ewing, F.B.S., Professor of Mechanism and Applied
Mechanics in the University of Cambridge, and W. Bosen-
hain, 1851 Exhibition Besearch Scholar, Melbourne Univer-
sity. Delivered May 18, 1899.

(Abstract.)

In a previous communication, read to the Society on March 16, a
preliminary account was given of some of the results the authors had
arrived at in studying metals by the microscopic methods initiated by
Sorby, and pursued by Andrews, Arnold, Behrens, Charpy, Osmond,
Boberts-Austen, Stead, and others. The present paper deals with a
development and extension of the same work. It relates chiefly,
though not exclusively, to the effects of strain, and the relation of
plasticity to crystalline structure.

It is well known that the etching of a polished surface of metal
reveals, in general, a structure consisting of irregularly shaped grains,
with clearly marked boundaries. Each grain is a crystal, the growth
of which has been arrested by its meeting with neighbouring grains.
This view, as Mr. Stead has pointed out, is strongly supported by the
appearance of the etched surface under oblique illumination, when the
several grains are seen to reflect light in a way which is consistent only
with the idea that on each there is a multitude of facets with a definite
orientation, constant over any one grain, but different from grain to
grain. The formation of such a structure is well exhibited, on a rela-
tively enormous scale on the inner surface of a cake of solidifying

The Crystalline Structure of Metals.

173

bismuth, from which the still molten metal has been poured away.
Another striking example of this structure is seen in steel containing
about 4J per cent, of silicon. The fractured ingot of this material
exhibits large crystals, and by deeply etching a polished surface
Mr. Stead has obtained a beautiful development of the regularly
oriented elements of which the crystalline grains are built up* on a
scale so large as to require but little magnification.

The authors have obtained much evidence that this structure is
typical of metals generally. Probably under no condition does any
metal cease to be crystalline.

The crystalline character of wrought-iron bars or plates is seen when
the polished surface is etched, not merely by the general appearance of
the grains under oblique light, but by the development of geometrical
pits on the surface. These pits have a definite orientation over each
grain, and the orientation changes from one grain to another. Usually
in the purest commercial iron their outline is that of plane sections of
.a cube, but occasionally they are apparently plane sections of an octa-
hedron. In some instances isolated and comparatively large pits only
.are seen ; in others nearly the whole surface of a grain, when viewed
under a magnification of 1000 or 2000 diameters, is found to be
covered with small as well as large pits, geometrically similar and simi-
larly oriented. Photographs of these are given in the paper.

For the purpose of producing smooth surfaces in the more fusible
metals, without polishing, the metal was poured in a molten state on a
plate of smooth glass. The surface produced in this way shows well
the boundaries between the grains, and in some cases it also exhibits
the crystalline character of the grains in a remarkable way by means
of geometrical pits, which are apparently formed on the surface in con-
sequence of the presence of small bubbles of air or, more probably, of
gas given out from the metal itself during solidification. Cadmium
shows these particularly well, and they are to be observed also in tin and
.zinc. These air-pits are seen, under 1000 diameters, to be negative
crystals, similar and similarly oriented on each grain, and, in cadmium,
to have outlines which suggest that they are sections of hexagonal
prisms. Their characteristics are exhibited in the photographs, which
also show how the boundaries between the grains are emphasised by
the collection there of air or of gas given off by the metal during
solidification. The true boundary is merely the trace of a surface on
a plane, but it may be broadened out in this way into a wide shallow
channel.

The effects of strain have been examined in many metals, using sur-
faces prepared either by polishing or by casting against a smooth plate.
When any metal is strained beyond its elastic limit in any way, the
surface of each crystalline grain becomes marked by one or more
* ' Journal of the Iron and Steel Institute,' 1898.

174

Prof. J. A. Ewing and Mr. W. Eosenhain.

systems of lines running in a generally straight and parallel fashion
over it. The direction of the parallel lines changes from grain to
grain. Thus these lines serve to mark out one grain from another in
a metal which, although polished before straining, has not been etched
to develop the boundaries. As straining proceeds, the lines become
more and more numerous and emphatic, and two, three, or four
systems appear on each grain.

The nature of these lines has been described in the authors' paper of
March 16. They are slips along cleavage or gliding planes in the
crystals. The effect of each slip is to develop a step on the polished
face. The short inclined surface forming this step looks black under
vertical illumination, but shines out brightly when oblique light of
a suitable incidence is used. These slip bands, as they were named in
the previous paper, are thus seen as narrow dark or bright bands,
accordingly to the nature of the lighting.

The authors have developed slip-bands in iron, copper, gold, silver,
platinum, lead, tin, bismuth, cadmium, aluminium, nickel, as well as
steel, brass, gun-metal, and various other alloys. So far as the observa-
tions go, they occur in all metals.

The slip-bands are in themselves an evidence of crystalline structure,
and, further, they show how such a structure is consistent with plas-
ticity, and how it persists after plastic strain has occurred. The
" flow " or non-elastic strain of a metal occurs through numerous finite
slips taking place on the cleavage or gliding surfaces in each of the
crystalline grains of which the metal is an aggregate. The elementary
pieces which slip on one another retain their primitive crystalline
character.

Further, if the movement of the pieces with respect to each other
in any one grain is a movement of translation only, their orientation
should remain uniform in each grain.

That this is actually the case is demonstrated by examining speci-
mens of metal which had been violently deformed without any subse-
quent annealing or heating. In metal that has been rolled or hammered
in the cold state, or deformed by tension or compression or strain of
any kind, however severely, the grains are still seen where a surface is
polished and etched. Their form is much changed by the strain which
the piece has undergone. But the fact that they have retained their
crystalline structure is demonstrated when, after polishing, the piece
is subjected to a slight additional strain of any kind, for the effect of
this additional strain is to develop slips of the same general character
as before. Further evidence to the same effect is given by the fact
that etching the polished surface of a very severely strained piece
develops geometrical pits, which are similar and similarly oriented
over the face of each grain, notwithstanding the great distortion which
the grain has suffered as a whole. The effects of oblique lighting in

The Crystalline Structure of Metals.

175

metal which is polished and etched after severe straining are referred
to as illustrating the same point. The persistence of crystalline struc-
ture is demonstrated by micro-photographs of the section of a bar of
Swedish iron which had been rolled cold from a diameter of f inch to
a diameter of J inch without subsequent heating. The outline of the
grains is much distorted, but the orientation of the crystalline ele-
ments remains constant within each individual grain.

The slips in metals which exhibit a cubical crystalline structure on
etching are in some instances parallel to the faces of the cubes, and are
very frequently inclined to the faces, apparently along the octahedral
planes. Stepped lines are frequently seen, and also lines which appear
curved probably in consequence of numerous steps which are unre-
solved even under the highest powers. In exceedingly plastic metals
such as lead, copper, and gold, the lines are particularly straight. A
piece of lead cast against glass to produce a smooth surface gives*
when slightly strained, a splendid display of slip-bands, and the
boundaries of the grains are sharply defined by the meeting of the
lines on one grain with those on its neighbours. Another way to get
a clear lead surface for the purpose of showing slip-bands is to press a
freshly cut piece of the metal with considerable force against a smooth
object. Photographs of slip-bands in iron, gold, silver, lead, copper,
and other metals are given in the paper.

"When a metal is fractured the grains do not as a rule part company
at their boundaries, but split along cleavage surfaces. It is to this that
the crystalline appearance, obvious in many fractures, is due.

In several metals the authors find that " twinning " takes place in
the crystalline structure as an effect of strain. Samples of copper,
which in the original cast state gave no evidence of the existence of
twin crystals, were hammered or otherwise wrought, and were then
found to be full of twins. The twinning produced in this way sur-
vived after the wrought copper had been raised to a red heat and
allowed to cool. Similar results were obtained in gold and in silver ;
the metal in the cast state did not show twins, but they were found
after the metal had been wrought and subsequently softened by
annealing. An example of twinning was observed in nickel after the
application of a somewhat severe strain. Twins were readily developed
in cadmium by strain, apparently as a result of the slight strain which
was applied for the purpose of developing slip bands. They were also
found in lead, zinc, and tin, either as a primitive feature in the crystal-
lisation or produced by straining. The twinning frequently takes the
form of a large number of parallel bands within a single grain, and a-
twin band due to strain in one grain is sometimes associated with a
twin band in neighbouring grains, the bands being continuous except
for a change in orientation in passing from grain to grain.

Photographs of twin bands in copper, gold, lead, and other metals are

176

The Crystalline Structure of Metals.

given, showing the twin bands as revealed by a cross-hatching of
parallel slip lines, the sets of lines being parallel to one another in
alternate bands of the twin. The twinning under strain which we
have observed in various metals is similar to that which is known to
occur in calcite. It may be regarded as a result of slip accompanied
by a definite and constant amount of rotation on the part of the
molecules.

From this point of view there are two modes in which plastic yield-
ing occurs in a crystalline aggregate. One is by simple slips, where
the movements of the crystalline elements are purely translatory and
their orientation is consequently preserved unchanged. The other is
by twinning, where rotation occurs through an angle which is the same
ior each molecule in the twinned group. Both modes are often found
not only in a single specimen of metal but in the same crystalline
grain.

At the suggestion of Messrs. Hey cock and Neville, the authors'
examination of the effects of strain has been extended to certain
•eutectic alloys. The structure of such alloys has already been
described by Osmond, with whose observations these are in agreement.
The alloy generally exhibits rather large grains, the structure of which
is very different from that of pure metals, for it consists of an intimate
intermixture of two constituents, one of which appears as separate or
dendritic crystals on a field formed of the other constituent. The two
.are seen forming an exceedingly minute and complex structure within
•each of the large grains of which the alloy is made up. Straining has
the effect of making this intimate structure more apparent, by causing
slips which set up differences of level between pieces of one and the
other constituent.

A study of the micro-structure of alloys suggests a possible explana-
tion of the peculiarities they present in regard to variation of electrical
conductivity with temperature. The two constituents may behave
individually as pure metals in this respect, but if their coefficients of
.expansion are different the closeness of the joints between them will *
depend on the temperature. Thus if the more expansible metal exists
as plates, or separate pieces of any form within the other, the effect of
heating will be to make the joints between the two conduct more
readily, with the result of reducing the increase of resistance to which
heating would otherwise give rise, and in extreme cases with the effect
even of producing a negative temperature coefficient. The high
resistance of alloys generally may be ascribed to the large number of
joints across which the current has to pass.

In casting metals against glass and other smooth bodies for the
purpose of getting a surface fit for microscopical examination, a surface
is occasionally produced which not only shows the true boundaries be-
tween the crystalline grains, but also additional markings which simulate

The Yellow Colouring Matters accompanying Chlorophyll. 177

boundaries in a very curious manner. These pseudo-boundaries are*
often polygonal in form, like the real boundaries, and have an intimate
geometrical association with them. Under low powers they are in
some instances difficult to distinguish from true boundaries ; but the
distinction is apparent under high powers, and it becomes obvious as
soon as slip-bands are developed by the straining of the metal. The
pseudo-boundaries are found to consist in small variations of level
in the surface of the grains in which they occur. Their form suggests that
they are projections upon the surface of real edges below. They occur
very conspicuously in cadmium, especially when it is cast on a cold
surface, and less conspicuously in zinc. It is probable that in the
strain set up by unequal cooling after the metal has solidified, the lower
edges of the crystalline grains project a sort of image of themselves on
the surface by slips, or possibly by narrow bands of twinning. The
effect resembles that of a Japanese " magic " mirror, in which slight
inequalities of the surface, corresponding to a pattern behind, cause'
light reflected from the mirror to produce an image in which a ghost of
the pattern may be traced.

The authors regard their experiments as establishing the conclusion
briefly stated in their previous paper, to the effect that the plasticity of
metals is due to the sliding over one another of the crystalline elements
composing each grain, without change in their orientation within each
grain, except in so far as such change may occur through twinning.

" The Yellow Colouring Matters accompanying Chlorophyll, and
their Spectroscopic Eelations." By C. A. Schunck. Com-
municated by Edward Schunck, F.E.S. Eeceived April 20,
— Eead May 18, 1899.

[Plate 6.]

The yellow colouring matters dealt with are those accompanying
chlorophyll in healthy green leaves and which are extracted along with
it by means of boiling alcohol.

This group of yellow colouring matters is generally known by the
name xanthophyll, a term first used by Berzelius, who was the first
observer to express the belief that a yellow colouring matter pre-exists
along with the green colouring matter in alcoholic extracts of green
leaves. The subject has subsequently received the attention of many
investigators — Fremy, Michels, Millardet, Miiller, Tinisnaseff, Ger-
land, Eaunenhoff, Askenasy, Stokes, Sorby, Tschirch, Kraus, Filhol,
Hansen, and Schunck. The principal results arrived at by these
investigators are as follows >. — Filhol noticed that by treating crude
alcoholic chlorophyll solutions with animal charcoal it is possible to

178 Mr. C. A. Schunck. The Yellow Colouring Matters

remove the green constituent of the mixture when a yellow coloured
solution remains, the colour of which he believes is evidently due to a
pre-existing colouring matter or matters associated with the green one.
Kraus — to whom we are indebted for a most elaborate study of the
physical properties of the yellow constituent of crude chlorophyll
solutions — confirmed the observations of Filhol, and added a number
•of new ones which lead, according to him, to an explanation of the
absorption spectrum of crude chlorophyll solutions which has hitherto
been universally accepted as the correct one. The author used
amongst other methods, for the purpose of separating the yellow
colouring matters from the green, their different solubility in alcohol
and benzol or, correctly speaking, benzoline. An alcoholic solution of
chlorophyll, treated with benzoline, retains, according to him, the yellow
colouring matter or mixture of colouring matters, while the benzoline
takes up the green constituent. By an investigation of the spectro-
scopic properties of these solutions, compared to the original one,
Kraus arrived at the result that the ordinary chlorophyll spectrum,
which has been described with considerable accuracy already by
Brewster, is a complex one, i.e., that some of the absorption bands are
due to the green constituent and some to the yellow. The former, he
says, is characterised by six bands, four of which (comprising the well-
known chlorophyll spectrum) are situated between the solar lines B
and E ■ the fifth between F and G, and the sixth in front of G. The
yellow constituent shows two bands, one at F or just behind it, and
the second in front of G. These observations, according to Kraus,
explain the constitution of the spectrum of crude chlorophyll solutions,
the first four bands of which being due solely to the green constituent,
the fifth to the yellow, and the sixth to a combination of the sixth
band of the green constituent and the second of the yellow. The fifth
band of the green constituent being very faint, and situated between
the fifth and sixth bands of the mixture, does not, according to him,
appear at all. These explanations, however, as will be shown, are
erroneous.

Sorby using carbon bisulphide as the separator in place of benzoline
states that along with chlorophyll in the crude alcoholic extracts of
the green leaves of the higher plants there are three accompanying
yellow colouring matters present which he names orange xanthophyll,
xanthophyll, and yellow xanthophyll, each showing a couple of bands
in slightly different positions in the more refrangible visible portion of
the spectrum, but none in the less refrangible part, and also that
there are other yellow colouring matters present, which he groups
under the term lichnoxanthine, which obscure the more refrangible
portion but exhibit no bands. He also states that chlorophyll (the
green constituent) of the higher plants is separable by the same means
into two colouring matters which he terms " Blue Chlorophyll " and

accompanying Chlorophyll, and their Spectroscopic Relations. 179

" Yellow Chlorophyll," the former being the chief constituent, the
latter being present in only a small relative quantity, and each give a
series of bands situated throughout the visible portion of the spectrum.

Hansen's method of isolating the yellow colouring matters is dif-
ferent from those of the previous observers. He treats the alcoholic
extracts of green leaves with caustic alkali, evaporates the liquor to
dryness, and extracts from the residue the yellow colouring matter by
means of ether, the study of which lead him to believe that the yellow
constituent shows only two bands.

Schunck obtains from all crude alcoholic chlorophyll extracts minute
sparkling red crystals which are deposited on standing, and to which
he has not applied a name, but which he considers identical with the
erythrophyll of Bougarel and the chrysophyll of Hartsen. On dilution
the yellow solutions of these crystals gave two absorption bands in the
more refrangible portion of the spectrum, but none in the less re-
frangible, and, though not in the same positions as the similar bands
(the fifth and sixth) shown by crude chlorophyll solutions, he considers
these latter bands not due to chlorophyll but to an accompanying-
yellow colouring matter.

Finally, Tschirch, who used Hansen's method for separating the
yellow from the green constituent, describes two yellow colouring
matters, to which he gives the names xantho-carotin, showing three
bands in the more refrangible part of the spectrum and to which,
according to him, the bands in the blue and violet shown by crude
chlorophyll solutions are due, and xanthophyll proper, which shows no
bands whatever but only a total obscuration in the violet region.

It will be seen that the results obtained by the various observers do
not agree, and a renewed study of the yellow constituent of crude
chlorophyll solutions appeared to be desirable. My own results differ in
many respects from the hitherto generally accepted ones ; they relate not
only to the physical nature of the yellow colouring matters in question,
but also enable us to characterise chlorophyll proper in a different
manner than was possible before. The preparation of pure chlorophyll
seems to baffle all attempts, but so far the physical properties of this
substance, the knowledge of which would guide an experimenter in
reaching the goal have, as it proves, not been known with sufficient
completeness.

I will now give the results of the experiments I have made, in the
endeavour to separate these yellow colouring matters from the accom-
panying chlorophyll, dealing more especially with their spectroscopic
relations as compared to those of chlorophyll in the violet and ultra-
violet region of the spectrum investigated by aid of photography — a
means which, with the exception of Tschirch, former observers have
not applied — and by which means I have been able to ascertain one or
two new facts, and have, I think, been able to clear up the much

180 Mr. C. A. Schunck. The Yellotu Colouring Matters

debated point whether the absorption bands in the violet and ultra-
violet region shown by crude chlorophyll solutions are due to chlo-
rophyll itself or to the accompanying yellow colouring matters.

The chlorophyll solutions experimented upon were obtained in the
usual manner by extracting the colouring matter from the leaves with
boiling alcohol. I have already shown* that chlorophyll solutions pre-
pared in this way show three characteristic absorption bands on proper
dilution, in the violet region of the spectrum, giving in the less
refrangible region the well-known spectrum of four bands which in
very pure solutions may be said to be reduced to three, so faint does
the fourth band appear.

If the extracts are concentrated enough one finds invariably on
standing for a day or so minute sparkling red crystals deposited on
the sides of the containing vessel or along with the fatty deposit,
coloured green by chlorophyll, which generally comes out of the extracts
on standing. These crystals are found in variable quantities, but
more often than not in a minute quantity.

This is the first yellow colouring matter one comes across and is the
erythrophyll of Bougarel, and the chrysophyll of Hartsen and
Schunck. f That chrysophyll is always to be found in chlorophyll
solutions proves that either it pre-exists as such along with chloro-
phyll in its alcoholic extracts, or that it is formed spontaneously
from one of the colouring matters, and is not, according to Hansen,:};
formed under certain conditions only by the decomposition of a deri-
vative. Chrysophyll thus obtained is not a very stable substance, and
in order to preserve it unchanged it should be placed in a glass tube
through which a current of hydrogen has been passed before sealing,
and kept in the dark. Its alcoholic solutions are bleached rapidly
when exposed to the air and sunlight, and even when kept in the
dark a change very soon takes place in its solutions as shown by its
spectrum, though there is no apparent change in colour. According
to Arnaud§ it is identical with carotin. Chrysophyll gives no absorp-
tion bands in the red, yellow, or green, but three very distinctive
bands in the violet region of the spectrum which, as I have shown, j|
are almost identical in position with those of carotin. They (Plate 6,
fig 5) occupy intermediate positions compared to the three bands
shown by crude chlorophyll solutions in the same region, being shifted
more towards the red end of the spectrum.

The method I have applied for separating the other accompanying
yellow colouring matters from the chlorophyll is that of treating the

* ' Roy. Soc. Proc.,' vol. 63, p. 393.

f ' Roy. Soc. Proc.,' rol. 4-4, p. 449.

% ' Die Farbstoffe des Chlorophylls.' 1889, p. 58.

§ ' Compt. Rend.,' vol. 102, p. 1119, and vol. 104, p. 1293.

|| 1 Roy. Soc. Proc.,' vol. 63, p. 393.

accompanying Chlorophyll, and their Spectroscopic Relations. 181

crude alcoholic extracts with an excess of animal charcoal in the cold
for about an hour, which removes all the chlorophyll and leaves a
yellow solution, which gives no absorption bands in the red, yellow, or
green, but four distinctive bands situated in the violet and ultra-violet
region of the spectrum. The prolonged action of animal charcoal has
the effect of ultimately absorbing all the colouring matters, leaving the
solution colourless. By this means the yellow colouring matters can
be obtained free from chlorophyll, but I have not been able to recover
the latter from the animal charcoal. This was tried by boiling the
charcoal with ether, the result being a greenish-yellow solution, giving
the four bands in the violet as before, but, in addition, a faint band in
the red, showing that it consisted of a portion of the yellow colouring
matters which had been absorbed by the charcoal together with a trace
of chlorophyll which caused the greenish colour, and the faint band in
the red region of the spectrum.

Yellow solutions obtained in this way from some chlorophyll
extracts — which, when freshly prepared, show signs of decomposition,
viz., the fourth band in the visible region of the spectrum darker
and the third fainter — show only the first two or three bands in the
violet region, the rest of the violet and ultra-violet being obscured.
In such cases a separation can be effected by ether, the yellow colouring
matter causing the obscuration remaining in the alcoholic portion, the
yellow ethereal portion now showing the four-banded spectrum as
before. This colouring matter causing obscuration no doubt belongs
to the lichnoxanthine group of Sorby,* and corresponds to the so-
called xanthophyll of Tschirch.f These yellow solutions deposit on
spontaneous evaporation an amorphous substance impregnated with
much fatty matter, which so far I have failed to remove and have been
unable to get in a crystalline form. It is insoluble in water, but easily
soluble in alcohol and ether, giving as before the distinctive spectrum
of four absorption bands in the violet and ultra-violet, but no bands in
the red, yellow, or green ; these bands, with the exception of the first,
which is almost if not quite identical in position with the first band
shown by crude chlorophyll solutions in the violet region, are in distinctly
different positions to the two remaining chlorophyll bands, or the three
due to chrysophyll (figs. 1 to 5). It appears to be much more stable
than chlorophyll, resisting the action of light and air to a greater
extent ; even from crude chlorophyll solutions which have been kept
for some time and show distinctly from their spectra the formation of
phyllocyanin, it can be obtained unaltered by the action of animal
charcoal. The solutions, when exposed to sunlight, gradually become
colourless, but at a less rapid rate than those of chrysophyll, and appa-
rently without the formation of products of decomposition) as is the
* ' Koy. Soc. Proc.,' vol". 21, p. 462.

f 'Ber. der Deutscli. Bot. Ges.,' vol. 14, part 2. p. 76. 1S96.
VOL. LXV. P

182 Mr. C. A. Schunck. The Yellow Colouring Matters

case with chlorophyll ; while away from the light it can be kept for a
considerable time without any apparent change taking place in its solu-
tions. Alkalis which induce so great a change in chlorophyll appear
to have no action upon it. From chlorophyll solutions which have
been boiled with potash or soda, and from which animal charcoal will
not now absorb any appreciable amount of colouring matter, ether
takes up this yellow colouring matter unaltered. Likewise, if its
alcoholic solutions be boiled with an alkali, no alteration is discernible.
On the other hand, if hydrochloric acid gas be passed through its
alcoholic solutions the colour changes to a dull dark-red, giving no
bands, but a general obscuration in the violet and ultra-violet region
of the spectrum. By this method I have obtained this yellow colouring
matter from two species of Ficus, from parsnip, clover, birch, and
Virginia creeper leaves, the extracts of the last-named, when even
freshly prepared, showing a near approach to the phyllocyanin spec-
trum, I have also examined the yellow colouring matter of autumnal
leaves, and find that it is identical with it both in properties and
spectrum (Plate 6, figs. 3 and 4) ; but in autumnal leaves invariably I
find the presence of the other yelloAv colouring matter which obscures
the spectrum, but which can be got rid of by separating with ether.
This supports the belief that the yellow colouring matter of autumnal
leaves is what remains after the chlorophyll of the healthy green leaf
has faded away, the latter being the less stable of the two (as has already
been shown to be the case), and fading first. I have examined the
colouring matter of some etiolated leaves, and here the presence of a
yellow colouring matter that obscures the spectrum in the violet and
ultra-violet region is undoubted. I endeavoured to get rid of it, as in
the former experiments, by separating with ether, but seemingly with
only partial success, the ethereal portion showing only three bands, but
in the same positions as the first three bands of the yellow colouring-
matter under review, the rest of the spectrum in the violet and ultra-
violet being obscured.

It is to this yellow colouring matter giving the characteristic spec-^
trum of four bands in the violet region, a spectrum not before observed,
I believe, and which I believe to be the predominating yellow colour-
ing matter accompanying chlorophyll in ordinary green leaves, and
also of faded autumnal leaves, that I would restrict the name xantho-
phyll, just in the same way that phylloxanthin, from being first applied
by Fremy to include all the yellow colouring matters accompanying
chlorophyll, as well as a yellowish-brown decomposition product of the
latter, has been applied by Schunck in a stricter sense to one of the
group only — the yellowish-brown decomposition product of the action
of acids. Crude alcoholic chlorophyll solutions can, as before stated,
be separated into a green and yellow portion by agitating with carbon
bisulphide — Sorby's method — or by benzoline, Kraus's method, the

accompanying Chlorophyll, and their Spectroscopic Relations. 183

carbon bisulphide and benzoline portions being coloured green, and
the alcoholic portion in each case yellow. I believe from the few ex-
periments I have so far made by these methods of separation, that the
xanthophyll, as I have defined it, passes along with the chlorophyll
into the carbon bisulphide or the benzoline, a point overlooked by
Kraus, but noticed by Sorby, while the alcoholic portion contains yet
another yellow colouring matter or matters, showing ill-defined absorp-
tion bands in the violet region, but in different positions again to either
the bands of chrysophyll or xanthophyll. May be we have here the
xanthophyll and yellow xanthophyll of Sorby, * whilst probably my
xanthophyll corresponds to his orange xanthophyll.

I hope after some further experiments by these means of separation
which I am now undertaking, to be able to throw some further light
upon the apparent complex nature of this group of accompanying
yellow colouring matters.

The Spectroscopic- Relations of Chlorophyll, Chrysophyll, and Xanthophyll.

The method of observing the absorption spectra by means of
photography is the same as I adopted in a former investigation, deal-
ing with chlorophyll and its derivatives,! quartz lenses and an Iceland
spar prism being used. The question whether the bands shown by
crude chlorophyll solutions in the violet region are due to chlorophyll
itself, or to an accompanying yellow colouring matter, has not so far
been answered in a satisfactory manner. Some observers consider them
due to the former, whilst others, the majority I believe, attribute them
to the latter.

On inspection of the plate (figs. 1, 2, and 5), it will be seen that the
chrysophyll bands are shifted towards the red end of the spectrum
compared to those of crude chlorophyll, so that if chrysophyll pre-
exists along with chlorophyll in its alcoholic extracts, the bands of the
two spectra would overlap, so as to produce one broad band extending
from F to K|3 ; but in all the freshly prepared normal crude chlorophyll
extracts I have examined, I have never found this to be the case, the
three bands always being visible on proper dilution, with no indication
of those due to chrysophyll.

We can therefore conclude that they are not due to chrysophyll, and
must assume that if chrysophyll pre-exists, its relative quantity com-
pared to chlorophyll must be small. In the xanthophyll spectrum, it
will be noticed (Plate 6, fig. 3) the first three bands do not coincide
with the chrysophyll bands, being shifted towards the ultra-violet, and
the spectrum is further distinguished by a fourth band situated between
K|3 and L, which is lacking in chrysophyll, the latter having the cha-

* ' Roy. Soc. Proc.,' vol. 21, p. 456.
f 'Roy. Soc. Proc.,' vol. 63, p. 391.

P 2

184 Mr. C. A. Schunck. The Yellow Colouring Matters

racteristic of allowing considerably more of the ultra-violet rays to pass
through its solutions. Comparing now the xanthophyll spectrum with
that given by crude chlorophyll solutions, it will at once be seen that
the character of the two spectra are quite different, and this points to
the fact that the bands are due to two distinct colouring matters.
This is supported by the fact that from crude alcoholic chlorophyll
solutions that have been acted upon with alkali, the xanthophyll can
be recovered unaltered by shaking up with ether, the chlorophyll being
so altered (though the colour of the solutions is unchanged) that animal
charcoal will not now absorb the green colouring matter ; this altera-
tion has been shown by Schunck* to be due to the formation of an
alkali compound of alkachlorophyll. Again, in crude chlorophyll solu-
tions which have been kept a little time, and which show signs of
decomposition, owing to the formation of phylloxanthin and phyllo-
cyanin, the bands in the violet are no longer discernible on proper
dilution, yet act upon such solutions with animal charcoal, and we
get a yellow solution giving the four-banded xanthophyll spectrum,
showing that the xanthophyll was there all the time unaltered, but did
not affect the spectrum. From the spectra of their alcoholic solutions
it will be observed the only xanthophyll band that could affect the
crude chlorophyll spectrum is the first, for it is the only one in a
similar position ; but I think I can show that this cannot be the case,
and that it is in no way connected with chlorophyll. If a crude chloro-
phyll solution be examined in ether compared to alcohol, it will be
found that in the former solvent, the bands are all shifted towards the
more refrangible end of the spectrum, while the xanthophyll bands in
both solvents are in identical positions (Plate 6, figs. 6 to 11), so that
we see now the first xanthophyll band no longer coincides with the first
band of crude chlorophyll, and therefore can not have any influence
upon the chlorophyll spectrum. And thus I came to the conclusion,
taking one experiment with another, that as with chrysophyll, xantho-
phyll is present in such a small relative quantity, compared to chlo-
rophyll, that the bands of its spectrum cannot be detected in the cruder
chlorophyll solution, and that the bands shown by the latter are due
to chlorophyll itself, and not to any of the accompanying yellow
colouring matters which the majority of former observers believed
they were due to. This belief may have arisen from the fact that
without the aid of photography, only the first two bands of chlorophyll
and the accompanying yellow colouring matters in the violet region
are discernible in alcoholic solutions to the eye, and they thus failed to
observe the complete spectrum which, from an inspection of the plate,
will be seen makes in each case a vast difference to the character of the
spectrum.

That all freshly prepared chlorophyll extracts show in every case
* ' Boy. Soc. Proc.,' vol. 50, p. 312.

Schunck.

Roy, Soc. Proc, Vol. 65, Plate 6.

accompanying Chlorophyll, and their Spectroscopic Relations. 185

the three bands in the violet region, is I think a proof in itself that
they are clue to the same colouring matter that produces the character-
istic and unmistakable spectrum in the less refrangible region, or to a
very intimately connected colouring matter so far unknown. Seeing
also that all the chlorophyll derivatives give a characteristic absorption
in the violet and ultra-violet region, it would be strange that
chlorophyll proper should prove the exception, and it is an interesting
fact that its two chief, and intimately connected decomposition pro-
ducts inter se — phyllocyanin and phylloxanthin — should have, as I
have shown,* bands in the violet region in identical positions with
these three chlorophyll bands, the first of them corresponding to a
faint, but distinct one, observable in pure phyllocyanin solutions on
proper dilution, the other two to the two characteristic phylloxanthin
bands in the violet region.

Summary.

1. I find in all crude alcoholic extracts of healthy green leaves, two
yellow colouring matters accompanying the chlorophyll. One chryso-
phyll which deposits out of the extracts on standing in lustrous red
crystals, but more often than not in minute quantities, the other
obtained by treating the extracts with animal charcoal in the cold, the
charcoal taking up the chlorophyll, and leaving a yellow solution
which deposits on spontaneous evaporation an amorphous substance
impregnated with much fatty matter, and to which I have restricted
the name xanthophyll. Another yellow colouring matter is sometimes
found along with xanthophyll which gives no absorption bands, but
only an obscuration in the violet and ultra-violet region of the
spectrum, and in such cases a separation can be effected by ether.
There is also evidence to show that other yellow colouring matters
may exist. Xanthophyll, however, I believe, is the predominating
yellow colouring matter accompanying chlorophyll in the healthy
green leaf, and I also find it to be identical with the principal yellow
colouring matter of the faded autumnal leaves.

2. Chrysophyll and xanthophyll each give a characteristic absorp-
tion spectrum in the violet and ultra-violet region • the former consists
of three bands, the latter of four, but in slightly different positions.
Crude chlorophyll solutions also give, in addition to the characteristic
spectrum of four bands in the less refrangible region ; three characteristic
bands in the violet, and from observations by means of photography
I come to the conclusion that these bands are due to chlorophyll
itself, and not to any of the accompanying yellow colouring matters
which the majority of former observers believed they were due to. I
also find that phyllocyanin and phylloxanthin have bands in identical
positions with these three chlorophyll bands.

* ( Eoy. Soc. Proc.,' vol. 63, p. 393.

186

Sir Norman Lockyer.

EXPLANATION OF PLATE.

A and B. Spectrum of a hydrogen vacuum-tube from which the reference lines,
F, G-', h, and Hx are obtained, with the potassium Kp line thrown in. The
solar lines, L, M, N, and P are obtained by measurement from a negative of
the solar spectrum.

Fig. 1. Crude chlorophyll in alcohol.
„ 2. „ ,, ,, diluted.

3. Xanthophyll from crude chlorophyll solutions in alcohol by the action of
animal charcoal.

4. Xanthophyll from faded autumnal yellow leaves in alcohol.
,, 5. Chrysophyll in alcohol.

,, 6. Crude chlorophyll in alcohol.

„ 7. „ „ ether.

„ 8. ,, „ alcohol diluted.

„ 9. „ » ether „

;, 10. Xanthophyll in alcohol.

„ 11. „ ether.

" On the Chemical Classification of the Stars." By Sir Norman
Lockyer, K.C.B., F.K.S. Eeceived April 27 —Bead May 4,
1899.

[Plate 7.]

In the attempts made to classify the stars by means of their spectra,
from Bntherford's time to quite recently, the various criteria selected
were necessarily for the most part of unknown origin ; with the excep-
tion of hydrogen, calcium, iron, and carbon, in the main chemical
origins could not be assigned with certainty to the spectral lines.
Hence the various groups defined by the behaviour of unknown lines
were referred to by numbers, and as the views of those employed in
the work of classifying differed widely as to the sequence of the
phenomena observed, the numerical sequences vary very considerably so
that any co-ordination becomes difficult and confusing.

Recent work has thrown such a flood of light on the chemistry of *
the stars that most definite chemical groupings can now be established,
and the object of the present communication is to suggest a general
scheme of classification in which they are employed, in relation to the
line of cosmical evolution which I have developed in former papers
communicated to the Society.

The fact that most of the important lines in the photographic region
of the stellar spectra have now been traced to their origins renders
this step desirable, although many of the chemical elements still remain
to be completely investigated from the stellar point of view.

The scheme is based upon a minute inquiry into the varying inten-
sities, in the different stars, of the lines and flutings of the under-
mentioned substances : —

On the Chemical Classification of the Stars.

187

Certain unknown elements (probably gaseous, unless their lines
represent " principal series ") in the hottest stars, and the new form of
hydrogen discovered by Professor Pickering (which I term " proto-
hydrogen " for the sake of clearness), hydrogen, helium, asterium,
calcium, magnesium, oxygen, nitrogen, carbon, silicium, iron, titanium,
copper, manganese, nickel, chromium, vanadium, strontium ; the
spectra being observed at the highest available spark temperatures.
The lines thus observed I term "enhanced" lines, and I distinguished the
kind of vapour which produces them by the affix " proto," e.g., proto-
magnesium, for the sake of clearness.*

Iron, calcium, and manganese at arc temperatures.

Carbon (flutings) at arc temperatures.

Maganese and iron (flutings) at a still lower temperature.

In a communication to the Society! I stated the results arrived at
recently with regard to the appearances of the lines of the above sub-
stances in stars of different temperatures, and the definition of the
different groups -or genera to be subsequently given are based upon
the map which accompanied the paper, together with more minute
inquiries on certain additional points, the examination into which was
suggested as the work went on.

So far as the inquiry has at present gone, the various most salient
differences to be taken advantage of for grouping purposes are repre-
sented in the following stars, the information being derived from the
researches of Professor Pickering J and Mr. McClean,§ as well as from
the Kensington series of photographs.

Hotted Stars.

Two stars in the constellation Argo (£ Puppis and y Argus ||).

Alnitam (e Orionis). This is a star in the belt of Orion shown on
maps as Alnilam. Dr. Budge has been good enough to make inquiries
for me, which show the change of word to have been brought about by
a transcriber's error, and that the meaning of the Arabic word is "a
belt of spheres or pearls."

* 'Boy. Soc. Proc.,' vol. 64, p. 398.

f ' Eoy. Soc. Proc.,' vol. 64, p. 396.

X ' Astro-phys. Journ.,' toI. 5, [). 92, 1897.

§ ' Spectra of Southern Star?.'

|| The spectrum of this star contains bright lines, but I show in a paper nearly
ready for communication to the Society, that when these occur with dark lines,
the latter alone have to be considered lor purposes of chemical classification.

188

Sir Norman Lockyer.

Stars of intermediate Temperature.

Ascending Series.

/3 Crucis.
£ Tauri.
Kigel.
a Cygni.

[ ]
Polaris.

Aldebaran.

Descending Series.

Achernar.

Algol.

Markab.

[ ]

Sirius.

Procyon.

Arcturus.

Stars of lowest Temperature.

Ascending Series.

Catalogue of

Antares, one of the brightest
stars in Duner's
Class Ilia*

[Nebulae.

Descending Series.

19 Piscium, one of the brightest
stars in Dnner's Catalogue of
Class III5.

[Dark Stars.]

In order to make quite clear that both an ascending and a descend-
ing series must be taken into account, I give herewith (Plate 7) two
photographs showing the phenomena observed on both sides of the
temperature curve in reversing layers of stars of nearly equal mean
temperatures, as determined by the enhanced lines. The stars in
question are : —

Sirius (descending). \
ling). J

a Cygni (ascending).

Procyon (descending
y Cygni (ascending).

The main differences to which I wish to draw attention are the very
different intensities of the hydrogen lines in Sirius and a Cygni, and the
difference in the width and intensities of the proto-metallic and metallic
lines in Procyon and y Cygni. These differences, so significant from a
classification point of view, were first indicated in a communication to
the Society in 18871, and the progress of the work on these lines has
shown how important they are. I have based the group — or generic —
words upon the following considerations.

As we now know beyond all question that a series of geological
strata from the most ancient to the most recent brings us in presence
of different organic forms, of which the most recent are the most com-

* 1 Sur les Etoiles a spectres de la troisieme classe.'
f ' Eoy. Soe. Proc.,5 vol. 43, p. 145.

On the Chemical Classification of the Stars.

189

plex, it is natural to suppose that the many sharp changes of spectra
observed in a series of stars from the highest temperature to the lowest,
bring us in presence of a series of chemical forms which become more
complex as the temperature is reduced. Hence we can in the stars
study the actual facts relating to the workings of inorganic evolution
on lines parallel to those which have already been made available in
the case of organic evolution.

If then we regard the typical stars as the equivalents of the typical
strata, such as the Cambrian, Silurian, &c, it is convenient that the
form of the words used to define them should be common to both;
hence I suggest an adjectival form ending in fan. If the typical star is
the brightest in a constellation, I use its Arabic name as root ; if the
typical star is not the brightest, I use the name of the constellation.

The desideratum referred has to a certain extent determined the
choice of stars where many were available. I have to express my
great obligations to Dr. Murray for help generously afforded in the
consideration of some of the questions thus raised. The table runs as
follows : —

Classification of Stars into Genera depending ufon their
Chemistry and Temperature.

Highest temperature, simplest chemistry.

Argonian.
Alnitamian.

Crucian.

Taurian.
^ Eigelian.
!ss Cygnian.

§ Polarian.
Aldebarian.
Antarian.

Achernian.

Algolian.

Markabian.

«5i

Sirian.
Procyonian.
Arcturian.
Piscian.

The chemical definitions of the various groups or genera are as
follows : —

Definitions of Stellar Genera.

Argonian.

Predominant. — Hydrogen and proto-hydrogen.
Fainter. — Helium, unknown gas (\ 4451, 4457), proto-niagne-
sium, proto-calcium, asterium. .

Alnitamian.

Predominant. — Hydrogen, helium, unknown gases (A 4089'2,
4116-0, 4649-2).

Painter. — Asterium, proto-hydrogen, proto-magnesium, proto-
calcium, oxygen, nitrogen, carbon.

190

Sir Norman Lockyer.

Crucian.

Predominant. — Hydrogen, helium, as-
terium, oxygen, nitrogen, carbon.

Fainter. — Proto -magnesium, proto-
calcium, unknown gas (A 4089'2), sili-
cium.

f Taurian.

Predominant. — Hydrogen, he-
lium, proto-magnesium, asterium.

Fainter. — Proto-calcium, sili-
cium, nitrogen, carbon, oxygen,
proto-iron, proto-titanium.

Rigelian.

Predominant. — Hydrogen, proto-
calcium, proto-magnesium, helium,
silicium.

Fainter. — Asterium, proto-iron}
nitrogen, carbon, proto-titanium.

Predominant. — Hydrogen, proto-
calcium, proto-magnesium, proto-
iron, silicium, proto-titanium,
proto-copper, proto- chromium.

Fainter. — Proto-nickel, proto-
vanadium, proto-manganese, proto -
strontium, iron (arc).

Polarian.

Predominant. — Proto-calcium,
proto-titanium, hydrogen, proto-
magnesium, proto-iron, and arc
lines of calcium, iron, and manga-
nese.

Fainter. — The other proto-metals
and metals occurring in the Sirian
, genus.

Achernian.
Same as Crucian.

Algolian.

Predominant. — Hydrogen, proto-
magnesium, proto-calcium, helium,
silicium.

Fainter. — Proto-iron, asterium,
carbon, proto-titanium, proto-cop-
per, proto-manganese, proto-nickel.

Marlcabian.

Predominant. — Hydrogen, proto-
calcium, proto-magnesium, sili-
cium.

Fainter. — Proto-iron, helium,
asterium, proto-titanium, proto-
copper, proto-manganese, proto-
nickel, proto-chromium.

Sirian.

Predominant. — Hydrogen, proto-
calcium, proto-magnesium, proto-
iron, silicium.

Fainter. — The lines of the other
proto-metals and the arc lines of
iron, calcium, and manganese.

Proeyonian.
Same as Polarian.

Lockyer.

Fig. 1.

Roy. Soc. Proc, Vol. 65, Plate 7
Fig. 2.

On the Chemical Classification of the Stars.

Aldebarian.

Predominant. — Proto-calcium, arc
lines of iron, calcium, and manganese,
proto-strontiuni, hydrogen.

Fainter. — Proto-iron and proto-tita-
nium.

Antarian.
Predominant. — Fhitings of manga-
nese.

Fainter. — Arc lines of metallic ele-
ments.

191

Arcturian.
Same as Aldebarian.

Piscian. -

Predominant. — Mutings of carbon.
Fainter. — Arc lines of metallic ele-
ments.

We may take for granted that as time goes on new intermediate
genera will require to be established ; the proposed classification lends
itself conveniently to this, as there are no numerical relations to be
disturbed.

A still more general chemical classification is the following, it being
understood that in it only the most predominant chemical features are
considered, and that there is no sharp line of separation between these
larger groups. The peculiar position of calcium and magnesium renders
this caveat the more necessary.

Classification of Stars.
Highest temperature.

f Proto-hydrogen stars
Gaseous stars <

Cleveite-gas stars ^ Saurian
Eigelian.

Proto-metallic stars <l Cygnian.

f Argonian.
\ Alnitamian.

Metallic stars

{Polar ian.
Aldebarian.
Stars with fluted spectra Antarian.

Lowest temperature.

Achernian.
Algolian.
Markabian.

Sirian.
Procyonian,
Arcturian.
Piscian.

The detailed chemical facts to be gathered from the definitions of
the several genera indicate many important differences between the
order of appearance of the chemical substances in the atmospheres of the
stars and that suggested by the hypothetical " periodic law." Special
investigations are in progress by which it is hoped some light may be
thrown on this and other points of a like nature.

192

Mr, J. S. Townsend.

" The Diffusion of Ions into Gases." By John S. Townsend, M.A.
(Dublin), Clerk Maxwell Student, Cavendish Laboratory,
Cambridge. Communicated by Professor J. J. Thomson,
F.B.S. Beceived April 25,— Bead May 18, 1899.

(Abstract.)

In the paper upon this subject, the principles upon which the theory
of interdiffusion of gases depends, are applied to the diffusion of ions
produced by Bontgen rays. When a gas has been left to itself,
the conductivity gradually disappears. When no electromotive forces
are acting, the loss of conductivity is due partly to positive and
negative ions coming into contact with each other, and partly to the
effect of the surface of the vessel, which discharges those ions which
come into contact with it.

In order to illustrate in a simple way the principles which are
involved, we take the case of a gas contained in a metal sphere, and
consider what happens to the ions after the gas has been removed from
the influence of the rays. For present purposes we may neglect the
effect of recombination.

The ions may be considered as constituting a separate gas, the
molecules of which may be either bigger or smaller than the molecules
of the gas in which they are immersed. When an ion comes into contact
with the surface of the sphere it loses its charge, so that the metal
may be regarded as a body which completely absorbs the ions. The
reduction in the conductivity by the diffusion of the ions to the sides,
is exactly analogous to the removal of moisture from a gas by
bubbling it through sulphuric acid. The more rapidly the water
vapour diffuses through the gas, the greater will be the number of
water molecules which come into contact with the acid round the
bubble. If the quantity of moisture which is removed be found ex-
perimentally, the coefficient of diffusion of water vapour into the gas ^
can be deduced.* It would be impracticable to use this method to
find the coefficient of diffusion of ions into a gas contained in a large
vessel, as the loss of conductivity due to recombination would be large
compared with the loss due to the sides.

The method which was employed was to pass a uniform stream of
gas through fine metal tubing, and to allow the rays to fall on the gas
immediately before entering the tubing. The bore of the tubing can
be so adjusted that the number of ions which come into contact with
the sides will be large compared wifh the number which recombine.
It is convenient to use tubing of such a length, that the conductivity
will be reduced to about one half its initial value.

* John S. Townsencl, ' Ph.il. Mag.,' June, 1898.

The Diffusion of Ions into Gases.

193

In order to obtain the coefficient of diffusion when the reduction in
the conductivity has been found experimentally, the following problem
presents itself : —

If a small quantity of a gas, A, is mixed with another gas, B, and
the mixture passed along a tube, the sides of which completely absorb
A, to find what quantity of A emerges from the tube with B.

It will be immediately seen that if the gases diffuse rapidly into
each other, a large proportion of the molecules of the gas A will come
into contact with the surface of the tube and will there be absorbed.
If on the other hand the rate of interdiffusion is very small, the
molecules of A will travel down the tube in straight lines parallel to
the axis of the tube, and practically none of them will come into con-
tact with the surface.

The complete solution of the above problem may be obtained from
the following equations : —

1 dp „

• = ~Tx+nXe'

1 dp TT

Kr dy

1 dp „ 1 ,,r

-pw = - -^ + n/je + -j)\\ ,

and the equation of continuity (pu) + ^- (pv) + ~ (pw) =0; where

n is the number of ions per cubic centimetre ; p their partial pressure ;
e the charge on each ion ; X, Y, and Z the electric forces at any point ;
u, v, and w the velocities of the ions ; W the velocity of the gas B
through the tube ; k the coefficient of diffusion of the ions into the gas B.

The partial differential coefficient with respect to the time, is omitted
from the equation of continuity, as we need only consider the steady
state.

The term dp/dz may be omitted from the third equation, as it is small
compared with the other terms.
2Y

W = —2(a2 - r2), where V is the mean velocity of the gas B,

defined by the condition 7ra2V/ = total volume of gas crossing any section
in a time t, a the radius of the tube, and r the perpendicular distance
of any point from the axis.

The forces X, Y, Z vanish since the electrification is too small to con-
tribute appreciably to the motion of the ions.

The boundary conditions are : —

p = 0 when r = a,

p = constant when z — 0.

194

Mr. J. S. Townsend.

The solution of the problem requires a somewhat lengthy analysis,
so that we give here only the final result.

The ratio of the number of ions (or molecules of the gas A)
coming out of the tube with B to the number which enter is

7-313ks 44-5kz

R = 4 [0-1952 € 2"2v + 0-0243 e + &c],

where z is the length of the tube.

The other terms of the series are too small to be taken into con
sideration.

Having determined the reduction in conductivity, due to tubes of
different lengths, the following values of the coefficients of diffusion
of ions into air, oxygen, carbonic acid, and hydrogen, were obtained.

Table of Coefficients of Diffusion of Ions in dry Gases.

Gas.

k for + ions.

k for — ions.

Mean value
of K.

Eatio of the
values of k.

Oxygen. .. .

0 -0274
0-025
0-023
0-123

0-042
0 -0396
0-026
0-190

0 '0347
0 -0323
0 -0245
0-156

1-54
1-58
1-13
1-54

Table of Coefficients of Diffusion of Ions in moist Gases.

Gas.

k for + ions.

k for — ions.

Mean value

Of K.

Ratio of the
values of k.

0-032

0-035

0 -0335

1-09

0-0288

0-0358

0 '0323

1-24

0*0245*

0 -0255

0-025

1-04

0-128 "

0-142

0 -1350

1 11

We should expect from the experiments, that the above numbers
were correct to 5 per cent.

Considering one of the equations of motion

1 dp „

we see that when clpjdx — 0, the velocity u due to the electric force
X is nKeK/p. If the potential gradient is one volt per centimetre X =
1/300 in electrostatic units, and the corresponding value of u is

Ke
300

The Diffusion of Ions into Gases. 195

Let N be the number of molecules in a cubic centimetre of a gas at
pressure P, equal to the atmospheric pressure, and temperature 15°
centigrade, the temperature at which u and k were determined.

The quotient N/P may be substituted for n/p in the above equa-
tion, and since the atmospheric pressure P is 106 in C.G.S. units, we
obtain

3 x 10%!

K

The following values of thus obtained for different gases are

Air NeA = 1-35 x 1010

Oxygen N^0 = 1*25 x 1010

Carbonic acid Ncc = 1'30 x 1010

Hydrogen NeH = 1"00 x 1010

The values of u were taken from the table of mean velocities given
by Professor Rutherford.* The values of k which were used, are the
mean values obtained for dry gases.

Experiments on electrolysis show that ' one electrodynamic unit of
electricity in passing through an electrolyte, gives off 1'23 c.c. of
hydrogen at temperatures 15° and pressure 10° C.G.S. units. The
number of atoms in this volume is 2*46 N, so that if E is the charge
on an atom of hydrogen in the liquid electrolyte,

2-46NE = 1 electrodynamic unit of quantity
= 3 x 1010 electrostatic units.
Hence NE = 1-22 x 1010,

the charge E being expressed in electrostatic units.

Since N is a constant, we conclude that the charges on the ions pro-
duced by Rontgen rays in air, oxygen, carbonic acid, and hydrogen,
are all the same, and equal to the charge on the hydrogen ion in a
liquid electrolyte.

Professor Thomson! has shown that the charge on the ions in
hydrogen and oxygen which have been made conductors by Rontgen
rays, is 6 x 10-10 electrostatic unit, and is the same for both gases.

Taking this value for the charge e, we obtain the number of mole-
cules in a cubic centimetre of a gas

N = 2 x 1019.

From this we deduce the weight of a molecule of hydrogen, p/N,
4*5 x 10~24 gram.

* E. Rutherford, * Phil. Mag.,' NoTeruber, 1897.
f J. J. Thomson, ' Phil. Mag.,' December, 1898.

196

Mr. D. Gill. On the Presence of Oyxgen

Since, as we have shown, the charge on an ion produced by Rontgen
rays is equal to the charge on a hydrogen ion in a liquid electrolyte,
this latter charge is also 6 x 10~10 electrostatic unit.

Although the value of N<? for hydrogen is 25 per cent, less than its
value for other gases, we are justified in including hydrogen in the
above general conclusion, as we should expect the value of u for hydro-
gen to be too small. Professor Rutherford makes no mention of
having corrected for the presence of air in his apparatus, or of having
used perfectly dry hydrogen. If we take the mean value of k for
moist hydrogen, we obtain N<?h = 1"15 x 1010.

In order to prove that the charge on the positive ion is equal to the
charge on the negative ion, the ratio of the coefficients of diffusion
must be shown to be equal to the ratio of the velocities. Professor
Zeleny* has shown that the negative ions travel faster under an elec-
tromotive force than the positive ions, the ratios of the velocities being
1*24 for air and oxygen, 1*15 for hydrogen, and 1*0 for carbonic acid.

The experiments on diffusion show that the ratio of the velocities
would be larger in dry than in moist gases ; but as this point has not
yet been examined by Professor Zeleny, we cannot expect a very close
agreement between the ratios which he gives for the velocities and the
ratios of the coefficients of diffusion.

We are led to conclude that the charges on the positive and nega-
tive ions are equal from another point of view. It has been proved
that the mean charge is the same as the charge on an ion of hydrogen
in a liquid electrolyte. If the charges differed, one of them would be
less than the charge on the hydrogen ion, whereas experiments on
electrolysis show that all ionic charges are either equal to the charge
on the hydrogen ion or an exact multiple of it.

" On the Presence of Oxygen in the Atmospheres of certain Fixed
Stars." By David Gill, C.B., F.R.S., &c, Her Majesty's
Astronomer at the Cape of Good Hope. Received April 14, — ^
Read April 27, 1899.

(Plate 8.)

In a paper read before the Society on April 8, 1897, and in a sub-
sequent paper,! Mr. Frank McClean draws attention to the group-
ing of lines other than those of helium and hydrogen in the spectra
of p Scorpii, p Canis Majoris, p Centauri and p Crucis, suggesting
that the close correspondence between the grouping of these extra
lines and the known lines of oxygen, points to the probable presence
of that gas in th.3 atmospheres of these stars.

* J. Zeleny, ' Phil. Mag.,' July, 1898.

t ' Roy. Soc. Proc.,' vol. 42, JNo. 386, p. 418.

in the Atmospheres oj certain Fixed Stars. 197

In the latter paper he writes, " The most remarkable correspondence
is in the case of the large group on either side of Hs. A slight
shift of about a tenth-metre is required to bring the groups into identi-
cal positions. However, the close similarity of the whole grouping of
the two spectra as they appear on the plate, admits of little doubt that
the extra lines actually constitute the spectrum of oxygen. If this be
established, the spectrum of the first division of helium stars would
be due to hydrogen, helium, and oxygen."

In his subsequent work* Mr. McClean concludes, " Taking everything
into account, the succession of coincidences between the extra lines of
/3 Crucis and the oxygen spectrum can only be accounted for on the
basis of the extra lines being in the main actually due to oxygen."

This conclusion does not, as yet, appear to have been fully accepted
by spectroscopists, partly because, from the low dispersion used, the
lines of the groups are not separately shown. It is very generally
known that the instrumental equipment of the Eoyal Observatory
at the Cape, has. recently been enriched by a complete equipment
for Astrophysical research, the whole being the munificent gift of Mr.
McClean, F.R.S.

The slit-spectroscope, for attachment to the photographic refractor,
reached the Cape in the middle of January last, and I resolved that
its first published work should deal with Mr. McClean's interesting-
discovery.

As a complete account of the instrument and its observatory will be
subsequently published, it may be sufficient for the present to state
that the object glass of the photographic telescope has an aperture of
24 inches and focal length of 22 feet 6 inches, its minimum focus
being, at present, for rays about midway between H/3 and Hy.

The collimator of the spectroscope has an aperture of 2J inches and
focal length of 22-J inches, so that a cylinder of parallel rays 2 inches
in diameter falls on the prisms, and the latter are of sufficient size to
pass the whole of the rays which form the image of the spectrum on
the sensitive plate. The instrument is provided with two camera-
telescopes of 2f inches aperture, one being of about 36 inches focal
length, the other of 16 inches.

Only the larger of the two camera-telescopes have been employed in
the after-mentioned observations.

There are two cast>-iron prism boxes ; one of them contains three prisms
of about 60° each, which for rays near Hy produces a deviation of 180°,
so that the camera-telescope becomes parallel in the reverse direction
to the slit-telescope. The other prism box contains a single prism of
62°. The prisms in both boxes are fixed, without screw adjustment,
in minimum deviation for Hy. The collimator is in the axis of a solid
drawn steel cylinder — the latter attaching by a flange at one end to
* ' Spectra of Southern Stars' (Stanford, London, 1898).

VOL. LXV. Q

198

Mr. D. Gill. On the Presence of Oxygen

the butt end of the telescope, a cast-iron plate attaching to the flange
on the other end of the cylinder carries either one or other of the
two prism boxes.

The slit-slide and 60° prism for reflecting the comparison spark on
the slit (made on the plan of the Lick spectroscope) are contained in
a strong cast-steel box which is permanently attached -to one end of
the collimator tube. This latter is a very strong solid drawn steel
tube, the external surface of which has been turned and finished to a
true cylindrical form, and it rests in proper geometrical bearings
formed in strong cast-iron diaphragms, which are fitted inside the
cylindrical body of the instrument.

A powerful slow motion permits the collimator to be slid along its
axis so as to focus the slit upon the image of the star at any required
reading of the focussing scale. The object-glass of the collimator is
mounted on the end of another steel cylinder which also rests on
geometrical bearings inside the outer collimator tube, and it is also
provided with fine slow motion and a focussing scale. Both these
scales are illuminated at will by small incandescent lamps, and are
read by microscopes which are accessible from the outside.

The whole instrument can be enveloped in felt to prevent any but
very slow change of temperature.

The comparison-spark apparatus is arranged with wide angle object-
glasses, in such a way that if the image of the spark shines on the slit
the object-glass of the collimator must be full of light. Numerous
trials in all positions of the instrument have invarably given photo-
graphs of the lines of the comparison spectrum of iron rigorously
coincident with the corresponding lines of the solar spectrum, the
latter being obtained by exposing the slit in diffuse daylight.

The camera end with its focussing and tilting adjustments can be
attached to either telescope by a flange with bayonet joint. The
focussing scale is divided to 1/10 mm., and the amount of tilt of the
plate-holder is measured on a graduated arc.

As the large telescope is fitted with an object-glass prism of 24 inches-
aperture (which, when the slit spectroscope is in use, is folded back in
the manner shown in the frontispiece of Mr. McClean's 1 Spectra of
Southern Stars '), heavy counterpoises are required to balance the tube
about the declination axis if the spectroscope is not attached.

In designing the slit spectroscope I was thus not limited by the neces-
sity for lightness in its construction. The complete instrument weighs
400 lbs., beiug almost exactly the equivalent of the counterpoises and
focussing slide and camera, which are removed for its adaptation. In
every detail the fittings of the spectroscope are designed with the
necessary geometrical limitations of freedom and no more, so that no
shake nor variation of adjustment can arise from imperfection of work-
manship ; that is to say, all the adjustments depend on adequate spring

in the Atmospheres of certain Fixed Stars.

190

pressure against the minimum number of necessary rigid points of
support.

The object-glasses of the spectroscope were made by Brashier, and
are all excellent. The three dense prisms were also made by Brashier ;
their definition is very fine, but the glass is rather yellow in colour, and
produces great absorption of rays more refrangible than Hy. The
single prism, by Steinheil, gives excellent definition, and the glass is
much whiter than in Brashier's prisms. The optical constants of the
prisms have not yet been determined.

The mounting was constructed by the Cambridge Scientific Instru-
ment Company to my designs, in the most careful and satisfactory
way. I am greatly indebted to Mr. Horace Darwin for much care in
supervision of the work and for some very ingenious and important
improvements in detail which he carried out.

Above all I am indebted to Mr. H. F. Newall, who has taken infi-
nite trouble in making and testing the permanent adjustments of the
instrument and in supervising the arrangement of its final details. To
him I owe the fact that the instrument arrived at the Cape practically
in perfect adjustment and ready for work. After a series of prelimi-
nary focussing trials by Ne wall's method,* a number of photographs of
star spectra were made with the three-prism box and long telescope.
The present paper deals chiefly with the results of measures of a
photograph of the spectrum of ft Crucis and of a comparison iron spec-
trum obtained on March 15. The plate was exposed to the comparison
spectrum of iron immediately before and immediately after the expo-
sure for the star spectrum. (See Plate 8.)

Lines of the iron spectrum cover the whole exposed length of the
plate from Fe X 4187*99 to 4563*99, the linear interval on the plate
between these lines being 70*367 mm.

As a preliminary step, the intervals between successive pairs of iron
lines were measured with the micrometer of the old Eepsold astro-
photographic measuring apparatus.! If

As is the interval between two adjoining lines in terms of revolu-
tions of the micrometer screw,
X1 and X2 their respective wave-lengths,
Kn = i(X\ + X2) and AA. = X1 - X2)

I found, to my surprise, on computing AA/As for many different values
of Xmi and plotting these on millimetre paper with AA/As as ordinate
and Xm as abscissa, that within the limits of error of plotting and
observation the resulting curve was practically a straight line : in other
words, the screw values can be represented by

* ' Monthly Notices, E.A.S.,' vol. 57, p. 572.

f Described by Bakhuyzen, 4 Bulletin du Congres Astrograpliique,' toI. 1, p. 169.

20{)

Mr. D. Gill. On the - Presence of Oxygon

As
As
As

or generally

where

= 71 + 2$,

AA.
As"

AX>

As

AXx-2
As

= 7?

=1 :— = ?i~ 2d

X0 =

XK - njd.

Such a law can only be applicable to a limited portion of the whole
spectrum, for it is obvious that for no value of X can A A. /As be really
= 0.

In order to test over what range of the spectrum this law might be
regarded as sufficiently rigorous, four selected iron lines were measured
in terms of the millimetre scale which is attached to the instrument,
the division errors of which are known for each 5th millimetre. For
convenience the measures are converted into screw revolutions of the
micrometer microscope which was used in measuring the spectra, viz.,
1 revolution = 0*5 mm.

Screw Eev.

X1 = Fe

4583-99

200-254

^2 = a

4442-52

159 130

h = „

4282-54

101-802

A4 = „

4187-99

59-520

Hence,

Xi + Xo . AA-4513.2G

for — - — we have — -r-—
2 As

3-4400

cl

0-0043053

3714-80

A.-3 + Xs
2

x3 + xd

As

— = 2-7906:

AA

■0-0043572 3722-08

4325-27

As

= 2-2361

Thus for interpolation between adjoining known iron lines the
second differences, d, are practically constant over the range of spec-
trum with which we have to deal, and consequently the logs of (X - X0)
vary proportionately to the measured intervals between the lines.

If this were strictly true for the whole range of our spectrum one
would obtain rigorously accurate interpolation as follows : —

in the Atmospheres of certain Fixed Steers.

201

Let mi and 1113 be the micrometer readings on any two known lines,
,, n> ■> ,, „ for an intermediate un-

known line,

„ Xi and A3 the corresponding wave-lengths of the known lines,
then to find X2 corresponding to m-2 we have

( [Aa - A0] - [A, - A0] ) [A3] = [A2 - A0] (1),

7/(3 — Till

where the square brackets denote the logs of the included quantities.

That the second differences of AX/ As for successive values of A, are
not strictly constant for the whole length of our spectrum is shown by
the -different values of el and AQ obtained above.

If we assume A0 a constant =3718 and interpolate the wave-lengths
of A2 and a3, by means of our formula (1) we obtain

Known. ' Computed. Known — Computed.

A, ;. 4425-52 4425-36 +0-16

A3 4282-54 4282-72 -0-18

The differences " known - computed " considerably exceed the prob-
able errors due to the observations and the small inaccuracies of the
determinations of the fundamental wave-lengths, thus showing that
errors have been introduced by neglect of the small variations in the
second differences of the successive values of A Xj As.

It is however very easy to take account of these variations by com-
puting an auxiliary table for different values of A0 with the argument
Xh Thus, on the assumption that A0 varies proportionally to X1 we
have : —

Mg.

A. A0.

4100 3732

4200 3727

4300 3722

4400 3717

4500 3712

4600 3707

Taking X0 from this table with the argument ' approximate wave-
length of the line, whose definitive wave-length is required,' we find
the intermediate wave-lengths accurately represented.*

As well known iron lines are found within 30 or 40 tenth-metres of
all the unknown lines, it is always sufficient for the purpose of inter-

* I find that the wave-lengths of solar lines as measured by Campbell (' A. P.
Journal.' October, 1898) are also very beautifully represented throughout by this
simple method of interpolation.

202

Mr. D. Gill. On the Presence of Oxygen

polation by our formula to employ the value of X0 from the above
table with the argument ^ ^"8*

In this way the wave-lengths of the lines in the spectrum of (3 Crucis,
given in the following table (pp. 203 — 204), have been determined : —

Column (1) gives the wave-lengths of all the known oxygen lines of
intensity 3 or brighter, between A. 4303 and 4575, according to
Neovius as well as Trowbridge and Hutchins.*

Column (2) gives the wave-lengths of all helium lines according to
Runge and Paschen,f contained within limits of the. spectrum under
observation.

Column (3) gives the results of my measurements of the negative
of the spectrum of f3 Crucis, those under head I being my first essays
n the measurement of any photographed spectrum, those under
head II being the results of my second measurement, including all the
lines which could be detected under most careful and repeated scrutiny.
Each result in series I depends on two pointings, each in series II on
four pointings.

The lines whose wave-lengths are given to two decimal places of the
tenth-metre, were measured with a magnifying power of fifteen diameters,
those give to one decimal place with a power of only three diameters
— the lines of the latter class being very faint and only certainly
visible under a very low power. When possible, different iron lines
were used in series I and II for determination of the wave-lengths of
the stellar lines.

The observations were not arranged for determination of motion in
the line of sight, but the exact coincidence of the star line 4417*06
with the air (oxygen) line, and the general agreement of the stellar
hydrogen and helium lines with their known wave-lengths, tend to
show that the relative motion of ft Crucis to the Earth on February 21
did not exceed ± 3 kilometres per second. On that date the Earth in
its motion round the Sun was moving towards /3 Crucis with a velocity
of 18 kilometres per second, and consequently /3 Crucis is probably
receding from the Sun with a velocity of 18 kilometres ± 3 kilometres
per second.

The whole of the known helium lines within the measured range of
spectrum are unquestionably present, as also are all known oxygen
lines stronger than intensity 4.

The exceedingly faint lines

P Crucis 4253*9 may be coincident with Neovius 4254*1 = 1<
and H. 4253*42, and
Crucis 4303*0 or 4304*0 may be coincident with Neovius
4304*4 = T. and H. 4303*8,

* Watts, ' Index of Spectra,' Appendix E.
t 1 A. V. Journal,' toI. 3, p. 10.

in the Atmospheres of certain Fixed Stars. 203

»

o a

00 **

s s

^ .-s

53 rr*. S

"Hi

-3

I

g

O .

43

2 ©

C8 Si H O ... N r-j .

^f^£ H OOO O WtaO O

do m

lOCOtCOOOHHN
(NN(MNMCCMCOM

Tfl T^l Tjfl ^jjl T^l

1> QO O N LO l> C5 r-l r-l CO CO
COCO^^T* tJI US CO CO CO

CO ors CO CO CO 00 co co co co co

Wo

00 CO
00 Oi
CO CO

CO 00 CO

OS

CO

CO

00

CO ■£■» 00

r-l

CO

Tj< CO

o

t> OS lO

l>

lO

as h

t>

00

HHN

CO

"#

CO

oo

CO CO CO

CO

CO

CO

co co

CO

CO

^ -^t

9

<s S 3 .
S «fl o a

^ C 2 co
• o

P3

^ g CO
§ 8

£1

J> OS ifl

CO CO CO CO
^ tJ<

•H Tfi CO CO CO

qs as co ih

H h IN (M CO
CO CO CO CO CO

^ Tf*

ITS l> t> OS r-t

rft "T* tJI kO
CO CO CO CO CO

CO CO 00 CO

00 CD CO

VO J>. OS rH

^J* "^T1 kO

CO CO CO CO

CO CO CO^ co

204 Mr. D. Gill. On the Presence of Oxygen

. A .

v CO
co 3

cs.2
.2-8

1 s
£ s

1= PI

'is

o o

Wo

O p!

05 J*

<p Q
§ 03.

a g

S3 o

o »

s *

lO fl S i-i CO «5 N IM
iH-rH - CO CO CO CO 1(5

t# » 03 TP

W > w

^ oo aq io co i>

rfl rfl lO K5 K5 1(5

«5©N
U5 lO 1(5

tJH rff -^+1

n 3 S H

9 03 <3 5 .

Pj 03^5 O A

© PI 2 .co

3^

bfi £

gco

O PI

£1

333

iO oo oi

t^h ^fi tJ^

3

in the Atmospheres of certain Fixed Stars.

205

b>ut these coincidences are very doubtful, and it is improbable that
such very faint lines would be represented, whilst the neighbouring
line 4327-3, intensity 3, is wanting on the photograph.

The oxygen lines of intensity 4 which are wanting, viz. : — Neovius
4465-4 and 4467*8 are in a portion of the .spectrum which is some-
what over exposed, and this fact probably accounts for the non-appear-
ance on the plate. Possibly also the relative intensities of the oxygen
lines at the temperature and pressure of the atmosphere of /? Crucis may
be different from their relative intensities in the conditions under which
Neovius determined the intensities of the air lines (spark spectrum).

There remains however not the slightest doubt, that all the stronger
oxygen lines are present in the spectrum of /3 Crucis, at least between
X 4250 and 4575, and this fact requires no further laboratory experi-
ments for its establishment. It is almost equally certain that there is
no trace of true nitrogen lines in this spectrum.

The only measured lines of /S Crucis near known nitrogen lines
.are : —

Neovius 4341-8 Intensity 15 ft Crucis 4342-6
4523-0 „ lb „ 4522-84

but it is improbable, although not impossible, that nitrogen lines of
intensity (1) should be present whilst the strong nitrogen line 4507*7,
of intensity 6, is absent.

Besides hydrogen, helium and oxygen, the spectrum of j3 Crucis
shows the probable presence of carbon (4267*2) and magnesiun
(4481-17). These lines were not included in the first measurement, in
the former case because the line could not be distinctly seen with the
higher power, in the latter because I was doubtful of the existence of .
the line on account of a slight defect in the film at that point, but
other negatives confirm its existence. A contact positive from the
original negative of the spectrum is sent herewith for reproduction.

The spectra of /S Crucis, p and e Canis Ma j oris, and probably
j3 Centauri are all practically identical. They all contain the three
unknown strong lines

4552*79
4567*09
4574-68

besides the probable magnesium line 4481*17, the lines of hydrogen,
helium, the stronger oxygen lines, and the probable carbon line 4267*2,
Farther investigations on this class of stars will be subsequently
communicated; in the meanwhile I forward also for reproduction a
contact positive from a negative of the spectrum of e Canis Ma j oris
taken on March 15 with the single prism, in which the slit has been
^ focussed for rays of about A 4080, and with a comparison spectrum
VOL. LXV. E

206

Proceedings and List of Papers read.

from an oxygen tube, a ley den jar and air space being introduced in
the secondary circuit of the Buhmkorff coil. (See Plate 8.)

This photograph shows the coincidence of stellar lines with the
group of three strong oxygen lines, viz. : —

Neovius. T. and H. Intensity.

4076-3 4076-19 9
4072-4 4072-34 9
4070-1 4070-24 8

and other neighbouring oxygen lines beyond the range of the three
prism train. The lines are all displaced towards the red by motion.

On March 15 the Earth by its motion round the Sun was receding
from e Canis Minoris, with a velocity of 17 kilometres per second,
which agrees with the direction of the displacement of the oxygen
lines. There has not yet been time to determine the constants of the
single-prism spectroscope, but this does not affect the question of |the
identification of the oxygen lines in the spectrum of e Canis Majoris.

The plates of the spectra of /3 Crucis and e Canis Majoris were
exposed and developed by my assistant, Mr. J. Lunt.

June 1, 1899.

Annual Meeting for the Election of Fellows.

Professor T. G. BONNEY, Vice-President, in the Chair.

The Statutes relating to the Election of Fellows having been read,
Dr. A. A. Common and Dr. J. H. Gladstone were, with the consent of
the Society, nominated Scrutators, to assist the Secretaries in the
examination of the balloting lists.

The votes of the Fellows present were collected, and the following
Candidates were declared duly elected into the Society : —

Barrett, Professor William.
Booth, Charles, D.Sc.
Bruce, David, Major B.A.M.C.
Fenton, Henry John Horstman,
M.A.

Gamble, James Sykes, M.A.
Haddon, Professor Alfred Cort,
M.A.

Head, Henry, M.D.
Hele-Shaw, Professor Henry Selby,
M.Inst.C.E.

Morgan, Professor Conwy Lloyd, <

F.G.S.
Beid, Clement, F.G.S.
Starling, Ernest Henry, M.D.
Tanner, Professor Henry William

Lloyd, M.A.
Threlfall, Bichard, M.A.
Tutton, Alfred E., B.Sc.
Windle, Professor Bertram Coghill

Allen, M.D.

Thanks were given to the Scrutators.

The Characteristic of Nerve.

207

June 1, 1899.

Professor T. G. BONNEY, Vice-President, in the Chair.

A List of the Presents received was laid on the table, and thanks
ordered for them.

Professor Ludwig Boltzmann, Professor Anton Dohrn, Professor
Emil Fischer, Dr. Georg Neumayer, and Dr. Melchior Treub, were
balloted for and elected Foreign Members of the Society.

The following Papers were read : —

I. " The Parent-rock of the Diamond in South Africa." By Professor
T. G. Bonney, F.E.S.

II. " Experimental Contributions to the Theory of Heredity. A.
Telegony." By Professor J. C. Ewart, F.E.S.

" The Characteristic of Nerve." By Augustus D. Waller, M.D.,
F.RS. Eeceived February 3 — Eead February 16, 1899.

The object of the present preliminary series of experiments was to
determine whether the excitability (or, as I should prefer to say in this
connection, the mobility) of living matter can be gauged by the rate of
impact of a mobilising stimulus ; in other words, whether, for various
kinds of more or less mobile protoplasm, an optimum stimulation-
gradient can be found, above and below which the curve of stimulation
is less perfectly adapted to the movement caused by stimulation. In,
e.g., the case of the rolling of a ship, there is an optimum wave-length
and wave-face outside the limits of which the " mobilisation " is less
than maximal. And just as the optimum wave-length giving greatest
movement by least energy harmonises with and gives therefore mea-
sure of the oscillation-period of a particular ship, so the optimum
gradient of mobilising energy producing greatest excitation by least
energy, might be expected to give measure of the excitation-period
proper to a particular tissue, and to characterise that tissue.

The best instrument at our disposal for an examination of this point
is obviously a condenser, of varied capacity, charged at varied pres-
sure. This method of excitation has been put into effect by several
previous observers. Chauveau(l) first systematically studied the " law
of contractions," and demonstrated upon frog's nerve that the make

R 2

208

Dr. A. D. Waller.

excitation is kathodic. Dubois (4, 5), of Bern, studied the law of con-
tractions upon human nerve, and came to the conclusion that the
greater and smaller effects of condenser discharges are governed by the
greater and smaller quantities of electricity in play. Hoorweg (11, 12),
Cybulsky and Zanietowsky (9, 10) studied the effects more closely
upon frog's nerve ; the first-named observer concluded that the effects
are a function of intensity ; while the last two observers, from very
similar observations, concluded that the magnitude of excitation is a
function of energy. Salomonsen (8), Boudet (2), and D'Arsonval (6, 7)
come to the same conclusion, and the last-named observer makes use of
an expression — " la characteristique de 1'excitation " — very similar to
the designation that I had been independently led to adopt, in ignor-
ance of its previous use in a different sense.

The problem is, I think, to be considered from the following a priori
point of departure : — A stimulus arouses excitable matter by reason of
its actual energy and not of its mere quantity. A weight per se is not
a stimulus, but a weight dropped from a height acts as a stimulus that
is to be expressed in terms of energy. One gram fallen through 1 cm.
strikes and stimulates with an actual energy of 1000 (or more precisely
981) ergs. Similarly, as regards the electrical stimuli afforded by the
(charge or) discharge of a condenser, it is neither the quantity (cou-
lomb) nor the pressure (volt) that alone gives measure of the stimulus,
but the energy (Joule or 107 ergs) of that quantity at that pressure.
A nerve (or other excitable tissue) may be struck and stimulated by
an electrical energy of so many ergs. And we are adequately
acquainted with the physiological value of a stimulus when we
know : — First, its absolute value in ergs or fraction of an erg. Second,
the rate at which such energy impinges upon and sets in motion the
excitable molecules under investigation.

The absolute value of a true minimal stimulus in ergs or fraction of
an erg, is ascertained by determining the optimum capacity and vol-
tage at which a minimal response is obtained. The energy E in ergs
= 5 FV2, where F is in microfarads and V in volts.
Its rate of impact depends upon the rate of (charge or) discharge

= x a constant, where E, denotes the resistance in circuit in
FIv

ohms, and can therefore be expressed by a number that will be higher
or lower according as the rate of discharge is greater or smaller.*

* With the same units as above, and with the unit of time = 10 — 6 second, the
constant is 0'8687. In the subjoined data it has been taken as 868700 for the unit
of time = 1 second.

For certain ends it is convenient to indicate the rate of discharge of energy by
stating the time necessary for its fall to any given fraction \\n of its original value,

according to the formula t — KE. ^ n- ; e.g., with a capacitv of O'Ol microfarad,

V. Iriry f

The Characteristic of Nerve.

209

The number of the true minimal stimulus at optimum capacity and
voltage, is what I propose to designate as the characteristic, or other-
wise stated — the characteristic is the constant of the curve of the
smallest discharge of energy that can provoke movement, or of the
discharge of energy provoking greatest movement.

By suitable adjustment of capacity and voltage we may obtain the
discharge of any desired energy — at any desired rate, if we also know
the resistance through which discharge is made. From a small con-
denser at high voltage, and from a large condenser at low voltage, we
may obtain the discharge (of equal quantities or) of equal energies,
but in the first case rapidly and in the second case slowly. We may
so adjust capacity and voltage, as to deliver constant energy in a
series of curves of varying steepness, and find by experiment the
more or less precise limits of steepness between which the motion or
excitation is greatest. The number denoting that steepness of energy
discharge by which a nerve is most economically mobilised, signifies
the optimum adaptation of excitation to excitability, and is its charac-
teristic. At that number the nerve is excited by the minimum of
energy. With a stimulus of higher or lower number, more energy is
necessary to produce an equal effect.

Taking either a series of stimuli of given energy, but of varying
gradients, as in experiments 1 and 9, we are to observe at what gra-
dient the maximum effect is produced. Or as in experiment 2, making
a series of trials with varying amounts of energy from varying
capacity and voltage, we are to observe at what minimum energy
discharge the smallest perceptible effect is produced, and thence cal-
culate the characteristic.

And even without actually determining the characteristic by means
of its upper and lower limits, we may often usefully ascertain whether
it is above or below a lowest or highest constant at the end of a series
of experimental data. Thus in experiment 6, a minimal energy has
not been reached, and we cannot therefore say that the characteristic
has been determined; we know however, that it is less than 10*4. In
experiment 4 the characteristic, as defined above, has not been actually
11382 and 0*6 since the stimulus has not been truly minimal; but
these values have been those of the constants of suitable stimuli not
far from minimal, and we know therefore that the characteristic has
been nearer to the higher value at high temperature, nearer to the
lower value at low temperature.

a resistance of 100,000 ohms, and 1/n = |, the time of discharge of f of the original
energy is 0 000693 sec. If the voltage under -which this discharge takes place is
0*2, the original energy is 0"002 erg, and the number indicating its rate of discharge
is 174. If 0*01 mf. and 0-2 v. should be the optimum capacity and voltage of a
minimal effect by excitation of a given nerve, i.e., a true minimal or optimal
minimal stimulus, then the number 174 is the characteristic of the nerve.

210

Dr. A. D. Waller.

Strictly speaking, the term minimal stimulus in, e.g., experiment 2,
applies only to the two values 0*001 134 erg and 0*002 erg in the two
groups respectively, since in these two cases only is the energy of
the stimulus at its minimum.

But in accordance with ordinary tests by induction currents, the
term " minimal " is applied to stimuli that are certainly not of minimal
energy, but that produce minimal effects. And in this loose sense all
the trials of the above groups give minimals, in the first group mini-
mals of effective capacity at the several voltages, in the second group
minimals of effective voltage at the several capacities.

The ambiguity may be provisionally neutralised by making use of
the expression " optimal minimal" for the true minimal stimulus qua
its energy value, and by avoiding use of the term minimal for stimu-
lation by induction shocks, or by condenser discharges of a gradient
above or below the optimum or characteristic gradient.

The method followed will be most readily understood by considera-
tion of the diagram.

(a) The apparatus for excitation, composed of battery, condenser,
and a Morse key, is connected with the nerve of a nerve-muscle pre-
paration by unpolarisable electrodes, the circuit being arranged so that
charge of the condenser is at contact a, and does not traverse the nerve,
while discharge is at contact b, and alone traverses the nerve. During
excitation the key Ki is closed, and the key K2 open.

(b) The apparatus for measurement of resistance, composed of
battery, Wheatstone bridge, and galvanometer, is connected with the
nerve by opening the key Ki (and closing the key K2).

To obtain fractions of a volt the disposition figured in the smaller
diagram! was adopted.

The Characteristic of Nerxe.

211

The muscular contraction was watched for, and occasionally recorded
in the usual way.

As regards the human subject, the circuit was as before, a Stohrer's
battery up to 60 volts being used to charge the condenser. One large
electrode (the "indifferent" electrode) was strapped to the abdomen;
a second small electrode (the " testing " electrode) was fixed over an
accessible nerve, such as the ulnar, median, or peroneal.

o-ooia

oocos

O-0CQQ

<

\

i
1

*

\

96

150

Time

1^
Ipoo

2^

3^
IfiOO

To illustrate rate of energy discharge and approximate gradients when the
constant has not less than two digits nor more than three digits. The shaded
area indicates range within which the nerve characteristic must have fallen.

The procedure was of three types : — (1) To take a series of com-
binations between voltage and capacity, so as to deliver constant energy
at various rates ; (2) to take a series of voltages and find for each
voltage the smallest capacity at which contraction was visible ; (3) to
take a series of capacities and find for each capacity the smallest voltage
at which contraction was visible.

The first plan has the considerable advantage of permitting the use
of a myograph. It is chiefly serviceable for the purpose of a pre-
liminary orientation, i.e., when it is desired to find quickly the range
at which the characteristic is to be looked for. Having obtained a
minimal effect with any given voltage and capacity, the next test is to
be made with half the voltage and four times the capacity, i.e., with
the same energy as before, but with that energy falling by a curve
with a constant equal to one-eighth of the constant of the previous case.
If the effect is manifestly greater, the voltage is again to be halved and
the capacity quadrupled from the starting point of a new minimal
effect. If, on the other hand, after the first step in reduction of the

212

Dr. A. D. "Waller.

constant, there is no effect at all, a test is taken in the opposite direc-
tion, with double the voltage and one-fourth the capacity, i.e., with
the constant of the original energy curve multiplied by eight.

The second plan affords a rapid survey of the whole field of stimula-
tion. When energy is varied by variation of voltage at any given
capacity, it varies as voltage squared, and the constant of its curve is
augmented directly as the voltage. The step from maximal to sub-
minimal stimulation is comparatively large when we test at a series of
diminishing voltages, and a voltage is soon reached at which no
increase of capacity, however great, can bring out an effective stimulus,
the constant of the energy curve being lowered as capacity is increased.

The third plan is useful only for closer determination of an optimum
gradient, when the range within which it is to be sought for is approxi-
mately known. When energy is varied by variation of capacity at any
given voltage, it varies directly as capacity, but the constant of its
curve diminishes as capacity is increased, i.e., a small increment of
energy is obtained at a reduced gradient, as compared with a large
increment of energy at a raised gradient obtained by increase of
voltage.

Exp. 1. — Nerve-muscle Preparation. Eesistance of Nerve + Electrodes

= 150000 w.

I

Capacity
in
micro-
farads.

Pressure
in
rolts.

Quantity
in
micro -
coulombs.

Energy
in ergs.

Con-
stants.

Time of
fall of
energy to
5 value.

Con-
traction.

F.

T.

FY.

5FV2.

C.

h-

0 -0004
0-0016

1-44
0-72

0-0005761
0 '001152 J

0 -00415

j 20850
\ 2606

0 -000042
0 -000168

small,
large.

0 -ooio

0 -0040

0-72
0 36

0-00072 1
0-00144 J

0 '00259

| 4170
1 521

0 -000104
0 -000416

small,
large.

0 '0025

o-oioo

0-36
0-18

0 -0009 \
0 -0018 J

0 -00162

f 834
\ 104

0 -000260
0 -001040

small,
large.

Opt.

0-0080
0 -0320

0-18
0-09

0-00144 \
0-00288 J

0 -00130

r 130

I 16

0 "000832
0 -003328

small,
none.

0 -0800
0 -3200

0-09
0-045

0 -0072 \
0 "0144 J

0 -00324

f 6-5
\ 0-8

0 -008320
0 -033280

small,
none.

4

0 -0400

o-oioo

0 -0025

0-09
0-18
0-36

0 -0036 ]
0 -0018 }
0 -C009 J

0 -00159

f 834
\ 104
I 13

0-000260
0 -001040
0 -004160

none,
large,
small.

Flie Characteristic of Nerve.
Exp. 1.

213

The constant of the optimal minimal stimulus (i.e., the characteristic)
is 130. The energy falls to \ of its original value in nearly y^Vo sec*

The second portion of this experiment, although less typical than
the first, exhibits a similar character. The minimal stimulus is higher
and of higher gradient, but further experiments will be required
before we may admit this latter change to be other than an accidental
effect. In other experiments there has been diminution of minimal
stimulus with lower gradient.

Exp. 1 (repeated.)

II

F.

V.

FV.

5FV2.

C.

Con-
traction.

0
0

•0005
•0020

1-44
0-72

0 -000720 \
0 -001440 J

0 -00518

f 16680
\ 2085

0 -000052
0 -000208

small,
large.

0
0

•0013
•0052

0-72
0-36

0 -000936 \
0 -001872 J

0 -00337

J" 3210
1 401

0 -000135
0 -000540

small,
large.

Opt.

0
0

•0035
0140

0-36
0*18

0 -001260 \
0-002520 J

0 -00227

/ 596
1 74

0 -000364
0 -001456

small,
none.

0
0

•0170
0680

0-18
0-09

0-0030601
0-006120 J

0 -00275

{ 6^

0 -001770
0-007080

small,
none.

0
0
0

0160
0040
•0010

0-18
0-36
0-72

0 -002880 1
0 -001440 [
0 -000720 J

0 -00259

f 4170
< 521
L 65

0 -000104
0 -000416
0 -001664

small,
large,
none.

214 Dr. A. D. Waller.

Exp. 2.— Frog. Sciatic Nerve. R = 80,000 a> *

Opt.

No effect.

Capacity
in micro-
farads.

F.

Pressure
in volts.

V.

1-44
0-72
0-36
0-18
0 '09
0 -045

Quantity
in micro-
coulombs.

FV.

Energy
in ergs.

5FT2.

Constant.
C.

Time of fall
of energy
to i of its
original
value.
t.

0 -00035
0 -00085
0 -00200
o -00700
0 -05000

10 -ooooo

0 '000504
0 -000612
0 -000720
0 -001260
0 • 004500
0 -450000

0 -003629
0 -002203
0 -001296
0 '001 134
0-002025
0 -101250

44700
9200
1950
279
19

0-05

0 -000019
0 -000047
0 -000111

0 -000388
0 -002770
0 -554000

0 '0001

5 -00

0-0005

0 -012500

543000

—

0 -000005

Opt.

0 -ooio

1 -oo

0 -ooio

0 -005000

10850

0 -000055

O "OIOO

O *20

0 -0020

O '002000

217

0 '000554

o-iooo

0-13

0-0130

0 -007450

14

0 -005540

1 -oooo

O'lO

0-1000

0-050000

1

0-055400

Exp. 2.

14 1-3

VoLt.

Influence of Temperature. — The characteristic of nerve is very sensitive
to alterations of temperature, being raised by high temperature (30°), and ^
depressed by low temperature (5°). Gotch and Macdonald have shown
that stimulation of nerve by break induction shocks is favoured by heat,
disparaged by cold, and that stimulation by the constant current is
favoured by cold, disparaged by heat (13).f The present experiments
show clearly that short stimuli (energy curve of high number) are

* In the first group of trials given voltages are taken, and the minimum effec-
tive capacities sought for. The characteristic = 279.

In the second group given capacities are taken, and the minimum effective volt-
ages sought for. The characteristic = 217.

f Gr. and M. allrde to condenser excitation, but did not employ it correctly.
They supposed that a " short " minimal stimulus coiild be obtained from a con-
denser of 0*5 microfarad.

The Characteristic of Nerve.

215

effective at high temperature, ineffective at low temperature, while long-
stimuli (energy curve of low number) are effective at low temperature,
ineffective at high temperature.

These results indicate further a point of probably considerable bio-
logical significance, inasmuch as the characteristic of frog's nerve at
high temperature is found to approximate to that of mammalian nerve.
Thus, whereas the characteristic of frog's nerve at room temperature
(16° to 18°) is represented by a number of three digits, and that of
human nerve at normal temperature (37°) by a number of five digits,
the approximate characteristic {i.e., the constant indicating the gradient
of a suitable stimulus) of frog's nerve at high temperature (30°) is also
expressed by a number of five digits. These and other data are
tabulated in the concluding summary.

Exp. 3. — R = 70,000 co*

T.

F.

V.

FV.

5FV2.

c.

30°
5

0 -0015
0 -5000

1-44
0-24

0 -0022
0 -1200

0-015
0-139

11910
6

0 -000073
0 -024250

f Abolished
\ by cooling.

f Abolished
\ by warming.

For the first trial at 32° the voltage is taken at 1*44, and the minimum
effective capacity is sought for and found to be a little below 0'0015.

In the second trial at 5°, the capacity is taken at 0*5, and the mini-
mum effective voltage is sought for and found to be a little below 0'24.

The first or short stimulus is immediately rendered ineffective by
cooling.

The second or long stimulus is immediately rendered ineffective by
warming.

* In experiments 3 and 4 the characteristic proper has not been determined,
but only the constants of a short and long curve of minimal but not optimal minimal
stimuli. An optimal minimal stimulus is rendered ineffective by heating and by
cooling. The alterations of resistance by alterations of temperature have not been
taken into calculation. Such alterations would, however, have only intensified the
contrast already apparent between short and long stimuli, in accordance with the
following numbers :—

T.

■ra

V.

FV.

5FV2.

c.

E.

sec.

30°

0 -0015

1-44

0 -0022

0-015

16700

0 -000052

50,000 w

5

0 -5000

0-24

0 -1200

0-139

4-17

0 -034600

100,000 w

30°
5

0 -0007

1 -oooo

1-44
0-09

o-ooi

0-090

0-007
0-040

17900

0-391

0 -000048
0 -139000

100,000 a>
200.000 w

216 Dr. A. D. Waller.

Exp. 4. — Effects of High and Low Temperature upon Characteristic of

Nerve.

R = 130000 w.

T.

F.

T-

FY.

5FV2.

c.

30°
4

0 -0007

1 -oooo

1-44
0'09

o-ooi

0-090

0-007
0-040

11382
0-6

0 -000063
0 -090100

J Abolished
[ by cooling.
J Abolished
\ by warming.

Procedure similar to that of previous experiment.
The characteristic is raised by heat, lowered by cold, as in the pre-
vious experiment.

* 1 microfarad, 0 '09 volt.

f 0 -0007 microfarad, 1 "44 volt.

The Characteristic of Nerve. 217

Exp. 5.— Cat. Sciatic Nerve. E - 17000 w.

Opt.

No effect |

F.

V.

FY.

5FV2.

C.

%

0 -0022
0 '0050
o '0170
0 -1700

0 -5000

1 -oooo

2-88
1-44
0-72
0-36
0-18
0-18

0-0063
0 -0072
0-0122
0-0612
0 -0900
0 -1800

0-071
0-052
0 -044
0-110
0-081
0-162

66910
14700
2165
108
18
9

sec.
0 -000026
0 -000059
0 -000202
0-002023
0 -005890
0 -011780

About 20 minutes later.

0:006

2-88

0 -0173

0-249

24530

0 -000071

0-008

1-44

0 -0115

0-083

9200

0 -000094

0*012

0-72

0 -0086

0-031

3066

0 -000141

Opt.

0 '030

0 -36

0 -0108

0 -019

613

0 -000353

0-500

0-18

0 -0900

0-081

18

0 -005890

Exp. 5.

/ 4 /tf I B H 10 09 O S 07 0-6 0-6 0-4 0-3 OB 01 00

Volt.

The method of trial was to start with given voltage and find mini-
mum effective capacities.

In the first group of trials the characteristic is 2165 ; in the second
it is 613.

The constants of minimum effective stimuli at various voltages are
higher in the first than in the second group.

Short stimuli are relatively more effective in the first group ; long
stimuli in the second group.

218 Dr. A. D. Waller,

Exp. 6. — Cat. Phrenic Xerve. E = 30000 w.

F.

Y.

FY.

5 FY2.

C.

h.

0 -0012
0 -0021
0-0060
0-0170
0-0300

o-iooo

1 '44
0-72
0-36
0-18
0-09
0-036

0 -00173
0 -00151
0 -00216
0 -00306
0-00270
0 -00360

0 -01 ?44v>
0 -005443
0 -003888
0 -002754
0-001215
0 -00064S

34700
9960
1740
307
87

10 '4

sec.

0 -000044
0 -000125
0 -000353
0 -000624
0-002080

A fe^v minutes later.

0 -0085
0-6000

1 -44 j
0-36

0 -01224
0 -21600

0 -08812S
0 -388800

4910
174

0-000177
0 -012500

Yoltage given, capacity looked for.

Exp. 6.

Voti.

Exp. 7.— Man A. Ulnar Xerve. E = 12000.

F.

Y.

FT.

5FY2.

C.

h.

mf.

TOitS.

me.

ergs.

sec.

0-8000

10

8 -00

400

905

0 -006650

0 -1400

20

2-80

280

10350

0 -001165

Opt

O "C^O

3°

i'6^

247 "5

39<oc

0 -0004.C8

0 -0350

40

1*40

280

82500

0 -000291

0-0250

50

1-25

312 -5

145000

0 -000208

0-0180

60

1 -os

324

241500

0 -000150

1

0 -0150

70

1 -05

367 5

337500

0 -000125

Tlu CJmracteristic of Nerve.
Exp. 8. — Man B. Ulnar Xerve.

219

F.

Y.

FT.

5FV2.

C.

h

mf.

TOltS

mc

ergs.

sec.

1-0000

10

10-00

500

725

0 -008320

0-1150

20

2-30

230

12600

0 -000956

Opt

o '04^0

3°

1 '35

202-5

482^0

0-000374

0 -0350

40

1 -40

280

82500

0 -000291

0 -0320

50

1 -60

400

113000

0 -000266

0 -0260

60

1-56

468

167000

0 -000216

0 -0255

70

1-78

615

200000

0 -000212

Exp. 8.

1

1

/

500
4G0
300
200
100

7oVoLte. 60

50

40

30

Exp. 9. — Man Peroneal Nerve.

20

E = 15,000 to.

F.

FY.

5FT-.

C.

H.

mf.

volts.

me.

ergs.

sec.

mm.

0 -5000

20

10-00

1,000

2316

0 -005200

0

0-3200

25

8-00

4525

0-003330

trace

Opt.

0 -2200

30

6-66

7825

0 -002310

1 -5

0-1633

12400

0 "001700

6 "5

0 -1250

40

5 00

18530

0 -001300

6-6

0 -09S8

45

4-45

26375

0 -0010300

4 -5

0 -0800

50

4-00

36200

0 -000832

4-0

0-0661

55

3 -64

48200

0 -000687

3-0

0 -0556

60

3-34

62500

0 -000578

2-5

25
•32

A-A

30

•22

35
■J65

11 JUL JUL JUL jul

40 45 50 55 60 volts.

125 I '08 066 055 mf.

220

Dr. A. D. "Waller.

Bibliography,

1. Chauveau. " Utilisation de la tension electroscopique, &c." Congres de Lyon,

1873.

2. Boudet. ' ^lectricite medicale.' Paris, 1880.

3. Marey. ' Methode graphique.'

4. Dubois (Bern). ' TJntersuclmngen u. d. physiologische Wirkung des Conden-

satorentladungen.' Bern, 1888;

5. Dubois (Bern). " Recherches sur Taction physiologique des courants et

decharges electriques." ' Arcliives des Sciences Physiques et Naturelles.'
G-eneve, 1891.

6. D'Arsonval. " Relations entre la forme de 1' excitation electrique et la reaction

nevro-musculaire," ' Arcliives de Physiologie,' p. 246, 1889.

7. D'Arsonval. "Rapport sur l'eiectro-physiologie," 'Arcliives de Physiologie,'

1890, p. 156.

8. Salomonsen. ' Ned. Tydschr. v. Geneeskunde,' 1891. .

9. Cybulsky and Zanietowsky. " Nouvelle methode d' excitation electrique, &c."

' Cbt. f. Physiologie,' vol. 6, 1892, p. 167. ' Bull, de l'Acad. des Sciences de
Cracovie,' Avril, 1891.

10. Cybulsky and Zanietowsky. " Ueber die Anwendung des Condensators zur

Reizung der Nerven, &c," ' Pfluger's Archiv,' vol. 56, 1894, p. 45.

11. Hoorweg. " Ueber die elektrische Nervenerregung," ' Pfluger's Archiv,'

vol 52, 1892, p. 87.

12. Hoorweg. " Ueber die Kervenerregung durch Condensatorentladungen."

' Pfluger's Archiv,' vol. 57, 1894, p. 427.

13. Grotch and Macdonald. " Temperature and Excitability," ' Journal of Physi-

ology,' vol. 20, 1896, p. 285.

Addendum received February 10, 1899.

Sensificcdory Stimuli.

In the attempt to. determine a "characteristic" of various modes
of sensificatory stimulation, I encountered the doubt, and, therefore,
difficulty, inherent to all investigations where the experimental
criterion is a subjective one — by appreciation of minimal sensation.
This broad distinction was, however, clearly apparent, that whereas
for series of stimuli of motor nerves the minimum exciting energy
decreased to a minimum from which it rose again (diminishing voltage
and increasing capacity), the energy of sensory stimuli of increasing
duration decreased indefinitely towards a minimum value. Within
the limits to which my experiments extended, decrease of energy was
made up for by increase of its duration.

Testing on the human subject for minimal felt effect, by looking for
minimum effective capacity at a series of given voltages, I found it
difficult to distinguish between apparently different qualities of
minimal sensation , In the case of cutaneous nerves, it was generally
impossible to say whether the smallest " something felt " was directly
cutaneous or directly subcutaneous, or indirectly subcutaneous, by

The Characteristic of Nerve.

221

reason of some muscular contraction. In the case of the eye it was
often difficult to tell whether the smallest " something felt " were a
slight common sensation or a slight phosphene.

In the case of the eye this point, however, came out with satisfactory
distinctness, viz., that low in the energy series, i.e., with longer stimuli
by lower voltage at greater capacity, the subject of experiment is
more conscious of a " flash " (optical sensation) than of a " stab "
(common sensation), whereas high in the energy series, viz., with
shorter stimuli by higher voltage at lower capacity, the subject feels
a "stab" (common sensation) more readily than a "flash" (optical
sensation). This indicates that longer stimuli are better adjusted to
the excitability of the retina, shorter stimuli to the excitability of
common sensory terminal organs.

Eyeball. E = 12000 ohms.

-F.

y.

FY.

! 5FY2.

C.

0-020

60

1 -2

360

217000

0-025

50

1 -25

312 5

145000

0-030

40

1 -2

240

96500

0-040

30

1-2

180

54300

0-080

20

1-6

160

18100

0-250

10

2 -5

125

2900

1

5

5-0

125

362

0-015

60

0-9

270

290000

0-020

50

1 -o

250

181000

0-030

40

1 2

240

96500

0-040

30

1-2

180

54300

0-060

20

1-2

120

24100

0-200

10

2 -0

100

3620

0-500

5

2-5

62-5

724

Pain.

Flash.

Pain

Flash.

000166
000208
000250
000333
000665
002030
008320

000125
000166
000250
000333
000499
001660
004160

Frgs

400

Pain to short stimuli, light to long stimuli.

V!

"FU

isti!

300
200
WO

60

Vctts.

50

40

<50

BO

10

This principal conclusion is confirmed by experiments conducted on
the first plan of procedure, viz., by testing with stimuli of equal energy
VOL. lxv. s

222

The Characteristic of Nerve.

at various gradients. To this end I found it convenient to first settle
upon a minimal sensation at a given mean voltage, and then to make
two further trials, one at half the voltage and four times the capacity,
the other at double the voltage and one-fourth the capacity. Each
group of three tests was thus at the same energy, but with the con-
stants of the three discharge curves in the proportion 1, 8, 64. The
minimal sensation having been settled with the constant at 8, it could
be inferred from equality or inequality of the effects at higher and
lower constants, whether the constant of a fitting stimulus lay above
or below either of the two extremes.

F.

V.

FV.

5FV2.

C.

Sensation.

Forearm,
common
sensation.

R = 45000

0 -4000

o-iooo

0 -0250

12

24
48

4-8
2-4
1 -2

ergs.
288
288
288

579
4632
34056

0 -001250
0 -000313
0 -000078

None.
Slight,
Marked.

Tongue,
common
sensation.
R = 7500

0 -0800
0 -0200
0 -0050

12

24
48

0-96
0-48
0-24

57 -6
57 -6
57 -6

17380
139040
1112320

0 -000416
0 -000104
0 -000028

Marked.

Slight.

None.

Optical

sensation.
R = 21000

0 -2400
0 -0600
0 -0150

12
24
48

2-88
1 -44
0*72

172-8
172-8
172 -8

2031
16248
129984

0 -003493
0 -000873
0-000218

Marked.

Slight.

None.

Movement,
median nerve.
R = .30000

0 '2000
0 -0500
0-0125

12

24
4S

2-4
1-2
0-6

144
144
144

1706
13648
109184

0 -004158
0 -001040
0 -000280

(Movement) .
None.
Slight.
None.

To verify the statement that at equal energy retinal stimulation pre-
dominates when the gradient is low, and common sensory stimulation
when the gradient is high, it is preferable to take stimuli above the
minimal and with larger difference of gradient, as e.g., in the following
experiment : —

Anode on Eyeball. Kathode on Abdomen. E = 9000 ohms.

F.

V.

FY.

5FY2.

C.

h. "

Sensation.

mf.
0-04
0-36
3-24

volts.
54
18
6

mc.
2-16
6-48
19-44

ergs.
583-2
583 -2
533 -2

130200
4828
179

sec.
0 -000250
0 -002245
0 -020200

Stab only.
Stab + flash.
Flash only.

Tlic Parent-rock of the Diamond in South Africa.

223

~ The Parent-rock of the Diamond in South Africa." By Professor
T. G. BoNNEr, D.Sc, LL.D., VJP.RS. Eeceived May 1 —
Read June 1, 1899.

So much has been written on the occurrence of diamonds in South
Africa that a very few words may suffice as preface to this communication.
References to many papers on the subject are given in ' The Genesis
and Matrix of the Diamond' (1897), by the late Professor H. Carvill
Lewis,* and others have been published since that date.f It may
suffice to say that the diamond, first discovered in 1867 in gravels
on the Orange River, was found three years later in certain pecu-
liar deposits, which occur locally in a region where the dominant
rock is a dark shale, sometimes interbedded with hard grits, or asso-
ciated with igneous rocks allied to basalt. These deposits occupy
areas irregularly circular in outline, and bearing a general resemblance
to volcanic necks. The diamantiferous material, near the surface, is
soft, yellowish in colour, and obviously much decomposed ; at a greater
depth it assumes a dull greenish to bluish tint, and becomes harder.
At the well-known De Beers Mine, near Kimberley, the works in 1898
had been carried to a depth of about 1,500 feet, and the diamantiferous
material, for at least the last 100 yards, was not less hard than an
ordinary limestone. It has a brecciated aspect, the dark, very minutely
granular, matrix being composed mainly of serpentine (about four-fifths
of the whole), and of a carbonate of lime (with some magnesia and a
little iron). In this matrix are embedded grains of the following
minerals : — Olivine, enstatite, smaragdite, chrome-cliopside (omphache
of some authors), a brown mica, garnet (mostly pyrope, but more than
one variety observed), magnetite, chromite, ilmenite, with several other
minerals much more] sparsely distributed.

Rock fragments are also present, variable in size, but commonly not
exceeding about an inch in diameter, as well as in quantity. These,
occasionally, but not generally, are rather abundant. In some cases
they are chips of the neighbouring black shale, but in others they are
greyish-coloured with somewhat of a porcelain aspect. The latter are
generally sub-angular in form and externally banded or bordered with
a darker tint ; crystalline rocks have also been noticed, though these
appear to be far from common, such as granite, cliorite, and varieties of

* Edited by the present writer.

t Jules Cornier, ' Qeol. Soc. South Africa Trans.,' 1897, p. 91 ; H. S. Harger,
ibid., p. 124. See also AY. G. Atherstone, Hid., 1896, p. 76 ; L. De Launay,
'Compt. Eend.,' vol.125 (1897), p. 335. The last author, in 'Les Diamants du
Cap' (Paris, 1S97), gives a very full account of the mines, but an even better one
will be found in Max Bauer, ' Edelsteinkunde ' (Leipzig, 1896. p. 208).

S 2

224

Prof. T. G. Bonney.

eclogite.* As to the genesis of the diamond, more than one opinion
has been expressed. Professor Lewis regarded the matrix as a porphy-
ritic form of peridotite, once a lava, now serpentinised,f in which the
diamond had been formed by the action of the molten rock on some
carbonaceous material (probably the Karoo shale). Others regarded
the matrix as a true breccia, comparing it with the agglomerates in
volcanic rocks. But among the latter, some thought that the diamond
had been produced in situ by the action of steam or hot water in a
subsequent solfataric stage of the volcano, while others (including
myself) held that it had been formed, like the garnets, pyroxenes, &c.,.
in some deep-seated holocrystalline mass which had been shattered by
explosions.]:

The specimens which I am about to describe were obtained at the
Newlands Mines, West Griqualand ; from 40 to 42 miles from
Kimberley, almost due N.W. Here the workmen occasionally came
across well-rounded boulder-like masses of rather coarsely crystal-
line rock, studded with garnets, which are sometimes about a foot
in diameter. Specimens of these were found or obtained by Mr. G.
Tmbenbach, the London manager of the Newlands Diamond Mine
Company, during a visit to the mines in 1897. His interest had
already been aroused by picking up a specimen, presently to be noticed,
in which some small diamonds occurred, very closely associated with
a garnet ; so the boulders were brought back by him to England. On
careful examination a small diamond was detected on the surface of one
of these. On breaking the boulder others were revealed. The most
interesting fragment was sent by Mr. Tmbenbach to Sir W. Crookes,
who showed it to me. Examination with a hand lens convinced me
that the rock could not be a concretion of the " blue ground," but was
truly holocrystalline and allied to the eclogites. Sir W. Crookes
generously waived his own claim to study the specimen, and obtained
for me permission from Mr. Tmbenbach to have slices cut from it. I
gladly take this opportunity of expressing my gratitude to both gentle-
men ; to Sir W. Crookes, for allowing me to carry out this interesting
investigation, and to Mr. Tmbenbach for his great liberality in placing at
my disposal a considerable suite of specimens (including other boulders)
from the Newlands mines, and for the trouble which he has taken in
affording me the necessary information.

* A. W. Stelzner, ' Sitzungsber. u. Abhandl. der Isis ' (Dresden), 1893 (April),
p. 71, calls attention to the fact that these show signs of attrition and that they
range in size from a few cubic millimetres upwards, being sometimes large boulders.
Among the materials (at Kimberley) he mentions both granite and eclogite.

f For the rock itself he proposed the name " kimberlite."

X In other words, that the volcano (as occasionally has happened) had ejected
little or no lava or scor.a, discharging only steam and hot water, with shattered rock.
This view is held by Max Bauer, in ' Edelsteinkunde,' p. 225, which, however, I
had not seen when this paper was written.

TJie Parent-rock of the Diamond in South Africa. 225

Prior to the discovery, just mentioned, one or two instances had
occurred at the De Beers Mine of a diamond apparently enclosed by
or projecting into a pyrope. One such, the garnet being the size of a
rather large pea, is in the collection at Freiberg (Saxony), to which it
was presented in 1892.*

The specimen foimd by Mr. Trubenbach at the Xewlands mine, was
a piece of blue ground, with a pyrope projecting from one angle.
A small, apparently broken, diamond seems embedded at the top.
The others (five) are well crystallised, two on one side, three almost in
contact on the other. The pyrope (which has a kelyphite rim) seems
to be indented by two, but to have once included the others, as they
are in contact with the unaltered mineral. We were thus brought so
far as to associate the diamond with the pyrope ; though this proved
no more than the presence of garnets in the parent rock of the diamond,
and thus made the eclogite (already known to occur) highly probable,
for, as observed by Professor E. Beckf, the specimen itself is blue
ground. In confirmation of his statement I pulverised a fragment,;
and find that the powder corresponds with the matrix of the blue
ground when similarly treated. The latest discoveries enable me to
complete the chain of evidence.

Eclogite Boulders containing Diamonds.

The first named, that containing several diamonds, is a fragment
(perhaps from a quarter to a third) of a boulder, which probably was
ellipsoidal in shape, two of the axes being nearly equal and the third dis-
tinctly the longest. AYe may infer that it was rounded from a roughly
rectangular block, since the curved surfaces are slightly flatter in the
middle parts. The axial lengths in the fragment (prior to removing a
piece from one end) were approximately 4 in. by 3 in. by 2 in.
The rock is coarsely granular, apparently composed of two green-
coloured minerals, one darker than the other (possibly only different
states of a single mineral), and of rich resin-pink coloured garnets,
varying in size from a hemp seed to a pea, with slightly irregular
distribution. The outer surface of the boulder, except for a very small
" step " on one side, is smooth, the garnets barely, if at all, projecting.
The latter are covered with a rather soft, dark skin, sometimes slightly
thicker than the thumb nail, which often has partly fallen off. This,
as can be seen on the broken surfaces, becomes less conspicuous in the
inner part of the boulder, and is sometimes invisible to the unaided eye.

* A. W. Stelzner, ' Sitzungber. der Isis zu Dresden,' 1893, s. 85, and K. Beck,
'Zeitsch. fur praktische Geologie,' 1898 (May), p. 163.
f Vt sujpra.

X I could not advise Mr. Trubenbach to bare a slice cut from tbe specimen, as I
feared it might be injured, but he kindly detached a little fragment from tbe oppo-
site end to that named above, -which I have thus examined.

226

Prof. T. G-. Bonney.

Two small diamonds are exposed on the curved outer surface, one
about half the other about one-fifth of an inch from the edge of the cross
fracture. On the latter surface, nearly an inch below the last named,
three small diamonds appear to lie in a line touching one another, and
near them are two others,* all four within a space about three-quarters
of an inch square ; an eighth diamond is about an inch and a half
away (on the same face) ; a ninth, about one-fifth of an inch from the
top edge ; and a tenth occurs on the larger cross-fractured surface, but
near to the edge of the other one. These diamonds are octahedra in
form, generally with stepped faces — one, at least, apparently twinned —
perfectly colourless, with brilliant lustre; the largest being quite 015
inch from apex to apex; the smallest not exceeding 0'05 inch. All
seem to be embedded in the green part of the rock. As the outer
part of the boulder looks rather more decomposed than the inner, I had
a piece removed from one end, thus enabling me to study the mass to
a depth of more than an inch from the surface, and examined a strip,
about 4 inches long, in a series of five slices.

The late Professor Lewis has given, in the volume already mentioned,
so full an account of the minerals which occur in the " blue ground,"
that it will be needless on the present occasion to do more than refer
to his descriptions,! only calling attention to any variations in the
mineral constituents and their association in these eclogites. These
constituents are : —

1. (ft) Garnet (Pyrope). — In the slice these appear a light tawny or
yellowish-red tint, retaining this tint (though much lighter) under the
microscope. £ They are generally clear, with frequent and irregular
cracks, but are occasionally traversed by wavy bands of minute en-
closures of a pale brown filmy mineral, which is rather irregular in
outline, very feebly pleochroic, and gives with crossed nicols fairly bright
polarisation tints. Similar minerals sometimes have formed along the
cracks. They are probably mica, or possibly chlorite, and indicate
incipient decomposition. The garnets towards the outside of the
boulder, as already said, are enveloped in a " skin," and the microscope
shows that it usually exists inside, though there it is thinner. In the
former case it is generally browner in colour and more distinctly
crystalline, corresponding in cleavage, pleochroism, &c, with a mica of
the biotite group ; in the latter it is greener and more filmy with an
aggregate habit and seems to project into the garnet. I regard it as due

* It is possible that these two form a twin crystal, but I think they are separate.
As the point is unimportant, I have not attempted to clear away the matrix.

f We must also not forget the paper by Professor Maskelyne and Dr. Flight
(*' Quart. Journ. Geol. Soc.,' vol. 30, p. 406), in which several of these minerals are
described, analysed, and identified. In fact the authors ascertained everything
that was possib]e -with the materials then obtainable.

X Unless it is expressly stated, the use of a 1-inch objective may be assumed.

The Parent-rock of the Diamond in South Africa.

227

Fig. 1. — Section of the diamond-bearing eclogite. Pyropes with, narrow kelyphite
borders, and chrome diopside intercrystallised. The clinopinacoidal cleavage
of latter risible in the lower part of the section.

to decomposition, a form of the well known kelyphite rim, sometimes
a mica, sometimes a chlorite, possibly now and then associated with
a little minute hornblende. In a few cases a " rim " is brown in the
outer part and green within. The constituents tend to a parallel
rather than a radial grouping. The garnets occasionally contain
minute branching root-like enclosures grouped in bands. Though these
act on polarised light, I regard them as empty cavities, and attribute
this to diffraction.

(b) Chrome-diopside. — The mineral described under that name by
Professor Lewis, and referred to by others as omphacite or sahlite. The
individuals are sometimes about a quarter of an inch long. In thin
slices it is a pale dullish green colour, inclining to olive ; under the
microscope, a pale sea-green, with a trace of pleochroism. It has one
strongly marked cleavage, not however nearly so close as in ordinary
diallage, and a second weaker, sometimes approximately at right
angles to it.* On examining flakes, obtained by crushing, I find the
strong cleavage to be clinopinacoidal and the other probably basal,

* One may give a general idea of their relative importance by comparing them
to the columns and cross- joints in some basalts.

228

Prof. T. G. Bomiey.

and obtain on a clinopinacoid an extinction of 35° with a prism edge.
It is in fact identical with the pyroxene described by Professor Lewis*
as chrome-diopside. In it (though rare) are small rounded enclosures
of a greenish mineral aggregate much blackened with opacite. I
regard them as alteration products of a ferriferous olivine. This
diopside, at the exterior and along cracks, is often converted into
a minutely granular to fibrous mineral, which gives a " dusty " aspect
to that part of the crystal, when viewed with transmitted light, and
a whitish-green one with reflected light. This often terminates in a
minutely acicular fringe, piercing the original diopside. Its grains
occasionally are a little larger, showing a cleavage, dull green in colour,
fairly pleochroic, and having the extinction of hornblende. A process
of secondary change, as in uralite, is no doubt indicated. Xow and
then a tiny film of brown mica occurs in this part or even in a crack
in the diopside.

It is this alteration product which gives the mottled aspect men-
tioned above as visible to the unaided eye, so this is not indicative of
a third important constituent in the original rock. In one of the
slices the mica just named attains a larger size (about 0*03 inch across),
has a fairly idiomorphic (hexagonal prism) outline, and is not restricted
to the margin of the garnet. In this case it is generally associated
with calcite,f which it tends to surround, and that in one place encloses
a radiating acicular mineral (? a zeolite) ; in another the calcite, or some
other carbonate, is mixed with a serpentinous material. Distinct
granules of iron oxide are practically absent from the slices, though
here and there it may be indicated by some opacite. I have not found
spinel, or rutile, or zircon, or pseudobrookite. In fact, putting aside the
diamonds, the rock in its unaltered condition was a coarsely holo-
crystalline mixture of chrome-diopside and garnet, with a few small
enclosures of olivine, in other words, it was a variety of eclogite and
of igneous origin .$

2. A fragment (probably about one quarter) of a flattish ovoid
boulder. — The two broken surfaces, which are nearly at right angles,
measure 5 and 5 J inches, roughly, and it is about 3J inches high. The
rock very closely resembles the one just described, except that mica occurs
rather oftener and in larger flakes ; perhaps the garnets (here also not
quite regularly distributed) are slightly more numerous. The outer

* Loc. cit., p. 21.

f From the facts I think it probably of secondary origin. It reminds me some-
times of the brown mica produced by contact metamorphism.

X I am, of course, aware that eclogite, in the past, has been regarded by some
geologists as a metamorphic rock. .Apart from the fact that several rocks once
assigned to this class are now, with good reason, regarded as igneous, I hare had
several opportunities of studying eclogite, and have no doubt as to its origin. Take
away the alkali from a magma with the chemical composition of a diorite, and the
result would be garnets in place of felspar, i.e., an eclogite.

The Parent-rock of the Diamond in South Africa. 229

•surface is not quite so well preserved, though enough remains to show
that it also has been smooth, and a few thin veins of a white mineral
(calcite 1) traverse the rock. On this surface, near the meeting of the
two fractures, and exposed by the removal of a little material
(i.e. it might originally have been just hidden) is a diamond (octa-
hedron), apparently about 0*1 inch in diameter. On one side it rests
against a pyrope, the adjacent surface of which is incurved, the
two minerals being parted by the dull green-colourecl kelyphite rim

Fig. 2. — Garnet and diamond (diagrammatic, nearly twice natural size).
(1) Diamond, (2) garnet, (3) kelyphite rim.

•of the latter, which is about 0*03 inch in thickness. Thin sections of
this boulder correspond almost exactly with those from the other, the
garnets showing precisely the same tints, though traces of a cleavage,
(roughly parallel throughout) are perceptible on close inspection, and
are distinct under the microscope. In garnet such a structure com-
monly indicates pressure, and the general parallelism accords with this
explanation, but the other constituents show no signs of crushing.
The " kelyphite " rims to the garnets are perhaps slightly broader and
the brown mica passes into a green (chloritic *?) mineral, and occupies
cracks in the garnet a little more frequently, but as before the con-
stituents tend to lie parallel rather than radially. One or two of the
diopsides show fine oscillatory twinning. The cracks are occupied with
calcite or some allied carbonate. There is no real difference between
this eclogite and the last-named one.

Eclocjite Boulders without Diamonds.

3. Part of a boulder, which must have been about a foot in
diameter (found at 250 ft. level). It presents a general resemblance to
the rocks described above, with, however, the possibility of a second
green constituent. This is not confirmed on microscopic examination.
The rock consists, practically, of pyrope and diopside, as already
described, except that negative crystals are rather unusually con-
spicuous in the latter. Into the details of these, as the point seems
not to have any bearing on the present investigation, I do not purpose
to enter.

4. A fragment, more irregular in form than the others, measures
very roughly, about 7 in. by 4f 'in. by 3J in. It retains a good piece
of the outer surface, whi^h, though now a little corroded, was once

230

Prof. T. G. Bonney.

smooth. The rock, which is rather decomposed and crumbly, consists
chiefly of three minerals : garnet, not quite so large, paler and more
pink in colour than the last-named; an emerald-green pyroxene, and a
yellowish or greenish-grey, platy to fibrous mineral, suggestive of
a second more altered pyroxene. In thin slices the paler and pinker
tint of the garnet is very perceptible, as well as the tendency to a
rude and generally parallel cleavage. But we find in it, under the
microscope, a few microlithic enclosures, of an apparently colourless
mineral, which occurs in long prisms crossed at about 70° by an
occasional transverse cleavage, and extinguishing at an angle of about
26° with the longer edge. Many of the cracks exhibit slight decom-
position, starting from them, and are sometimes occupied by calcite.
The pyroxene, under the microscope, hardly differs from the one
already described, except that the green tint is slightly richer and one
or two crystals contain the small dark brown negative crystals, com-
mon in hypersthene and diallage. The dominant cleavage, as before, is
along the clinopinacoid.* The thud mineral proves to be an altered
enstatite, but I leave the details for the present, as it is better pre-
served in another rock. A fourth constituent is also present, but more
sparingly, viz., a pale brown mica, only moderately pleochroic
(phlogophite 1). It occurs generally in plates, averaging about 0*1 inch
long. The minerals appear to have formed in the following order :
(a) garnet, (b) diopside, (c) mica, (d) enstatite. As before, iron oxides
are very inconspicuous ; there may be a grain or two (small) of ser-
pentinised olivine. The marked presence of enstatite distinguishes
this rock from the others, but it differs from the eulysites by the substi-
tution of that mineral for olivine, and so links those rocks to the more
ordinary eclogites. The occurrence of a little mica indicates the
presence of a small amount of an alkali in the magma. If necessary
we may name it newlandite, but personally I should prefer to call it an
enstatite-eclogite, for I think the coinage of fresh titles more often a
bane than a boon to science.

5. This boulder is almost perfect, except that the general flatness of .
one side indicates either traces of an old fracture or considerable loss
by crumbling. The surface has been smooth, but it has suffered from
unequal weathering of the minerals. Its girth, in three directions at
right angles, is approximately 20 J in. by 19 J in. by 17 J in. It appears
only to differ from the last-described in having its garnets a shade more
purple, and in an approach to a banded structure ; the diopside being
rather more abundant in a middle zone, the garnet in one, the enstatite
in the other of the outer zones. Being satisfied that it is merely a

* As noticed by Professor Lewis, ut supra, p. 22, in the diopside the prism
cleavage has practically disappeared, and a clinopinacoidal cleavage replaces the ^
orthopinacoidal usual in diallage.

The Parent-rock of the Diamond in South Africa. 231

variety of the last-described rock, I have preferred to leave it as an
intact boulder.

6. The next fragment, measuring about 3 in. by 2\ in. by 2 in., and
retaining part of its smooth outer surface, is labelled " found in the
yellow ground of No. 2 mine,* 50 feet deep." Though it is much more
decomposed than the others, the purplish garnet, the emerald-green
pyroxene, the altered enstatite (here very rotten), and a flake or two
of phlogophite (?) are easily made out. It is obviously a more
decomposed specimen of the rock represented by the two preceding-
specimens.

7. The last of this group of specimens is a rock fragment,! measuring
about 3 J in. by 2 in. in length and breadth, and slightly exceeding an
inch in greatest thickness. Its outline is irregular, being determined
by the fracture of the predominant diallage-like mineral. The crystals
of this run large, an inch or more in length, breadth, and thickness.
It is greyish-green in colour, having one dominant cleavage, with a
sub-metallic lustre, and close subordinate cleavages, giving a somewhat
fibrous aspect to that surface. Between these large crystalline lumps,,
numerous small, ill-defined garnets (pyrope) seem crowded, so as to
form fairly continuous partings, generally hardly 0T inch in thickness.
As the readiness with which the rather soft pyroxenic constituent split
away made it improbable that a good slice could be cut, and I was
reluctant to injure the specimen, I contented myself with detaching a
few flakes of this constituent for microscopic work, since the deter-
mination of its identity was sufficient for my purpose. These show the
mineral to have one easy cleavage and a rather fibrous structure ; they
give straight extinction parallel with this. As the usual rings and
brushes can be seen on the face of easy cleavage, the mineral belongs to
the bastite group. The same is true of the enstatite in boulder (4),
though, as it is slightly more fibrous, and not in quite so good a condi-
tion, the optical picture is less distinct. Thus we may name the rock
from which the present specimen has been broken, a garnet-bearing
bastitite.

8. This specimen, said to be a fragment of a boulder, is very different
from the rest. It is a compact greenish-grey rock containing enclosures,
which give it the aspect, at first sight, of a pebbly mudstone. Micro-
scopic examination shows it to be a compact felspathic diabase, with
vesicles, which have been filled up with calcite, chlorites, and other
secondary minerals (probably zeolites), but not to have any special
interest. Its relations appear to be with the rocks occurring in a con-
glomerate which we shall mention in a later paragraph.

* The others come from another mine (No. 1).

f I am informed that this was not part of a boulder, but came out of the " blue
ground " nearly in its present condition.

232

Prof. T. G. Bonney.

The " Blue Ground " and Associated Bocks.

Two areas of cliamantiferous rock are now being worked at the
Newlands Mines. The shape of the one which supplied most of the
specimens described in this paper is irregular, and, so far as I know,
exceptional. Its outline at the surface may be roughly compared to a
rounded triangle into the base of which the point of a rather short
shuttle is thrust, the greatest breadth of the two being about equal.
Exploratory workings at a depth of 300 feet show that the former
area rather quickly narrows, and the latter terminates in clefts. The
" blue ground," in fact, appears to fill a fissure, broadening in two
places to vents, which has been traced for some distance underground
southwards from the principal mass of diamantiferous rock, as repre-
sented in the annexed section, where the latter is dotted.

Fig. 3.

An igneous rock (i) occurs on either side. It is compact, a greenish-
grey in colour, not unlike some of the less acid Welsh feist ones.
Under the microscope it is found to be much affected by secondary
mineral changes, the iron oxides alone being in good preservation.
A few small crystals of decomposed felspar are scattered in a yet more
decomposed matrix, of which the minor details are uninteresting. The--
rock may be classed with the compact, rather felspathic, diabases.
These walls of diabase, farther to the south, turn off rather sharply to
east and west. In the interval, about 12 feet in width, between them,
ribs of the " blue " and a mudstone alternate, the thickest one of the
former being from 3 to 4 feet in width, and the inner part of it is in
better preservation than the outer. Specimens have been examined from
the heart of the mass (vii), a part outside it (vi), and the exterior portion
(v). The first (vii) in texture, hardness, and colour reminds me a little
of the dark serpentine found north of Cadgwith, in Cornwall. In this
matrix roundish spots occur, some darker than it, others a yellow-
green colour, besides a few angular whitish spots. The block is
traversed by two or three thin calcareous veins. Specimen (vi) while

The Parent-rock of the Diamond in South Africa. 233

generally similar is more decomposed, and apparently contains some
fragments of shale. Specimen (v) has a stratified aspect, being a dull
grey faintly mottled rock, with streaky, dark, rather carbonaceous-
looking bands ; the origin being doubtful, till it is seen under the
microscope. A fourth specimen (iii) shows the mudstone traversed by
a vein of rather pale-coloured decomposed " blue," not exceeding an
inch in thickness. A fifth (ii) is from near the diabase on the western
side, a dark compact rock, faintly mottled, here and there presenting a
slight resemblance to a " blue " traversed by thin veins of a carbonate ;
and a sixth (iv) from a like position on the opposite side is a generally
similar rock, but with wider veins filled with more coarsely crystalline
calcite. A seventh specimen represents the " blue " in the " neck," a
few yards to the north and at the same level (300 feet). This, inferior
in preservation to the first-named, includes numerous rounded frag-
ments a little darker than the matrix, with others, angular to sub-
angular, some also darker and some lighter than it.

A brief summary of the results of microscopic examination may
suffice, as these rocks do not materially differ from specimens obtained
in the De Beers Mine, of which I have published a full account.*

The matrix is a mixture, in slightly variable quantities, of granules
of calcite or dolomite, serpentine, pyroxene, and iron oxides, in which
occur flakes with fairly idiomorphic outlines of a warm-brown mica,
moderately pleochroic, corresponding with that described! in one or
two specimens from De Beers Mine. The prisms are about 0'002 inch
in diameter, and sometimes nearly as thick. This mica, which, as
stated in a former paper, I consider a secondary product, occurs abun-
dantly in all the specimens, but in that from the interior (on the whole
the best preserved rock) it is locally assuming a green colour, no doubt
by hydration. In the specimens from the thick rib, the one last
named contains mineral grains and rock fragments. Except for a few
flakes of the usual mica, the former are a mixture of two fibrous
minerals, the larger part corresponding with actinolite ; the rest,
giving lower polarisation tints, may be serpentine. This fact, and
structures suggestive of the former presence of a cleavage more regular
than that of olivine, make it more probable that diopside was the
original mineral. Though iron oxide is present in specks and rods
(especially in the worse preserved specimen), this occurs either in the
outer part, or as though it had been deposited along cleavage planes.
In the thin rib of " blue " (iii), some of the grains are composed partly
of a fibrous mineral, as above described, and partly of a clear one, which
often affords rather rich polarisation tints, and presents some resem-
blance to quartz. Its precise nature is difficult to determine, owing
to the absence of distinctive characters, but I believe it to be of second-

* « Geol. Mag.,' 1895, p. 492 ; and 1897, p. 448.
f < Geol. Mag.,' 1697, pp. 450, 451.

234

Prof. T. G. Bonne}'.

ary origin. Rock fragments are not common in the first (interior)
specimen (vii) ; one, however, is probably an altered shale, and another
possibly a limestone. This is bordered by a pale, pyroxenic mineral
piercing into the grains of calcite. In the second specimen (vi) frag-
ments are rather common ; among them are those of diabase, ranging
from fine to coarse, one specimen of the latter, originally, perhaps, an
inch in diameter, showing an ophitic structure ; felspar and augite both
being rather altered, seemingly by infiltration, and one small fragment
resembles a subcrystalline limestone. Specimen (v) does not materially
differ, but seems to contain more carbonate than the others. The dark
streaking is due to grains of iron oxide or serpentine with much opacite ;
rock fragments are few and small. Specimen (iii) from the thin vein
contains a few very small rock fragments, mudstone or shale, more or
less altered, possibly also a compact diabase. The " country rock " is a
mudstone, consisting of small chips of quartz and felspar, variable
in size, embedded in a dusty matrix, including a carbonate, which is
more abundant within about a fiftieth of an inch from the junction.
This part is slightly stained, but I was unable to detect any signs of
contact metamorphism. Specimens (ii) and (iv) are generally similar,
but the former contains some small rounded bits of varieties of diabase,
and one may represent a crystalline limestone. The veins are filled
with calcite and other secondary products, and are bordered by a very
thin film of a brown micaceous mineral, like that described as often
permeating the "blue." Both specimens suggest micromineralogical
changes, such as might be produced by the passage of hot water.

Other specimens of the sedimentary rock in the immediate neigh-
bourhood of the blue have been forwarded to me by Mr. Trubenbach.
One, from the adit on the southern side of the section mentioned above,
is a grey mudstone, containing a flattish rectangular pebble, of a dark
green compact rock. Two others are from No. 2 mine, or about 700
yards to the south-west. One, struck in the shaft at a depth of 200 feet,
is a conglomerate, composed of well-rounded rock fragments, with some
scattered grains of quartz. Each of the former is bordered by a zone
of a crystalline carbonate (impure calcite), and the interstices are
filled, sometimes by a clearer variety of the same, but more often by
some minutely granular secondary product. Of the rock fragments, one
is a subcrystalline dolomitic limestone ; two, perhaps, are chalcedony; the
remainder are igneous ; the majority being varieties of diabase, some-
times rather decomposed ; the rest trachytes, mainly andesites. Their
general aspect and the not unfrequent presence of vesicles (now filled
with viridite) suggest that they have been furnished by lava-flows.
Another specimen, obtained in the same working at a depth of 400 feet,
is a rather f elspathie diabase, not unlike one of the varieties in the
conglomerate. It is a good deal decomposed, is not improbably from
a lava-flow, but does not call for a minute description.

The Parent-rock of the Diamond in South Africa. 235

Conclusion.

Thus the diamond has been traced up to an igneous rock. The
" blue ground " is not the birthplace either of it or of the garnets,
pyroxenes, olivine, and other minerals, more or less fragmental, which
it incorporates. The diamond is a constituent of the eclogite, just as
much as a zircon may be a constituent of a granite or a syenite. Its
regular form suggests not only that it was the first mineral to crystal-
lise in the magma, but also a further possibility. Though the occur-
rence of diamonds in rocks with a high percentage of silica (itacolumite,
granite, &c.) has been asserted, the statement needs corroboration.
This form of crystallised carbon hitherto has been found only in
meteoric iron (Canyon Diablo),* and has been produced artificially by
Moissan and others with the same metal as matrix. But in eclogite
the silica percentage is at least as high as in dolerite • hence it is diffi-
cult to understand how so small an amount of carbon escaped oxidation.
I had always expected that a peridotite (as supposed by Professor
Lewis), if not a material yet more basic, would prove to be the birth-
place of the diamond. Can it possibly be a derivative mineral, even
in the eclogite ? Had it already crystallised out of a more basic
magma,f which, however, was still molten, when one more acid was
injected, and the mixture became such as to form eclogite 1 But I
content myself with indicating a difficulty, and suggesting a possi-
bility ; the fact itself is indisputable : that the diamond occurs, though
rather sporadically, as a constituent of an eclogite, which rock, accord-
ing to the ordinary rules of inference, must be regarded aw its birth-
place.

This discovery closes another controversy, viz., that concerning the
nature of the "Hard blue" of the miners (Kimberlite of Professor
Lewis), in which the diamond is usually found. The boulders de-
scribed in this paper are truly water worn. The idea that they have
been rounded by a sort of " cup and ball " game played by a volcano
may be dismissed as practically impossible. Any such process would
take a long time, but the absence of true scoria implies that the explo-
sive phase was a brief one. They resemble stones which have travelled
for several miles down a mountain torrent, and must have been derived
from a coarse conglomerate, manufactured by either a strong stream or
the waves of a sea from fragments obtained from more ancient crystal-
line rocks.]; The "washings," a parcel of which I received from

* P.S. — The Novo Urei meteorite (not wholly iron) was forgotten.

f This, however, cannot have been very rich in iron, because diopside does not
contain much of that constituent.

t As these eclogites are very coarsely crystalline, we are justified in assuming
they were once deep-seated rocks, and so much more ancient than the date of the
conglomerate. To prevent any misunderstanding I may repeat that the matrix

236 The Parent-rock of the Diamond in South Africa.

Mr. Trubenbach, also show that the boulders are really waterworn.
Besides two unworn pieces of pyrite and a rough bit of eclogite, about
three-quarters of an inch in diameter, the pyroxenic constituent of
which was a bright emerald-green (? smaragdite), I find part of a sub-
angular fragment of chrome-diopside associated with two or three
flakes of the usual mica, a well rounded garnet fully 06 inch across,
and half a well worn pebble of eclogite, about one inch long and half an
inch thick. The rounded waterworn look of the great majority of
the smaller constituents (chiefly garnets and pyroxenes), about the size
of hemp seed, is very obvious. I had suspected some of the grains in
washings from the De Beers Mine to have been similarly treated ; but
here it is indubitable, indeed many of the dark green specimens are so
smooth outside that they could only be identified after fracture. The
ordinary diopside can, however, be recognised, with some of a clearer
and brighter green. Most of the garnets are pyropes, but a few re-
semble essonite. I find also some grains of iron oxide and of vein
quartz. Thus, the presence of waterworn fragments, large and small,
in considerable abundance, shows the "blue ground" to be a true
breccia, produced by the destruction of various rocks (some of them
crystalline, others sedimentary, but occasionally including waterworn
boulders of the former)— i.e., a result of shattering explosions, followed
by solfataric action. Hence the name Kimberlite must disappear from
the list of the peridotites, and even from petrological literature, unless
it be retained for this remarkable type of breccia.

Boulders, such as we have described, might be expected to occur at
the base of the sedimentary series, in proximity to a crystalline floor.
The Karoo beds in South Africa, as is well known, are underlain in
many places by a coarse conglomerate of considerable thickness and
great extent, called the Dwyka conglomerate, which is supposed to be
Permian or Permo-carboniferous in age. It crops out from beneath
the Karoo beds at no great distance from the diamond-bearing district,
and very probably extends beneath it. If this deposit has supplied
the boulders, the date of the genesis of the diamond is carried back^
at the very least, to Palaeozoic ages, and possibly to a still earlier era
in the earth's history.

from which these boulders were taken (at various depths, from nearly 100 to about
300 feet) cannot be any alluvial deposit, but is the typical "blue ground,"
practically identical with that in the Kimberley mines.

Photographic Researches on Phosphorescent Spectra. 237

" Photographic Kesearches on Phosphorescent Spectra : on Vic-
torium, a New Element associated with Yttrium." By Sir
William Ceookes, F.RS. Eeceivecl May 2, — Eead May 4,
1899.

[Plate 9.]

It has long been known that certain substances enclosed in a vacuous
glass bulb phosphoresce brightly when submitted to molecular bom-
bardment from the negative pole of an induction coil. The ruby,
emerald, diamond, alumina, yttria, samaria, and a large class of earthy
oxides and sulphides, emit light under these circumstances. Examined
in a spectroscope the light from some of these bodies gives an almost
continuous spectrum, while that from others, such as alumina, yttria,
and samaria, gives spectra of more or less sharp bands and lines. Since
1879 I have been working on these phosphorescent spectra, chiefly in
connection with the earths of the yttria group, and by chemical frac-
tionation I have succeeded in separating from this group bodies whose
phosphorescent spectra consist chiefly of single groups of lines, other
groups being absent. For the last six years the research has been
extended beyond the visible spectrum, and photographs of the ultra-
violet portion of the spectra are now being taken with a spectrograph
with a complete quartz train. Some of the results of this investigation
were exhibited at the soiree of the Koyal Society, on the 3rd of May.
A preliminary mention of the discovery of a new element was made in
my address to the British Association in September last, when I pro-
visionally called it Monium ; but for several reasons I now consider the
name Victorium more appropriate.

The complicated scheme of fractionation carried on for so many years
is illustrated in the accompanying diagram (Plate 9). This must be
considered only as an indication of the methods employed, and not as
an actual representation of every operation through which the material
has passed. Crude yttria, from samarskite, gadolinite, cerite, and
other similar minerals, is the raw material. The first operation is to
free it roughly from earths of the cerium group — an operation effected
by taking advantage of the fact that the double sulphates of potassium
and the yttrium metals are easily soluble in saturated potassium
sulphate solution, while the corresponding double sulphates of the
cerium group of metals are difficultly soluble.

After this preliminary treatment the crude yttria is converted into
nitrate, represented by the topmost circle on the diagram. The nitrate
is exposed to heat until it fuses to a clear liquid, care being taken to
distribute the heat uniformly through the mass. Presently the liquid
mass commences to decompose, giving off red vapours. After this has
proceeded for a little time the fused mass is carefully poured into

VOL. LXV. T

238

Sir TV". Crookes

water, and the liquid well boiled. A white precipitate of basic nitrate
forms, while the undecomposed nitrates remain in solution. These are
separated by filtration, the precipitate going to the right and the solu-
tion to the left. The basic nitrate is dissolved in nitric acid, and the
right and left solutions are then evaporated to dryness and fused as
before. Partial decomposition by heat again divides each of these
portions into two lots, soluble and insoluble. The soluble from the
left-hand lot goes still further to the left, and its insoluble portion to
the right. The soluble from the right-hand portion goes to the left,
where it mixes with the insoluble from the other portion, while its
insoluble goes still further to the right. This series of operations is
continued for as long a time as the material will hold out.* From a
description the process seems to be more complicated than it really is,
but a study of the diagram and the direction of the arrows make it
clear. The number of times this operation is performed varies with
each lot of earth fractionated. The portions submitted to fusion
rapidly diminish in quantity, and the operation is continued until the
material becomes too scanty.

The last horizontal line of fractions spectroscopically examined in a
radiant matter tube shows differences in the visible spectrum. For
many years I recorded these differences in coloured drawings, which
have served on several occasions to illustrate papers before this Society.!
In the year 1893 I commenced to record the differences between the
various spectra by photographing them in a spectrograph having a
complete quartz train, and since that time attention has chiefly been
directed to the variations in the number, character, and positions of the
lines and bands in the ultra-violet spectrum : these are more striking
than those which are visible, and as they are self-recording, results are
more rapidly attained. A description of this instrument is given
further on.

On placing the photographed spectra of one of the horizontal lines
of earths in order, several differences are detected. One striking
difference is seen in the behaviour of a group of lines in the ultra--
violet. It is nearly absent in the end fractions, gradually becoming
stronger towards the middle, and attaining a maximum in the fractions
situate about two-thirds towards the right. This shows that at least
three different bodies are present : one, the great bulk, having a nitrate
difficult to decompose ; another whose nitrate is easiest to decompose ;
and a third body, occupying an intermediate position, whose nitrate
decomposition occurs at temperatures between that required by the
others, but nearer that of the nitrate easiest decomposed.

* " On the Methods of Chemical Fractionation," British Association, Birming-
ham Meeting, 1886; ' Chemical Xews,' vol. 54-, p. 131.

f "On some ~Ne\v Elements in Gadolinite and Samarskite detected Spectro-
scopically, 1 Roy. Soc. Proc.,' No. 245, 188G, vol. 40, p. 502."

Photographic Researches on Phosphorescent Spectra. 239

The above method of fractionation is not so effectual if more than
two bodies are present. In that case the process fails, in any reason-
able time, to yield practically pure specimens of more than two out of
a group of closely allied earths. Thus, if there are three earths — say,
A, B, and C — whose positions in reference to the chemical process em-
ployed are in the written order of sequence, we may get a specimen of
A as nearly as we please free from B and C, and a specimen of C as
nearly as we please free from A and B, but we cannot get a specimen
of B practically free from A and C. The law seems to be that to ob-
tain practically pure specimens of three closely allied earths, it is essen-
tial to have recourse to at least two different chemical processes. The
mere continued repetition of the same process will not do, unless indeed
the operations are repeated such a vast number of times as to make the
approximate expressions no longer applicable, even though the sub-
stances are chemically very close.

For this and other reasons it is advisable to change the method of
fractionation after one process has been in operation for some time.
It is evident that any process of fusion, crystallisation, or precipitation
can only divide the mass of material into two parts, a soluble and
insoluble portion, crystals and mother liquor ; and after a time a
balance of affinities seems to be established, and further fractionation
appears to do little good. It is better then to change the operation.

Following the diagrammatic scheme, the portions of earths contain
ing most victorium are collected together and fractionated by the crys-
tallisation of the oxalates from a solution strongly acidulated with
nitric acid in the following manner : —

To a boiling aeid solution of the nitrate a small quantity of hot
solution of oxalic acid is added. The solution remains clear, and it is
only after vigorous stirring that a small quantity of insoluble oxalate
is formed. The whole is thrown on a hot-water filter and slightly
washed with boiling water. To the boiling nitrate a fresh lot of hot
solution of oxalic acid is added, and stirred till more insoluble oxalate
comes down. This is again filtered off, and the operations of precipi-
tating, stirring, filtering, and washing are repeated, always keeping
the temperature as near the boiling point as possible, until the whole
of the earths are precipitated. Generally the initial earth is divided
by this method of fractionation into from six to twelve portions. Each
of these oxalates is dried, ignited, dissolved in nitric acid, and the
above-described operations repeated. Photo-spectroscopic tests are con-
stantly taken during the progress of this fractionation, and portions
are mixed together according to the data thus obtained, as shown on
the diagram by the lines joining the fractions ; the object being to
avoid lateral spreading as much as possible, and, while concentrating
the special line-giving earth, to prevent its too great diffusion over a
large number of fractions. When the fractionation by the oxalate

T 2

240

Sir W. Crookes.

method has proceeded for a considerable time, the fractions rich in
victorium are collected together and submitted to another mode of
treatment.

These fractions are converted into nitrates, and a small quantity is
thrown out by partial decomposition by heat, according to the method
already described. The filtrate is evaporated to dryness and again
fused so as to throw out a little more. This operation is repeated as
long as any soluble nitrate is left. Generally from six to twelve por-
tions are thus obtained. These form a regular series, differing accord-
ing to the stability of the nitrate under heat. On testing, the
victorium is found to concentrate in the centre portions, being less
easily decomposed than the earths of the cerium group and more
easily decomposed than those of the yttrium group.

The fractions rich in victorium are converted into sulphates and
mixed with a hot saturated solution of potassium sulphate. The pre-
cipitate is dissolved in boiling water and mixed with a further quantity
of solution of potassium sulphate. This produces a small quantity of
a precipitate. The filtrate from the first precipitate is also mixed with
fresh potassium sulphate, and the operations are repeated, mixing the
centre solutions to one lot and the side solutions to another, as shown
by the lines on the diagram. It is found on photo-spectroscopic ex-
amination that the earths thrown out on each side are poorest in
victorium, whilst those in the middle are richest. After a time no
further concentration is effected in this manner, all the earths that can
be removed as being more or less soluble in potassium sulphate having
been eliminated.

In thus describing the method of fractionation, my object has been
not so much to give a description of the plan actually carried out in
the laboratory — for the details have varied with each operation — but
to give an intelligible idea of the general manner in which a very com-
plicated operation is effected. *In the diagram I am supposing that one
particular substance, victorium, is to be separated, and I have endea-
voured to show its migrations and gradual concenoration as the work ^
progresses, by tinting the fractions where it mostly would concentrate ;
the depth of tint representing the amount of concentration.

In the purest condition yet obtained victoria is an earth of a pale
brown colour, easily soluble in acids. It is less basic than yttria and
more basic than most of the earths of the terbia group. In chemical
characters it differs in many respects from yttria. From a hot nitric
acid solution, victorium oxalate precipitates before yttrium oxalate and
after terbium oxalate. On fractional precipitation with potassium
sulphate the double sulphate of victorium and potassium is seen to be
less soluble than the corresponding yttrium salt, and more soluble
than the double sulphates of potassium and the terbium and cerium
groups. Victorium nitrate is a little more easily decomposed by heat

Photographic Researches on Phosphorescent Spectra. 241

than yttrium nitrate, but the difference is not sufficient to make this
reaction a good means of separating victorium and yttrium. Fusing
the nitrates can, however, be employed advantageously to separate
mixed victoria and yttria from the bulk of their associated earths.

On the assumption that the oxide has the composition VC2O3, the
atomic weight of victorium is apparently not far from 117.

The photographed phosphorescent spectrum of victoria consists of
a, pair of strong lines at about X 3120 and 3117 ; other fainter lines
are at 3219, 3064, and 3060. Frequently the pair at 3120 and 3117
merge into one, but occasionally I have seen them quite distinct.
The presence or absence of other earths has much influence on the
sharpness of lines in phosphorescent spectra, and it is probable that
these lines will be sharp and distinct when victoria is obtained quite
free from its associates.

The best material for phosphorescing in the radiant matter tube is
not the earth itself, but the anhydrous sulphate formed by heating the
earth with strong sulphuric acid and driving off the excess of acid at
a red heat. The sulphate thus produced, probably also containing
some basic sulphate, is powdered and introduced into a bulb tube fur-
nished with a quartz window, and a pair of thick aluminium poles
sealed into the glass with stout platinum wires. The tube is well
exhausted, keeping the current from a good induction coil going all
the time. The pumping and sparking must continue until the earth
glows with a pure light free from haze or cloudiness, and continues so
to glow during the passage of the current without deterioration.
The exposure in the spectrograph usually occupies an hour.

I give a diagrammatic plan of the two-prism spectrograph used in
this research. It is furnished with two quartz prisms, quartz lenses,
and condensers. The slit jaws are of quartz, cut and polished accord-
ing to the method I described in the ' Chemical News,' vol. 71, p. 175,
April 11, 1895.

The prisms are made in two halves, according to Cornu's plan ; one
half of each being right-handed and the other half left-handed. One
of the lenses also is right-handed and the other left-handecl. By this
device the effect of double refraction is so completely neutralised that
with a five-prism instrument it is impossible, under high magnifying
power, to detect any duplication of the lines.

The lenses are each of 52 mm. diameter and 350 mm. focus. The
focus of the least refrangible rays is longer than that of the most
refrangible rays, and the sensitive film must therefore be set at an
angle to get the extreme rays into focus at the same time. But this
alone is not sufficient. The focal plane is not a flat surface, but is
curved, and the film must therefore be curved,* and it is only when

* ' Chemical News,' vol. 72, p. 87, August 23, 1895 ; and vol. 74, p. 259, Foyem-
ber 27, 1896.

242 Photographic Researches on Phosphorescent Spectra,

Experimental Contributions to the Theory of Heredity. 24^

both these conditions are fulfilled that perfectly sharp images of
spectral lines extending from the red to the high zinc line 2138*30 can
be photographed on the same surface. Celluloid films are used, glass
not being sufficiently flexible.

Using the middle position showing the whole spectrum on a plate,
the angle is 403, and the curvature is 190 mm. radius.

The condensers are of quartz, and are piano-cylindrical — one being-
double the focus of the other. The object of this, when spark-spectra
are being photographed, is to concentrate on the slit a line instead of
a point of light, as would be the case if ordinary lenses were used.

^lien photographing phosphorescent spectra — or, in fact, any spectra
the wave-lengths of which are either unknown or require verification
— I always photograph on the same film a standard spectrum, usually
of an alloy of equal molecular weights of zinc, cadmium, tin, and
niercury. This forms a hard somewhat malleable alloy, giving
throughout the whole photographic region lines the wave-lengths of
which are well known. The chief objection to this alloy is its volatility
— the poles requiring frequent adjustment. Recently I have used
pure iron for this purpose \ this has the advantages of giving a great
number of fine lines whose wave-lengths are accurately known, and not
being very volatile, the poles do not rapidly wear away. If the poles
are kept about 1 mm. apart there is little or no interference from
air lines.

The most simple method of applying the standard lines to an
unknown spectrum is by the successive employment of two slightly
overlapping diaphragms immediately behind the slit, one being used
for the experimental and the other for the standard spectrum. In this
way, without disturbing the instrument, the two spectra can be recorded
on the plate one over the other ; the overlap of 1 mm. being in the
optical centre of the train. The resulting negative is then transferred
to a micrometer measuring machine of special construction, having a
screw of 1 100 of an inch pitch, and a means of accurately determining
1 1000 of its revolution — thus measuring directly to the hundred-
thousandth of an inch. In this way, in a five-prism spectrograph
having lenses 700 mm. focus, it is possible to determine wave-lengths
of photographed lines to the sixth figure.

Experimental Contributions to the Theory of Heredity. A. Tele-
gony/' By J. C. Ewaet, M.D., F.E.S., University of Edin-
burgh. Eeceived May 29 — Bead June 1, 1899.

I. Introductory.

The belief in telegony, or what used to be known as the " infection
of the germ " or " throwing back " to a previous sire, has long prevailed

244

Dr. J. C. Ewart.

It may for all we know be as old as the belief in " mental impressions/'
which has had its adherents since at least the time of the patriarchs.
During the eighteenth century the " infection " doctrine was frequently
discussed by physiologists, and since Lord Morton, in 1820, addressed
a letter to the Royal Society on the subject, believers in " infection "
have been increasing all over the world, with the result that one seldom
now hears of breeders or fanciers who are not influenced by the doc-
trine, while physicians and others interested in the problems of here-
dity either as a rule take telegony for granted or see nothing improbable
in the " infection " hypothesis.

It must, however, be admitted that, notwithstanding the criticisms of
Weismann and others, very different views are entertained by the
believers in telegony, not only as to the cause, but as to the results, of
" infection." By some telegony is confounded with simple reversion or
atavism, while the better informed generally assume that " infection "
invariably results in the subsequent offspring repeating more or less
accurately the characters of the first or of a previous sire. In a
breeders' journal of some standing there appeared recently under the
heading " Colour of Animals " the following sentence : — " Greys show
in breeding a great tenacity of assertion, as they are few in comparison
to other colours in the Stud Book, but they reappear and no doubt go
back to the Arab, and prove telegony to be a fact."* This shows
simple reversion is sometimes mistaken for telegony. In support of
the view that " infection " is commonly supposed to lead to " throwing
back " to a previous sire many instances could be given, but the follow-
ing from an article on telegony by De Varigny will suffice. De Varigny
states that an ordinary cat which had kittens to a tailless Manx cat
subsequently produced several tailless kittens to a normal cat of her
own breed, f

An extended series of experiments with various kinds of animals has
led me to the conclusion that if there is such a thing as telegony, it is
more likely to result in the subsequent offspring "throwing back " to an
ancestor of the " infected" dam than to a previous mate. This view of
telegony (which has not been insisted on hitherto) will be made at once
evident by an example. A sable collie crossed with a Dalmatian pro-
duced three pups which in their coloration are extremely like young-
foxhounds ; instead of numerous small spots each has a few large
blotches. According to the common view of telegony this collie, if
infected, should next produce with a dog of her own breed one or more
Dalmatian-like pups. If, however, the offspring of a collie and a Dal-
matian are like foxhounds, the subsequent offspring to a collie of the
same colour and strain could hardly be expected to present Dalmatian
characters, i.e., show numerous small spots. But if " infection " as a

* ' Live Stock Journal/ May 12, 1899, p. 588.
f ' Journal des Debats,' September 9, 1897.

Experimental Contributions to the Theory of Heredity. 245

rule results in the subsequent offspring " throwing back " either to the
ancestors of the sire or the clam, it will be extremely difficult, if not
in many cases impossible, to distinguish telegony from simple rever-
sion.*

But though " infection," if it does take place, is likely, as a rule, to
lead the subsequent offspring to resemble the ancestors of the dam, it
may in certain cases possibly lead to their " throwing back" to a pre-
vious sire. This result might follow if the previous sire happened to
be highly prepotent. For example, Highland heifers often produce to a
Galloway bull hornless black offspring indistinguishable from pure
Galloways. If infected by the Galloway bull, these heifers might
afterwards produce Galloway-like calves when mated with long-hornecl
bright coloured bulls of their own breed.

It is now commonly believed that if there is such a thing as telegony
it results from the unused germ cells of the first (or previous) sire
infecting — blending with — the unripe germ cells in the ovaries of the
dam. Were this possible, the subsequent progeny would in all proba-
bility in a mild way resemble the previous sire, but if this is impos-
sible, then infection — due perhaps to some obscure change in the con-
stitution or reproductive system of the dam — is more likely to lead to
more or less marked reversion to the ancestors of the dam. All my
observations point to its being impossible in the Equidse for the
unused male germ cells of the first sire to infect the unripe ova. The
spermatozoa lodged in the upper dilated part of the oviduct of the
mare are dead, and in process of disintegrating, eight days after insem-
ination ; they probably lose their fertilising power in four or five days.
There is no reason for supposing that in the Equidse they survive
longer in or around the ovary. Further, though at the time of fertili-
sation there may be several large Graafian follicles in each ovary con-
taining maturing ova, all these follicles disappear long before the
period of gestation is completed. The subsequent foals are developed
from successive new crops of ova into the composition of which it is
inconceivable any of the spermatozoa of the first sire could by any
chance enter. A study of the ovaries hence tends to confirm the view
that " infection " (if there is such a thing) is as likely to cause rever-
sion to a former ancestor of the dam as a " throwing back " to a pre-
vious sire.

Having made these general observations, it will be well next to con-
sider critically the case of " infection " communicated in the letter to
the President of the Eoyal Society in 1820 by the Earl of Morton.

* That reversion ever occurs has been questioned by Bateson (' Materials for
the Study of Variation') and others, but I have already (' Mature,' February 9,
1899) proved beyond doubt that reversion can be easily induced by intercrossing
distinct types, and I have recently heard of several instances of spontaneous reversion
— reversion not induced by intercrossing.

246

Dr. J. C. Ewart.

Though many other instances of supposed " infection " have been
recorded, Lord Morton's mare may be said to still hold the field — the
theory of telegony still mainly rests on the assumption that this
historic mare was " infected " by a quagga some years before she
passed into the hands of Sir Gore Ouseley and produced three " colts
to a black Arabian horse. One might even go further and without
much exaggeration assert that the telegony hypothesis at the present
moment mainly rests on an allegation by Sir Gore Ouseley's stud
groom.

It has been generally assumed that Lord Morton's mare (a nearly
purely bred chestnut Arab) was " infected " for two reasons (1) because
the subsequent offspring were of a yellowish-brown colour and more or
less striped, and (2) because, according to Sir Gore Ouseley's stud
groom, the mane of one of the striped foals had always been upright,,
while in another it arched to one side clear of the neck. The presence
of stripes in the subsequent offspring has never been questioned, nor
yet is there any doubt that when Lord Morton in 1820 inspected the
" colts " the mane in the filly was upright as in the quagga, while that
of the colt resembled the mane of Lord Morton's quagga hybrid.
There is, however, an absence of reliable evidence that the filly's mane
had always been upright as alleged to Lord Morton by Sir Gore
Ouseley's stud groom.

Were the evidence in support of this allegation satisfactory, there
would I think be no escape from the conclusion that Lord Morton's
mare was " infected " by the quagga. Hitherto the presence of stripes
on the " colts " has generally been looked upon as affording strong
evidence of " infection." Believers in telegony admit that stripes are
not uncommon in Norwegian and certain other breeds of horses, but,
with Mr. Darwin, they have taken for granted that they never or very
rarely occur in Arabs.

I find, however, that though in Arabia dun-coloured horses are dis-
liked and never used for breeding, stripes even in the most renowned
strains are not so uncommon as is generally supposed. I have now a
purely bred Arab filly of about the same colour as Lord Morton's filly f
but, unlike the filly we have heard so much of, both the fore and hind
legs are marked with distinct dark bars, and there are faint indications
of stripes across the withers and a distinct dorsal band. The history
of this filly (bred by Mr. Wilfred Sea wen Blunt at Crabbet Park,
Sussex, and very kindly presented to me) is well known for many
generations ; none of her ancestors could possibly have been " infected 'r
by a zebra. The dun colour and stripes are doubtless the result of
simple spontaneous reversion, for, unlike Lord Morton's mare, there is
no history of a cross in her pedigree. This filly proves that even in
high-caste Arabs of the best desert blood a dun colour and stripes may
unexpectedly appear.

Experimental Contributions to the Theory of Heredity. ' 247

As to the occurrence of stripes in other breeds I could give, were it
necessary, many instances. A year ago I had in my possession a light
bay (or yellow dun) pony, which showed nearly as many stripes on
the trunk as the Gore-Ouseley filly, and in addition had several in-
terrupted narrow stripes on the forehead.* Moreover, the stripes on
the Gore-Ouseley " colts," while agreeing with stripes occasionally seen
in horses, differ in their arrangement from the stripes in the quagga..
The stripes themselves are evidence of reversion, but nothing more,
and seeing that pure bred horses sometimes show quite as many stripes,
we are not justified in assuming that but for the dam of the " colts "'
having been first mated with a quagga the stripes would not have
appeared.

Hence unless it is proved that the mane in the filly and colt were
naturally erect, or nearly erect, the case for the " infection " of Lord
Morton's mare will be lost. It may be well to quote the passage from
Lord Morton's letter referring to the mane. It is as follows : — " That
of the filly is short, stiff, and upright, and Sir Gore Ouseley's stud
groom alleged it never was otherwise. That of the colt is long, but so
stiff as to arch upwards and to hang clear of the neck, in which circum-
stance it resembles that of the hybrid. This is the more remarkable
as the manes of the Arabian breed hang lank and closer to the neck
than that of most others."!

I am not prepared to accept the allegation as to the manes for the-
following reasons : —

1. I have had twelve zebra hybrids under observation, and in
each case the mane, though erect to start with, always after a time
arched over to one or both sides. The stud groom's statement, it
seems to me, proves too much. If in the quagga hybrid and in all
my horse hybrids the mane, sooner or later, falls to one side, it is
a little remarkable that in the pure bred two-year-olcl filly it had been
always upright.

I may here mention that the hair of the mane of zebra hybrids is
shed annually ; it is for this reason that the mane in hybrids is never
long enough to hang close to the neck.

2. The mane in the drawing of the filly by Agasse is not represented
as upright, but as lying to one side. If the mane had remained erect
during the first two years, by virtue of shedding its hairs, it could not
very well have lost this habit and fallen completely over to one side
subsequently, say, during the fourth year. From the mane being erect
in 1820, and hanging to one side in 1821 or 1822, when Agasse's
drawing was made, the presumption is that the mane of the " colts "
had been cut some time before they were examined by Lord Morton.

Two years ago I had a bay Arab with a mane which was to start

* See Fig. 36, ' The Penycuik Experiments,' A. and C. Black, 1899.
f < Phil. Trans.,' 1821.

248

Lord Arthur Cecil and Dr. J. C. Ewart.

with short, stiff, and upright, some months later it arched freely to
one side, as in my zebra hybrids, and later still it hung lank and close
to the neck.

3. There is always an intimate relation in the Equidse between the
mane and the tail, when the mane is short and erect the upper third or
so of the tail is only covered with short hairs, which, like the hairs of
the mane, are annually shed. Lord Morton noticed nothing peculiar
about the tail of the " colts," and the tail of both the colt and filly in
Agasse's drawings is the tail of a high-caste Arab. This seems to me
to warrant the conclusion that the filly's mane had been hogged some
time before Lord Morton's visit.

It thus appears that the evidence in support of the belief that Lord
Morton's mare was " infected " by the quagga is at the best far from
satisfactory. The same may be said of the evidence in support of all
the other supposed cases of telegony in the Equidse — of, amongst
others, Lord Mostyn's mare, referred to by Darwin •* of the mule-like
mare in the Paris Gardens, referred to by Tegetmeier and Sutherland ;f
and of the African ass {E quits asinus), still in the Zoological Gardens
(London), which now and then has a reddish-coloured foal, like the
cross-bred foal she produced in 1883 to an Asiatic ass (E. hemionus).

Although I am now satisfied that Lord Morton's case throws little
light on the telegony hypothesis, like many others I had no very
decided views on the subject some years ago, and hence when ar-
ranging in 1894 to make a collection of horse embryos, I decided to
repeat, as far as circumstances permitted, what is commonly called
Lord Morton's experiment. For this purpose, I procured early in
1895 three zebras and a number of mares. Two of the zebras died
during the winter of 1895, but the third — a handsome stallion of the
Chapman variety (E. burchelli v. chapmctni) still survives and is now
thoroughly acclimatised.

During 1895 I only succeeded in mating the zebra with one mare,
and hence there was only one hybrid born in 1896. During the last
two years however, quite a number of hybrids have made their appear- %
ance, and the dams of several of the hybrids have subsequently pro-
duced pure-bred foals. The time has hence come, when some of the
results of the experiments may with propriety be communicated
to the Royal Society.

" II. Experiments with West Highland Ponies." By Lord Arthur
Cecil, Orchardmains, Kent, and J. C. Ewart.

The first mare mated with the zebra was a black, West Highland
pony (Mulatto), set apart for the telegony experiments by Lord

* ' Animals and Plants,' vol. 1, p. 435, 1875.
t ' Horses, Asses, and Zebras,' p. 81.

Experimental Contributions to the Theory of Heredity. 249

Arthur Cecil. The better bred West Highland ponies are supposed to
have descended from " Armada " horses, and are hence perhaps related
to Mexican and Argentine horses, so often dun-coloured and partially
striped. Mulatto's hybrid (Eomulus, born 12th August, 1896) is on
the whole more a zebra than a pony both mentally and physically.
He is especially remarkable in being more profusely striped than his
sire (the zebra Matopo), in having a heavy semi-erect mane, which is
shed annually, and in having a mule-like tail, from the upper third of
which the longer hairs are periodically shed. The body colour of the
hybrid varies from a dark orange colour to a mouse dun ; the stripes
of a reddish-browD colour on the head are dark brown or nearly black
on the trunk and limbs.

In the number and plan of the stripes, the hybrid agrees more
closely with the Somali zebra than with any of the Burchell zebras.
Over the brow, e.g., there are narrow rounded arches instead of some-
what broad pointed arches as in his sire, the neck and trunk have
quite double the number of stripes found in the sire, while over the
croup in the position of the " gridiron " of the mountain zebra there
were at birth irregular rows of spots which in course of time blended
to form somewhat zig-zag, narrow, transverse bands. The ears are
nearly as large as in the sire, while the eye-lashes are longer and dis-
tinctly curved. In his movements the hybrid resembles his sire, and
like his sire he is always on the alert, very active and suspicious of
unfamilar objects. Further in his call he agrees far more with his sire
than his dam. In being profusely striped, Romulus differs greatly
from the quagga hybrid bred by Lord Morton, in which the stripes
were fewer in number than in many dun-coloured horses.

Mulatto's second foal arrived in July, 1897, the sire, Benazrek,
being a high-caste grey Arab horse. Like Lord Morton's colts,
Mulatto's foal by the Arab horse, in make, action, and temperament,
agreed with ordinary foals, but it differed from the majority of
foals in presenting quite a number of indistinct stripes — subtle mark-
ings only visible in certain lights. These stripes differed but little
from the body colour, which varied from dark bay to brown. Though
few references have been made to the occurrence of stripes in foals,
they are, we find, far from uncommon. As is well known, Mr.
Darwin once bred a striped foal by putting a cross-bred bay mare to
a thoroughbred horse. This foal was for a time marked nearly all
over with obscure dark narrow stripes, plainest on the forehead, but
also distinct over the croup.*

There is no figure of Mr. Darwin's striped foal, but from the descrip-
tion given, there can be little doubt that the markings were more
abundant than in Mulatto's second foal. In this foal (as in Mr. Dar-
win's) the stripes became more and more indistinct, and by November
* ' Animals and Plants,' to! . 1, p. 60.

250

Dr. J. C. Ewart.

they had almost vanished. Unfortunately, the foal died when about
five months old, and hence it is impossible to say whether any of the
stripes would have persisted. It will be evident that Mulatto's second
-foal helped but little to clear up the vexed " infection " problem.
Mulatto missed having a foal in 1898, but she recently produced at
Knole, Kent, her third foal. The sire (Loch Corrie) of this foal
belongs to the Island of Rum section of the West Highland ponies,
and closely resembles Mulatto. The third foal has about as many
stripes as the second. As in the second, they are most distinct over
the croup and hind quarters ; and as in the second, they differ both
from the markings in the previous sire, the zebra, and from the
markings of the hybrid Romulus.

This third foal, which was born on the 6th of May, 1899, seemed
like the second, to lend some support to the " infection" hypothesis.
Fortunately, since it made its appearance, two other West Highland
mares have had foals to Loch Corrie. These foals put all doubt as to
the nature and significance of the stripes on Mulatto's second and
third foals at an end.

One of the clams is of a brown colour, the other is nearly black.
Though neither the brown dam nor the black has ever seen a zebra,
both foals are marked in very much the same way as Mulatto's, and
some of the stripes in one of the new foals look more like persisting
than the stripes on Mulatto's third foal. Hence, in order to account
for the markings on Mulatto's foal to the grey Arab, and on her foal
to the black West Highland pony, it is unnecessary to fall back on the
" infection " hypothesis.

" III. Experiments with Shetland, Iceland, Irish, Thoroughbred,
and other Ponies." By J. C. Ewart.

An effort was made to cross four Shetland ponies with the zebra
stallion, but I only succeeded in obtaining one hybrid. The dam
(Nora) of this hybrid closely resembles, except in size, the Island of^
Rum ponies — she is a small edition of Mulatto. Her first foal, by a
black Shetland pony, was of a dun colour and nearly as striped as
Sir Gore Ouseley's filly ; her second is the most zebra-like of all my
hybrids ; her third closely resembles her sire, a bay Welsh pony. For
some time after birth there were faint indications of stripes over the
hind quarters of this foal, but now it is a year old there are no mark-
ings or any other suggestions of a zebra. It is not a little suggestive
that the foal bred from this pony before she was mated with the zebra
was distinctly striped, while the subsequent pure bred foal has no
stripes.

Of five Iceland ponies put to the zebra only one produced a hybrid.
This hybrid was faintly striped, and showed less of the zebra than any

Gill.

Roy, Soc. Proc, Vol. 65, Plate 8.

Experimental Contributions to the Theory of Heredity. 251

of the others. The dam, a prepotent yellow and white (skewbald)
pony, had first of all a light bay foal to an Iceland pony. Her third
foal, by a bay Shetland stallion, is a skewbald, and in the size and
arrangement of the brown patches closely resembles the dam. There
is no hint whatever that the Iceland pony has been "infected" by the
zebra.

Several Irish mares were put to the zebra, and two of them (bays)
have first produced hybrids and subsequently pure bred foals. A
cream coloured Irish-Canadian mare unfortunately died before her
hybrid foal was born. One of the bay mares had a bay hybrid richly
striped ; the other a hybrid with but indistinct stripes. The subse-
quent foals — one by a chestnut thoroughbred horse (Tupgill), the
other by a hackney pony (Mars Royal) — are bays, not only devoid of
stripes but affording no indication whatever that their dams had been
previously mated with a zebra.

Although I experimented with seven English thoroughbred mares
and an Arab mare, I only succeeded with one, a small chestnut. This
mare produced twin hybrids last summer ; yesterday she had a foal to
a thoroughbred chestnut horse (Lockstitch). One of the twins died
soon after birth, the other, richly but unobtrusively striped, in its
colour and make strongly suggests its dam. The chestnut mare's new
foal neither in make, colour, nor action in any way resembles a young
zebra nor a zebra hybrid. In 1897 a bay mare by a bay Arab horse
(Hadeed) was for some months in foal to the zebra. Since she
miscarried in 1896 she has had two foals to a thoroughbred horse
(Lockstitch). Neither of these foals in any way suggests a zebra. In
this case the unused germ cells of the zebra had presumably a better
chance of reaching the ovum from which the first of the two pure-bred
foals was developed than is usually the case.

Attempts were made to cross Welsh, Exmoor, New Forest, Nor-
wegian, and Highland ponies with the zebra without success, and
though a cross-bred Clydesdale has twice had a hybrid, she has not yet
produced a pure bred foal. The experiments as far as they have gone
afford no evidence in support of the telegony hypothesis.

VOL. LXV.

u

252

Dr. W. M. Haffkine.

June 8, 1899.

The LORD LISTEE, F.R.C.S., D.C.L., President, in the Chair.

Meeting for Discussion.

Subject : — Preventive Inoculation.

Professor William Osier (elected 1898) was admitted into the
Society.

Mr. Henry J. H. Fenton, Dr. Henry Head, Professor Conwy Lloyd
Morgan, Mr. Clement Reid, Professor H. S. Hele-Shaw, Dr. Ernest
Henry Starling, Professor Henry W. Lloyd Tanner, and Professor
B. C. Allen Windle, were admitted into the Society.

A List of the Presents received was laid on the table, and thanks
ordered for them.

The Discussion was opened with the following communication : —

" On Preventive Inoculation."* By W. M. Haffkixe, CLE.
Communicated by Lord Lister, P.R.S. Received June 5,
—Read June 8, 1899.

My Lord and Gentlemen, — The most important modern methods of
prophylactic treatment are based upon the fact that an attack of
disease from which an individual recovers leaves in him a condition of
relative immunity to another attack.

The method of turning this result to advantage for the protection
of whole communities was first demonstrated to us by Mahomedan c-
physicians, to whom the world thus owes what proved to be one of the
most fertile principles of modern science.

The successes of Jenner and Pasteur, who utilised cultivated virus
for preventive treatment, have led to a general conception that there is
the possibility of creating artificial immunity to diseases by treating
the organism with morbid virus rendered by some special means
harmless.

This view involves a generalisation which led to a considerable

* The paper as published here, after final revision by the author, differs in some
details from what had appeared in the ' Lancet ' of .Tune 24, and 1 British Medical
Journal,' of July 1, 1899, under the title of " A Discourse on Preventive Inocula- *
tion, delivered at the Koyal Society, London, on June 8, 1899."

On Preventive Inoculation.

'253

amount of disappointment, as the application of the principle in a
number of instances did not give the expected results.

Derivatives from Microbic Tims and their Effect.

When we cultivate a pathogenic micro-organism in a liquid medium,
two different elements are obtained mixed together : the bodies of the
microbe and the liquid which it has modified, and into which it has
secreted its own products.

A modification of the entire preparation, as represented by this
mixture, can be first of all obtained by filtering and separating the two
-elements just mentioned, and considering each of them by itself.

Or else the two can be left together, and only the vitality of the
microbe destroyed by some physical or chemical agent.

Or the constitution and the properties of each, or of both of these
elements can be, to a desired degree, altered by the admixture of
chemicals, or by subjecting them to physical processes.

Or, the vital and pathogenic properties of the microbe can be mocli-
-fied by artificial breeding, and then the microbe itself, or the products
of such a modified microbe, used for treatment.

The immediate effect which a given virus or its derivate produces
on an animal differs with the kind of virus taken, the process of
modification to which it has been subjected, and the species of animals
upon which it is used. The following instance may give an idea of
these variations.

The ordinary Indian grey, as well as the brown, monkey are sus-
ceptible to the plague virus, and may contract a fatal disease from being
simply pricked with an infected needle. The rabbit and guinea-pig are
also susceptible to the disease. The horse, on the contrary, contracts
no fatal disease after being infected even with large doses of the
living virus.

If, however, a plague culture be heated and the microbes killed in it,
the relative susceptibilities of the monkey and the horse seem to be
reversed, the guinea-pig remains comparable with the monkey, while
the susceptibility of the rabbit is now like that of the horse, and not,
as previously, like that of the monkey.

It will require, according to several observers, a very large dose of
virus so treated to produce in the monkey or in the guinea-pig any
marked rise of temperature, or any alteration of the skin at the seat
of injection; while the horse answers to the injection by almost as brisk
an attack of fever as if the virus were a living one, and at the seat of
inoculation a tumour is produced which, if the dose be at all con-
siderable, may lead to a complete mortification of the tissue. The
Tabbit similarly answers to the injection by an attack of fever and by
the formation of a hard tumour at the seat of injection.

As varied as is the immediate effect of different forms of virus upon

u 2

254

Dr. W. M. Haffkine.

different animals, so varied is the result of the application of such
virus from the point of view of immunity conferred by it.

There are animals in whom the inoculation leaves no lasting effect
whatever. In others a very temporary immunity is created, vanish-
ing entirely in a few days. In other cases again, a condition appears
that produces the impression that, after the treatment, the animal has
become more susceptible to a subsequent infection than is a normal
animal not so treated.

And lastly, there may be found animals in which the same virus
will produce a firm and enduring immunity.

In general I believe it to be admissible that in the case of every
disease, and with regard to every species of animal, a form of pro-
phylactic treatment may be found that will produce immunity in that
particular case ; but that same method of treatment may or may
not be applicable to another animal or to another disease affecting the
same animal.

It is the failure of taking into account this variation of circumstances
that, I believe, more than anything else has checked the success of
a number of experimenters.

Immunity from Attack, and Resistance to actual Symptoms of Disease.

The study of the anti-cholera inoculation in India has revealed a new
problem in the subject of prophylactic treatment.

The particular character of cholera epidemics, which appear un-
expectedly, do not last, and in places where they are permanent, are
spread and scattered over large areas, makes the study of that disease
and the demonstration of the effect of a preventive treatment in its
case a matter of much greater difficulty than is the case in localised
contagious diseases, like smallpox, or in plague ; and although a large
amount of material has been collected already, it is desirable that
further observations be still added to the present ones confirmatory of
the results obtained.

The information collected permits, however, already of pointing out
very important features in the working of the anti-cholera inoculation.

The most extended and continuous observations on the subject were
organised by the Municipality of Calcutta, upon the enlightened
initiative of Professor W, J. Simpson, late Health Officer, and under his
continuous supervision, as well as my own. These observations refer
to the cholera stricken suburbs of Calcutta, the so-called " busties "
or groups of huts situated round the tanks, where rain-water is collected
during the monsoon.

Some 8000 people were inoculated in those localities, and for two
years observations were made and the results collected as to the occur-
rences of cholera in the huts inhabited by the inoculated.

On Preventive Inoculation.

255

In the vast majority of cases there lived in the same families
members who had not been inoculated, together with others inoculated,
and the possibility thus presented itself of comparing the incidence of
the disease in individuals of the same households, exposed as much as
it is possible to the same chances of infection.

During the time under observation cases of cholera occurred in
seventy-seven huts. The interval which elapsed between the applica-
tion of inoculation in each particular hut and the occurrence of cholera
in it was as follows : —

Among uninoculated members of the families cholera occurred — 1, 2,
3, 4, 5, 6, 9, 12, 13, 15, 17, 22, 34, 37, 44, 57, 62, 63, 71, 95, 99,
109, 114, 118, 119, 120, 129, 132, 139, 143, 162, 189, 191, 203, 240,
251, 271, 281, 284, 300, 309, 318, 319, 334, 356, 359, 362, 370,
372, 378, 383, 384, 389, 391, 393, 394, 401, 404, 408, 416, 433, 446,
448, 453, 472, 493, 498, 673, 720, 723, 724, and 738 days after the
inoculation of the other members of these families. Among the
inoculated members of the families cholera occurred — 0, 2, 3, 4, 219,
421, 459, 512, 688, 735, and 738 days after their inoculation.

Thus for a period of 738 days, cases of cholera occurred among
the uninoculated, so to say, at all intervals after the date of inocula-
tion ; whereas the figures referring to the inoculated showed a striking-
variation of the incidence when compared at various distances from
the time of inoculation. Cases continued to occur among the
inoculated for a period of four days after the treatment, and then for
416 days they practically remained free from the disease, only one
death from cholera having occurred among them during that time.
From the 421st day up to the end of the observations six cases occurred
among them again.

The relative immunity in the inoculated considered separately
during those three periods shows that during the first four days the
inoculated had proportionately T86 times fewer deaths from cholera
than the uninoculated.

During the period between the 5th and 420th days, i.e., for a period
of nearly fourteen months, the number of deaths among the inoculated
was 22-62 times smaller than amongst the uninoculated. And for the
rest of the time under observation the proportion in their favour fell
to 1 to 1-54.* The plan has since been formed of trying the effect
of larger doses and of stronger vaccines, in order to obtain a more
lasting protection.

While thus the absolute number of cases and deaths from cholera
appeared so strikingly influenced by inoculation, the peculiarity that
became apparent from the observations in Calcutta as well as in other

* Vide Health. Officer's Report to tlie Chairman of the Calcutta Municipal Cor-
^oration, reprinted, in the ' Indian Medical Gazette,' toI. 31, No. 8, August, 1896.

256

Dr. W. M. Haffkine.

places was that the proportion of deaths to cases was not changed by
the treatment.

Thus, in the observations made in a camp of coolies of the Assam-
Burmah Eailway Survey, out of thirty-three attacked among the
uninoculated portion of the camp twenty-nine died, and of four
attacked among the inoculated all four died.

In the Durbhanga prison out of eleven uninoculated attacked all
eleven died, while five inoculated attacked lost three.

In the Gaya gaol twenty uninoculated attacked lost ten, and eight
inoculated attacked lost five.

In a group of tea plantations in Assam 154 cases in uninoculated
had sixty deaths, fifteen cases in inoculated had four.

In the East Lancashire Eegiment in Lucknow 120 uninoculated
attacked had 79 deaths, and 18 inoculated attacked had 13.

This circumstance, the non-reduction of the case mortality by a
treatment which influenced unmistakably the case incidence, appears
an astonishing divergence from the result of small-pox vaccination,
where both the number of attacks and their fatality are reduced by the
treatment.

The new aspect of the problem of preventive inoculation which thus
presented itself in these observations on human communities consisted
in the possibility of a prophylactic treatment being directed separately
towards the reduction of the number of attacks, leaving the fatality of
the disease unchecked, and towards the mitigating of the character of
the disease and the reduction of the case mortality in those who get
attacked.

Possible relation between Immunity from Attack and Resistance to actual
Symptoms of Disease.

In analysing the nature of this particular result, the following two
facts well known in laboratory practice presented themselves to me as
of essential significance.

In patients who recover from an infectious disease the pathogenic
microbe does not disappear from their body for a considerable time
after their recovery. It does not do them harm any longer, though
when transferred to another animal it may still cause a fatal attack.
Similarly, as in the case, for instance, of a guinea-pig inoculated with
the bacillus of chicken cholera, a naturally immune animal can breed
for weeks, in an abscess, microbes of an intense virulence without in
the least suffering in its own general health.

A condition seems to set in in the convalescent patient, or to exist in
naturally immune animals, by virtue of which they do not suffer from the
results of activity of a pathogenic microbe, i.e., from its morbid products ;
and from that time the presence of the microbe in the system, even in

On Preventive Inoculation.

257

the tissues, becomes innocuous. Immunity against morbid symptoms
generated by the products of microbes does not seem to imply neces-
sarily the ridding of the system of such microbes. It is known now,
since the discoveries of Behring and Kitasato, that such a resistance
to these products can be originated artificially, by gradually treating
the system with increasing quantities of toxines. The system reacts by
developing anti-toxines tending to neutralise the effect of the toxines.

On the other hand, Gamaleia first drew attention to the fact that it
is possible to create in an animal resistance to lethal doses of virulent
microbes without that animal acquiring any resistance to a dose of the
products prepared from the same microbes in the laboratory.

One seems justified therefore in considering separately two kinds of
immunity : One against the living microbe, which would prevent it
from entering the system and causing an attack ; and another against
the fatality of the symptoms of the disease caused by the products of
the microbe when the latter overcomes the initial resistance and does
invade the system.

In the inoculation against cholera, which is done with the bodies of
microbes, the first result alone is obtained.

These considerations were confirmed by a set of laboratory experi-
ments by PfeifTer and Kolle, intended to verify our Indian results,
and in the course of which they detected in the serum of men inocu-
lated with only one dose of cholera vaccine an extremely high pro-
tective power, equal to that which, in goats for instance, could be
created only after a very prolonged treatment, extending over five
or six months, and including injections with gigantic doses of cholera
vaccine.

On analysing, however, in detail the properties of that serum they
found that it possessed an intense power of destroying the cholera
microbes, but exhibited no antitoxic properties capable of neutralising
the effect of the products of those microbes.

The Plan of A nti-plague Inoculation.

When, in 1896, I was confronted with the problem of working out a
prophylactic treatment against the plague, I determined to put to test
the ideas originated by the observations on our cholera patients, and to
attempt, in the new preventive inoculation, to obtain at once a lower-
ing of the susceptibility to the disease, and a reduction of the case
mortality.

This I resolved to obtain by treating the system with a combination
of the actual bodies of microbes and of the concentrated products of
their activity.

In presenting the above considerations, I beg that they may be con
sidered as provisional, subject to modification or to complete refutation.

258

Dr. W. M. Haffkine,

There may exist already facts unknown to me, which are opposed to
the guesses implied in them. It was those guesses which led to the
results obtained in the plague inoculation ; but, in giving the reasoning
which I passed through while working out the method, I am yielding
only to a demand expressed to me to that effect, as I considered that
part of my communication unnecessary ; the more so that the theo-
retical conjectures above enumerated are not shared even by very
eminent experimenters, such as Pfeiffer himself, to whose results I owe
some of my premises ; and the correctness of the composition of the
plague prophylactic, with regard to the extracellular toxines which I
have added to it, the so-called supernatant fluid of the plague prophy-
lactic has been the subject of an animated dispute.

It is certain that no theoretical views conceived by one experimenter
are binding, or need even be interesting, to others. What is obligatory
is the acceptance of the results obtained.

The Plague Prophylactic.

In order to accumulate for the plague prophylactic a large amount
of extra-cellular toxines, the bacilli are cultivated on the surface of a
liquid medium where they are suspended by means of drops of clarified
butter or of cocoanut oil. "

The bacilli grow down in long threads into the depth of the liquid,
and produce what I have termed a stalactite growth in broth, an
appearance quite peculiar to this microbe, and which, I hope, will be
accepted as the specific diagnostic feature of this microbe.

The products of their vital exchange — the toxines — are secreted by
the stalactites into the liquid and accumulate there.

The growth is periodically shaken off the surface of the broth, after
which a new crop appears underneath that surface.

Thus a large quantity of the bodies of microbes is collected at the
bottom of the cultivation vessel, and the liquid itself gets gradually
permeated with increasing quantities of toxines.

The process is continued for a period of five to six weeks, at the end
of which the bodies of the microbes become extremely deteriorated.

It will be seen from this that, in my eagerness to put to test our ability
to influence the case mortality I may have, perhaps, paid less atten-
tion than I might have done to the problem of reducing the number
of attacks; and I have now sketched out a simple plan whereby to
test this circumstance, and to try to improve our results from this
point of view.

In order to render harmless the inoculation of the virus above
described, I deterr lined to kill the microbes by heating the material up
to 65 to 70° C. The virus so treated, differing from what is observed
in some other instances, loses at once, for the animals susceptible to the

On Preventive Inoculation.

259

■disease, almost all its pathogenic power; and it was a question to
determine, whether it contained the qualities that were sought for,
viz., the power of creating in man a useful degree of resistance to
plague.

The plan has been contested by a number of experimenters who tried
a material similarly prepared on different animals and failed to detect
in it any immunising properties.

Among other forms of plague virus which were tested by us, and by
other experimenters, a large number were found to be too dangerous'
to use ; in other instances the mode of application was inadmissible in
the case of men ; in others, again, the effect appeared too transient to
be of practical use.

The Properties of the Plague Prophylactic.

The immunising effect of the plague prophylactic, as above described,
was worked out .on domestic rabbits, and its actual efficiency was
verified and confirmed by a number of investigators who experi-
mented on the infection with virulent plague of protected and unpro-
tected rabbits.

Comparing the rabbit with other l?Jboratory animals, such as the rat,
the guinea pig, the mouse, the monkey, one may consider the rabbit as
the one that perhaps required the least amount of protection, as its
natural resistance to plague is relatively high. The most altered virus,
i.e., such as was rendered the most harmless of all, was found to confer on
the rabbit a very considerable degree of immunity, enabling it in a few
days to resist ten or fifteen-fold lethal doses of virulent plague microbes.
The same treatment applied to animals of a more susceptible nature
would, on the contrary, in many instances fail.

The Questions which were to be solved by Experiments on Human Beings.

At the end of our laboratory experiments a set of questions stood
before us that were to be solved by direct experiment on human
beings. Those questions were : —

1. Would man behave with regard to the prophylactic like the
animals upon which its protective power had been worked out ?

2. If it so happens that the answer is affirmative, what would the
dose of the prophylactic, and the method of administering it be ; and
would not the dose required be so high, and the reaction to be pro-
duced so severe, or the number of inoculations to be repeated so great,
as to render the treatment inapplicable to men, or impracticable 1

3. How many days counting from the date of inoculation would it
take to produce in man a useful degree of immunity 1

4. How long would that immunity last ?

280

Dr. W. M. Haffkine.

And lastly, there followed two questions, to which my experience
of the anti-cholera inoculation entitled me to give a reassuring answer,
but the correctness of which it was necessary to demonstrate afresh in
the case of plague, viz. :

5. During the period of reactionary fever and all the other symptoms
produced by inoculation, will the resistance of the inoculated exposed
to plague be, for the time being, reduced, or remain the same, or be
increased *? i.e., would it constitute a danger to apply the inoculation
in localities actually affected with plague 1 and

6. When a man who happens to be incubating the plague, or to have
initial symptoms of the disease already, chances to be inoculated,
would it aggravate his condition, or have no effect, or on the contrary,
help him ?

Demonstration of the Harmlessness of the Treatment.

The perfect harmlessness of the inoculation was first of all demon-
strated by the officers of the Laboratory, the Principal and Professors
of the Grant Medical College, a large number of leading European and
native gentlemen of Bombay, and their families and households, being
inoculated. And then, in the last week of January, 1897, when the
plague broke out in Her Majesty's House of Correction at Byculla,.
in Bombay, the option of inoculation was offered to the prisoners.

The Experiment in Her Majesty's Byculla House of Correction, Bombay.

The Byculla House of Correction is a long-term prison. There are
no children, nor very young people among the inmates, there being in
Bombay a separate establishment, the Sassoon Reformatory, where
minor criminals are sent.

The prisoners of the House of Correction present a well-fed, well
clad, regularly worked, and almost as uniform a set of people as can be
seen in a regiment, amongst whom one could scarcely see a single
infirm or very aged individual.

At the appearance of plague the prisoners numbered 346 souls.

The inoculation was introduced after nine cases of plague had
already occurred, five subsequently ending fatally; there remained
thus 337 individuals to be dealt with. Of these, 154 only volun-
teered for inoculation, and 183 remained uninoculated.

On the 30th January, in the forenoon, before the inoculation was
applied, six more cases occurred, of which three afterwards proved
fatal. The inoculation was applied in the afternoon, and afterwards
it was discovered that one more prisoner had already a bubo on him
when inoculated, while two prisoners developed buboes the same evening
after their inoculation. These three cases, attacked on the day of
inoculation, proved also fatal.

On Preventive Inoculation.

261

After that, the difference which was observed in the fate of the two
groups, the inoculated and uninoculatecl, is seen from the subjoined
table

Occurrences in uninoculated.

Occurrences

in inoculated.

Date of occurrence
of plague.

Number of
uninoculated
present.

Cases.

Fatal.

Number of
j inoculated
present.

Cases.

Fatal.

1897.

23rd to 29th January,
previous to the day
of inoculation.

9

5

t S
© vs

rForenoon, be-
fore inocula-

6

3

30.1.97, tl
of inocul

-

Afternoon,
after inocula-

Q

1st day after inocula-
tion, 31.1.97

2nd day after inocula -

177

172 .

2
1

1
1

151
150

1

3rd day after inocula- 1
lion, 2.2.97

5th day after inocula-
tion, 4.2.97

6th day after inocula-
tion, 5.2.97

7th day after inocula- ,
tion. 6.2.97 i

173
171
169
169

1

1

9

5

1

. 1
1
1

146
146
146
146

1

Total after the day
of inoculation ....

172

uninoculated,
average daily
strength.

12

cases.

6

deaths.

i

147

inoculated,
average daily
strength.

2

cases.

No
deaths.

* During the time under observation the number of uninoculatecl was reduced
on the second, fifth, and sixth day after inoculation by three, one, and one dis-
charged prisoners, whose terms of confinement expired on those days ; and it was
increased on the third and seventh day by two and two prisoners newly admitted
into the jail. The number of inoculated was reduced on the third day after inocu-
lation by four released prisoners.

No information as to the subsequent history of these few released (uninoculated
and inoculated) prisoners reached the authorities or the officers of the Laboratory,
but the question had no interest for the experiment, since their conditions of life
and their exposure to plague ceased to be comparable from the time the prisoners
were discharged into a large city, with various chances of infection, different in its
different quarters. The observations referred only to the prisoners who remained
exposed to plague under the conditions of the jail.

262

Dr. W. M. Haffkine.

With the exception of the fourth day, cases of plague continued to
■oGcur among the uninoculated group for seven days after inoculation,
their average daily strength throughout the week being 173; altogether
twelve cases occurred among them with six deaths; while in the 148
inoculated there was one case on the next day after inoculation, who
rapidly recovered, and one on the last clay of the epidemic, who
recovered also.

Analysis of the Results of the Byculla Jail Experiment,

A glance at the above table will show the progress which was made
in our information by that initial experiment, and how far it carried
us ahead from the state of uncertainty which surrounded the question
originally.

The dose of prophylactic administered to the prisoners was 3 c.c.

They all had the customary attack of fever from the operation, with
the discomfort accompanying that condition, — a headache in many
cases, nausea, loss of appetite for a couple of days, a feeling of fatigue
and lassitude, recalling a milcl attack of influenza, with swelling and
pain in the inoculated side. Did, however, all this make them more
susceptible to the disease than were their non-inoculated fellow inmates 1
It is certain that the results testify unmistakably to an opposite
effect.

Further, the incubation period in plague appears to be on the average
five days, extending, however, not unfrequently up to ten.

As to the few newly admitted (uninoculated) prisoners mentioned above, none
were subsequently attacked, and the cases of plague stated in the table refer all to
old residents, who were present in the jail on the day, and long before the day, of
inoculation. Had a case of plague occurred in any of the uninoculated new-
comers, it would not have been included in the comparison with the occurrences
among the inoeulated inmates, since such new-comer had been exposed outside the
jail to conditions of infection other than those which existed inside the jail. Not-
withstanding this, it was permitted that the number of the uninoculated new-
comers should be added by the jail authorities to the strength of the uninoculated
as quoted in the table, since this tended to weaken, and not to exaggerate, the pro-
portion of immunity obtained by inoculation.

In a similar way, in the outbreak of plague in the second (Umerkadi) Bombay
jail, described below, one of the three attacks among inoculated persons, as reported
by the jail authorities, namely in a man named Sital Vary, prisoner No. 7542,
occurred after he had been set at liberty in the city. Had it been an attack in an
uninoculated person, it would not have been admitted into the comparison with the
inoculated, for the reason mentioned above, i.e., because the conditions of infection
outside the jail were not comparable with those in the jail. But as this was a case
in an inoculated subject, and tended to present the results in a less favourable light,
it was included in the record.

Every prisoner attacked with plague and taken away to the isolation or plague
hospital was excluded on the subsequent day from the strength of those who
remained in the jail and were still liable to an attack.

On Preventive Inoculation.

263

Of the twelve prisoners in the uninoculated group who developed
plague during the next few days after the date of inoculation a large
proportion, if not all, must have been already incubating the disease
on that day ; and seeing the perfect similarity of conditions under
which the inoculated and the uninoculated, who belonged to the same
crowd of people, were living, one could infer safely that a similar group
of individuals incubating plague was present among the inoculated
also at the time when the inoculation was performed on them.

The inoculation, however, did not aggravate their condition, as the
number of inoculated who developed plague, counting from the first
twelve hours of inoculation, was proportionately five times smaller
than the corresponding number among the uninoculated ; and the two
cases that appeared among the inoculated, one on the very next morning
after inoculation, both ended in recovery.

As far as that first experiment went, therefore, men behaved like the
laboratory animals upon which the prophylactic properties of the drug
had been worked out.

For communicating that protection one injection of prophylactic
appeared sufficient, with a dose of 3 c.c, which dose, however, in our
subsequent operations was further reduced to 2J c.c.

The difference in favour of the inoculated appeared within some
twelve hours after the operation ; but the man who was inoculated with
plague on him, and the two who developed clear symptoms of plague
the same evening, did not benefit by the operation.

This completed the first information gathered with regard to five of
the six questions enumerated above. No answer could be given as to
the final duration of the effect of inoculation, except that the operation
appeared to be useful in a localised already existing epidemic, extending
over seven days.

The Experiment in the Umerkadi Common Jail, Bombay.

In the next case the strictness of the conditions of the last experi-
ment was enhanced further. This was in the second Bombay jail, the
Umerkadi Common Jail.

The plague broke out there at the end of December, 1897, and by
the 1st of January three prisoners were attacked and all of them sub-
sequently died.

In the interval between the operations in the two jails, some 8000
people in the free population of Bombay had already availed them-
selves of the inoculation.

This time the whole of the prisoners, numbering 401, appeared willing
to undergo the preventive treatment. In view of the novelty of the
operation, however, and of our responsibilities before the government
and the public, and the necessity of demonstrating clearly the effect of

264

Dr. W. M. Haffkine.

inoculation, the prisoners were not allowed to undergo the treatment
in a body, and it was resolved that only one half of them should be
permitted to do so.

The manner in which that half was selected guaranteed the elimina-
tion of all possible errors usually inherent to observations on human
communities.

The population of a jail in India is gathered into several groups,
the largest being the ordinary convicts divided into simple prisoners
and hard labour convicts; then there is a group of civil prisoners,
debtors \ a group of under-trial prisoners, of convict warders, of cooks,
bakers, men employed in the infirmary, &c. ; and a separate group of
female prisoners.

On the morning of the 1st of January, 1898, in the presence of
Major Collie, I. M.S., and Dr. Leon, medical officers, Mr. Mackenzie,
the superintendent, and of all the officials of the jail, the above groups
were brought one after the other into the jail-yard, and asked to seat
themselves in rows; and after all were seated, every second man
without further distinction was inoculated, excepting two of them, who
did not volunteer for the treatment.

From this moment the even numbers, the inoculated, were left to live
with the uninoculated, under conditions identical with those under
which they were living before. They had the same food and drink,
the same hours of work and of rest, shared with them the same
yards and buildings, &c.

In this case fatal attacks continued to occur in the jail for thirty
days, during which time an almost equal number of prisoners, inocu-
lated and uninoculated, were discharged from jail, and thus excluded
from further observation.

The average daily strength of the uninoculated who remained in the
jail up to the end exposed to the plague was 127 ; and of the inocu-
lated, 147.

In the smaller uninoculated number, ten cases of plague occurred,
six of them proving fatal ; while the larger inoculated number pro-
duced three cases, of whom all recovered.

In these three cases, however, in the inoculated, the character of the
disease was so much mitigated that the authorities of the Government
hospital at Parel, Bombay, where they were sent, hesitated to return
them as plague, and the Director-General of the Indian Medical Service,
who examined two of them, diagnosed them as mumps. They were
returned as cases of plague in order that no possibility of error in
favour of the inoculation should be admitted.

The Experiment in the Dharwar Jail.

On the third and last occasion, when the plague broke out in a jail,
the authorities did not feel justified in withholding the inoculation

On Preventive Inoculation.

265

from any of the inmates, and all were permitted to be inoculated.
This was in Dharwar, at the end of October and the beginning of
November, 1898, during the terrible outbreak of plague in that town
and district, the news of which must have reached you even here.

Five cases of plague, of which one was imported, and four in old •
residents of the jail, occurred, all five ending fatally.

The prisoners, then numbering 37 3, submitted in a body to inocula-
tion, and only one case followed, in a man attacked two days after
inoculation, and he recovered, the only one of the six.

The Experiment of Undhera, in a Free Population.

The most carefully planned out and precise demonstration of the
working of the prophylactic system among a free population, exposed
to a great amount of infection, was that made in the village of
Undhera, six miles from Baroda.

The following was the mode of operation adopted : —

A detailed census was made by the authorities of all the inhabitants
of the place, and on the 12th February, 1898, when a committee of
British and native officers arrived to carry out the inoculation, the
people were paraded in the streets, in four wards, family by family.

Major Bannerman, I. M.S., of the Madras Medical Establishment,
and myself, accompanied by the Baroda officials, went from one house-
hold to another, and, within each, inoculated half the number of the
male members, half that of the females, and half that of the children,
compensating for odd figures that happened to be in one family by odd
figures in another. I, personally, and the officers who were with me,
directed special attention to distributing the few sick in the two groups
of inoculated and uninoculated as equally as our judgment permitted
us to do.

The plague, which had carried off, before inoculation, seventy-nine
victims, continued afterwards in this instance for forty-two days and
appeared in twenty-eight families, in which the aggregate number of
uninoculated was sixty-four, and of inoculated seventy-one.

The total number of attacks in those families was thirty-five, and
they were distributed as follows : —

The 64 uninoculated had 27 cases with 26 deaths ;
The 71 inoculated „ 8 „ 3 „ ,

thus showing 89*65 per cent, of deaths fewer in the inoculated members
of the families than in the uninoculated.

There were no deaths from other causes in the inoculated of the
village, while among the uninoculated there were three deaths attri-
buted to causes other than plague.

The subjoined figures show the number of days which elapsed

•266

Dr. W. M. Haffkine.

between the date of inoculation and the occurrence of a death from
plague in the families. The first row of figures refers to occurrences in
uninoculated members, the second to occurrences in inoculated, while
the small figures show the number of deaths which occurred in each
group on those days : —

Deaths from plague occurred in uninoculated —

32 41 53 72 83—103 113 12i 151 i6i 191 201 211 241 321 and 421,

and in inoculated

— 91 I21andl41

days after date of inoculation.
There had elapsed therefore eight days, during which eleven deaths
from plague occurred among the uninoculated members of the families,
before the first death took place in an inoculated case. The inocula-
tion has again acted, so to say, immediately ; or, to use the mode of
expression which we have adopted, has exercised its protective effect
within the time necessary for the subsidence of the general reactionary
symptoms produced by the inoculation.

The investigation in this village was carried out by Surgeon-General
Harvey, the Director-General of the Indian Medical Service, and a
committee of British and native officials. Every member of the famih^
who survived was seen, his particulars verified from the documents,
and every detail was confirmed from the registers kept at the time,,
and from the testimony of the whole of the villagers, who were present
throughout the inquiry.

Experiments op a Large Scale, Average of the Results obtained,

I have dwelt so long upon the description of the above experiments
not because they were the largest in volume or the most striking
which were made, but because they were the most precise of all, and,
so far as I am aware, free from any possible loophole of mistake.

I made prolonged and detailed observations in very severely affected
communities of Lanowlie, in a population of 700 people, and among the
followers of the artillery at Kirkee, numbering at the time 1530. Very
complete data were collected by Professor Eobert Koch and Professor
Gaffky, of the German Government Plague Commission, by Major
Lyons, I. M.S., of the Bombay Medical Establishment, and by myself,
in the Portuguese colony of Damaon, in a population of 8230 indivi-
duals, during a frightful outbreak of plague there, which lasted more
than four months, in 1897. A minute investigation extending over
four months, was made by me in the Khoja Mussulman community of
Bombay, numbering some 12,000 people, where about half of the total
number were inoculated under the auspices of His Highness the Aga
Khan. A most comprehensive inoculation campaign, and with widely

On Preventive Inoculation.

267

reaching and most satisfactory results, was carried out, under Mr.
Cappel, late Collector of Dharwar, by Captain Leumann, LM.S , Dr.
(Miss) Corthorn, Dr. Hornabrook, Dr. Foy, Dr. Chenai, and others, in
three adjacent small towns of Hubli, Dharwar, and Gadag, where
80,000 people were inoculated. The latter was the most magnificent
piece of work done, from the point of view of practical application of
the method, and of the testing of its general efficiency.

With the extension of the number of inoculated the exactitude and
precision of observation certainly suffer. A number of questions and
objections arise with regard to points of detail, which it is not always
possible to answer with certainty. Such wholesale observations are,
however, required to enable us to judge whether the application of the
method as a general measure answers to the expectations formed;
whereas the exact extent of the results is only to be gathered from
mathematically precise experiments, imitating the conditions of labora-
tory practice, such as were those which I have detailed above.

The difference in. the mortality from plague in inoculated and unin-
oculated sections of communities was estimated to average over 80 per
cent., approaching often 90, as was the case in Undhera described
above. The lowest proportion ever observed in the experiments which
I made personally was 77*9 per cent. ; this was at Kirkee.

Effect on the Case Mortality.

A very accurate set of data were collected in almost all the larger
hospitals where inoculated plague cases were admitted, upon the fatality
of the disease in inoculated. These were to the effect that the case
mortality among them was some 50 per cent, lower than among
uninoculated plague cases. A number of documents on this point has
been collected by the Indian Plague Commission and will, I trust,
appear in their records.

Minimum Duration of the Effect of the Plague Inoculation.

As to the duration of the effect of the plague inoculation, the only
statement which can be made for the present is that it lasts at least for
the duration of one epidemic, which, on the average, extends over
four to six months of the year.

The Government of India have recognised the inoculation certifi-
cates, entitling the holder to exemption from plague rules, as being
valid for a period of six months ; on the understanding that if accurate
data are forthcoming as to the effect lasting longer, the holders
will be permitted to renew their certificates for another period, without
being reinoculated.

VOL. LXV.

x

268

Mr. W. M. Haffkine.

The further Problems pursued in the Bombay Plague Research Laboratory.

The task which lies now before the Plague Laboratory in connection
with prophylactic inoculation comprises the following problems : —

The perfecting of processes for turning out large and uniform quan-
tities of material, and avoiding the variations due to the character
of the plague microbe, and to the differences in the composition of
the cultivation media ;

The further investigation of the different constituents of the plague
prophylactic, with a view of intensifying those which produce
definite and beneficial results ;

The possible mitigation of the reactionary symptoms after inocula-
tion; and

The study of the effect of antiseptics used for preserving the prophy-
lactic ;

while the most important general problems concerning plague relate to
the study of the curative treatment and to the life-history of the
plague microbes in nature.

The Typhoid Inoculation.

The inoculations against cholera and plague, which are the outcome
of the work of Jenner, Pasteur, and Koch, and their admirable succes-
sion of pupils, contain in their turn a promise of development and
success, which, I trust, will be only enhanced at every subsequent
step, and which, it seems to me, already warrants the application of
the same kind of effort to other diseases and epidemics.

In this order of ideas permit me to enter a plea in favour of a
new inoculation campaign, which has been inaugurated already, and
which I hope will be carried out successfully, for the benefit of a large
number of soldiers of this country residing in India, and of white men
in general in all tropical countries.

The problem of typhoid inoculation has quite a special interest for
Europeans, as much as cholera has for the natives of India. Typhoid
proved to be a more difficult disease to eradicate from military canton-
ments than cholera. It is possible that the explanation of this lies in
what is already known of the character of the microbes of these
diseases. ,

The typhoid bacillus when subjected to different chemical and physical
agents, such as acids or antiseptics, or a high temperature, or desicca-
tion, or the admixture of other microbes, appears far more resistant
than the cholera microbe.

Such a character would ensure for the typhoid bacillus an existence
in more varied media, under more various climates, and a greater inde-

On Preventive Inoculation.

269

pendence from seasonal or local changes, than is the case with the cholera
microbe.

Outside the endemic area cholera remains in one and the same place
but for a few weeks, and in any given part of a town often for a few
days only. It is rare that it visits the same barrack more than once in
five, sometimes ten years, and when it occurs, a temporary evacuation
of the place puts a stop to the disease.

The typhoid virus, on the contrary, sticks to an infected locality for
years, and causes a continuous incidence of the disease for which occa-
sionally nothing short of a complete desertion of the station is
effective.

At the same time, while the cholera infection seems to be almost
exclusively confined to the water supply, in typhoid the improvement
of the water appears to leave intact a large number of other sources of
danger which up to the present have escaped detection.

While thus differing in their life-history in nature, the bacilli of
cholera and typhoid present important common features in the manner
in which they behave in the human and animal body.

The chief centre of infection in both instances is the intestinal canal,
the circulatory system remaining free from invasion. When inoculated
into animals, both microbes admit of the same kind of transformation
by passages from animal to animal ; and against both immunity can
be created in laboratory animals by the same preparation of virus as
used in the inoculations for cholera ; while, when examining the tissues
of immunised animals, the same modifications are detected in them as
those observed after the anti-cholera inoculation.

These considerations have led us to expect from the typhoid inocula-
tion in man a similar protective effect to that observed in the inocula-
tion against cholera ; and seeing that the period of life during which the
newcomers to India remain susceptible to typhoid extends only over a
few years, it would seem that the application of the system, when
properly organised, is likely to prove of a very high practical value.

Inoculation and General Sanitary Measures.

The anti-cholera inoculation, the inoculation against plague, and that
against typhoid thus came to put themselves on the same line as
vaccination, and represent attempts at dealing with epidemics on a
plan differing from measures of general sanitation. During the last
few years the question has, therefore, been frequently debated as to
the relation in which the two stand to each other.

It is scarcely necessary to say that inoculation cannot be substituted
for a good water supply, the draining, cleansing, or improvements in
the building of cities, or for the admission of a larger amount of light
and air into over-crowded localities, for all those measures to which the

x 2

270

Mr. W. M. Haffkine.

nations owe the marked improvement in health which has taken place
during the present century.

Only, injustice would be done to the sanitarian by calling him in
when a patient lies already on his sick-bed, or when an epidemic
actually breaks out in a community, and by asking him to stay the
sickness, or the epidemic, to improve the health of the population, so to
say, while you wait.

To be dealt with, epidemics, like individual diseases, require specifics,
promptly administrable remedies, and measures of general sanitation
can be no more advised for arresting a sharp outbreak of cholera or
plague than an individual patient be directed to build for himself a new
house, or to dry up the marshy lands or to cut down the jungle round
his habitation when he requires a dose of quinine to arrest an attack
of ague.

The part of vaccination and of preventive inoculation in combating
epidemics stands in the same relation to general sanitary measures as
therapeutics and the art of the healing physician stand to domestic
hygiene and sanitation. It is certain that neither of these can ever be
substituted for the other.

Inoculation and the Segregation-Disinfection Method.

A comparison of another kind now very actively discussed is that
between the methods of combating epidemics by separation of sick and
healthy, and disinfection, on the one hand, and by preventive inocula-
tion of the people, on the other.

From this point of view the following distinction between infectious
diseases is to be made : —

When we take some affected tissue from a leper, or a pustule from a
small-pox patient, or virulent saliva from a rabid animal, or some
syphilitic matter, and throw it into milk, broth, or any organic sub-
stance such as is to be found in the ordinary surroundings of men,
it produces no modification in the medium, and in the course of time
loses its infective properties and dies out. When, on the other hand,
we repeat the experiment with cholera, or plague, or typhoid products,
instead of dying out, the contagion begins to grow and multiply,
spreads in the medium and soon transforms the whole of it into one
mass of infectious matter.

It is evident that such a distinction — the strictly parasitic nature
of one microbe and the capacity of the other to lead both a parasitic
and saprophytic life — must influence most directly the ways in which
these diseases spread and assume epidemic forms, and also the measures
which are likely to be effective in combating them.

In the first instance the infection must remain confined entirely, or
almost so, to the body of the patient, and the disease can be propagated

On Preventive Inoculation.

271

only directly from individual to individual, or by means of their im-
mediate belongings. It is the inability of a virus to grow in lifeless
nature that communicates to that disease a strictly contagious cha-
racter.

In the second case, provided the surrounding conditions be favour-
able to it, the virus will spread widely around the original focus, and
the sources from which infection reaches fresh individuals will grow in
number rapidly.

From the point of view of preventive measures, therefore, in diseases
like rabies, or syphilis, or small-pox, or leprosy, where infection is to be
found in the patient alone, precautions of isolation taken with regard
to the sick and their immediate surroundings, must affect directly the
prevalence and propagation of the disease ; whereas in typhoid, cholera,
or plague, where the patient is only one, and proportionately a limited,
source of danger, his isolation, and the destruction of his belongings,
leaves unaffected the vast cultivations of infection which are going on
in nature besides-. Measures taken for circumscribing the prevalence
of an epidemic by isolating and destroying the foci of infection are less
likely to succeed in this category of diseases ; attempts at eradicating
an epidemic or at protecting individuals by ways which appear effec-
tive in merely contagious diseases will be in this case easily eluded ;
and the necessity of personal protection by means of a prophylactic
treatment will soon be urgently felt and acknowledged.

Conclusion. Tlie Officers who assisted in the Bacteriological Investigation in

India.

My Lord and Gentlemen — Permit me, before I leave this place, to
pay a tribute of gratitude for assistance and co-operation in the inves-
tigation work in India to the Officers of the Indian and Bombay
Governments, to the Director-General of the Indian Medical Service, to
Lieut. -Colonel Owen, Major Bannerman, Major Lyons, Major Herbert,
Captain Thorold, Captain Hare, Captain James, Captain Vaughan,
Captain Maynarcl, Captain Earle, Captain Stevens, Captain Green,
Captain Clarkson, Captain Milne, Captain Leumann, of the Indian
Medical Service, to Dr. (Miss) Corthorn, Dr. Gibson, Dr. Marsh, Dr.
Balfour Stewart, Dr. Ransome, as well as to Dr. W. J. Simpson, Dr.
Powell, Dr. Mayr, Dr. Surveyor, Dr. Paymaster, Mr. E. H. Hankin,
the distinguished bacteriologist of the North- West Provinces of India,
to His Highness the Aga Khan, and to a number of other European as
well as Indian gentlemen, happily far too numerous to permit of my
mentioning them all.

272

•

Proceedings and List of Papers read.

June 15, 1899.

The LOED LISTER, F.E.C.S., D.C.L., President, in the Chair.

Professor E. W, Brown (elected 1898) was admitted into the Society,

Professor W. F. Barrett, Dr. Charles Booth, Professor A. C. Haddon,
Mr. Richard Threlfall, and Mr. A. E. Tutton were admitted into the
Society.

A List of the Presents received was laid on the table, and thanks
ordered for them.

A communication on the Echelon-grating Spectroscope was made by
Professor A. A. Michelson, of the University of Chicago.

The following Papers were read : —

I. "A Comparison of Platinum and Gas Thermometers, including
a Determination of the Boiling Point of Sulphur on the
Nitrogen Scale : an Account of Experiments made in the
Laboratory of the Bureau International des Poids et
Mesures, at Sevres." By Dr. P. Chappuis and Dr. J. A.
Harker. Communicated by the Kew Observatory Com-
mittee.

II. " A Preliminary Note on the Morphology and Distribution of
the Organism found in the Tsetse Fly Disease." By H. G.
Plimmer and J. Rose Bradford, F.R.S.

III. " The Colour Sensations in Terms of Luminosity." By Captain

Abney, C.B., R.E., F.R.S.

IV. " On the Orientation of Greek Temples, being the Results of

some Observations taken in Greece and Sicily, in May,
1898." By F. C. Penrose, F.R.S.

V. "The Absorption of Rontgen Rays by Aqueous Solutions of
Metallic Salts." By Lord Blythswood and E. W.
Marchant. Communicated by Lord Kelvin, F.R.S.

VI. " On the Oxidation of Carbon at ordinary Temperatures by
means of Atmospheric Oxygen with the Production of
Electrical Energy." By W. E. Case. Communicated by
Sir William Preece, F.R.S.

Proceedings and List of Papers read.

273

VII. " The Conductivity of Heat Insulators." By C. G. Lamb and
W. G. WILSON. Communicated by Professor EwiNG,
F.E.S.

VIII. " On Simultaneous Partial Differential Equations and Systems
of Pfaffians." By A. C. Dixon. Communicated by Dr.
J. W. L. Glaisher, F.E.S.

IX. "On the Comparative Efficiency as Condensation Nuclei of
Positively and Negatively charged Ions." By C. T. E.
Wilson. Communicated by the Meteorological Council.

X. "Data for the Problem of Evolution in Man. II. A First
Study of the Inheritance of Longevity and the Selective
Death-rate in Man." By Miss Mary Beeton and Professor
Karl Pearson, F.E.S.

XL " Collimator Magnets and the Determination of the Earth's
Horizontal Magnetic Force." By Dr. C. Chree, F.E.S.

XII. " On the Resistance to Torsion of certain Forms of Shafting,
with special Eeference to the Effect of Keyways." By
L. N. G. Filox. Communicated by Professor M. J. M.
Hill, F.E.S.

XIII. "On the Waters of the Salt Lake of Urmi." By E. T.

Guxther and J. J. Maxley. Communicated by Sir John
Murray, K.C.B, F.E.S.

XIV. " On the Application of Fourier's Double Integrals to Optical

Problems." By Charles Godfrey. Communicated by
Professor J. J. Thomson, F.E.S.

XV. " On Diselectrification produced by Magnetism. Preliminary
Note." By C. E. S. Phillips. Communicated by Sir
William Crookes, F.E.S.

XVI. " On the Orbit of the Part of the Leonid Stream which the
Earth encountered on 1898, November 15." By Dr. A. A.
Eambaut. Communicated by Dr. G. J. Stoxey, F.E.S.

XVII. " On the Numerical Computation of the Functions G0(«),
Gi(£), and Jn(x Ji) ; and on the Eoots of the Equation
Kn(x) = 0." By W. Steadman Aldis. Communicated by
Professor J. J. Thomsox, F.E.S.

XVIII. "Eesearches in Absolute Mercurial Thermometry." By (the
late) S. A. Sworx. Communicated by Professor H. B.
Dixox, F.E.S.

The Society adjourned over the Long Vacation to Thursday,
November 16, 1899.

274 Mr. H. G. Plimmer and Prof. J. Eose Bradford.

" A Preliminary Note on the Morphology and Distribution of the
Organism found in the Tsetse Fly Disease." By H. G.
Plimmer and J. Eose Bradford, F.E.S., Professor Super-
intendent of the Brown Institution. Eeceived June 12, —
Eead June 15, 1899.

(From the Laboratory of the Brown Institution.)

These observations are the result of an inquiry entrusted to us by
the Tsetse Fly Committee of the Eoyal Society, at a meeting of the
Committee on March 16, 1899.

The material for our investigations was obtained in the first place
from a dog and a rat, inoculated with the blood of a dog suffering
from the disease, by Mr. H. E. Durham, at Cambridge.

The organism found in the Tsetse Fly disease was discovered by
Major Bruce, E.A.M.C., F.E.S., and was classed by him as a Trypano-
soma. These belong to the order Flagellata, and, according to
Butschli, to the sub-group Monadina.

We will, in the first place, describe the adult form of the organism,
such as is met with most frequently in the blood of a susceptible animal
affected with the disease.

A. Description of the Adult Form of the Trypanosoma.

In freshly drawn blood examined as a hanging drop, or as a very
thin layer in a cell, the adult form of the Trypanosoma can be easily
studied. The latter method is the better, as the organism can be
better seen and more accurately examined, in the thin, uniform layer
of fluid than in the rounded drop. The easiest method of examining
the blood in this way is to make, with a red-hot platinum loop and a
small piece of paraffin, a thin ring of paraffin on an ordinary glass
slide ; the drop of blood is placed in the centre of the ring and a
cover-glass placed on it, the thin layer of paraffin preventing pressure.
If it be desired to keep the blood for continuous examination, it should
be drawn into a graduated Pasteur pipette, and one-tenth part of a
5 per cent, solution of sodium citrate should be drawn up after it,
then the blood and citrate solution should be carefully mixed in the
bulb ; the tube should then be sealed up, and drops can be taken from
it as desired.

Under ordinary conditions of illumination the Trypanosoma, as seen
in the blood, appears to consist of a uniform, homogeneous mass of
protoplasm, of worm-like form, with at one end a thick, stiff extremity,
and at the other a long, wavy flagellum. It is generally in active
motion, and this is seen to be caused by the rapid lashing movement
of the flagellum, and by the rapid contractions and relaxations of the

On the Organism found in the Tsetse Fly Disease. 275

mass of protoplasm forming the body, and by the movements of an
undulating membrane which is attached to one surface of the body,
and which appears to undulate synchronously with the contractions of
the protoplasmic body. This membrane is, excepting at the free edge,
very transparent, and can be seen much better in citrated blood which
has been thickened by the addition of a small drop of 1 per cent,
gelatine solution, when its contour and attachments can be much
better made out, owing to the slower rate of vibration effected by the
thickened medium.

The general shape of the Trypanosoma, when rendered quiescent by
this means but not killed, is that of a long oval, with one end blunt
and the other continued into the flagellum ; the membrane is then seen
to be attached to one side of the body ; it begins a little in front of
the blunt end of the organism, and is continued at the end into the
flagellum.

But with better illumination, such as a very oblique pencil of rays,
or, better still, with monochromatic light (green or blue), the proto-
plasm is seen not to be homogeneous. The organism appears then as
a highly refractive body, and near the middle, or between it and the
flagellate end, is seen a large dark body much more refractive than the
rest of the protoplasm; this is the macronucleus. Near the thick,
stiff end of the body a tiny still more refractive body (with mono-
chromatic light nearly black) is seen, which is the micronucleus. The
addition of a drop of 5 per cent, acetic acid makes both of these bodies
much more distinct. At the stiff end of the Trypanosoma, in varying
relation to the micronucleus, is seen a vacuole. There is no suggestion
of a mouth or of any organs, but the protoplasm with the most careful
illumination appears not to be uniform, which suggests an alveolar
structure, as described by Biitschli. With the ordinary simple stains
(hematoxylin, fuchsin, methylene-blue, thionin) the differentiation is
not much better than can be observed by careful illumination of living
unstained organisms, as these stains are with these, and similar
organisms, too diffuse to be of any service. Acting on a method
which Ehrlich originated in 1889, and which Eomanowsky modified in
1891, and which has still been further elaborated by Ziemann in 1898,
we have used a mixture of methylene-blue and erythrosin, which has
enabled us to follow the different stages of the Trypanosoma with
certainty. This method depends on the fact that when a basic and an
acid stain are mixed together in certain proportions, a third neutral
body is formed, which has a specific colour reaction with chromatin.
By the use of this method we have been able to trace the various
stages of the organism in the blood and organs of the affected animals,
which is not possible with the ordinary stains, these being useless for
many of the forms to be presently described. With this method the
macronucleus of the Trypanosoma is stained a clear, transparent, crim-

276 Mr. K G. Plimmer and Prof. J. Eose Bradford.

son lake, the micronucleus a deep red, and the protoplasm a delicate
blue; these reactions are constant throughout all the stages of its
life-history.

The protoplasm of the adult Trypanosoma does not stain uniformly,
as does that of some of the other forms, but there are parts faintly
stained and parts unstained which is again in favour of the alveolar
structure mentioned above. The vacuole is quite distinct as a clear
round space, when the organism is stained by this method.

The macronucleus is generally of an oval or elongated shape, and it
may be either unform in colour, or in the form of fine threads ; this
latter is seen especially in those forms which show other signs of division.
The micronucleus is seen as an intensely stained round dot, or as a short
rod, this latter form again being seen in those' forms which show other
signs of approaching division. With the highest powers (1*5 apochro-
matic objective and 18 compensating eyepiece of Zeiss) we have not been
able to make out any special structural characters in this body. The
flagellum is not stained by this method, but if the preparation has
been well fixed, it is easily visible; the vibratile membrane also is
unstained, and can be generally better studied in specimens stained by
simple stains, preferably thionin.

As regards the movements of the organism, in preparations where no
pressure is exercised, they can be seen moving either with the flagellum
or with the blunt end in front ; but we think that the commoner mode
of progression is with the flagellum forward.

The size and length of the body varies very much with the period of
the disease at which the blood is examined, and with the kind of animal.
The largest forms we have seen have been in rat's blood, just after death,
and the smallest in rabbit's blood, early in the disease.

B. Distribution of the Trypanosoma.
1. In the body of Normal Animals.

(a) In the Blood. — We have found the flagellate form in the greatest
numbers in the blood of the mouse, towards the end of the disease.
In the rat also they occur in great numbers, and in both these animals
they can be found in the blood on the fourth or fifth day. In the dog
large numbers can be seen in the 1 blood from the sixth day. In the
cat they are fewer in number in the same lapse of time than in any of
the animals before mentioned.

The rabbit seems to be the most refractory animal of any we have as
yet used, and the Trypanosoma are found in the blood in small num-
bers only, and at very uncertain intervals.

(b) In the Lymphatic Glands. — In the superficial glands nearest to the
point of inoculation the flagellate organism can be found earliest. In
the rat the Trypanosoma can be found in the nearest superficial gland

On the Organism found in the Tsetse Fly Disease. 277

in twenty-four hours after inoculation. We have not found that
generalisation of the organism in the lymphatic glands occurs until
nearly the end of the disease, when the organism is present in very
large numbers in the blood. In the rabbit, in which the organisms are
few or rare in the blood, the glands do not show any marked change,
and the Trypanosoma are not readily found in them. Many other
forms are found in the glands, to which reference will be made
below.

(c) In the Spleen, — The adult Trypanosoma is found in but small
numbers in the spleens of the various animals we have examined ; but
other forms are found there which will be described later. The en-
largement of the spleen is post mortem the most obvious fact in the
morbid anatomy of the disease ; it may attain even to four or five times
the average volume — this is especially the case in the rat.

(d) In the Bone-marrow. — We have found either very few flagellate
organisms or none at all, in the bone-marrow of the various animals we
have worked with. The marrow is altered in colour and structure,
but there does not seem to be a greater number of Trypanosoma than
can be accounted for by the blood in the marrow.

In the other organs and parts the number of organisms present
depends upon the relative quantity of blood in the part.

2. In the Body of Spleenless A nimals. — As the spleen in the ordinary
animals is the organ which is most obviously altered in this disease,
we have made a series of inoculations into animals (dog, cat, and
rabbit) from which the spleen had been removed a year ago. In the
dog, the adult forms of the Trypanosoma are not found so early in the
blood of spleenless as in that of ordinary animals (seventh day as com-
pared with fourth day after inoculation). The glands, after death,
are much more generally enlarged, and are reddish in colour, and con-
tain many more organisms than in the normal animal. Both the blood
and glands contain, however, numerous other forms to be described
below.

This marked difference in the colour of the glands of spleenless
animals is probably clue to the removal of the spleen, and the glands
consequently taking on some of the splenic functions.

The bone-marrow is much altered, and in it likewise are found a
large number of Trypanosoma : both flagellate and what are termed
below "amoeboid" forms.

In the cat the conditions of experiment were altered, the blood (1 c.c.)
from the infected animal being introduced, with every precaution to
avoid contamination of the tissues, direct into the jugular vein. In
this case the organism appeared in the blood in numbers on the fourth
day, and the animal died on the twelfth day. As the Trypanosoma
were introduced into the blood stream direct, there was no marked
glandular enlargement, but the glands were all reddish in colour, the

278 Mr. H. G. Plimmer and Prof. J. Ptose Bradford.

change in colour being due to the splenectomy. A few adult organisms
were found in the glands and in the bone-marrow.

In the spleenless rabbit a few Trypanosoma have been found in the
blood on two occasions, but the animal lived nearly two months, and
notwithstanding the failure to detect adult flagellate forms in the
blood on numerous occasions, the blood was always infective, and con-
tained numerous forms termed " amoeboid " and " plasmodial " below.

C. Infectivity.

(a) In Ordinary Animals. — The blood and organs of an animal dead
of the disease lose, before twenty-four hours after death, their infective
power. This is apparently due to the rapidity with which decomposi-
tion sets in after death, as we have found living Trypanosoma in film
preparations, made as described above, as long as five to six days after
removal of the blood from the body; and we have also found that
large quantities (200 c.c.) of blood removed from the body into a
sterile vessel and kept in an atmosphere of oxygen, retain their virulence
for at least three days, notwithstanding the fact that the flagellate form
cannot be demonstrated.

We have found that the blood of the dog is infective at least two
days before any adult Trypanosoma can be seen in the blood ; and we
have also found that the blood of the spleenless rabbit, in which we
have only on two occasions seen any adult forms, is invariably infec-
tive. This of course suggests the idea that the organisms must be
present in another form, and we have been able, by the use of the
method of staining described above, to demonstrate the presence of
other forms in the blood and organs, and have shown, by the experi-
ments just mentioned, that the infectivity of the blood, in cases where
there are no flagellate forms discoverable, depends in all probability
upon the presence of one of the other forms which the Trypanosoma
assumes.

Although a differential staining method, such as the one we have
used, is necessary for following and demonstrating the various stages
of the life-history of the Trypanosoma, still these stages can be seen in
unstained living specimens, with very careful illumination. As a
matter of fact, our first observation of them was in unstained prepara-
tions.

In the blood of the dog, cat, rabbit, rat, and mouse, besides the adult
forms as described above, which, as mentioned, are very various in
size, there are adult forms undergoing division, both longitudinal,
and transverse, to which reference will be made later. Also two
organisms are sometimes seen with their micronuclei in close apposi-
tion, or fused together, with more or less of their bodies also
merged together. Such forms we believe are conjugations. Again, there
are other large forms, with or without a flagellum, in which the chro-

On the Organism found in the Tsetse Fly Disease. 279

matin of the macronucleus is broken up into a number of tiny granules,
not bigger often than the micronucleus. Besides these there are other
forms, which we call for convenience here " amoeboid " forms, by which
term we mean single, small, irregularly shaped forms, with or without a
flagellum, but always with a macro- and micro-nucleus. These nuclear
structures are generally surrounded by a very delicate envelope of
protoplasm, of greater or lesser extent, but occasionally forms are seen
which seem to consist only of chromatin, with or without a flagellum.
Besides these, again, there are other forms which we call, also for con-
venience, " plasmodial " forms, meaning thereby an aggregation or
fusion of two or more amoeboid forms. In the blood these plasmodia
are not generally very large, but may show evidence of from two to
eight separate elements. Signs of division are very common ; but in
the blood one does not often meet with a plasmodium dividing up into
more than four organisms of the adult shape. The plasmodial form
also retains intact the two nuclear structures — the macro- and micro-
nucleus — which we believe divide in the plasmodium, thus increasing-
its size.

In the spleenless animals the blood may contain no forms but the
amoeboid and plasmodial, such as is the case in the rabbit, yet this
blood is infective ; moreover, in the dog, before the adult organism
appears in it, the blood is infective, and therein, at this period, these
plasmodial forms can be demonstrated. In the glands these plasmodial
forms are found, but only in quantity in those animals from which the
spleen has been removed.

The spleen is the organ which shows these forms in the greatest
abundance. The whole spleen is crammed in every part with plas-
modia, which are wedged in between the splenic cells in every direc-
tion : many anoeboid forms, and also immature flagellate forms are also
seen, but the most striking thing is the enormous quantity and uniform
distribution of the plasmodia. The great enlargement of the spleen,
which we have found constant in all the animals we have used, is
caused by this mass of plasmodia, which we have found in the spleen
within forty -eight hours from the time of inoculation.

In the marrow these plasmodial forms are only found, so far as onr
experience goes, in those animals from which the spleen has been
removed. In these cases there are both plasmodial and amoeboid forms
in the marrow, the latter the more abundant.

The principal differences in the distribution of the plasmodial forms
in animals with and without spleens is this : that in the animals with
spleens the organ of choice for the plasmodia is the spleen, but they
are also found constantly in the blood, and in less quantity in the
glands, whereas in animals from which the spleen has been removed
the plasmodial forms are plentiful in. the blood, the glands, and the
bone-marrow.

280 Mr. EL G. Plimmer and Prof. J. Eose Bradford.

D. Life-History of the Trypanosoma " Bmcii."

Besides the forms mentioned above, we have seen in the blood and
in the organs divisions of the adult form, both longitudinal and trans-
verse, the former the more frequent ; but we think that this direct
mode of reproduction is far less common than the indirect by means
of conjugation (probably), breaking up of chromatin, production of
amoeboid forms, with subsequent division of these amoeboid forms, and
the formation of plasmodia by the aggregation or fusion of the
amoeboid forms, and these finally giving off flagellate forms, at first
small, and gradually increasing up to the normal adult form.

So that we should tentatively summarise the life-history of the
Trypanosoma found in Tsetse Fly disease, which we think might properly
be called " Trypanosoma Brucii" in recognition of the work done in
connection with it by its discoverer Major Bruce, F.R.S., as follows : —

1. Reproduction by division, this being of two kinds : —

(a) Longitudinal, the commoner.

(b) Transverse, less frequent.

2. Conjugation, consisting essentially, so far as our observations go,
of fusion of the micronuclei of the conjugating organisms.

(a) After this we are inclined to place those forms mentioned above,
in which the chromatin is broken up, and scattered more or less uni-
formly through the whole body of the Trypanosoma, since this occurs
after conjugation in other organisms not far removed biologically from
this one. The next stage in our opinion is the amoeboid ; we think
that the flagellate form becomes amoeboid perhaps after conjugation,
but also probably apart from this process.

(b) Amoeboid forms. These are found with and without flagella, of
various shapes and sizes, but always possessing a macro- and micro-
nucleus. These forms are constantly seen in the process of division,
and sometimes are very irregular in shape, with, in this case, an
unequal number of macro- and micro-nuclei, the latter being the more
abundant. The amoeboid forms then fuse, or aggregate, together to
form —

(c) The plasmodial forms. Whether these are true plasmodia, or
whether they are only aggregations of amoeboid forms it is not yet
possible to say, but as many related organisms form true plasmodia
we are inclined to look upon these masses, provisionally, as true plas
modia. In the spleen these plasmodia reach a large size. From these
again are given off' —

(d) Flagellate forms, which increase in size, and become the ordinary
adult form. Small flagellate forms are not infrequently seen in pro-
cess of separation from the margin of these plasmodial masses.

Besides these forms we have observed frequently, especially in rat's

On the Organism found in the Tsetse Fly Disease. 281

blood after death, the adult forms arranged in clumps. They appear,
upon watching them for a considerable time, to get tangled together
to form a large writhing mass ; then the movements become gradually
slower in the centre of the mass, and are only seen at the periphery.
At this stage, if the specimen be fixed, the mass appears to be made up
of a quantity of macro- and micro-nuclei, as the protoplasm does not
stain, except in the organisms at the periphery, i.e., those which have
arrived latest. Eventually these, too, become motionless, and the mass
becomes an indistinct collection of granular matter, which is not
infective, so that we look upon these tangles as a proof of death.

Since these observations were made, there has been published an
important paper on the Eat Trypanosoma, by Lydia Eabinowitsch and
Walter Kempner in the ' Zeitschrif t f iir Hygiene,' vol. 30, Part 2. We
have been able to confirm many of the observations and statements as
to the morphology and reproduction of the Trypanosoma made by
these writers. But there is no mention made of the plasmodial stage,
or of any reproductive stage elsewhere than in the blood; and the
writers recognize only three methods of reproduction, namely, longi-
tudinal and transverse division, and division by segmentation. This
segmentation, they consider, arises from one organism, and they state
that it may divide up into as many as ten to sixteen elements. This
segmentation form would seem to correspond to our plasmodial stage,
but we have seen much larger masses than those mentioned above, and
they do not notice the enormous masses of plasmodia which infil-
trate the spleen in every direction, and which can be found also in
glands, and marrow. Moreover, their amoeboid stage (Kugelform)
would precede the segmentation form, and therefore the " Kugelform "
should be much larger than the ordinary adult form, but we have
observed that as a rule, our amoeboid forms are very much smaller
than the adult forms, some not being visible with any but the highest
magnifying powers ; so that we have been unable to accept this form of
division by segmentation, except in the form in which we have
described it above, i.e., our plasmodial stage.

VOL. LXV.

Y

282

The Colour Sensations in Terms of Luminosity.

" The Colour Sensations in Terms of Luminosity." By Captain
W. de W. Abney, C.B., D.C.L., F.B.S. Beceived February 23,
—Bead June 15, 1899.

(Abstract.)

This paper deals with a determination of the colour sensations
based on the Young theory by means of measures of the luminosity
of the three different colour components in a mixed light which
matches white. At the red end of the spectrum there is only one
colour extending to near C, and there is no mixture of other colours
which will match it, however selected. At the violet end of the spec-
trum, from the extreme violet to near G, the same homogeneity of
light exists, but it is apparently due to the stimulation of two sensa-
tions, a red and a blue sensation, the latter never being felt unmixed
with any other. Having ascertained this, it remained to find that
place in the spectrum where the blue sensation was to be found
unmixed with any other sensation except white. By trial it was found
that close to the blue lithium line this was the case, and that a
mixture of this colour and pure red sensation gave the violet
<of the spectrum when the latter was mixed with a certain quantity
of white. The red and the blue sensation being located, it remained
to find the green sensation. The complementary colour to the
red in the spectrum gave a position in which the green and blue
sensations were present in the right proportions to make white, and
a point nearer the red gave a point in which the red and blue sensa-
tions were present in such proportions as found in white, but
there was an excess of green sensation. By preliminary trials this
point was found. The position in the spectrum of the yellow
colour complementary to the violet was also found.' The colour of
bichromate of potash was matched by using a pure red and the
last-named green. To make the match, white had to be added to the
bichromate colour. A certain small percentage of white was found
to exist in the light transmitted through a bichromate solution with
which the match was made, and this percentage and the added white
being deducted from the green used, gave the luminosity of the pure
green sensation existing in the spectrum colour which matched the
bichromate. Knowing the percentage composition in luminosity of
the two sensations at this point, the luminosity of the three sensations
in white was determined by matching the bichromate colour with the
yellow (complementary to the violet) and the pure red colour sensa-
tion. From this equation and from the sensation equation of the bi-
chromate colour already found, the composition of the yellow was
determined. By matching white with a mixture of the yellow and the

The Conductivity of Heat Insulators.

283

violet, the sensation equation to white was determined. The other
colours of the spectrum were then used in forming white, and from
their luminosity equations their percentage composition in sensations
were calculated. The percentage curves are shown. The results so
obtained were applied to various spectrum luminosity curves, and
the sensation curves obtained. The areas of these curves were found,
and the ordinates of the green and violet curves increased, so that
both their areas were respectively equal to that of the red. This gave
three new curves in which the sensations to form white were shown by
equal ordinates.

A comparison of the points in the spectrum where the curves cut
one another, and of those found by the red and green blind as matching
white, show that the two sets are identical, as they should be.
The curves of Koenig, drawn on the same supposition, are called attention
to, and the difference between his and the new determination pointed
out.

The red below the red lithium line, as already pointed out, excites
but one (the red) sensation, whilst the green sensation is felt in greatest
purity at A. 5140, and the blue at A 4580, as at these points they are
mixed only with the sensation of white, the white being of that
whiteness which is seen outside the colour fields.

"The Conductivity of Heat Insulators." By C. G. Lamb, M.A.,
B.Sc, and W. G. Wilson, B.A. Communicated by Professor
Ewing, F.B.S. Beceived May 3 —Bead June 15, 1899.

The comparative efficiency of materials used as insulators has been
the subject of several investigations ; the majority of these have been
conducted at fairly high temperatures, and it may be questioned
whether the results can be applied to the same materials when used as
a lagging to protect bodies at low temperatures. Moreover the
methods adopted do not seem susceptible of any considerable accuracy.
The method to be described was devised with the object of using
lower temperatures and smaller ranges than had been used in previous
experiments, to attain a perfectly steady state of heat transference,
and allow of greater accuracy and simplicity in the measurements.

The substances selected for experiment so far have been those which
could easily be tested in the dry state, without being made up into
cements; they include air, sawdust, charcoal, pine shavings, paper,
asbestos, sand, silicate cotton, hair felt, rice husks, and a heat insulator
known as " kapok."

The method employed consisted in placing the material under test in
the space between two cylindrical copper pots, kept at a definite distance
apart by pieces of vulcanised fibre ; the inner pot contained a small

Y 2

284

Messrs. C. G-. Lamb and W. G-. Wilson.

motor with a fan attached to the axis ; a tin-plate cylinder, open at the
top and with holes at the bottom, was put inside to direct the currents
of air over the inner surface of the inside pot, in the direction of the
arrows shown in Fig. 1. Energy was supplied electrically to a heating

Tig. 1.

coil within as well as to the motor : this constituted an internal supply
of heat, which maintained the temperature within the pot at any
decided upper limi^. The motor and heating coil were connected in
series, and leads were carried through a small hole in the lid of the
pots to measure the current and potential difference, and thus the

The Conductivity of Heat Insulators.

285

power expended on internal heating was measured. The outer pot
was immersed in a tank kept overflowing from the water main, the
lid of the pot being made into a sort of saucer, into which the incoming
water ran, thus the outer pot's surface was kept at a uniform and
constant temperature. The resulting temperature differences were
measured by means of thermo-electric junctions of copper and iron, in
a way shortly to be described. The current was allowed to pass steadily
into the inner pot, driving the motor and fan, until the inner thermo-
electric junction arrived at a steady value ; this usually occurred in
about three hours or so; when this was the case, the supply of energy
by the current was just equal to the heat conducted through the insu-
lator and carried off by the water. Knowing the temperature gradient
and the number of watts supplied and the dimensions of the system,
we can deduce the specific conductivity of the material.

The general arrangement is shown in Fig. 1. (P) is the outer pot
which stood in the tank j its approximate dimensions were 8 inches in
diameter and 1 6 inches high ; (Q) is the inner pot, with one inch clearance
Tjetween it and the outer one ; outside are shown the voltmeter (V),
.ammeter (A), and adjustable resistance (?•) ; inside the pot (Q) are a
fixed resistance (R), the motor (M), and fan (F). To a point about the
middle of each pot, inside the outer, and outside the inner, are soldered
the copper-iron junctions B, C, which are brought outside to a three-
way plug and a galvanometer, a is a junction placed in a vessel of
water at a known temperature ; g is a galvanometer of the Crompton-
D'Arsonval pattern. The junctions used were always made of exactly
the same length to keep the total resistance constant, and on calibra-
tion were found to give a linear relation between temperature difference
and deflection on the galvanometer within the ranges which were to
be used, and during each experiment a check calibration at two known
temperatures was made by means of a third junction in water at
another known temperature. When a steady state was attained, the
temperatures of B and C above A were measured by the deflections
and calibration tests. The reason for the above indirect method of
measuring the difference of temperature between B and C was to avoid
possible leakage currents. The flow of heat per second from the inside
pot to the outside one will be given by the expression H = XcO, where c
is the specific conductivity, $ is the temperature difference, X is a constant
depending on the size of the pots, being the area in square centimetres
of two plane surfaces, distant one centimetre apart, that would permit
the same flux of heat as the actual arrangement employed under the
same conditions of heat transference and temperature gradient. The
value of A. was calculated from careful measurements of their dimen-
sions, on the assumption that the flow of heat could be taken as radial
for the sides, and from the top and bottom of the inner, to the bottom
and top of the outer, pot ; this leads to the expression

286

Messrs. C. G. Lamb and W. G. Wilson.

ri

for this quantity, where ?x and /2 are the lengths, ri and r2 the radii of
the pots. This gives for A the value 1560 when the lengths are
measured centimetres. If W denote the rate of supply of energy in
watts, the rate of transference of heat in gram-C "-units when the
steady state is attained will be H = 0239 W. Hence, if c is the
specific conductivity, and 6 the temperature gradient,

H = 1560 c$;

therefore c = 0-000153 — .

6

Two points had to be settled before the determination could be
evaluated : — (1) Whether the temperature was uniform over both
copper pots ; (2) whether any part of the temperature gradient was
due to a sudden drop at the surface of separation. As regards the
outer pot, since it is wholly surrounded by water, its temperature*
must be uniform, and the inside and outside can only differ by the
drop through the copper, which is eliminated by putting the junction
inside that pot: to test the question as regards the inner pot (and
partly as regards the outer one) junctions were placed at various
points on the surfaces, as well as at certain other points, as shown in
Fig. 2. The result of this experiment is given below.

Junction.

1

2

3

4

5

6

7

8

9

Temperature
relative to S

25 -6

25-6

25 4

25 2

13 -8

0-6

0-6

o

28-1

It will be seen that the temperature over the surfaces of the pots
was practically imiforni, and hence the determination of the gradient
could be satisfactorily determined from readings taken with one junction
on each pot, as at first described.

The second point was tested as follows : — A third pot was provided,
having a clearance of 1 inch with regard to the former larger pot ;
under similar conditions as to character of insulator and power sup-
plied the temperature gradients in the case of the two combinations of
small and medium, and small and large, pots were measured ; the ratio
was found to be 0-5-±. The value of A for the new arrangement of
small and large pots was calculated, and found to be S±0 in centimetre
measure, being thus 0*538 of the standard combination. The close

The Conductivity of Heat Insulators.

2S7

agreement between these values shows that no appreciable drop occurs
at the separating surfaces.

These points being settled, the various substances named above were

Fig. 2.

Material.

Tempera-
ture

differences.

Watts
supplied.

Watts
per 1° C.

c.

19 2

25 -0

1-30

0 -000200

19-8

31 -7

1-58

0 -000242

Pine shavings

19 -5

20-8

1-06

0 -000162

Brown paper (crumpled up) . .

20-2

22 -0

1-09

0-000167

25 -8

24-5

0-95

0-000145

Hair felt in two sheets, % inch

i . , i ; . yi J i

27 '2

18 -9

0-69

0 -000106

14 9

29-0

1-94

0 -000297

27-8

27 2

0-98

0 -000150

8-0

39-0

4-85

0 -000740

14-0

13 -7

0-98

0 -000150

23 -2

21-8

0-94

0 -000144

27-8

22 -3

0-80

0 -000122

24 -7

24-5

0-99

0 -000151

288

On the Orientation of Greek Temples.

placed between the pots, and the watts and temperature gradients
were determined in the manner described. At least two tests of each
material were made, the mean temperature throughout being about 40°
centigrade. The results are given in the table (p. 287).

It will be noticed that hair felt was the best insulator that was tested.
The insulation in the case of brown paper was practically that of air
with subdivided spaces, as the paper occupied relatively a small volume :
a comparison of this with insulation by air only will show how great
an improvement in air-lagging such a simple expedient will give.

In repeating the experiments with wider ranges and a higher mean
temperature, indications were observed tending to show that the con-
ductivity is a function of the temperature. It is hoped to continue
the investigation as regards this point, and to extend it to other insu-
lators.

" On the Orientation of Greek Temples, being the Eesults of some
Observations taken in Greece and Sicily, in May, 1898." By
F. C. Penrose, M.A., F.E.S. Eeceived May 5, — Eead June
15, 1899.

(Abstract.)

The orientation of the Cabeirion Temple, near Thebes, of which the
angle has been disputed (see p. 46 in my paper of 1897), was re-
measured with the theodolite in May, 1898, and the previous observa-
tions confirmed. An additional example is added from an archaic-
Temple of Neptune in the Isle of Poros, introducing the employment
of the bright zodiacal star Eegulus, which I had not before met with.

In Sicily the re-examination of the temples at Girgenti, where, in
my former visit, I had relied for azimuth on the sun's shadow and the
time, has enabled me to give to the elements some amendments in
detail, the only point of consequence being, that the orientation date
of the temple named Juno Lacinia is brought within the period of
the Hellenic colonisation of that city.

The most interesting point in the paper seems to be, that in the
case of two Athenian temples, namely the Theseum and the later
Erechtheum — i.e., the temple now partially standing — it is shown that
the days of those months on which the sunrise, heralded by the star,
illuminated the sanctuary, coincided exactly, on certain years of the
Metonic cycle, with the days of the Athenian lunar months on which
three important festivals known to be connected with at least one of
those temples were held. The years so determined agree remarkably
well with the probable dates of the dedication of those temples ; and
in the case of the first mentioned, the festival, which was named The
Thesea, seems to leave little doubt that the traditional name of the
temple, which has recently been much disputed, is the correct one.

Efficiency of Ions as Condensation Nuclei.

289

" On the Comparative Efficiency as Condensation Nuclei of posi-
tively and negatively charged Ions." By C. T. E. Wilson,
M.A. Communicated by the Meteorological Council. Ee-
ceived May 11,— Eead June 15, 1899.

(Abstract.)

The experiments described in this paper were undertaken with the
object of throwing some light upon what appeared to be fundamental
questions in connection with the electrical effects of precipitation. It
was hoped in this way to make some advance towards an understand-
ing of the relation between rain and atmospheric electricity.

It was pointed out by Professor J. J. Thomson* that if positive and
negative ions differed in their power of condensing water around them,
drops might be formed upon one set of ions only, and separation of
positive and negative electricity would then take place by the falling
of the drops, the work required for the production of the electric field
being due to gravity.

To make this process worthy of consideration as a possible source of
atmospheric electricity, it would be necessary to show reason for
believing (1) that atmospheric air in the regions in which rain is
formed is likely to contain free ions, (2) that positively and negatively
charged ions differ in their efficiency as condensation nuclei.

It is mainly with the second of these questions that this paper deals.
The result of this part of the investigation was to prove that water
condenses much more readily on negative than on positive ions. The
experiments consisted in measurements of the expansion required to
cause condensation in the form of drops in air initially saturated and
containing ions, alternately nearly all positive and nearly all negative.
The ratio of the final to the initial volume being indicated by v2/vi,
then, to cause water to condense on negatively charged ions, the super-
saturation must reach the limit corresponding to the expansion v2/vi =
T25 (approximately a fourfold super saturation). To make water con-
dense on positively charged ions, the supersaturation must reach the
much higher limit, corresponding to the expansion v2/vi = 1*31 (the
supersaturation being then nearly sixfold).

We see, then, that if ions ever act as condensation nuclei in the
atmosphere, it must be mainly or solely the negative ones which do so,
and thus a preponderance of negative electricity will be carried down
by precipitation to the earth's surface.

Incidentally it was proved that the difference between the effects as
condensation nuclei of the positively and negatively charged ions is
not to be explained by supposing that the charge carried by the nega-

* ' Phil. Mag.,' December, 1898.

290

Miss M. Beeton and Prof. Karl Pearson.

tive ions is, say, twice as great as that carried by the positive ions, for
equal numbers of positive and negative ions are produced by the
ionisation of the neutral gas.

Attempts were now made to find an answer to the first question
suggested above — Is there any evidence that ions are likely to be
present under normal conditions in the atmosphere 1

Former experiments furnished a certain amount of evidence in
favour of an affirmative answer.

When moist dust-free air is allowed to expand suddenly a rain-like
condensation always take place if the maximum supersaturation exceeds
a certain limit. This limit is identical with that required to make
water condense on ions ; the identity is in fact so exact as to furnish
what is at first sight almost convincing evidence that ordinary moist
air is always to a very slight extent ionised. The number of these
nuclei is too small to make the absence of sensible electrical con-
ductivity in air under ordinary conditions inconsistent with the view
that they are ions.

All attempts, however, to remove these nuclei, by applying a strong
electric field such as would have removed ordinary ions almost as fast
as they were produced, have failed, even when a differential apparatus
was used, such as would have made manifest even a slight diminution
in the number of the nuclei by the action of the field. The same is
true of the similar nuclei produced by the action of weak ultra-violet
light on moist air.

Such nuclei, therefore, in spite of their identity as condensation
nuclei with the ions, cannot be regarded as free ions, unless we suppose
the ionisation to be developed by the process of producing the super-
saturation. This question requires further investigation.

" Data for the Problem of Evolution in Man. II. A First Study
of the Inheritance of Longevity and the Selective Death-rate
in Man." By Miss Mary Beeton and Karl Pearson, F.Pt.S.,
University College, London. Keceivecl May 29, — Read June
15, 1899.

1. According to Wallace and Weismann* the duration of life in any
organism is determined by natural selection. An organism lives so long
as it is advantageous, not to itself, but to its species, that it should live.
But it would be impossible for natural selection to determine the fit
duration of life, as it would be impossible for it to fix any other
character, unless that character were inherited. Accordingly the
hypothesis above referred to supposes that duration of life is an

* See Weismann, ' On Heredity,' Essays I and II, and especially Professor
Foulton's note as to Wallace, p. 23, of first English edition.

Data for the Problem of Evolution in Man.

291

inherited character. So far as we are aware, however, neither the
above-mentioned naturalists nor any other investigators have published
researches bearing on the problem of whether duration of life is or is
not inherited. We are accustomed to hear of a particular man that
" he comes of a long-lived family," but the quantitative measure of the
inheritance of life's duration does not yet seem to have been deter-
mined. This absence of investigation appears the more remarkable
as a knowledge of the magnitude of inheritance in this respect would,
we should conceive, be of primary commercial importance in the con-
sideration of life insurance and of annuities. The biological interest of
the problem, as we have already noticed, is very great.

2. It will be well in the first place to point out that the problem is
by no means an easy or a straightforward one. The ages at death of
even close relatives must be found from records of some kind, or else
collected ab initio. Now if we take records like the Peerage, Baronetage,.
Landed Gentry, family histories, and private pedigrees, we find various
serious omissions. -In the first place the ages of the women are rarely
given, pages of the Peerage or Landed Gentry may be examined before
a single record is found of the ages of two sisters at death, or even of a
mother and a daughter. Further, the census and other returns show how
liable we are to find women's ages given erroneously. Family histories
and pedigrees suffer in the same way, the pedigrees are mostly taken
through the male line, and the women's ages can only be found in rare
cases. An exception must be made in the case of the Quaker family
histories, such as those of the Backhouse, Whitney, and other families.
Here the data are as full for the women as for the men, but naturally
the history of a single, even much-branched family, does not provide
anything like the material that the Peerage and Landed Gentry do in
the case of men. For this reason our first study is confined to inherit-
ance of longevity in the male line. We hope eventually to collect
enough data for inheritance in the female line, but it will take a longer
time to amass, and we fear will scarcely be as homogeneous.

In the second place, the sources to which we have referred omit
more or less completely all record of the ages at death of infants and
children. The Quaker records give better results than the Peerage,
but even here the great bulk of child deaths appears to remain un-
recorded. Out of 1000 males born in this country more than 300 die
before they are 20 years of age. But when 1000 cases of ages of
father and son were taken out of the Landed Gentry, only 31 cases
showed the death of a son before 20 years of age. Of 2000 brothers
from the Peerage in 1000 pairs, only 21 individuals died before 20
years of age. In the Quaker histories we found about 16 per cent, of
deaths before 20 years of age. Clearly such early deaths are not
represented in anything like their proper proportions. They will
have to be found from other sources ; possibly by direct inquiry, and

292 Miss M. Beeton and Prof. Karl Pearson.

the issue of data cards. We were thus compelled to limit this first
• study to the case when both relatives die at a greater age than 20.
In the case of fathers, when we are dealing with the correlation
between father's and son's ages at death, this is practically no limita-
tion at all, as no father dying under 20 years of age was met with.
In the case of the offspring, however, the limitation cuts off' the dis-
tribution somewhat abruptly with a finite ordinate at 20 — 25, five
years being our unit of grouping.

3. Now duration of life is a very different character to eye-colour, or
to some extent to the size of organs in adult life. Eye-colour is fairly
well determined, it may change with old age slightly, but it cannot
transform itself from light blue to brown. Again, nourishment and use
undoubtedly affect the size of organs, but they are likely to influence
father and son, or, at any rate, brother and brother, in much the same
way, for they are members of the same family and the same class. On
the other hand, death depends not only on inherited constitution but
on innumerable chance elements of environment and circumstance.
The environment both of home and period is much more alike for two
brothers than for a father and son; food, sanitation, habits of life,
change considerably in a generation, and two brothers have more equal
chances of life than a father and son. But even with two brothers,
one may live on the family estates and the other ruin his health in
Africa or India. Hence, while the non-differential death-rate will not
materially alter the correlations between most characters in relatives,
it must seriously affect the correlation between the durations of life in
father and son, and to a lesser extent between brother and brother.
A good stock may be better protected against death than a weak one,
but no stock at all can resist certain attacks. Hence if we look upon
death as a marksman, p per cent, of his shots are, we may say, sure to
be effective whatever they hit, this is the non-differential death-rate,
the remaining 100 —p per cent, of his attacks will only be successful
on the weaker stocks. Now the effect of this conception of death's
action is that the correlation table for ages at death of any pair of
relatives must be looked upon as a mixture of uncorrelated material —
deaths due to the non-differential death-rate, and correlated material —
deaths due to the differential or selective death-rate. At different
periods of life also one of these death-rates may give more material to
the table than at another. In the case of fathers and sons we should
expect the non-differential death-rate to be more numerous in its con-
tributions than in the case of brother and brother.

Now it has been shown by one of us that when correlated material
is mixed with uncorrelated material, the result is approximately to
reduce the coefficient of correlation in the ratio of the amount of corre-
lated to the total amount of material.* Hence, if we assume that the
* ' Phil. Trans.,' A, vol. 192, p. 277.

Data for the Problem of Evolution in Man.

293

actual correlation between constitutional strengths to resist death
would be given, at any rate approximately, by the values determined
for other characters in a memoir on the Law of Ancestral Heredity,*
we have clearly a method of to some extent ascertaining the proportion
of the selective and non-selective death-rates in man. In the sequel
it will be shown that from the age of 20 to the end of life our tables give
a correlation between the duration of life of father and son of about
0*12 to 0-14, and between brother and brother of about 0'26. Accord-
ing to the Law of Ancestral Heredity we should expect these quantities
to be about 0*3 and 0*4. Hence we conclude that the amounts of
correlated material in the two cases are 40 to 50 per cent, and 65 per
cent. But if be the number of cases in which the death-rate is
selective for N individuals, y2N will be the number of cases for which
it is selective when we take pairs of individuals. In other words the .
selective death-rate in the first casef is about 63 to 70 per cent, and in
the second about 80 per cent, of the total death-rate. Without laying
great stress on the actual numbers just stated, we think that they are
sufficiently close to demonstrate that a substantial selective death-
rate actually exists at work on mankind, and that with like environ-
ment it may amount to as much as four times the non-selective death-
rate.:}: In other words, having demonstrated that duration of life is
really inherited, we have thereby demonstrated that natural selection
is very sensibly effective among mankind. The natural selection we
are here dealing with is not in the first place, of course, a result of any
struggle of individual with individual, but of individual with environ-
ment and with the defects of personal physique.

4. In order to show the biological importance of investigating the
inheritance of duration of life, we have cited the results obtained for
correlation between the ages at death of father and son and brother
and brother. But the method by which these results were obtained
requires further discussion. We have already seen the need to exclude
deaths under 20 years, but even then we have not got in the case
of father and son two like groups of material. The father has been
more severely selected than the son. He has lived to become a father,
and he is strong enough to be the father of a son who lives to be

* ' Koy. Soc. Proc.,' vol. 62, pp. 397 and 400.

t This selective death-rate from the data for father and son must be interpreted
in the sense indicated above. The drop from 80 to 65, say, per cent, is in itself a
measure of the change of environment of the two generations.

t The correlation on which this determination is based might be illusory, if
families were reared under very individual environments ; the correlation in
duration of life of brothers, for example, might then be a result of their individual
family environment. But the environment when we take comparatively homo-
geneous classes like the Peerage or Landed G-entry must be very similar, and we
think this source of error, suggested to us by Professor Weldon, while very real
has been sufficiently provided against.

294

Miss M. Beeton and Prof. Karl Pearson.

20 years of age. Evidence of this selection is to be found in the
facts that (1) fathers have a mean age 5 to 7 years greater than
that of their sons ; (2) the variability of their age at death is very
sensibly less than the variability of their sons' age, i.e., as 2'9 to 3*5,
or (3) by noticing for example that in our first table 82 sons as
against 20 fathers die before 32*5 years of age, and that in our second
table some 100 sons as against 20 fathers die before 35 years of age.
Clearly the group " son " is a much weaker type than the group
" father." As will be shown in a memoir on the effect of selection on
correlation, this want of likeness in fathers and sons itself tends to
modify the correlation between them.*

While this selection occurs only in the case of fathers and sons and
not in the case of brethren, still the general character of the correla-
tion surface is alike in both. It is known that the curve of frequency
of death at different agesf is by no means normal. It is probably
compound, and only approximates to normality round three score
years and ten. It would hardly, however, fulfil a useful purpose to
deal only with the correlation of ages of death of relatives both dying
under the old age mortality group, even if on the sunny side of 70 we
could distinguish old age from middle age mortality. But in dealing
with correlation and regression in such cases as this, we must throw
entirely on one side any notion of normal surface and curves of error,
and go simply to the kernel of the affair.

What we want is the law connecting the mean age at death of one
relative when another relative has died at a given age. When the
given age of the latter and the mean age of the former are plotted to
form a curve, this curve is the regression curve whatever be the form
of the frequency surface. The line of closest fit to this curve is the
regression line, and Yule's theorem^ tells us that the slope of this line
is found in exactly the same way as if the frequency surface were a
normal distribution. The slope of this line has nothing whatever to
do with the particular form of surface, and may be found even if we
cut off a portion of the surface parallel to one axis, e.g., if we take the
regression line for fathers or sons we get the best fitting lines in pre-
cisely the same manner whether we take all sons dying from infancy
to old age, or only those from 20 years onwards. If, of course, the
regression curve is sensibly linear, then the regression line is the true
curve of regression. Everything proved in the memoir, " On the Law
of Ancestral Heredity "§ holds for such linear regression equally
well ; we need not suppose normal correlation. Now the reader has

* Not of course very largely, still with the values given in the first series of
fathers and sons, the correlation would be reduced about 0-86 to 0-9 of its value by
the selection of fathers

f ' Phil. Trans.,' A, vol. 186, p. 406 and plate 16.

% ' Roy. Soc. Proc.,' vol. 60, p. 480.

§ ' Roy. Soc. Proc.,' vol. 62, p. 336.

Data for the Problem of Evolution in Man.

295

only to look at our regression diagrams, in particular at that for
brethren, to assure himself that no curve will serve for practical pur-
poses substantially better than a straight line. Now, if <tx be the
standard deviation of the relative whose mean age at death is taken?
and o-y of the relative of a given age at death, and r be the correlation
defined by

r = S{.z(x-mx)(ij-my)}l(N(rxo-y),

where z is the frequency of deviations x-mx and y-my from the
means mx and my in the total observations N, then the line of closest
fit, the regression line, passes through mx and my, and has rcrx/o-y for
its slope. All this is independent of any theory of frequency distribu-
tion, and the vanishing of r with the correlation simply flows from the
fundamental problem that the chance of a combined event is the
product of two independent probabilities. Our conclusions in this paper
are deduced from the above value of r and from the slope of the regres-
sion line, and they involve no further assumption than the approximate
linearity of the regression curves. Our appeal to the memoir, "On
the Law of Ancestral Heredity " makes also no greater demands.

5. We now turn to the material itself. Our data consist of three
series, from which all deaths recorded as accidents, an exceedingly small
proportion of the whole, were excluded. In excluding these we of
course slightly, but very slightly, reduced the non-selective death-rate.
In the first series, 1000 cases of the ages of fathers and sons at death,
the latter being over 22*5 years of age, were taken from 'Foster's
Peerage'; in the second series a 1000 pairs of fathers and sons, the
latter dying beyond the age of 20, were taken from ' Burke's Landed
Gentry'; and in the third series the ages at death of 1000 pairs of
brothers dying beyond the age of 20 were taken from the Peerage.

The first series was obtained by grouping all fathers dying between
22-5 and 27*5, 27'5 and 32*5, &c. We started at 22*5 because this
was the earliest recorded death of a father among those extracted from
the Peerage, and to have sons dying in the same range they were also
.started at 22*5 years. In extracting the ages at death, they were
taken to the nearest whole year, and consequently in the subsequent
grouping we were spared decimals. In the second and. third series we
•originally took all deaths from birth onwards also to the nearest
whole year, and then grouped in five-year periods ; thus fractions were
introduced when a death fell on a five-year division. Subsequently we
eliminated the few deaths occurring before 20 years of age.

The aggregate material for the three series is given in Tables I, II,
and III ; and the means of the arrays of fathers' ages at death for sons
dying at a given age, i.e., the regression polygons of fathers' on sons'
age at death in figs. 1 and 2 ; the regression polygon for brethren is
given in fig. 3.

In the case of brothers, we have rendered the original distribution,

29G

Miss M. Beeton and Prof. Karl Pearson.

which was nearly symmetrical, absolutely symmetrical, by entering
into the table each pair of brothers twice, an individual first appearing
-as a first brother and then as a second brother. Thus the mean age at
death and variability of age at death of both sets of brothers appears
the same, and we have a nominal 2000 instead of a 1000 entries. Of
course in calculating the probable errors of the constants, 1000 has
been taken as the number of observations. We shall now consider
these diagrams and tables a little at length.

Fig. 1. — Diagram giving Mean Age of Fathers at Death for Sons dying at a given
age. First Series, 1,000 cases.

I

a \h

y

\

\

\

h

\

afbf

fa

r

)=AcU
irGdl

Ms

ygon.

\ i

— jtti

:=

ressi

7/7 Lit

e,20L

, 50i

'OIOO.

)oioo.

\ I

*A

0

~X —

-| —

— &-

W'
Ml

-iW

X"

J

4

\

-%

— ^ — > —

— tt

\ 1

\

\

A —

1

\

i

-rH
\k

m

45 50 55 60 65 70 75 80 65 90 95 100

Mean Age of Fathers at Death.

Data for the Problem of EvolvMon in Man.

297

6. First Series. — The means of the arrays of fathers for a given age
at death of the son, are shown by the broken line abcdefg in tig. 1.
The point a for the group of sons dying between 17*5 and 22*5 years
was put in from a few observations not afterwards included in the
table. Beyond the group 82*5 to 87*5 years, there were not sufficient
observations to form a reliable mean at all : yy gives the mean age of
all the 1000 fathers observed, and represents 65*835 years, xx. gives
the mean age of the 1000 sons, and represents 58*775 years. The
former may be taken as the mean age at death of all fathers, the
latter was only the mean age at the death of so as who live more than
22*5 years. The regression curve is a somewhat broken polygon, but
one or two points may be deduced at once from it.

(a) It is entirely to the left of yy above xx and entirely to the right
of yy below xx. Thus there is certainly correlation between the ages
at death of father and son. A son dying below the mean age will
have on the average a father dying below the mean age, and a son
dying above the mean age will have on the average a father dying-
above the mean age. Graphically we see that correlation must exist.
The straight line which best fits the regression polygon is given on the
diagram by Id. The Law of Ancestral Heredity would give hn with a
slope of 0*3. It is clear that with a quite sensible regression there is
a quite sensible divergence from the law of inheritance, in other words,
the death-rate is only in part selective.

Quite similar results are to be observed in fig. 2 ; there is again a
very sensible correlation, but it is sensibly less than that required by
the Law of Ancestral Heredity. The lines are lettered the same.
Numerically, if Ms, Mf be the mean ages at death of sons and fathers,
o-s, o-f their standard deviations, *»*sf their correlation, RSf = ?"sf o*s/SF,
RFS = rsF o"f/o"s the regression coefficients of son on father and father
on son, we have —

i

First Series.

Second Series.

1 Peerage.' Fathers and sons, .

25 years

' Landed Gentry,' Fathers and sons,

and on.

20 years and on.

65 "835 years

Mf

65 '9625 years

58 775 „

Ms

60 -9150 .,

14-6382 „

<7F

14 -4308 „

17 "0872 „

c-s

17 -0986 .,

0-1149 ±0-0210

'/"ST

0-1418 ± 0-0209

0-0985 ±0*0182

Rfs

0-1 196 ±0-0178

0-1341 ±0-0367

Ksf

0-1682 ±0-0371

Now these results extracted from very different records are in good
accordance. The values of the correlation and regressions are 5 to 7

VOL. LXV. Z

298

Miss M. Beeton and Prof. Karl Pearson.

times the magnitudes of their probable errors, and they agree within the
probable error of their differences. The only significant difference
is the mean age of deaths of sons in the Landed Gentry, which is some
two years higher than in the Peerage. This is the more noteworthy
in that we have begun our peerage record at 25 and not 20. Clearly
the sons of the Landed Gentry are longer lived. We have undoubtedly
correlation, say somewhere about 0*12, sensible and definite in
amount, but clearly considerably below the 0*3 required by the law of
inheritance.

(b) A second point may be noticed by looking at the diagrams (1)
and (2), namely, that from about the age of 32*5 to 52*5 the regres-
sion line is sensibly vertical, or when the son dies in middle life, the
mean age of death of the father is sensibly uncorrelated with it. In
other words, we have the remarkable result that the mortality which
in a paper on skew variation by one of us,* has been termed that of
middle life is largely uninherited. It is during this period of life that
the non-selective death-rate is chiefly predominant. After this period
the regression curve becomes sensibly steeper, although not fully up
to the steepness of the line given by Galton's Law. This is more
properly the inheritance of longevity. The inheritance of duration of
life may not be continuous.

If we seek the best fitting straight line for the regression polygon
from 50 years onward we find : —

First Series.

Second Series.

' Peerage,' 52 "5 years of son and on.

,

' Landed G-entry,' 50 years of son
and on.

66-680 years
69 -686 „
14 -6734 „
9 -6148 „
0-1156 ±0-0232
0-1764± 0-0380

Mf
Ms

<TF
<*S

Rfs

66-878 years
68-960 " „
14-3273 „
10 -4055 „
0-1125 ±0-0243
0-1549 ±0-0333

Results such as these are as close as we could expect, and they mark
an increase in the steepness of the regression line from about 0*11 to
0-17, an undoubtedly substantial increase of the selective death-rate as
we approach old age. The regression line for this old age mortality
is marked as jk in diagrams (1) and (2), and we see the advance
towards the Galtonian value.

(c) Below 32*5 years the regression line in figs. 1 and 2, especially
the former, seems to indicate increased correlation again, but unfortu-

* ' Phil. Trans.,' A, vol. 186, p. 408, and Plate XVI.

Data for the Problem of Evolution in Man.

299

nately our records do not give enough data to determine its form in a
reliable manner.

Fig. 1 seems to indicate a great approach to the Galtonian value
towards youth, and we should not be surprised to find the selective
death-rate in youth and infancy even more predominant than in old age.
This would be the inheritance of the reverse of longevity, of "brachy-
bioty." The regression curve for this portion of life cannot be deter-
mined from our present statistics, but we hope to return to it in a second
study when more data have been collected.* So far as we are able to
judge at present the inheritance of the duration of life breaks up into
two parts, an inheritance telling its tale in youth and another after
middle life. It is the former part which seems to us to have most
bearing on the fertility and survival of stocks, most individuals having
reproduced themselves by 50 years of age. It is the latter part only,
the true inheritance of longevity, to which it would appear that
Weismann and Wallace's arguments apply : — f

"For it is evident than when one or more individuals have pro-
vided a sufficient number of successors, they themselves, as consumers
of nourishment in a constantly increasing degree, are an injury to
those successors. Natural selection therefore weeds them out, and in
many cases favours such races as die almost immediately after they
have left successors."

7. We now turn to the third series, giving the correlation between
the ages of death of brothers. The data give the following numerical
results : —

MB = 60-971 years.
<rB = 16-8354 „
rBB = 0-2602 ±0*0199.
EBB = 0-2602 ±0-0216.

The results here are in good agreement with those for sons in the
Landed Gentry, i.e., in the second series. The mean age of one of
a pair of brothers is slightly greater and the variability of one who has
a brother slightly less than in the case of sons. But this is exactly what
we might expect, considering that "brothers" are a selection from
"sons," and a brother is likely to have greater vitality than a son. The
group sons covers sons of fathers who did not live to have more than one
son, and who therefore came of any early dying stock, while brothers
denotes at least two sons, and therefore on the average some years
more life than is necessary for one son.

The values of the coefficients of correlation and regression are some
13 times their probable errors, and we have a substantial correlation,.

•

# This collection has already commenced, and we hope shortly to give more-
definite information on this point,
f Wallace, loc. cit., svpra.

z 2

300

Miss M. Beeton and Prof. Karl Pearson.

-Diagram giving Mean Age of Fathers at Death for Sons dying at a given Age.
Second Series, 1000 cases.

20

25

50

35

40

50

•

s

oo

*«

60

1

65

"is

70

ft]

75

80

85

30

35

100

105

i

\<pa

y

\

\

u

\

c,dAf
4

m~G

CtUdt
dUot

Reg,

'essic
iw.

nPoL
j

tgon.

J

k =

"egres

sion

Line,

20 U
50 U

uoo.

)I00.

v|

0

X

X

x'~

VA

e

\
\

\ —

H

\

\\

%

y

— L

\

k

m

45 50 55 60 65 70 75 80 35 90 95 100 105

Mean Age of Fathers at Death.

approaching much closer than in the case of sons to the value
demanded (0*4) by the Law of Ancestral Heredity. The diagram shows
(i) how substantial is the correlation; (ii) how much more nearly the
regression line JcJc given by observation approaches the theoretical line
Im • and (iii) how very nearly the regression curve is truly linear.
The reason of this closer approach to the theoretical value of heredity
is owing to the diminution in the non-selective death-rate, the environ

Data for the Problem of Evolution in Man.

ment of brothers during their lives being as a rule much more alike
than that of father and son. It must be noted that the predominance
of the non-selective death-rate in middle life, so marked in the latter
case, no longer appears in the case of brothers. This would suggest
that the environments of father and son differ most in middle life and
are then much more unlike than those of brothers.

8. We conclude this first stud}- by putting on record formulae for
estimating the age at death of a man, using the theory of multiple
correlation as developed in a memoir* by one of the present writers,
and taking as basis the second and third series, which seem to us to
present the best results.

Let P be the probable age in years at death of a man, F be the age
at death of his father, Si of his first son, S2 of his second son, Bx of his
first brother, Bo of his second brother. Then we have the following-
cases : —

Prediction of Age at Death. All Deaths after 20 Years.

(a) From age of father at death —

P = 49-8201 +0-1682 F, 2 = 16-9259.

(b) From age of brother at death —

P = 45 1063 + 0-2602 Bl5 ^ = 16 2555.

(c) From age of son at death —

P = 58-6771 +0-1196 Sx, 2 = 14-2850.
((7) From ages of father and brother at death —

P = 37-6647 + 0-12685 F + 0-24502 BT, 2 = 16-4099.
(e) From ages of father and son at death —

P = 48-7991 +0-15706 F + 0-11168 S, 2 = 14-1573.
(J) From ages of two brothers at death —

P = 35-7930 + 0-206475 (Bx + B2), 2 = 15-9052.

(g) From ages of two sons at death —

P = 54-3928 + 0-09497 (Si + S2), 2 = 14-1987.

(h) From ages of brother and son at death —

P = 44-2601+0-1046 S + 0-2514 B, 2 = 13-8508.

Here 2 is the standard deviation of the array of men for each group.
Such formulae f seem to us to give a quantitative accuracy to much

* " Contributions to the Mathematical Theory of Evolution. III. Kegression,
Heredity, and Panmixia," ' Phil. Trans.,' A. vol. 187, pp. 253 — 318.

f In obtaining the formulae for prediction from the age at death of Uvo relatives,
certain assumptions have had to be made. Thus the correlation of ages of a man
and his grandfather and of a man and his uncle at death, being at present unknown,
were taken to be half the correlation of father and son. This cannot be far wrong,
but the actual values ought to be found. We did not feel justified in assuming the

302

Miss M. Beeton and Prof. Karl Pearson.

Fig. 3.— Diagram giving tbe Mean Age of Man at Death for a Brother dying at
given Age. 1000 cases.

1

\y

1

\

b

a,b,c,

1,7?

•UdLt, i

Lborii.

?epre

LcLV*

ssior

;

PoLy

§0/7.

Q=Mt

M

)dtlAc

o

►= — —

gress
eAU

ion L
drotk

ine.
ers a

£ Dec

\ \c

* \\

d

X

0

i

— —

~~x

X

! ^

\

\

V

45 50 55 60 65 70 75 80 85 30

Mean Age of First Brother.

35 100 105

that is allowed rather indefinite weight at present in the actuarial and
medical professions. Based on a wider mass of data and a larger series
of relationships we cannot but believe they would be of much help to

variability of grandfathers, which must be less than that of fathers, or their mean
age at death, which musk be greater than that of fathers, in order to determine the
probable age at death of a man from that, say, of his grandfather and father, which
would be of much interest. We only wish to draw attention to what we believe to
be a new and important field of enqxiiry, and to indicate tbe nature of its problems.

Data for the Problem of Evolution in Man.

303

the physician and actuary. If their importance were once recognised
by the insurance offices, we believe that the necessary data would be
readily forthcoming. As illustrations take the following : —

(i) A man's father dies at 40, and his brother at 25. What is the
probable reduction in his own life % Answer : 12 years.

(ii) A man has two brothers, who die young at 25 and 29. How
much will this shorten his probable duration of life % Answer : 14 years.

(iii) A man's father died at 40, and his brother, his senior by one
year, died at 50, twenty-two years ago. A life estate now accrues to
the man, whose whereabouts are unknown. What is the probability
that he is still alive, and should he return and claim the estate, how
long is he likely to enjoy it. Answer : The man belongs to an array of
men of mean age 54*99 years at death, and standard deviation 16*41
years. Hence the odds against his living beyond 71 years are 835 to
165, and, accordingly, the odds against the possibility of his return are
about 21 to 4. Should he be alive and return, he is as likely as not
to hold it for 6*8 years, and 8*7 years is his expectancy of life, so that
the contingency of his being alive and enjoying the estate is worth only
about 1*4 years' income of the estate.

Clearly such problems can be extended in a great variety of ways,
which might be serviceable in actuarial practice.

I.— Correlation Table for the Inheritance of Longevity from Father to

Son.

First Series.

Age of father at death.

20

25

30

35

40

45

50

55

60

65

70

75

80 85 90

1

95

100

105

Totals.

25

1

1

3

1

1

9

4

3

5

4

3

1

1

37

30

1

3

1

4

1

6

2

6

3

10

5

2

1

45

35

1

2

4

6

4

5

10

6

6

5

7

1

57

40

1

5

2

3

5

6

6

11

7

17

11

2

1

77

45

1

2

2

3

3

5

6

11

11

8

6

4

1

63

50

1

2

2

1

3

7

13

10

4

8

14

11

7

1

84

55

1

2

3

5

5

8

10

7

6

12

9

8

5

2

83

60

1

1

2

2

6

9

6

7

17

13

10

2

2

1

79

65

1

2

5

5

4

11

10

5

18

18

14

10

4

107

70

2

3

1

4

10

9

12

23

9

20

12

10

3

1

119

75

3

1

2

4

3

9

6

10

8

20

23

13

10

6

2

120

80

1

3

1

3

2

6

3

6

9

10

6

8

1

59

85

1

2

1

3

2

6

6

10

8

8

2

1

1

51

90

1

2

2

2

2

2

3

14

95

2

1

1

4

100

0

105

1

1

Totals

9

11

28

30

44

64

99

85

108

133

165

115

77

25

6

0

1

1000

Miss M. Beeton and Prof. Karl Pearson.

2

I

§ r§ 1 S S 2 | 5 js & ^

9

1

1 1 1 ! 1 1 rt 1 1 1 - 1 1 « *1 It 1

•CO.

1 I 1 t 1 1 1 1 1 ? £ ? 1 f !

O O —1 r-<

| j | 9 9 -P | 'P ?> ? £ £ o | j

rH r-l O O l-t CO C9 «

9

'p

OG

I j , e 9 « -g g * o s e | | o. o |. ,

■P

S

9 9 ? £ o ? & 12 £ £ £ ? £ £ £ j |

9

s

© * © o o $ .= ? £ § 3 S © « £

9

s

'? ? 'P = 9 'P £ £ 9 £ T- '? «T 9 IP ■

'P

'P ? ? ? ? ? P f 'P , ,

9.

a

OOM « £ £ 9 « 9 ? «c >o p 8 * o o

j t- rH cm «c co o o o rH -m o ^ cc .— 1 1

9

,|

9

3-SSc

f ? ? 'P 'P P ^ 'P 'P ? 9 ? $ 'P 9 | |

'P

S

9 £ £ 'P f f 'P « £ 9 £ 'P '#i •? |

1 1

P
El

§£3

9 9 9 9 'P 9 ? ? 'P 9 £ ? 9 9 !|

1 1

9

&3§

'P 9 'P 9 'P | ? « 'P 9 * 'P 'P 'P jp | |
rn^irtcc^ eo oj oo © i-h co o p

9

?i

83&

| | | H | >P ! | 9 V V 9 9 | |

WOO 00 iH irt <M rH

1 V

9

| | | $ « ? | | <? « ? | 'P | |

O O Ol rH C<l t-H o

1 1

•P

!

i i - i i i i i i i i i i i i ;i i

ZBSZSSSSZSSSSSZSS

i

•t?:p?dp itj uos jo sSy

Data for the Problem of Evolution in Mara.

305

i

© w o us p © o « p o p .p p p p

1

1?
i ©
§

(N

i i i i i i i i i i i i i i i i ,i.

1

Ml

i i i i i i i i i i i i i i i i i

1

8'ss

o 10 its o ic ><; 'O £ 10 £ © ©

r-i 1 © © —i 1 o . ih iti ea i-< op w w 1 1 1

iC

©

ic ia o o o t-> »c t- t- £ © i(i © o
...... • . . • - . . . i

© rH 00 I-) iH csi *a SO t- 00 • © W ICS dt> cn 1 1

l(t

k

1- !- ic o o is o c o l> n iO i(? ©

......... . .... 1

Hi

k

op© p £ p £ ?i p p © p £ © £

oo «, » « « s $ i $ g <». * 1 1

'p

ao 'N p p ? s M ? ? ? ? * ? s * . i *
w h ^ a s * s a s s s M 1 1 §

BOO ? ? ? .9 ? « P P .P § « ? S ?

B<3*. H ^ « « * H g g g- « H « * ffl CO 1 1

9

P 9 9 9 9 ? 9 « 'P P 'P 9 £ « , j
* w « 31 * 2 £ 2 S3 8 ?1 S S l~ rt 1 1

S

p £ ?> © S S 9 9 'P P P "p 'P 'P . |

'P

9 ? ? M ? ■? ?. ? ? ? ,
« s 2 " 2: ?] 2 2 a ° 1 1

9

1

P p p £ p p g S £ £ Si p p £

9

§

P = £ = « 8 © © = o £ £ © © ©

« w • l- 2 - 2 J- s • a « * " * 1 1

1 ■

S3?

9 'P ? 9 9 <T P f» 9 9 © p © © p

*
s

p © p £ £ p £ £ o o p © p © p

9
%

p © © p © p © p © £ p © £ p t

'P
3

| 'P 'P 9 'P 'P 'P | 'P P 'P 9 ? 'P 9 | :

9

20 to 25
25 to 30
30 t o 35
35 to 40
40 to 45
45 to 50
50 to 55
55 to 60
60 to 65
65 to TO
70 to 75
75 to SO
80 t o 85
85 to 90
90 to 95
95 to 100
100 to 105

I

"tf^uai.) %v. .t31[^ojq pxxoogs jo aJjY

306

Collimator Magnets. Thermal Expansion.

" Collimator Magnets and the Determination of the Earth's Hori-
zontal Magnetic Force." By C. Chree, Sc.D., LL.D., F.E.S.,
Snperintenclent of the Kew Observatory. Communicated by
the Kew Observatory Committee of the Eoyal Society.
Eeceived May 31, — Bead June 15, 1899.

(Abstract.)

During the last forty years, there have been examined at Kew
Observatory upwards of 100 collimator magnets used in observing the
horizontal force and declination.

The " constants " of these magnets — temperature and induction
coefficients, and moment of inertia — have been determined at the
Observatory, and the tables based on these determinations have
served to reduce magnetic observations at a large number of the
leading magnetic observatories.

The present paper deals with the data recorded in the Observatory
books for the constants specified above, and with other quantities —
such as the " permanent " magnetic moment — which are deducible from
the records. It determines the mean values of the several quantities
for the instruments of the leading English makers, and investigates
whether relations do or do not exist between them. It then deduces
from the records the probable errors in the values of the several
quantities, proceeding on the hypothesis that the methods of deter-
mining them are correct. It next examines, from a mathematical
standpoint, the accuracy of the formulae employed in reducing hori-
zontal force observations, and, from a physical standpoint, the possibility
of differences between the quantities determined at the Observatory
and the quantities actually concerned in horizontal force observations.

The various sources of uncertainty are dealt with, and an attempt
is made to ascertain to what extent they may affect the values found
for the horizontal force.

The results of the paper are of too technical a character to admit
of their being summarized briefly in an intelligible way.

" The Thermal Expansion of Pure Mckel and Cobalt/' By A. E.
Tutton. B.Sc. Communicated by Prof. Tilden, D.Sc, F.B.S.
Eeceived April 18, — Eead May 5, 1899.

The following are the numerical experimental data of the eighteen
individual determinations of the coefficients of expansion of pure nickel
and cobalt, referred to in the abstract previously published (p. 161,
supra). Full explanations of the signs employed in the tables will be
found in the memoir " On the Thermal Expansion of certain Sul-
phates.""55"

* 'Phil. Trans.,' A, vol. 192, p. 455.

The Thermal Expansion of Pure Nickel and Cobalt. 307

o s

CO

n

33

s

s

OS

GO
OS

GO

x>.

rH

o

vo

rH

co

OS

ID

00
CO

©

co

co

CO

vO

vo

rH

rH

rH

©

©

©

©

©

o

(M

©

©

Cor

o
1

o

1

o
1

o
1

o
1

o

1

©

1

©
1

©
1

OS

o

CO

©

rH

OS

v§

CM

CO

CO

lO

VO

rH

CO

co

?

VO

OS
rH

o

rH

rH

VO
VO

rH

CO |

s

CO

to

CO

i>

*>

lO

lO

vO |

d

©

©

©

o

©

©

©

©

© 1

O

6

]

o

1

o

1

©

1

©

1

©

1

©

1

? ;

x>

CM

cm

CM

CO

CO

CM

VO !
CO

CO

1>

t>

VO

CM

00

CM

©

©

CO

©

CD

©

VO

©

CO

©

1>

, ©

vO CO ©
iO

*> l> J>

lO 00
CD © ©
t«. .!> J>

3

CO VO CO ©
^ «5
1> J> X> J>

CO VO 00

© © ©
i> 1> 1>

©

CM

00

t>

©

1>

©

0 ©

©

GO

CO

©

©

rH

©

GO

t— 1

rH

CM

rH

rH

CO

CM

1^

CM

00

GO

©

10

TP

kO

LO

VO

CO

©

CD

©

©

©

©

©

©

©

©

l>

CM

00

° CM

CM

GO

GO

©

CO

CM

rH

!

rH

rH

v.

308 Mr. A. E. Tutton,

Calculated Expansions.

Diminution of thickness

Expansion

of tripod

Expansion of nickel

of air-

layer.

screAvs.

block.

For t, - tx.

For ts - tY.

u.z - uv

L*3 — L#t.

'0*0020933

0 -0045639

0-0046086

0-0095348

0 -0067019

0 -0140987

21064

45869

46240 j

95572

67304

141441

21163

45213

46433 '

94602

67596

139815

23591

49248

51185 !

101953

74776

151201

23623

49707

52006 i

103355

75629

153062

23328

48986

50927 !

101993

74255

150979

18210

37863

38958 !

76927

57168

114790

17750

37338

38229

76231 :

55979

113569

17521

36945

37869

75638

55390

112583

Calculated Linear Coefficients of Expansion.

e. $ ■ j

0,

a.

b.

f 0-000 121 54

0-000 000 072 7

9-7985 j

0 -000 012 40

0*000 000 007 4

1 121 74

72 6

9-7986 i

12 42

74

[ 121 91

69 7

9-7985

12 44

71

f 129 45

72 8

10-2679 !

12 61

71

1 128 37

79 9

10 -2679

12 50

78

L 129 26

72 6

10-2678

12 59

71

f 097 74

59 9

7 -8327

12 48

76

1 097 56

60 6

7 -8328

12 46

• 77

L 097 43

612

7 -8328

12 44

78

Mean values . .

0 -000 012 48

0 -000 000 007 4

The mean coefficient of linear expansion, a + bt, of pure nickel,
between 0° and f , is thus found to be

0-000 012 48 + 0-000 000 007 4/, or 10-s(1248 + 0-74/).

The true coefficient, a, of linear expansion at f , or the mean coeffi-
cient between any two temperatures whose mean is /, is oc = a + 2M>
that is

0-000 012 48 + 0-000 000 014 8/, or 10"s(1248 + 1-48/).

The order of agreement of the nine individual determinations
must be regarded as highly satisfactory, and those for each series of
three referring to the same direction particularly so. The slight
differences in the value of a for the three directions, possibly due to
slight internal strain, fully justify the author in having carried out

The Thermal Expansion of Pure Nickel and Cobalt.

43

P3
o

'zc
Pi
eg
ft

1

!

t>

30

rH

-f

00

as

?

o
CM

rH

CM

i>

CO

CD

lO

SO

rH

rH

o

—
o

©

CM
O

CM

o

cm
o

CM

o

CM
O

CM

o

1

o
i

o

1

©

1

©

1

o

1

o
1

o

1

o

1

o

1

LO

as

rH

X

05
00

co

o
o

CM

CM
CM

co

CM

O-

CD

CO

CO

r— 1

rH

S

CO

GO

o

CM

OS

CO

as

co

lO

as
as

b-

CO
CO

as

CO

GO

x>»

CO

SO

ID

ID

CM
O

CM
O

CM

o

rH

o

rH

o

rH

o

©

rH

O

rH

o

Cor

o

1

©

1

o
T

o

1

©

1

o

1

o

1

O
1

o

1

to

30

CM

t— 1

30

&

r>

o

o
o

»o
r>

CO

©

I>

30

i>

CO

b-

00

ID

00

O0

CO

a>

cm

©

O

eii

goo

a x>

OS
lO

x>

o

CO

*>

CO
CO

b-

lO
CD
1>

00

co
i>

o

CO

I>

to
J>

$

1 ^9

co

CO

rH

00

IS

ID

£ oo

as

as
us
x>

-3

t>

CO

co
i>

iffl

CO
b-

30
CO
l>

o

CO

b-

t>

1 -

o

30

1^

CM

as

9

1 a oo

d *o

as
so
x>

8
t>

CO
CD

1>

Hi
CD
I>

CO
CD
t>

rH

CO

b-

'f

»D
b-

00

TP

b-

CM

9P

CM

CO

X

LO

©

CM

CM

o O
CM

33
rH

©
CM

co

00

as

rH

00
rH

X

rH

X
rH

SO

co

co

©

CO

»D

ID

CD ■

cq

co

CO

ID

co

lO
CO

lO
CD

co

CD

SO
CD

SO
CO

co

00

as

CM

©

00

X

CM
rH

a

rH

O

rH

00

to

30

CO

©

_

y —

©
rH

1

V

H-

g

30
CD

©

X

©

O

©

©

;

g

rH

CO

rH

CO

rH
rH

as
i>

CD
X

310

Mr. A. E. Tutton.

determinations for all the three directions ; the mean, however, can be
regarded with the fullest confidence as expressing the true coefficient
at 0°. The agreement of the values for the constant b is really
remarkable, considering the extreme smallness of the constant, and is
to be attributed to the perfection of the polished surfaces of the nickel
block ; the mean undoubtedly expresses the true semi-increment per
degree of temperature.

Calculated Expansions.

Diminution of thickness

Expansion of tripod

Expansion of cobalt

of air-layer.

screws.

block.

/-2V2.

For t2 - tx.

For *3 - tx.

- w

f 0-0025690

0 -0056138

0 -0060802

0 -0125913

I 0-0086492

0 -0182051

{ 26904

55352

65423

126876

92327

182228

L 25559

55022

63120

126710

88679

181732

22836

48789

56324

112972

79160

161761

\ 24804

50560

58150

114269

82954

164829

[ 26216

52431

60980

• 118065

87196

170496

f 18833

37403

43540

84246

62373

121649

< 18472

36748

44799

86203

63271

122951

L 18669

■

36781

44653

•

85905

63322

122686

Calculated Linear Coefficients of Expansion.

f.

<p. L0.

a.

b.

f 0-000 156 93

0-000 000 092 3

12 -9740

0 -000 012 10

0-000 000 0071

< 155 10

98 9

12 -9743

11 95

76

L 153 79

1061

12 -9742

1185

82

137 52

88 6

11 -5876

11 87

76

< 141 21

691

11 -5878

12 19

60

[ 141 89

65 4

11 -5882

12 24

56

105 65

45 3

8 -5981

12 29

53

1 103 95

48 3

8-5983

12 09

56

[ 104 77

42 9

8-5983

1218

50

Mean values

0 -000 012 08

0 000000006 4

The mean coefficient of linear expansion, a + bL of pure cobalt,
between 0° and f, is thus found to be

0-000 012 08 + 0-000 000 006 4/, or 10-8(1208 + 0-64/).

The true coefficient a of linear expansion at t\ or the mean coefficient
between any two temperatures whose mean is t, is a = a + 2bt, that is

0-000 012 08 + 0-000 000 012 8/, or 10-8(1208 + 1'28/).

The Thermal Expansion of Pure Nickel and Cobalt.

311

The agreement of the individual values is not quite so good as in the
ease of nickel, owing to the impossibility of obtaining such absolute
perfection of the surfaces of the cobalt block as was obtained in the
case of the nickel block. In the case of the constant a the differences
only amount to 3 per cent., and the whole amount of /; is so minute that
one is fortunate in finding the agreement so good. These differences,
however, from the nature of their cause, are bound to be on both sides
of the truth, and the mean of so large a number as nine is sure to be
very near the true value.

It will now be interesting to compare these results with those of
Fizeau. The latter were published in very brief form in the £ Comptes
Eendus,' for 1869* and also in ' Poggendorff 's Annalen,' for the same
year.f In neither of these publications are any further details given
beyond the values of the coefficient of expansion for 40° and the incre-
ment per degree, (= 2b), which occur in a table of similar-
quantities for various metals ; together with the information that the
specimens of nickel and cobalt employed had been reduced by hydrogen
and compressed, and that the range of temperature of the observations
was from 10° to 80°. The values in question are —

a 40°. Aa/Atf.

Nickel 0-000 012 79 071

Cobalt 0-000 012 36 0'80

It will be observed that the values of the coefficient for 40° now
presented are higher than those of Fizeau ; in the case of nickel the
difference is 1307 - 1279 = 0028, and in the case of cobalt 1259 - 1236
= 0023. The author's increments are likewise higher, 148 and 128
against 71 and 80 respectively. The fact that the author's increment
for nickel is twice as great as Fizeau's might suggest the possibility of
a mistake between b and 2b. The author has certainly not made any
such mistake, for the mode of calculation employed yields b directly,
and the values afforded were 74 and 64 respectively. The increment
Aa/A0 (Fizeau's 6 being the author's t) is equally certainly 2b. More-
over, the author's value of the increment for aluminium, 2*12, calcu-
lated in precisely the same manner, agrees fairly with the value given
by Fizeau, 2*29, in the same table in which the values for nickel and
cobalt are published. It may be that Fizeau inadvertently gave the
value of b instead of 2b in the particular cases of nickel and cobalt, but
it is much more likely that the numbers are correctly given, and that his
results were not very concordant with those now given. For Fizeau could
certainly not have possessed specimens of nickel and cobalt of the same
degree of purity as those supplied to the author by Professor Tilde n. The
recent discovery of nickel carbonyl has afforded an incomparable means

* Vol. 68, p. 1125.
f Vol. 138, p. 30.

312

Messrs. E. T. Giinther and J. J. Manley.

of separating the two metals, and also of isolating nickel from other
metallic impurities. Further, the discrepancy between the increment
values of the author and of Fizeau for these metals is only of the same
order as that between the concordant values of the author and of
Benoit, 0*46, on the one hand, and of Fizeau, 0*76, on the other, for
the 10 per cent, alloy of platinum-iridium, the value for which Fizeau
gives in the same table referred to.

Taking, therefore, the values published by Fizeau for the increments
of nickel and cobalt as correctly representing the results of his experi-
ments, his values of the coefficients at 0°, the constants a, calculated by
use of the increment, are as under : —

Nickel a = O'OOO 012 51

Cobalt a = O'OOO 012 04

j» Percentage difference 3*8.

At 100° the coefficients would become —

Nickel a = 0*000 013 22 1 ^ ., r„.

Cobalt a = 0-000 012 84 j PercentaSe dlfeence 2'9'

The values thus calculated for the expansion at 0° from Fizeau's
data are almost identical with the author's values. But the consider-
able difference between the values and the order of the increments now
given and those of Fizeau introduces a different order of progression
with rise of temperature. According to Fizeau the difference between
the coefficients of the two metals is a diminishing one, the percentage
difference having fallen from 3*8 at 0° to 2*9 at 100°; whereas the
author's determinations indicate that the difference is an accelerating
one, rising from 3*2 per cent, at 0° to 4*3 at 100°.

" On the Waters of the Salt Lake of Urmi." By E. T. Guhther,
M.A., and J. J. Manley, Daubeny Curator, Magdalen College.
Communicated by Sir John Mueeay, F.E.S. Eeceived June
8— Eead June 15, 1899,

In June, 1897, a portion of the Government Grant was allotted to
one of the authors by the Committee of the Eoyal Society, for the
investigation of the fauna and flora of the great salt lake of Urmi, in
Persia, as well as of the relations of that fauna and flora to its environ-
ment. The present research was undertaken with the view of placing
on record some of the conditions prevailing in the lake at the present
day.

The extraordinary changes which the level of the waters of the lake
has undergone, and is still undergoing, enhance the importance of
periodical examinations of the nature of the waters. The advisability
of the preservation of such records was urged upon the Eoyal Society

On the Waters of the Salt Lake of Urmi.

313

by its Secretary, 184 years ago. Edmund Halley, in his "Short
Account of the Cause of the Saltness of the Ocean, and of the several
Lakes that emit no Rivers," expressed himself as follows : — " I recom-
mend it therefore to the Society, as opportunity shall offer, to procure
the Experiments to be made of the present degree of Saltness of the
Ocean, and of as many of these Lakes as can be come at, that they may
stand upon Eecord for the benefit of future Ages."* At the present
day there are additional reasons for recording the properties and com-
position of such salt lakes as are known to be inhabited by life, because
Schmankewitschf and many others of the modern school of Eniwickel-
ungsmechanih (Morgan, Loeb, Vernon, &c.) have proved that every
change in the salinity of the waters is accompanied by definite, rapid,
and corresponding changes in the anatomical structure of certain of
their halophilous fauna, and especially of the species of Artemia, one
of which occurs in Lake Urmi.|

The superficial area covered by Lake Urmi is about 1750 square
miles, or about four times that of the Dead Sea. For so large an expanse
of water its depth is inconsiderable ; the greatest soundings do not
exceed 40 feet, and since much of the lake is extremely shallow, its
average depth is probably under 20 feet. When viewed from the
commanding heights of one of its islands, its waters show that brilliant
deep blue colour which is so characteristic of salt lakes, but as seen
from a boat, the light which is reflected from the light grey mud at the
bottom is green.

The temperature of so small a volume of water, which is at the same
time so extended as Lake Urmi, must necessarily vary considerably with
the seasonal changes of temperature. During the months of July and
August the temperature of the surface waters varied from 27*8° C. to
25*8° C, and the temperature of the bottom water at a depth of about
25 feet was 25° C. The specific gravity of the water, as measured on
the spot by an ordinary hydrometer, was I'll, whether the water was
drawn from near the bottom or from the surface of the lake ; it may
therefore be assumed that the waters of the lake remote from the
mouths of the fresh water tributaries were of a fairly uniform density,
a result which was probably due to the thorough mixing of the waters
produced by the strong south-easterly breezes prevalent at the time.

The total quantity of water available for the more detailed examina-
tion was brought home in two glass wine bottles, holding about f litre
a-piece. The samples A and B were collected on September 16, 1898,
near the base of the Bezau Daghi, on the western shore of the lake,
where there is comparatively deep and clear water close in shore. The
bottles were carefully corked, and it is, I think, fair to assume that no

* ' Phil. Trans.,' vol. 29, p. 299, 1715.
f - Zeitschrift Wiss. Zoologie,' vol. 25, 1875.
X Griinther, ' Nature,' vol. 58, p. 435.
VOL. LXV. 2 A

314

Messrs. E. T. Glintlier and J. J. Manley.

great changes have occurred in the interval between the bottling of the
samples in Persia and their examination in Magdalen College Labo-
ratory, in Oxford, seeing that the discrepancies between the analyses
are very small.

The examination was both physical and chemical.

Physical Examination.

Determination of Specific Gravity.

The specific gravities of the two samples of water (A and B) were
determined by Mr. H. N. Dickson, according to the method of Sprengel,
with the following mean results : —

A. B.

Specific gravity, at 15° C 1-11338 1-11389

0-3° C 1-11891 1-11945

Difference .. 0-00553 0-00556

Determination of Refractive Index.

The refractive indices (/x) were determined by means of a hollow
quartz prism of 60° 6' refracting angle, and a large spectrometer*
reading to 2" of arc. The water was at a temperature of 12 -2° C.
during the readings.

A. B.

Angle of minimum deviation of D line 25° 50' 4" 25° 50' 35",

whence p = 1-36110 1-36122

It is believed that similar optical measurements will be found to be
applicable to ordinary sea waters and will be found to give a more
accurate and a more readily obtained indication of the physical nature
of the water than the ordinary specific gravity methods.

Determination of Boiling Point.

It has long been customary to record the boiling points of strongly
saline natural waters, but in only too many cases, owing to the lack of
description of the conditions of the experiment, the records have only
a small value. Many trials have convinced us that the boiling points
of brines properly determined under similar conditions yield as reliable,
although less minute, information concerning the degree of salinity
as specific gravity determinations.

The salt water was boiled in a platinum bottle, to which an inverted
condenser containing ice-cold water was attached, in order to prevent

* The spectrometer, which was constructed for Dr. Bedson, and the prism em-
ployed are the property of the Royal Society.

On the Waters of the Salt Lake of JJrmi.

315

loss of water vapour and consequent concentration. The temperature
was measured by a form of platinum resistance thermometer, which
reads to 0*01° C. Three readings were always taken. Firstly, the
temperature of the steam from ordinary boiling water in a steam jacket
and under atmospheric pressure; secondly, the temperature of the
boiling salt water ; and finally, the temperature of steam once more.
If the first and last readings were identical, it was considered that the
conditions of the experiment had remained constant. As a matter of
practice it was found that when once the boiling point of the salt
water had been reached, the water continued to boil at that tempera-
ture for any length of time, so long as the pressure remained constant.

A. B.

Boiling point under normal pressure 103-84° C. 103*88° C.

! It will thus be seen that the three results of the physical examina-
tion of the two samples A and B are all mutually confirmatory, in so
far that they indicate that sample B had become a little more concen-
trated than A during its journey from Persia to Oxford.

Chemical Examination.

The method adopted was that of Dittmar, as described in the
Beport on the Composition of Ocean Water.*

For the estimation of the lime and magnesia, 20 c.c. of the water,
weighing approximately 22*2 grams, were measured off, and the quan-
tities used in the determination of the potash and total salts were half
that amount.

Examination of the Correctness of Dittmar 's Factor 0*91 for " Crude "

Lime.

Forty c.c. of the water were measured off and weighed. In accord-
ance with Dittmar's recommendation, the calcium was precipitated as
oxalate, filtered, washed, and finally weighed as oxide. The " crude "
oxide obtained amounted to 0*0319 gram: this was then redissolved
and again precipitated and weighed as " pure " oxide ; the weight was
found to be 0*0284 gram. If we multiply the weight of crude lime,
0*0319 gram, by Dittmar's factor 0*91, we obtain 0*0290 gram as the
weight of " pure " lime. This amount only differs from that actually
found jby + 0*0006 gram, thus affording confirmatory evidence of the
correctness of the factor.

The quantities of pure lime given below were determined by repre-
cipitation and repurification. The magnesia was precipitated by
* ' Challenger Keports,' "Physics and Chemistry," vol. 1.

316 Messrs. E. T. Giinther and J. J. Manley.

sodium phosphate instead of by ammonium phosphate as Dittmar
recommends, because when the latter reagent was used the magnesia was
found to come down very slowly and to adhere inconveniently to the
sides of the vessel.

The soda was determined by the method in which all the bases are
converted into normal sulphates, and the weight of the mixed sul-
phates is diminished by the subtraction of the weights of the potas-
sium, calcium, and magnesium sulphates. The potash was determined
by precipitation from the mixed sulphates by chloride of platinum,
according to Dittmar's third and final method,'55' and his observations
upon the appreciable solubility of the finely divided platinum by the
cold dilute hydrochloric acid employed for washing were confirmed.

From the known amounts of lime, magnesia, and potash (see below)
were deduced the following weights of normal sulphates in 100 grams
of the mixed sulphates : —

A. B.
Grams. Grams.

Potassium sulphate 0'258 0-259

Calcium sulphate 0-146 0-171

Magnesium sulphate , 1 -870 1-871

Sodium sulphate (by difference) ... 15*547 15*606

Total sulphates 17*821 17*907

The quantities of the principal saline components dissolved in 100
grams of the water of Lake Urmi are —

A. B.

Lime(CaO) : 0*0603 0-0706

Magnesia (MgO) 0-6265 0*6266

Potash (K20) 0-1394 0*1402

Soda(Na20) 6*788 6*814

Chlorine (CI) 8*496 8*536

Sulphates (SOs) 0*6205 0*6312

16*7307 16*8186

Oxygen equivalents of the chlor-
ine to be deducted 1*9167 1*9258

Total salts in 100 grams of water 14*814 14*893

* e Challenger Keports,' loc. cit., p. 16.

On the W aters of the Salt Lake of Urmi.

317

Or, recalculated for 100 parts by weight of total salts, we have —

Chlorine (CI)

Sulphates (S03)

A.

B.

57-315

4-189

4-238

0-407

0-474

4-229

4-207

0-941

0-941

45-822

45-753

- 12-939

-12-931

100-000

99-997

The hypothetical proximate composition of 100 parts of the total
salts was calculated, with the following results : —

A.

r

i.

N

ii.

Sodium chloride

86-332

86-203

86-203

Magnesium chloride

6-661

6-816

6-816

„ sulphate . . .

4-211

4-150

3-915

Calcium sulphate

0-988

1-151

1-151

Potassium sulphate

1-741

1-741

1-741

99-933

100-061

99-826

Result B i was obtained by calculating the magnesium sulphate
from the residual sulphate (S03). Result B ii from the residual mag-
nesium.

It is a remarkable fact that notwithstanding the occurrence of lime-
stone rocks and pebble beaches in the lake, no combined carbonic acid
could be detected in the water ; indeed, there would be no base for it
to combine with. On the other hand, small quantities of free carbon
dioxide were present dissolved in the water, and were estimated by
Tornoe's method.*

A. B.
Free carbon dioxide in solution... 0'028 per cent. 0-017 per cent.

Result A was the mean of two determinations which agreed to
within 0-002 per cent., and result B was obtained twice by the use of
different standard solutions ; the results were identical.

* Before applying Tornoe's method for the estimation of carbon dioxide to the
samples of water A and B, two determinations of combined carbon dioxide in a
dilute and standard solution of sodium carbonate were carried out, in order to
ascertain the degree of accuracy one might hope for. In the first, the carbon dioxide
found only exceeded that known to be present by 0*0003 gram, and in the second
by 0-0001 gram.

318 Mr. C. Godfrey. On the Application of

No iodine or bromine could be detected in the small quantity of
water available for examination.

' Spectroscopic examination revealed the presence of a trace of
barium. The quantity present would have been quite unweighable,
and although estimated with the calcium could not have vitiated the
results.

The results given under the heading B are regarded as those which
most nearly represent the true condition of the lake, and consequently
no attempt has been made to strike an average between the two series
of results. The A results are given in extenso in order to demonstrate
the degree of reliability of the B results, a matter which will be of
importance in the future, when, after an interval of some years,
another investigation of the water of Lake Urmi is made.

" On the Application of Fourier's Double Integrals to Optical
Problems." By Charles Godfrey, B.A., Scholar of Trinity
College, Isaac Newton Student in the University of Cam-
bridge. Communicated by Professor J. J. Thomson, F.R.S.
Received June 12, — Bead June 15, 1899.

(Abstract.)

The propagation of plane plane-polarised light in the direction of z
is governed by the equation Y2^ = where V kis the velocity

of light.

The most simple solution of this equation is —

(!)•

This may be interpreted as a train of waves of amplitude R, period
2ir/u, and phase ^, travelling with velocity V. This train of waves is
without beginning or end. Most of the results of physical optics have
direct application to a disturbance of the above form.

No radiation is found in nature which has the properties of the
above function. The fact alone that all natural radiations have
beginning and end would suffice to render (1) an inadequate repre-
sentation.

It is required to solve the problem how to represent any natural
radiation faithfully without losing the conveniences connected with
the form (1). The problem recalls the familiar process of harmonic
analysis. This p-ocess is applicable only to periodic functions,
whereas such a motion of the aether as constitutes white light is
non-periodic. In this connection it has been pointed out by Gouy

Fourier's Double Integrals to Optical Problems.

319

that Fourier's theorem of double integrals enables us to express a
wide class of functions in terms of circular functions. In fact, sub-
. ject to certain limitations,

It is proved in the course of the present paper that this process is
always legitimate when f(t) is such a function of time as can occur in
a physical problem.

The above theorem reduces f(t) to the limit of a sum of simple
circular function of time, the element being du{C cos ut + S sin ut). If
we write this du . E cos {ut + ^), a simple vibration of amplitude J&du,
period 2tt/u and phase \j/ is suggested. To connect this analysis with
the physical analysis of light into a continuous spectrum is tempting.
The present essay is an attempt to prove that, in certain very general
cases, such a connection exists. The proof depends upon the two
principles (i) that we can have no cognizance of instants of time, but
can observe only the contents of small intervals of time ; (ii) that,
in spectrum analysis, we do not deal with definite wave-lengths, but
rather with small ranges of wave-length.

, The fruitfulness of this calculus is illustrated by several applications.
The radiation of an incandescent gas is discussed. The trains of waves
emitted by molecules are continually being terminated by collisions.
It is held that, in dealing with the limiting widths of spectrum lines,
this effect must be included in the same investigation with the Doppler
effect first pointed out by Lippich and Lord Eayleigh.

Other problems shown to be within the grasp of this method are :
the connection between Eontgen rays as explained by Professor
Thomson, and ordinary light ; and the effect of radiative damping of
the molecular vibrations in widening the lines of the spectrum. All
these investigations are based upon a theorem for dealing with a radia-
tion composed of a vast aggregate of similar pulses distributed at
random. The theorem is due to Lord Eayleigh.

It is usual to examine the theory of dispersion by considering the
action of a simple periodic force upon a simple vibrator. Since no
light is simply periodic, it is necessary to extend the examination.
This is done below. We have also inquired whether fluorescence can
be due to natural vibrations of the molecules aroused by the non-
periodic quality of light. It is shown that so long as the equation of
motion is linear, no such explanation is possible.

f{t) = (C cos ut + & sin uf)du,
Jo

where

320 On Diselectrification produced by Magnetism.

" On Diselectrification produced by Magnetism. Preliminary
Note." By C. E. S. Phillips. Communicated by Sir
William Crookes, F.B.S. Eeceived June 13, — Eead June 15,
1899.

The writer has found, that, under certain conditions, an electrified
body rapidly loses its charge when in the neighbourhood of a magnetic
field. Nor does it, so far, appear essential that there be any relative
motion between the lines of magnetic force and the charged body
itself.

Preliminary experiments have been made with apparatus consisting
of a glass tube six inches long and one inch in diameter, at the centre
of which there was cemented upon both the inner and outer surfaces,
a strip of tin-foil one inch wide. Suitable connections were then
arranged for the purpose of charging either of these metallic layers
by means of an electrical machine. The pole-pieces of a powerful
electro-magnet projected into each end of the glass tube, through an
air-tight flange, and in such a manner as to ensure the production of
a strong magnetic field at the central portion of the tube. A Sprengel
air-pump was used to rarefy the gas within the tube, and, in the first
instance, the inner coating of tin-foil was charged positively.

This charge gave rise to the well known free positive and a bound
negative charge upon the outer tin-foil coating, the presence of the
former being indicated by the divergence of the leaves of an electro-
scope connected to that coating. While the pressure of the gas
within the glass tube was varied over a» range of from atmospheric
pressure to that represented by 0*2 mm. of mercury, the charge upon
the inner coating being either positive or negative showed no appreci-
able indication of being affected by the turning on or off of the
magnet. But at pressures lower than 0*2 mm., and when the inner
coating was positive, the sudden collapsing of the electroscope leaves
pointed to the removal of the charge through the action of starting
or stopping the magnetic flux. Although the effect was more power-
ful at the moment of making or breaking the magnet circuit, it per-
sisted in a modified degree as long as the magnetic field existed. No
such effect was observed when the inner coating was negatively
charged, nor was there any action even in the first case when the
magnetic pole-pieces, projecting into the tube, were magnetised so as to
be either both north or both south.

The leaves of the electroscope were then connected to one of the
internal pole-pieces, and it was seen that if sufficient positive elec-
tricity were supplied to the inner coating while the magnet was excited,
it became rapidly withdrawn, and ultimately resided upon the pole-
pieces themselves.

On the Orbit of Part of the Leonid Stream.

321

" On the Orbit of the Part of the Leonid Stream which the Earth
encountered on the Morning of 1898, November 15." By
Akthur A. Eambaut, M.A., D.Sc., Eadcliffe Observer. Com-
municated by G. Johnstone Stoney, M.A., D.Sc, F.E.S.
Eeceived June 14, — Eead June 15, 1899.

For an accurate prediction of the return of the Great Leonid swarm
of meteors, it is of the highest importance to determine as accurately
as possible the orbit in which each part of the swarm is moving.

As has been pointed out by Drs. Stoney and Downing, * the denser
part of the stream, with which we are chiefly concerned, being now
drawn out to such a length that it takes more than two years to pass
any point of the orbit, it results that the perturbing effect of the
several planets will not be the same on different parts of the stream.

Hence it follows that the orbit deduced from the observations made
during the great shower of 1866, however reliable they may have been,
must not be assumed to represent accurately the track of the meteors
which we shall meet when we pass through the node of the orbit next
November, even when allowance is made for the perturbations which
that part of the stream has suffered in the meanwhile.

But although the determinations of the radiant point of the shower
made in 1866 are entitled to a high degree of confidence, owing to the
large number of meteors upon which they depend, yet the discrepancies
existing between the results of different observers, and the fact that
the varying effect of the earth's attraction upon the position of the
radiant at different zenith-distances was generally overlooked by the
observers at the time, introduce an element of uncertainty into the orbit.

The extent of these discrepancies may be estimated from the " List
of Observed Places of the Eadiant Point in 1866," as given by Professor
A. S. Herschel in the 'Monthly Notices of the Eoyal Astronomical
Society,' vol. 27, p. 19, and the effect of this uncertainty on the
resulting orbit may be illustrated by comparing the two sets of
elements deduced by Adams and Schiaparelli, respectively, from English
observations of the radiant point on that occasion. These are —

Ada

ms.f

Schiaparelli. X

Period (assumed)

P =

33*25 years.

33*25 years

Mean distance

a =»

10-340

10-340

Eccentricity

e =

0-9047

0-9046

i -

16°

46'

17° 44'-5

v —

51°

28'

51° 28'

Longitude of perihelion...

7T =

58°

19'

56° 26'

* ' Eoy. Soc. Proc.,' No. 410.
' f 'Monthly Notices,' vol. 27; 'The Scientific Papers of John Couch Adams,
vol. 1, p. 269.

X ' Entwurf einer Astronomischen Theorie der Sternschnuppen,' von J. Y. Schia-
parelli, p. 57.

322 Dr. A. A. Bambaut. On the Orbit of the Part of the

The part of the stream through which the earth will pass this year
lies between the position of those meteors which the earth encountered
last year and that of the meteors of 1866. It is, therefore, of import-
ance to investigate how far the orbit of the meteors observed in
November, 1898, agrees with that deduced from the observations of
1866.

Unfortunately the observations obtained in 1898 are fewer than
could be desired. Unfavourable weather prevailed almost universally
in England. On the Continent meteors were observed at a few stations,
but they appear to have been outlying members of the stream, and the
denser portion was not encountered until about 19 hrs. G.M.T., on the
morning of November 15 (civil time). For observations at the time
that the earth was passing through the densest part of the stream, we
are wholly dependent upon the American observers. The time of maxi-
mum display is, owing to the paucity of meteors, somewhat indefinite,
and seems to have differed by several hours at different stations. The
most reliable accounts, however, agree in placing it between 19 hrs,
30 min. and 22 hrs. G.M.T. on the morning of the 15th, and from a
discussion of all the separate determinations, I have been led to adopt
20 hrs. 45 min. as the most probable time of maximum. This agrees
exactly with Professor Young's result, who writes, " The maximum
was about 3 hrs. 45 min. (Eastern Standard Time) when for about
20 minutes the meteors averaged two or three a minute."*

In Table I are given — the name of the observer, his observatory or
station, its longitude and latitude, the Greenwich time, and the K.A,
and declination of the radiant point.

In this connection I may remark that in order to contribute to an
accurate determination of the orbit, the G.M.T. corresponding to the
observation, and the approximate position of the observer are just as
essential as the co-ordinates of the radiant itself. In several of the
accounts before us I have had to assume that the position of the
radiant corresponds to the time given as that of maximum display, or
to the middle of the time over which the watch extended.

The longitudes and latitudes of the observers' positions may also
in some cases be in error to the extent of several minutes, but they are
sufficiently exact to enable me to compute the " zenith-attraction."

In Table I, I have included only those observations which seem to
relate to meteors belonging to the dense, or central, part of the stream.
They are all contained in the interval between 18 hrs. 58 min. and
23 hrs. 2 min. G.M.T.

Each of these separate results has to be corrected for the influence
of the earth's attraction on the paths of the meteors during their
approach, and for the influence of the earth's motion on the apparent
position of the radiant.

* ' The Observatory,' December, 1898, p. 459.

Leonid Stream which the Earth encountered in 1898. 323

a>
ft

o

"a
>

<a

o

o

-2

o

lO
rH

i>
rH

o

CQ

o

o

CO

©

o

CO
rH

O
CO

m

O

CO

O
CO

GO
rH

■43

©
P

o
CO

CO
CO

rH

CO

CO

CO

co

CO

o

CO

(SI
CO

o

CO

o

CO

1— 1

CO
CO

CO
CO

g
5

+

+

+

+

+

+

+

+

+

+

+

+

+

3

_

lO

H*

o

O

O

CO

O

O

o

o

CO

co

O

so

CO

CO
CO

CO

m

o

CO

X
'H
rH

in

rH

o

lO
rH

OS
H*
rH

rH

XQ

in

rH

o

in
rH

Ci

HH
rH

00

HH

rH

05
rH

$

rH

o

lO
rH

rH

lO

rH

EH

N

a oo

in

CO

co

m

m

in

CO

o

CO

i> CO

in

CO

rH

o

6

_rj CO
^ rH

o

rH

CO
CO

rH

CO

o

CO

Ci

o

o

CO

OS O
rH CO

co

CO

rH

CO

CO
CO

CO
CM

CO
CO

CO

CO

rH

CO

t>

in

t>
m

JO

TP

CO
CO

m

CO
CO

Lai

o

CO

CO

o

X

CO

OS

co

CO

hh
CO

+

+

+

+

+

+

+

+

+

4-

+

bp
U

in

CO

in

a
m

05

rH

a

H^

CO

CO
rH

O

r£l

m

m

CO

m

CO

r>

CO

be

•a

r-H ?H CD

5J S

CJO

P*

O

rH

CO

CO

m

CO

CO

OS

o

rH

CO

CO

rH

rH

rH

324 Dr. A. A. Rambaut. On the Orbit of the Part of the

The former varies with, the zenith-distance of the radiant and its
elongation from the " apex," or the point of the heavens towards which
the earth is moving at the time. It has always the effect of displacing
the radiant towards the observer's zenith, and hence has been called
by Schiaparelli the " zenith-attraction." The amount of this displace-
ment (rj) is given by the expression,

w-u

tan -hrj = tan -kz.

21 w + u 2

in which z is the apparent zenith-distance of the radiant, u is the velo-
city of the meteors relatively to the earth before the influence of the
earth's attraction has become sensible, while w is the accelerated
velocity with which the meteor encounters the earth.

This displacement, which has been too frequently overlooked by
meteor observers, may, as pointed out by Schiaparelli, amount in
extreme cases to as much as 25° 38'. In the case of the Leonids, how-
ever, it happens that the elongation of the radiant from the " apex " is
so small (in the case before us not exceeding 11° at anytime) that the
effect of the zenith-attraction never amounts to half a degree, its
greatest value being 29'.

For computing the value of w we have the expression

w2 = u2 + 2gr,

r being the earth's radius, and g the acceleration of gravity at the
surface, or, expressing the velocities, as is convenient, in terms of the
mean velocity of the earth in its orbit,

= «2 + 0-141587.

In computing the value of 2gr in the above expression the Sun's
parallax has been taken to be 8" "80, and the ratio of the earth's mass to
that of the Sun equal to 1/331,100.

We thus have for computing w

w = u + [8-84999] x i - [7-39895] x 1 ,

the figures in brackets being the logarithms of the coefficients.

For determining u we may with quite sufficient precision adopt
Adams's orbit of 1866. Also if U denotes the velocity of the earth at
any time expressed in terms of its mean velocity, and R its distance
from the Sun, then,

Leonid Stream which the Earth encountered in 1898. 325

or, neglecting the second power of the eccentricity we may, without
loss of accuracy, write U = 1 /E.

In Table II are given the values of the Sun's longitude (©), the
longitude (/), the right ascension (a) and the declination (d) of the
apex, and the orbital velocity of the meteors (v), all of which can be

Table II.

Time.

©

■

Log E.

•

I.

a.

d.

v.

1898, Nov., 14 -0
14*5
15 -0

232° 9' -8

232 40-0

233 10-3

9 -99514
•99510
•99505

142° 52' -8
143 22-7
143 52-7

145° 13' -7

145 42 -7

146 11-9

+ 13° 53' -8
44-0
34 -1

1 -3877
•3878
•3879

computed from the ' Nautical Almanac ' or from Adams's orbit of the
meteor stream. These are given for three epochs, viz. : — November
14'0d., 14'5d., and-15-0d., from which their values at the time of each
observation may be obtained by interpolation. All of these quantities
are needed in the subsequent reduction of the observations, or for
deducing the elements of the orbit.

We next compute the quantities u, w, and ?/, exhibited in Table III.

Table III.

No.

u.

w.

'/■

1

2-3779

2 -4075

22'

2

•3718

•4014

19

3

•3782

•4078

10

4

•3680

•3977

15

5

•3725

•4021

15

6

•3748

•4044

29

7

•3845

•4140

22

8

•3767

•4063

15

9

•3857

•4152

29

10

•3821

•4116

20

11

•3799

'4094

11

12

•3736

•4032

27

13

•3724

•4020

15

Applying the corrections for the earth's attraction, da. and d8, to the
E.A. and declination from the formulae

da = 7] sin p sec S • d8 = —tj cosp,

p being the parallactic angle, we find the corrected E.A. and declination
of the radiant, as given in the second and third columns of Table IV.
In the next two columns of the same table are found the longitude (L')
and latitude (B') of the same points, and in the sixth and seventh

326 On the Orlit of Part of the Leonid Stream.

Table IV.

1> 0.

/

a .

d'.

i T,'

!

B'.

T
JU.

B.

W Xi.

1

1 AQ° A'

+ 22°

2

+ 8

54

14U> ol

+ 15

22

U O

o
£l

lOJ. UD

22

6

l^bU o

9

53

1 1'7 <^d.

17

3

1 U

q

10U O /

21

22

1 A,^ 1 1

i^o 11

8'

46

1 A(K 1 9
l*tO 1Z

15

6

U O

**

1^

23

51

1 ASK A

30

38

1 zL9 31
1*2 ol

18

22

U o

K
O

lOJ. JLO

22

21

1 AK 99

9

53

1 A(K Qfi

17

4

A 'A

6

151 26

21

43

145 47

9

21

147 23

16

9

0-3

7

150 19

19

47

145 29

7

11

146 50

12

24

0-2

8

149 48

22

2

144 14

9

8

144 37

15

47

0*6

9

148 23

20

11

143 38

6

56

143 36

11

59

o-i

10

149 52

20

41

144 46

7

53

145 32

13

37

0*3

11

149 43

21

21

144 24

8

28

144 50

14

38

0-3

12

151 8

22

7

145 23

9

38

146 37

16

38

0-2

13

151 42

22

7

145 52

9

50

147 26

16

53

0-5

columns are given the longitude (L) and latitude (B), of the true
radiant corrected for the effect of the earth's orbital motion.

The quantities L and B define the direction of the tangent to the
orbit of the meteors at the point where the earth intersects it, and
from the mean of these separate determinations, the position of the
earth in its orbit at the time, and an assumption with regard to the
period of the meteor stream, the orbit is to be determined.*

The only difficulty lies in deciding on the best mode of combining
the various observations, or in laying down a rule for determining the
weights. In this part of the work a certain amount of arbitrariness
is, I think, unavoidable. From a careful consideration of all the cir-
cumstances of each case, as far as they are recorded, the experience or
inexperience of the observer in this class of work as far as it is stated,
the number of meteors observed, and the size of the area from which
the meteors appeared to radiate, I have been led to adopt the weights
given in the last column of Table IV, which represents, I think very
fairly, the relative value of the individual observations. It will be
noticed that I have given the two photographic results (Nos. 2 and 13)
an importance out of all proportion to the number of trails photo-
graphed, viz., four trails in the case of No. 2, and two trails in the
case of No. 13. This is, I think, justified by the superior accuracy of
photographic results in this class of observations.

I thus find as the definitive position of the radiant of the Leonid
meteors of 1898,

145° 49'±20'-5; +16° 2'±19'*9,

corresponding to thu epoch November 14*864 (astronomical time).

* See ' Handworterbucli der Astronomie,' heransgegeben von Dr. W. Yalentiner,
vol. 2, Breslau, 1898.

A Comparison of Platinum and Gas Thermometers. 327

From the researches of the late Professor H. A. Newton, and the
results of the investigations of Drs. Stoney and Downing on the per-
turbations of Adams's orbit, the most probable value of the period of
revolution would appear to be at present about 33*49 years, corre-
sponding to mean distance of 10*39.

Any admissible variation in the length of the period, however, makes
but a small change in the other elements of the orbit, as is evident
from Table V, in which the elements in each column have been com-
puted with the value of the mean distance which is contained in it.

Table V.

I.

II.

III.

Longitude of descending node v =

33*25 yrs.

10*34
64° 46'
16 3
53 2
58 40

33 *49 yrs.

10 39
64° 50'
16 3
53 2
58 40

33 "73 yrs.

10*44
64° 54'
16 3
53 2
58 40

If we adopt the value 10*39 for the mean distance, as being on the
whole the most probable, we have the orbit in column II representing
the result of the observations of 1898, as far as they have been
published.

" A Comparison of Platinum and Gas Thermometers, including a
Determination of the Boiling Point of Sulphur on the
Nitrogen Scale : an Account of Experiments made in the
Laboratory of the Bureau International des Poids et Mesures,
at Sevres." By Drs. J. A. Harker and P. Chappuis. Com-
municated by the Kew Observatory Committee. Eeceived
June 8 — Bead June 15, 1899.

(Abstract.)

In 1886, Professor Callendar drew attention to the method of
measuring temperature, based on the determination of the electrical
resistance of a platinum wire. He showed that the method was
capable of a very general application, and that the platinum resistance
thermometer was an instrument giving consistent and accurate results
over a very wide temperature range.

T Callendar pointed out that if E0 denote the resistance of the spiral
of a particular platinum thermometer at 0°, and Ei its resistance at
100°, we may establish for the particular wire a temperature scale,

323 A Comparison of Platinum and Gas Thermometers.

which we may call the scale of platinum temperatures, such that if R be
the resistance at any temperature T°, this temperature on the platinum
R

scale will be — — - x 100 degrees. For this quantity Callendar

Ki - Ro

employs the symbol pt, its value depending on the sample of platinum
chosen.

In order to reduce to the standard scale of temperature the indica-
tions of any platinum thermometer, it is necessary to know the law
connecting T andpt. These are, of course, identical at 0° and 100°,
but the determination of the curve expressing the relationship between
them is a matter for experiment.

The work of Callendar had established for a particular sample of
platinum the relation

1 |_\ioo/ looj

over the range 0° to 600°, T being measured on the constant pressure
air scale.

Later experiments by Callendar and Griffiths showed that this
relation holds for platinum wires generally, provided they are not
very impure. They propose that the value of 8, the constant em-
ployed in the formula, should be determined by taking the resistance
of the thermometer in the vapour of sulphur. A new determination
of this point on the air scale made by them gave 444*53°, as the
boiling point under 760 mm. pressure.

The present paper is the outcome of the co-operation of the Kew
Observatory Committee and the authorities of the International
Bureau of Weights and Measures at Sevres, for the purpose of
carrying out a comparison of some platinum thermometers with the
recognised international standards.

A new resistance-box, designed for this work, and special platinum
thermometers together with the other accessories needed were con-
structed for the Kew Committee, and after their working had been
tested at Kew, were set up at the laboratory at Sevres in August,
1897. The comparisons executed between these instruments and the
standards of the Bureau may be divided into several groups. The
first group of experiments covers the range - 23° to 80°, and consists
of direct comparisons between each platinum thermometer and the
primary mercury standards of the Bureau. Above 80° the mercury
thermometers were replaced by a gas-thermometer, constructed for
measurements up to high temperatures. The comparisons between 80°
and 200° were made in a vertical bath of stirred oil, heated by different
liquids boiling under varying pressures. For work above 200° a bath
of mixed nitrates of potash and soda was substituted for the oil tank.
In this bath comparisons of the two principal platinum thermometers
with the gas-thermometer were made up to 460° ; and with a third

Remits of Experiments on Permanent Grass-land 329

thermometer, which was provided with a porcelain tube, we were able
to go up to 590°. Comparisons of the platinum and gas-scales were
carried out at over 150 different points, each comparison consisting
of either ten or twenty readings of the different instruments.

By the intermediary of the platinum thermometers a determination
of the boiling point of sulphur on the nitrogen scale was also made.
The mean of three very concordant sets of determinations with the
different thermometers gave 445°'27 as the boiling point on the scale
of the constant volume nitrogen thermometer, a value differing only
0°'7 from that found by Callendar and Griffiths for the same tem-
perature expressed on the constant pressure air scale.

If for the reduction of the platinum temperatures in our comparisons
we adopt the parabolic formula, and the value of 8 obtained by assum-
ing our new number for the sulphur-point, we find that below 100° the
differences between the observed values on the nitrogen scale and those
deduced from the platinum thermometer are exceedingly small, and
that even at the highest temperatures the differences only amount
to a few tenths of a degree.

Full details as to the instruments employed and the methods adopted
are given in the paper.

Agricultural, Botanical, aud Chemical -Results of Experiments
on the Mixed Herbage of Permanent Grass-land, conducted for
many Years in succession on the same Land. Part III. — The
Chemical Results." By Sir John Bexnet Lawes, Bart.,
D.C.L., ScJD.. F.R.S., and Sir J. Henry Gilbert, LL.D.,Sc.D.,
F.R.S. Received August 10, 1899.

(Abstract.)

The experiments were commenced in 1856, and are still in progress,
so that the present is the forty-fourth year of their continuance. There
are about twenty plots, two of which have been continuously un-
manured, and the remainder have respectively received different
descriptions or quantities of manure of known composition. A report
on the " Agricultural Results " was published in the ' Phil. Trans.,'
Part I, 1880 ; and a second on the "Botanical Results " in the ' Phil.
Trans.,' Part IV, 1882. The present paper deals with a portion of the
" Chemical Results."

In all cases, of both first and second crops, the dry matter and the
ash. and in most the nitrogen, have been determined. In selected cases
determinations have been made of the amount of nitrogen existing as
albuminoids, and in some of the amount of " crude woody fibre," and
of crude fatty matter. More than 200 complete ash-analyses have
also been executed.

VOL. LXV. 2 V,

330

Sir J. B. Lawes and Sir J. H. Gilbert.

It was found that the chemical composition of the mixed herbage
was very directly dependent, not only on the seasons and on the sup-
plies within the soil, but very prominently also on the description of
plants encouraged, and on the character of their development ; so
that it was essential to a proper interpretation of the variations in the
chemical composition, to bear in mind the differences in the botanical
composition. Hence a summary table was given showing the character-
istic differences in the botanical composition under the different con-
ditions as to manuring, the influence of which on the chemical com-
position it was sought to illustrate.

As the investigation involved the consideration of the chemical
composition of the mixed produce of about twenty plots over forty or
more seasons, including the discussion of the results of more than 200
complete analyses of the ashes of the separated or the mixed herbage,
attention was called to the state of existing knowledge as to the role or
function in vegetation of the individual constituents found in the
ashes of plants ; and this was seen to be very imperfect. Further,
in calculating the percentage composition of the " pure ash," the plan
usually adopted was to exclude not only the sand and charcoal, but
also the carbonic acid. The authors considered, however, that the
presence and the amount of carbonic acid associated with the fixed
constituents in plant-ashes was a point of considerable significance ;
and they entered into some detail as to the methods of determining
the carbonic acid in ashes, and as to the results obtained.

In order to throw some light on the connection between the growth
of the crops and their mineral composition, results relating to the
separated gramineous, the separated leguminous, and the separated
" miscellaneous " herbage of the mixed produce, grown without manure
and by different manures, were first discussed. To obtain more
definite evidence illustrating the connection between character and
stage of growth and the composition of the products — especially the
ash-composition — results relating to the bean plant, taken at succes-
sive periods of growth, and also to the first, second, and third crops of
clover, were next considered. Lastly, in further illustration, results
as to the nitrogen and the ash-composition of crops of three different
natural orders— wheat representing the Graminese, Swedish turnips the
Cruciferse, and beans and clover the Leguminosese — were given.

The general result was, that there were very characteristic differ-
ences in the composition of the ashes of different crops according to
the amounts of nitrogen they assimilated. Red clover, for example,
yields large amounts of nitrogen over a given area, part of which is
due to fixation, but much is certainly taken up as nitrates from the
soil ; and the results show, that the greater the amount of nitrogen
assimilated the more is the ash characterised by containing fixed base
in combination with carbonic acid ; presumably representing organic

Me&idts of Experiments on Permanent Grass-lancl. 331

acid in the vegetable substance before incineration. The conclusion
was that, independently of any specially physiological function of the
bases, such as that of potash in connection with the formation of
carbohydrates, for example, their office was prominently also that of
carriers of nitric acid, and that when the nitrogen had been assimilated,
the base was left as a residue in combination with organic acid — which
was represented by carbonic acid in the ash. Further existing knowledge
— as to the condition in which combined nitrogen is found in soil waters,
as to the action of nitrates used as manures, as to the presence of
nitrates in still-growing plants, and as to the connection between the
nitrogen assimilated and the composition of the ash as had been illus-
trated— pointed to the conclusion that, at any rate a large amount of
the nitrogen of the chlorophyllous vegetation on the earth's surface was
derived from nitrates ; whilst, so far as this was the case, the raison
d'Mre of much of the fixed base found in the ashes of plants would
seem to be clearly indicated.

The various results and conclusions above referred to were found to
afford material aid in the interpretation of the differences in the
chemical composition of the mixed herbage of the different plots which
was next considered, so far as the first crops over the first twenty
years were concerned.

For the purposes of the illustrations the differently manured plots
were arranged in four groups as follows : — 1. Plots without manure or
with farmyard manure. 2. Plots with nitrogenous manures alone.
3. Plots with mineral manures alone. 4. Plots with nitrogenous and
mineral manures together. Average results for each plot, generally
for a period of eighteen years, 1856 — 1873, and including the per-
centages of nitrogen, crude ash, and pure ash, in the dry substance of
the produce ; also the percentage composition of the pure ash were
brought together in a table, and are discussed in detail. The close
dependence of the chemical composition of the mixed herbage on its
botanical composition, and on the character of development of the
plants, was throughout illustrated. It was further shown, that the
mineral composition of the mixed herbage was very directly dependent
on the supplies available to the plant within the soil. Indeed, when it
was considered that the mixed herbage of permanent grass land
includes plants of very various root-range and root-habit, and that
some of them vegetate more or less almost the year round, it was not
surprising to find that the composition of the produce was, upon the
whole, a somewhat close reflection of the available supplies within the
range of the roots. It was, in fact, much more so than in the case of
individual crops grown separately. Within certain limits, this was the
case even with the constituents of, so to speak, less functional im-
portance than those which more obviously determined the description
of plants encouraged and the character of their development. It was at

332 Results of Experiments on Permanent Grass-la nd.

the same time obvious, that when the more functionally important
constituents are available in relative abundance, those which are of less
importance in this respect were taken up and retained in less amount
than they otherwise would be ; the result being determined in great
measure by the character of growth induced.

For example, if potash be liberally available the produce is much
more stemmy, and the amount of soda, of lime, and to some extent of
magnesia also, will be less relatively to the potash. In defect of suffi-
cient potash, on the other hand, more of soda, or of lime, or of both,
will be taken up and retained ; but the herbage will at the same time
be more leafy and immature. That is to say, the constituents are not
mutually replaceable in the processes of growth, but accordingly as the
one or the other predominates, so will the product of growth be
different.

There can be no doubt, that luxuriance or vegetative activity is
intimately associated with the amount of nitrogen available and taken
up. Further, it may be stated that chlorophyll formation to a great
extent follows nitrogen assimilation. But the results relating to the
increased amount of non- nitrogenous substance yielded in the mixed
herbage under the influence of the various manures clearly indicated that
the nitrogen being taken up, and the chlorophyll formed, the carbon
assimilation, and the carbohydrate formation, depended essentially on
the amounts of potash available. It may be stated as a matter of
fact that, in practical agriculture, artificial nitrogenous manures are
chiefly used for crops containing a comparatively low percentage of
nitrogen in their dry substance, and yielding comparatively low
amounts of nitrogen per acre. Indeed, they are mainly used for the
increased production of the non-nitrogenous bodies — the carbohydrates
— starch and cellulose in the cereals, starch in potatoes, and sugar in
the sugar-cane and in root crops, for example. And now, in the case
of the mixed herbage of grass land, it was seen that, provided the
mineral constituents, and especially potash, were abundantly available,
a characteristic effect of nitrogenous manures was to increase the
production of the non-nitrogenous bodies.

Orientation of the Pyramids and Temples in the S4cldn. 333

u On the Orientation of the Pyramids and Temples in the Sudan."
By E. A. Wallis Budge, M.A., LittJD., D.Lit., E.S.A. Com-
municated by Professor Sir NqkmaxT,ockyer, K.C.B., F.R.S.
Received April 14, 1899.

In the year 1897 I was sent on a mission to the Sudan by the
Trustees of the British Museum, and in 1898 I was again sent to that
country to complete the work in the places which I could not reach the
year before on account of the unsettled state of that unhappy land.
By the favour of Viscount Cromer and Lord Kitchener, the Sirdar of
the Egyptian army, I was enabled to visit sites which had not been
visited by Europeans for a great many years, and, by the unusual
facilities which these gentlemen afforded me, to make notes on matters
of scientific interest which have, in recent years, been widely dis-
cussed. Besides the examination of the ruins of temples and the copy-
ing of the inscriptions which the hand of time had spared, my wish
was to collect, so far as possible, accurate information concerning the
orientation of the pyramids in the Sudan, and to obtain measurements
of them with special reference to the work which Professor Sir Norman
Lockyer and • Mr. Penrose have done on the temples of Egypt and
Greece respectively.

It will be remembered that a few years ago Sir Norman Lockyer
promulgated the theory that Egyptian temples and pyramids were
oriented to certain stars, which were sacred to certain Egyptian
divinities, and to the sun at certain points of his course. Having
worked through all the available material which had been collected by
himself and others, he came to the conclusion that his theory was
correct, and that with accurate data in his hands concerning a given
temple or pyramid, the astronomer would be able to supply the archaeo-
logist with a tolerably correct idea of the date when the site was first
covered by a religious or fimeral edifice. In the ' Dawn of Astronomy '
a number of test cases were discussed with results which convinced me
of the truth of the theory ; and Mr. Penrose, working on the same
lines, applied it to the temples of Greece with such remarkable results
that my conviction was strengthened. It must, however, be admitted
that several difficulties still remain to be cleared away, but I think
that these will disappear when the temples and pyramids of Egypt
have been measured and surveyed according to modern requirements.
For no one can fail to notice that the plans published, even those in
the great work of Lepsius, present inaccuracies of a serious kind,
especially when we consider that a variation of a few degrees will
wreck the most careful calculation. The object of the present paper
is to inquire if, and how far, the pyramids of the Sudan are oriented

vol. LXV. 2 c

334

Dr. E. A. Wallis Budge. On the Orientation

according to any definite plan, and to put on record for the use of
those interested in the subje:t such notes and figures as I was able to
make.

The pyramid fields of the Sudan may be enumerated as follows : —
(1) Kurru, (2) Zuma, (3) Tankassi, (4) Gebel Barkal, (5) Nuri or
Nawari, (6) Merawi, i.e., the Meroe of the Greeks. The pyramids
which stood upon these sites of which any remains at all exist are in
number about two hundred, and it is quite certain that those which
have been destroyed may be reckoned at another two hundred at least.
But a pyramid field to be useful for working out the theory of orienta-
tion according to a certain plan must possess certain characteristics,
such as the following : — (1) The pyramids upon it must be in a good
state of preservation at their bases, and all should not, if possible, be
oriented in the same direction. (2) One or more temples should be
either on or near the pyramid field, so that the direction of the orienta-
tion of both kinds of buildings may be readily compared. Now, every
pyramid which I have seen in the Sudan, with the exception of those
of Nuri, consists not of a solid mass of cut stones carefully built up
with a funeral chamber inside it, but of a core formed of a mixture of
stones, sand, and lime which has been surrounded with a casing of
stones, each measuring about 18 inches by 12 inches by 10 inches. It
seems to me that the core was first built, and the casing of cut stones
put round it afterwards. Curiously enough, every pyramid, with the
exception of those at Nun, is truncated, and it is this peculiarity
which has worked its ruin. For the rain has run through the flat
layer of stones at the top in large quantities, and in passing between
the stones at the sides, which are built without mortar, has taken with
it the lime and sand from the inside a hollow has thus been formed
round the core, and the stones, aided by the furious winds which rage
in the Sudan at certain times of the year, have by their own weight
fallen in upon it. Sometimes the casing has been built at too steep an
angle, and the upper parts of the sides have fallen in or fallen out, as
the case may be.

Yet another reason for the ruin of the Sudan pyramids must be
mentioned. The stones of which the sides are built, unlike the stones
which form the pyramids of Egypt, are relatively small, and the
natives have found them to be admirably adapted for certain purposes.
As a result they have been filched from their places, and used to make
the foundations of water-wheel supports and of houses, and also to line
the shallow trenches in which the Muhammadans have buried their
dead for countless generations. Thus the pyramids have, one by one,
been stripped of their stone coverings, and the wind and rain together
have beaten the cores so much out of shape that it is sometimes diffi-
cult, if not impossible, to distinguish them from small natural hills.
The pyramids which have been built in the mountains, or at any great

of the Pyramids and Temples in the Silddn. 335

distance from cultivated land, are the best preserved, and this is only
what might be expected. When a native wanted stones for any pur-
pose, he went for them to the pyramid which was nearest to him, and
the result is that the pyramids which stood near the villages or culti-
vated land have in some districts quite disappeared. Thus at Tan-
kassi, about seven miles from Senem abu-Ddm, where the Egyptian
troops were encamped about eighteen months ago, it is most difficult
to identify the cores of the pyramids wliich once stood there. At
Gebel Barkal the pyramids which were nearest to the cultivated land
have disappeared, and the same may be said ©f dozens of the small
pyramids which stood at Nuri. At Meroe the pyramids, which were
built near the temple that stood only about a mile from the river, and
were in consequence close to the main road which has been the high-
way to Khartum and the south for countless generations, have also all
but disappeared.

In this way the six pyramid fields of the Sudan become reduced to
three, for those of Kurru, Ztima, and Tankassi may well be left out of
consideration. It -is, however, tolerably clear from the general dispo-
sition of the pyramid remains at these places, that the system of
orientation employed by the builders of the pyramids there resembled
that found to have been in use at Gebel Barkal, Nuri, and Meroe.
With the view of showing the present condition of the pyramids of the
three principal fields in the Sudan, I took about fifty photographs,
one of which is reproduced in this paper. Some such record was abso-
lutely necessary, for if the lithographic landscape views printed by
Lepsius, in his work the ' Denkmaeler,' be compared with these photo-
graphs, the serious deterioration in the condition of the remains since
his time will at once be clear. #

In the summer of 1897 I arrived at the village of Senem abu-D6m,
which is situated on the left bank of the Nile, about sixteen hundred
miles from the Mediterranean ; on the opposite bank lie the villages of
Shibba, Merawi, and Barkal, and on the same side as these, viz., the
east bank, a few miles to the south, rises the magnificent rock of sand-
stone called Gebel Barkal. Before I began serious work at Gebel
Barkal, I visited the pyramids . of Nuri with a view of finding out
which was the more promising site. I could not visit the pyramids of
Meroe that summer, because all the country round about was in the
hands of the Dervishes; I therefore had to content myself with the
pyramid fields of Nuri, Barkal, Tankassi, &c. ; and with the hope that
I might visit Meroe later, I decided that, for several reasons, the
pyramid field of Gebel Barkal suited my purpose best, and so began
work there.

Gebel Barkal is a huge rock about three hundred feet high ; it is
three-quarters of a mile long, and is about half a mile wide in its
widest part. The widest end has served as a quarry, and all the

2 C 2

336

Dr. E. A. Wallis Budge. On the Orientation

stones used in the casings of the pyramids for 10 miles north and
south have come from it. Close under the almost perpendicular end
of the mountain are the remains of a temple built by Eameses II, King
of Egypt, about B.C. 1330; and those of a temple built by Piankhi,
King of Egypt and Ethiopia, about B.C. 730 ; and those of another
built by Tirhakah, King of Ethiopia, about B.C. 680. To the south of
the mountain lies the pyramid field, and the remains of the ancient
city of Napata must be sought for some five or six miles further south.
On the western bank of the Nile there must have stood a great city,
with many temples, palaces, and other great buildings, for on several
occasions when the Egyptian troops have had to build block houses
and other military works, portions of large columns, pottery, &c, have
been found in digging out the foundations. The site of this city is
probably marked accurately by the modern village of Senem abu-Dom ■
and the tombs which were made for the nobles thereof are to be found
away back in the desert, at a distance of about two hours from the
river, in a range of low sandstone hills.

A Pyramid at Gebel Barkal.

Of the pyramids at Gebel Barkal some are in ruins and some are
tolerably complete; the former are useless for purposes of measure-
ment, because the broken sides and the debris round the bases make it
impossible to get accurate compass bearings. I therefore made no

of the Pyramids, and Temples in the Sudan.

337

attempt to deal with the remains of the pyramids which are scattered
about on the rocky plateau on the south side of the mountain, and I
limited my inquiry to the seven which stood on the top of it. In
Plan I these are set out to a scale of 30 inches to the mile, and thanks
to the kindness of Colonel the Hon. M. G. Talbot, K.E., their position
is very accurately indicated.* The bearings were taken with a pris-
matic compass, the variation of which was determined by comparison
with an astronomical azimuth; but owing to the irregularities of the
masonry, they cannot be relied upon to nearer than 2°. The distances
were paced. The variation between the true north and the magnetic
north was estimated at 5JC, and this estimate has been confirmed from

Plan I.

^Vi>a°from T.N.
(Lepsius 13)

Pi&nkhi's TsmpLe (Lepsius L) izffrom T.N.
Tirhak<zh's~fernpLe(LepsiusA)i43afromT.N.

Scale 2772 , or 30 inches to l Kibe.
Yds.io so 10 so 30 40 so 60 to eo 90 looYds.

The Pyramids of Gebel Barkal.

an examination of an Admiralty map,f which I have been so fortunate
as to have had placed at my disposal. On looking at the plan, we
notice that one pyramid is oriented at 129° from the true north, three

* My friend Colonel Talbot, who was employed by the Egyptian Government to
make the triangulation and general survey of the Sudan for military purposes,
obtained astronomical azimuths from time to time, to determine the variation of
his compass bearings, and he made use of such data in preparing the two plans
which accompany this paper. He employed an azimuth compass. I did not use
plumb-lines for finding the alignments, but a steel tape stretched horizontally
along the general surface of the pyramid was taken as the direction of ea^h side.

t The Admiralty map here referred to was specially prepared in the Hydro-
graphic Department for the use of Professor Sir Norman Lockyer, and it i- now in
his possession. It is not a rough reconnaissance made with a prismatic compass,
but contains the lines of declination as determined by compass observations
made in the Mediterranean, and the Eed Sea, and the adjacent waters of the Indian
Ocean.

338 Dr. E. A. Wallis Budge. On the Orientation

at 140°, one at 122°, one at 147°, and one at 139°. Of all the temples
which have been built at the foot of Gebel Barkal only two have any-
substantial remains, viz., those of Piankhi and Tirhakah ; the orienta-
tion of the former is 127° from the true north, and that of the latter
143°. Now it seems to me that we may fairly assume that both these
kings, with their ancestors and successors, were buried near their
temples; indeed from a common-sense point of view there was no
place more suitable for their tombs than the neighbouring hill or
mountain slope. I searched diligently, hoping that I might find some
trace upon some of the blocks of stone which had formed the shrines
or funeral chapels that had stood in front of their pyramids, but with-
out success. Every pyramid at Gebel Barkal must have had such a
shrine or chapel, but the size of the chapel depended upon the import-
ance of the man whose tomb the pyramid was intended to cover.
Pyramids Nos. 6 and 7 on the plan must, judging by the ruins, have
had very large shrines enclosed by walls, and we may assume, from the
absence of similar buildings in the fronts of the other pyramids, that
they were royal tombs. The masonry, however, and the general appear-
ance of them somehow suggest that they were not the oldest of the
group, though, arguing from archaeological considerations, they should
be as old as B.C. 700.

Passing to the most northerly end of the pyramid field of Gebel
Barkal, we find lying there the remains of a " step " pyramid, and as it
has an entirely different orientation from that of the other pyramids
there, and the masonry is of a better class of work, and the whole
building is on a scale mi generis, it is clear that it belongs to a
different period. It is a striking fact that archaeological considera-
tions indicate that the pyramids which have different orientations
belong to different periods, and at the end of this paper it will be seen
that the results deduced from astronomical considerations point the
same way.

The official instructions which I received before I went to the Sudan
in 1897 allowed me to make a trial excavation of one pyramid. I
therefore selected No. 5 of my plan, as on its shrine there were sculp-
tured scenes in which funeral offerings were being made to the royal
personage who had the pyramid built, by priests, by Anubis the god
of the dead, and by a number of gods whom it is not easy to identify.
This royal personage was assumed to have identified himself with
Osiris, and the goddesses who usually attended the god were here seen
attending the king or prince. That he was royal there is no doubt,
for one relief showed him in the act of grasping the hairy scalps of
representatives of a number of captive or subdued nations, and brand-
ishing a Sudani club over them, much in the same way as the kings of
the XVIIIth and XlXth dynasties of Egypt are depicted on the walls
of their tombs and palaces. To indicate the greatness and power of

of the Pyramids and Temples in tlia Sudan.

339

the king his figure had been made very large, a custom common among
•savage or semi-savage tribes ; the figures of the vanquished were small,
and were huddled together.

To make certain that the mummy chamber was not in the pyramid
itself, the stones from one corner, about half way up, were removed,
and a boring was made to the length of several feet ; but it was soon
evident that the core of the pyramid was not made of masonry, but of
stone, sand, lime, &c, roughly mixed together. This being so, the
hole was filled up, and the stones replaced in the casing work.

As soon as this trial work was done, by the help of the British
officers, about forty men were collected, and, provided with a few tools
and a good number of baskets, we began to search for the pit which
led to the mummy chamber, which seemed to be below the ground. A
trench was dug round all four sides, and at length a large flat slab of
hard stone, 10 feet by 6 feet by 10 inches, set in lime, was found on
the S.E. side of the pyramid ; this was broken through, and the layers
of lime and sand .which came beneath showed that we had reached the
mouth of the pit or shaft. We toiled through 60 feet of rough con-
crete in about four weeks, and at length reached a rectangular chamber
about 9 feet cube, hewn, like the pit, out of the solid rock ; the roof of
this chamber was supported upon square pillars. A narrow passage on
the S.E. side of the chamber led into another chamber which had
square pillars likewise ; both chambers were half filled with sand. On
the sand at the foot of the shaft or pit we found some bleached
bones and a broken wine jar, upon which were inscribed the words
01N02 P0AI02, i.e., "Khodian wine." The bones were like the
bones of a small sheep, but they fell into dust when touched, and I
could not therefore bring them home. The broken wine jar is a most
interesting object, for it enables us to arrive at the date when it was
put into the tomb. We know from several sources that Ehodian
wine was used extensively during the latter half of the second
century B.C., and the shape of the letters on the jar-neck points to
that date. The jar, then, must have been taken up to Gebel Barkal
from Alexandria, probably by boat, between B.C. 150 and B.C. 50, and
broken in the chamber, which no doubt served as a funeral chapel, at
the last feast of offerings held there. I do not think that this date is
the date of the building of the pyramid, for it is a well-known fact that
commemorative offerings were made in these chapels as long as funds
for the purpose were forthcoming. Tombs of royal personages and of
people of high rank were kept open by the priests for the express
purpose of inducing relatives and friends to contribute offerings,
chiefly in kind, at stated seasons of the year, and if a proof be wanted of
this statement it is sufficient to refer to Diodorus Siculus,* who says in
his history that he visited the halls and the chapels of the royal tombs
* He yisited Egypt B.C. 57.

340 Dr. E. A. Wall is Budge. On the Orientation

in the Valley of the Kings at Thebes. Now we know that several of
these tombs were built 1,300 years before the time of this historian,
and yet they were open to the inspection of visitors, even Greeks, at
that late date. What the broken wine jar shows us is that the pyra-
mids of Gebel Barkal were not built by late native rulers who nourished
under the rule of the Greeks and Bomans between B.C. 300 and
A.D. 350, but by native kings of the earlier dynasties who followed
ancient funeral customs and rites that were known and practised as far
back as the Xllth dynasty of Egypt, about B.C. 2500.

But to return to the excavation. Having removed the sand from
the chambers, it became evident that neither the pillars nor the walls
thereof had ever been inscribed or sculptured. A diligent search
convinced me that the mummy chamber must be situated somewhere
under the pyramid, and, judging from the analogy of several tombs
which I excavated with General Sir Francis Grenfell, K.C.B., at
Aswan in 1886-87, it ought to be exactly under its apex, if such a
term may be applied to a truncated pyramid. Therefore, returning
to the end of the chamber which had been hewn out of the rock
immediately beneath the place where the shrine had stood, we
searched for an entrance which would lead us in under the pyramid.
At length we found that the wall of the chamber at this end con-
sisted of a slab of stone about 7 feet by 6 feet by 1 foot 2 inches,
embedded in lime, and when this was broken through we found a
sort of vault, but there was nothing in it. The only thing that
remained to do was to cut into the floor of the vault, and when we
had done this we found that we had arrived at the mouth of a
second pit or shaft, by which the mummy in its coffin must have
been lowered into the chamber wherein it was to rest finally.
We worked away at clearing out the pit by candle light under great
difficulties, and when we were about 100 feet below the base of the
pyramid, we began to come near to the short passage at right angles to
the pit, which should have led us into the mummy chamber. At this
point, however, the sides of the pit became very damp, and everything
which we dug out was thoroughly wet ; a foot or two lower down the
men found themselves in standing water, and at length we had to stop
work. We had reached a depth which had brought us down to the
level of the Nile, and found that its waters had forced their way by
infiltration into the shaft, and presumably into the miunmy chamber
also. This discovery was most disappointing, and one which I think
could hardly have been foreseen. We were, therefore, obliged to be
satisfied with having demonstrated the plan upon which the pyramid
tomb in the Sudan was built, and its analogy to the rock-hewn tombs
of Egypt of the firrt twelve dynasties.

Now it is clear that, when the builders of the pyramids at Gebel
Barkal selected the site for their tombs on the edge of the sandstone

of the Pyramids and- Temples in the Sitddn. 341

plateau to the south, it must have been well out of the reach of the
infiltration of the Nile waters, for they knew from long experience
that the presence of damp was fatal to the preservation of the mummied
body. And besides, it is impossible that they should have gone on
building pyramid after pyramid at various places on the plateau and
on the side which slopes down to the river without finding out that the
water of the Nile was making their labour ^vain. How then can the
presence of the water be accounted for 1 The true answer seems to be
that the bed of the Nile has gradually risen since the time when the
pyramids were built, and it may be that the river has also somewhat
changed its course. A small annual deposit in its bed would easily
cause both the rise and the change, of course, and as a result certain
sites which originally stood well above the level of the waters of the
highest inundation would be flooded from time to time. The cause-
way leading to the river and the foremost courtyard of the temple of
Piankhi were flooded by the Nile when I was at Gebel Barkal in 1897,
and that year the inundation was not by any means one of the highest.
Had there been any risk of this happening in Piankhi's time he would
never have built, or rebuilt, the temple, which must have been one of
the glories of the city of Napata, near the site whereof its remains
now lie. Thus we see that in the Sudan, as in Egypt, the same geolo-
gical changes have been at work.

Passing now from Gebel Barkal, we have to consider the pyra-
mids of Nuri, or Belal, which stand on the west bank of the Nile, just
opposite to the now famous village of Kassingar, and at about a distance
of six miles as the crow flies from Gebel Barkal. The value of the
pyramid field of Nuri for a discussion on the orientation of pyramids
is not so great as that of the field of Gebel Barkal ; but from other
points of view it is of very considerable interest. Its pyramids are
not so well preserved, and there is not so much variety of orientation
as is usually found in groups of pyramids elsewhere in the Sudan.
The largest of the Nuri pyramids must, with the exception of the
" step " pyramids, have been the largest in the Sudan, for when com-
plete they cannot have been less than 100 feet high. Several of them
are built of hewn stones throughout, and the excellence of the mate-
rial and the handy sizes of the stones have tempted the natives to use
them freely for building tombs for their sheikhs and houses and graves
for themselves. Speaking roughly, the group at Nuri consists of two
rows of pyramids, the older row being that in which stand the remains
of a large " step " pyramid (see Plan II). The next oldest pyramids stand
in a somewhat irregular row about 200 yards to the S.E., and between
the two rows a number of pyramids were built at a later date, great care
being taken by their builders to place them in such positions that the
light from the celestial body to which they were oriented should not be
obstructed by the older buildings. The first row of pyramids is in

342

Dr. E. A. Wallis Budge; On the Orientation

Plan II.

Yds.M50 10 20 SO 40 50 60 70 BO 90 lOOYds.
L-U — I — i t i — i — i i i i I

The Pyramids of Nuirf.

ruins ; with the exception of the " step " pyramid all are small. Of
the pyramids in the second row, six are sufficiently well preserved to
afford us a good idea of what they were like when complete. They
must have stood upon a slightly elevated site, and they could not have
been much less than 100 feet in the side; on the S.E. side each pyra-
mid had a shrine or chapel, into the innermost part of which the light
from the celestial body to which it was oriented could enter. At each
end of the pyramid field was a temple, the size of which it is impossible

of the Pyramids and Temples in the Sudan.

343

to guess at without clearing away the tens of thousands of tons of sand
with which the whole site is covered. That at the X.E. end of the
held is of considerable interest, for it shows traces of a rebuilding,
during which the orientation angle was altered. We have seen that
tihe pyramid field of Gebel Barkal represented the royal necropolis of
the city of Xapata on the north ; to what city, then, did the pyramids
of Xuri belong 1 In the absence of the definite knowledge which can
oryly be obtained by excavating, we shall probably be right in regard-
ing the pyramid field of Xuri as the royal necropolis of the city the
remains of which now lie beneath the sand at Senem-abu-D6m. This
city was certainly older than Xapata, and therefore the pyramids
which form the tombs of its royal rulers are older than those of Gebel
Barkal or of any of the other sites which lie further to the south. A
remarkable fact is that all the pyramids at Xuri, of which tolerably
accurate compass bearings can be taken, are oriented 129° from the
true north, and that is the angle of orientation of the " step " pyramid
at Gebel Barkal. - From archaeological considerations the pyramids,
including those built with steps, which have this angle of orientation
should be older than those which have a different angle. The temples in
Nubia and other countries, including Egypt, which have this angle of
orientation, are as old as the period which lies between the Xllth and
XVIIIth dynasties, and I therefore think that the pyramids of Xuri, at
least the oldest of them, are considerably older than all the pyramids
at Gebel Barkal, with the exception of the "step" pyramid. There is
another point to mention : the pyramids at Xuri are on the left bank
of the Xile, like all the ancient pyramids of Egypt, and I think this
is a strong proof of their great antiquity. With the making of measure-
ments of the pyramids, and a careful examination of the ruined
shrines in the hope of finding inscriptions, my work at Xuri in 1897
came to an end.

Towards the end of last year I left England for the Sudan, and
"before the end of December, thanks to the facilities afforded me by the
Sirdar, I arrived at the Atbara Biver. In due course I was sent on to
a place called Begrawiyyeh, from where I was able to examine the
remains of the temples and pyramids which mark the site of the
ancient city of Meroe, and of the home of the great queens who ruled
there under the name of " Candace." From the spot where I landed,
the ruins of the great temple of Meroe are about one and a half miles
distant ; by the side of them, running nearly due north and south, is
the old Khartum road, and in a line almost due east lie two groups of
pyramids, at a distance of about four and a half miles from the river
as the crow flies. The shekh of the district, Muhammad Amin, who
had been an officer of some rank under General Gordon, gave me
every assistance in his power, but the country had been so ravaged
and wasted by the Dervishes led by Mahmud under the Khalifa's

344 Dr. E. A. "Wallis Budge. On the Orientation

orders, that it was with the greatest difficulty that a couple of donkeys
were found for us to ride upon. The country is depopulated, and we
saw hundreds of well built houses falling into ruins.

The first site visited was that of the temple of Meroe, of which
portions of several pillars in situ still remain. Before any satisfactory
measurements of angles of orientation could be obtained as regards
both the temple and the pyramids, I saw that a good deal of excava-
tion and clearing away of debris would have to be done ; but as no
men could be found to do the work — there being none in the country,
thanks to Dervish rule — I had to abandon that idea, and get the best
results I could from the examination of the ruins only. The temple
appears to have been surrounded by a wall which was built at some
considerable distance from it, and a very large number of people could
have assembled in the space between the wall and the temple. The
temple was, like most of the Sudan temples, dedicated to the Sun-god
Amen-Ra, whose visible type upon earth was the ram, and clearly the
ram peculiar to that portion of the Sudan. Of the shrine nothing
remains, but a figure of a ram in hard, bluish-grey stone lay among the
ruins of the pillars. The pyramids, over a hundred in number, which
are situated at no great distance to the north-east of the temple, are in
ruins, and the masonry which still remains is of the same class and
style as that of the most recent portions of the temple. At two or
three places in the plain round about are remains of buildings of the
Roman period, and near one of these was found a small Greek in-
scription, which I brought home ; it is now in the British Museum.

Passing from the riverside ruins we made our way due east towards,
a chain of very low hills that lay in the distance, and after an hour's
ride we arrived at a crescent-shaped eminence which stood with its.
convex side towards us. The top of the eminence was about eighty
feet above the pebbly plain. When I had walked round the pyramids
it was easy to see that they must be divided for purposes of examina-
tion into three groups. The first group stood on the crescent-shaped
eminence mentioned above, and it seems as if this site originally con-
sisted of a series of low hills, from which the tops had been cut off and
thrown into the hollows between to make a level base for the pyramids.
The first group consisted originally of about twenty-five pyramid?.
The second group stood away to the south-east, and consisted origin-
ally of about twenty-two pyramids. The third group lay close to the
second, and the pyramids which belong to it are about twenty in
number. Of the pyramids of the first group there are abundant
remains, and it is easy to see that they reproduce all the characteristics
and all the angles of orientation with which we are familiar from the
pyramids of Gebel Barkal and Nuri, together with some others. Here,
as at Gebel Barkal and Nuri, and the pyramid field to the north-east
of the ruins of the temple of Meroe, we find a " step " pyramid larger.

of the Pyramids and Temples in the Sudan. 345

better built, and better planned than the other pyramids which stand
near, and in each place this " step " pyramid has an orientation angle
of 129°.

In Northern Egypt, as is well known, all the pyramids are oriented
east and west; and observations show that in Southern Egypt the
" step " pyramids are oriented not east and west, but in another
azimuth facing south-east, with such an amplitude that it could not
have been a question of the sunlight entering the shrine. We are
therefore driven to star worship. Now, the chief pyramids in
Northern Egypt date from B.C. 3800 to B.C. 2600; we may expect
therefore that, as the idea of pyramid building was introduced into the
south from the north, the building of the " step " pyramids must
have taken place at a very early date. Looking to this amplitude it
has already been shown that some of the chief temples of Southern
Egypt were oriented to cc Centauri. Taking the " step " pyramid at
Gebel Barkal with an azimuth of 129°, we have an amplitude of 39°
south of east, and, not taking into account refraction and the heights
of the hills on the horizon, which would tend to neutralise each other, we
find the declination of a star thus observed to be 35° 58' south, a posi-
tion which the star occupied B.C. 2700. These " step " pyramids then
were probably built under the influence of the kings of the Xlth and
Xllth dynasties, who were famous for their building operations. This
date will also suit admirably from both an astronomical and an
archaeological aspect the temple of Piankhi, which has a nearly identi-
cal amplitude. We have to assume therefore that Piankhi, like many
other kings, simply restored a Xllth dynasty temple. With regard to
azimuths from 143° to 150°, I find that the same star, a Centauri,
might still have been observed, but in this case the building of both
temple and pyramid must have taken place about the period which lies
between B.C. 1200 and B.C. 700. Here, in my opinion, the astro-
nomical determination agrees with the archaeological requirements.

At Nuri, as at the other places mentioned above, we find numbers
of pyramids with the orientation angles of 122°, 140°, and 147°; and
the characteristics of the pyramids having the same angle in one place
are the same as those having the same angle elsewhere. At Nuri and
at Begrawiyyeh or Meroe, a considerable number of pyramids have
exactly the same angle of orientation, viz., 129°, and the number is so
great that it is clear that we are not dealing with a question of acci-
dent but of design. The great Lepsius came to the conclusion that
the pyramids of Nuri looked older, and were older, than the pyramids
of Gebel Barkal, and my own observations made on the spot convince
me that his view was right ; and he might have added also that the
pyramids at Meroe having the same angle of orientation as those at
Nuri are older than those which have a different angle. It has been
said above that of the shrines which originally stood before the

346 Dr. E. A. Wallis Budge. On the Orientation

pyramids of Gebel Barkal and Nuri, few remains exist, but this is not
the case with the pyramids of Meroe, where we have the greater por-
tion of many of their shrines still standing in situ. These remains
show that the shrines consisted of two, and sometimes three, cbajnbers
with narrow doorways which served, like the various sights and sections
of a telescope, to direct the rays of light from the celestial body to a
given spot, that spot in the case of a pyramid being the centre of the
shrine where a figure of the deceased was placed. On the walls of the
shrines are cut in outline figures of kings, armed with bows and
arrows, and swinging clubs over the heads of a mass of people who
belong to captive races, and in some few cases we have been fortunate
enough to find preserved the names of the kings who built them. Now
these names help us to assign a date to the pyramids on which they
are found, and it is thus possible to compare the results derived from
astronomical calculations based upon angles of orientation, and tkose
which are derived from arch geological experience.

For purposes of convenience the sb-called kings of Ethiopia have
been divided into four groups : —

1. The dynasty of Piankhi which ruled in the eighth century B.C.

2. The dynasty of Tirhakah which ruled about a century later.

3. The earlier kings of Meroe who ruled from about B.C. 500 to the

end of the Ptolemaic period.

4. The later kings of Meroe who ruled from the beginning to the

middle of the Eoman domination over Egypt.

Of each of these groups of kings monuments, i.e., temples and
pyramids, have been found, and there can be no doubt whatever about
this, for a number of royal names belonging to each of these four
groups have been found inscribed upon them. We have already seen
that some of the temples and pyramids of Gebel Barkal belong to the
period which lies between B.C. 800 and B.C. 600, and it is now clear
that some of the pyramids of Meroe belong to the period which lies
between B.C. 500 and a.d. 200. Here, again, the orientation theory
shows that any shrine built about that time would have been directed
to the important south star Fomalhaut. Now we may see from Dr. O.
Danckwortt's important inquiries concerning the precessional change
of place of forty-six fundamental stars from B.C. 2000 to A.D. 800,*
that at zero time the declination of Fomalhaut was 39° south, and that
it was slowly changing. So that a difference of 1° to 40° south would
give us B.C. 300, and to 38° would give us A.D. 200. After what I
have said as to the difficulties, almost impossibilities, of determining
exact azimuths, in my mind there now remains no doubt as to the
time when these shrines were built. We have now to consider the

* See ' Vierteljahrsschrift der astronomischen Gesellschaffc,' Leipzig, 1881, p. 76.

of the Pyramids and Temples in the S4ddn. 347

date of the pyramids of Nun, on which no inscriptions have been
found, and the date of some of the pyramids of Meroe on which also
no inscriptions have been found ; but before we can arrive at any con-
clusion on these points, we must briefly consider the question of the
ancient civilisation of the Sudan and its origin.

In the first place we must put aside the name Ethiopian which is so
often applied to it, because there is no evidence whatever to show that
it is of Ethiopian origin ; the term " Ethiopian " has been loosely
applied to the ancient peoples of the Eastern Sudan and their works,
just as to any object the source of which was unknown the name
"Phoenician" or "Hittite" has been applied in our own day. The
ancient tribes who lived on the east bank of the Nile from Wadi Haifa
to Khartum were not negroes, and they had but little in common with
the tribes who lived south and west of Khartum ■ indeed they were
not a black race at all. The colour of the men's skins was red, not
black, and that of their women, who did not expose themselves to the
sun's rays, was of a yellowish-red. Between these peoples and the
ancient Egyptians a more or less friendly intercourse existed from the
earliest times ; otherwise how could the Egyptian officials who were
sent to the Sudan, to the district of " big trees," to bring back huge
tree trunks to cut up for coffins and sarcophagi for their royal masters,
have succeeded in their enterprises 1 Men sent upon missions of this
kind must have followed the course of the river, for the shorter desert
routes were quite impossible, at any rate on the return journey for
men laden with baulks of timber. Still more remarkable is the fact
that a high official, called Ba-ur-Tattu, in the reign of Assa, about B.C.
3300, travelled as far south as the land of the pygmies, and brought
back one of these folk for the king; and eighty years later Heru-
Khuf, a governor of Abu, or Elephantine, did the same thing. It is
difficult to imagine that Egypt could have exercised any great power
over the country south of Wadi Haifa, but that there should have
been a relatively brisk intercourse between Egypt and the Sudan for
trading purposes is only natural. Again, if the Egyptians made any
colonies in the south, the introduction of Egyptian civilisation and
religious ideas would inevitably follow, and intermarriages between
Egyptian strangers and natives would take place. Moreover, the
natives who visited Egypt would bring back with them new ideas,
which in course of time they finally adopted, adding such modifications
as their opinions dictated. The Egyptian viceroys and Sudan princes
would naturally build temples and tombs (i.e., pyramids) after the
manner of those they found in Egypt, but the orientation of these
would be altered in accordance with the religious views of those who
built them. But suitable sites for temples and pyramids are more rare
in the Sudan than in Egypt, and priests in both countries were unwil-
ling to abandon a spot which had become associated with sacred beliefs

348

Dr. E. A. Wallis Budge. On the Orientation

or religious worship. For this reason they were driven to various
shifts in order to make an old temple suitable for modern require-
ments : and when no amount of modification would suffice they even-
tually pulled it down, in whole or in part, and rebuilt it on the same
site as the old one.

Since constant intercourse existed between Egypt and the Sudan in
very early times, and since the people of the one country were influ-
enced greatly by those of the other, it is clear that there is no reason
why certain of the Sudan pyramids should not be as old almost as those
of Egypt. On this point both astronomy and archaeology agree, and
I find, on comparing the tables in the ' Dawn of Astronomy ' and the
deductions which the author has made from them with my own obser-
vations made from an archaeological standpoint, that our conclusions
are identical. I may therefore say finally that we seem to be in the
presence of three different sets of structures which were built at three
different times. The oldest dates from the Xllth dynasty, when
a Centauri was used as a warning star ; the second from B.C. 1200 to
B.C. 700, the same warning star being used ; and, finally, a third group
much later, when the star Fomalhaut could be observed on the horizon
with the identical amplitude first employed. Tables have been pre-
pared showing the various azimuth amplitudes and declinations, with
corrections for refraction and for hills on the horizon, which are esti-
mated at 1° or 2° in height ; but it is not necessary to give them in
this place because of the local difficulties in determining the azimuths,
to which I have already referred. <fl *

I have very great pleasure in expressing my thanks to Viscount
Cromer, Lord Kitchener (Sirdar of the Egyptian Army), Colonel Sir
Francis AYingate, Colonel Sir Budolf Slatin Pasha, Major-General Sir
Leslie Bundle, Colonel the Hon. M. G. Talbot, B.E., Colonel W. H.
Drage, D.S.O., and to many other British officers for assistance and
help in the course of my work. Notwithstanding the incessant and
laborious duties which devolved upon them as officers of a frontier
field force, they readily and freely found time to forward my investi-
gations, and but for their many acts of personal kindness I should
have found it impossible to have completed my mission.

The following table of amplitudes, &c, I owe to the kindness of
Professor Sir Norman Lockyer, K.C.B., F.B.S.

of the Pyramids and Temples in the Sudan.

1

-6

i

I

1

!

ii

fc55$89S3?

83938853.83

M CO N X H 00
N O iM (M N N

CD O O O O N

so ^ co ^ ^ xa

§

352S2S8SSS
Sg393«&99£

8889*83

No
correal ion.

£93988993

8S8

§gS3§§

-9

888S8599&

888

C5 tJ» 05 C5 O t>
CO lO M ^ ifl lO

Orientation
from north
through east.

&389S589SI-1

1— II— l!—lf— ll— ll— Ii— !l— II— (

SIS

i-l rH i-l i-l i-i i— (

350 Dr. H. M. Yernon. The Effect of Stateness of the

" The Effect of Staleness of the Sexual Cells on the Development
of Echinoids." By H. M. Yernon,. M.A., M.D., Fellow of
Magdalen College, Oxford. Communicated by W. F. R
Weldon, F.B.S. Eeceived June 27, 1899.

The effect of varying degrees of staleness of the ova and sperm of
an organism upon subsequent development appears to have been very
little studied, though such a condition must obviously be a factor of
frequent occurrence under natural conditions. Thus in most of the
Coelentera, Echinoderms, and in some of the worms, it would seem to
be a matter of chance whether the ova and spermatozoa come into
contact when freshly shed, or only many hours after extrusion. In
some mammals also, especially in man, the relative degree of freshness
of ovum and spermatozoon at the time of fertilisation is entirely a
matter of chance.

The chief connection in which the question of staleness has been
hitherto studied is that of polyspermy. Thus O. and R. Hertwig
found* that on crossing certain species of Echinoderms, as the ova of
Sphcerechinus granulans wibh the sperm of Strongylocentrotus lividus,
more and more of the ova were fertilised up to a certain point if they
were kept for an increasing number of hours in sea water, but that
after this point they began to undergo polyspermy in an increasing
degree, and to develop abnormally, To what precise extent is this
tendency present, however, and how is it affected by the staleness of
the ova on the one hand, and of the sperm on the other ? Also, do
the normally developing ova of stale sexual products continue to
develop equally well with those from fresh products,- or not 1 Such
are the questions it is attempted to answer in this paper.

The method of experiment was very simple. The ovaries and testes
of ripe specimens of the Echinoid Strongylocentrotus lividus were shaken
in jars of water, and portions of the contents of these were mixed, either
immediately, or after a given number of hours. The mixed solutions
were allowed to stand for an hour, and were then poured into beakers
and diluted with about ten times their volume of water. Twenty-four
hours later, some of the stirred up contents were introduced into a
small glass cell, and a drop of corrosive sublimate solution added to
kill the blastulaB and make them sink to the bottom. The numbers of
normally developing blastulse, and of abnormally developing and
unsegmented ova were then counted, 300 to 500 being usually
enumerated, in order to get an accurate estimate. In all cases the
mixed ova from two or more ripe specimens were used, and were

* 4 Experimentelle Untersuclmngen uber die Bedingungen der Bastardbefruch-
tuug.' Jena, 1885.

Sexual Cells on the Development of Echinoicls. 351

fertilised by the mixed sperm of two or more specimens, in order to
get as average results as possible. In the experiments to be sub-
sequently described, however, in which the stale ova and spermatozoa
were mixed several different times at a few hours' interval with fresh
sperm and ova, as often as not only one fresh specimen was used in
•each case.

In the subjoined table are given the results obtained in one of the
most complete experiments. In this case parallel series of determina-
tions were made, in which the ova and sperm were kept in, and after
fertilisation diluted with, respectively tank water from the Aquarium,
and pure water collected several kilometres from the shore of the Bay
of Naples.

Time of
fertilisation.

Tank water.

Pure sea water.

. Per cent,
blastulse.

Per cent,
diminution
per hour.

Per cent,
blastulse.

Per cent,
diminution
per hour.

98-5

96-9

After 6 hours ....

95-3

0*5

95-6

0-2

„ 21 „ ....

83 -2

0-8

97-2

nil

n 24 „ ....

77 '9

1-8

92-7

1-5

1 „ 27 » ....

73 2

1-6

66 -5

8-7

„ 30| „ ....

55-7

5-0

0-25

18 9

„ 33 „ ....

36 -0

7*9

o-o

o-i

„ 35^ „

2-2

13 5

„ 46 „ ....

0-0

0-2

It will be seen that the ova survived better in the tank water than
in the pure sea water, though in two other similar series of experiments
the reverse relationship, which one would naturally expect, showed
itself. Of the ova fertilised immediately after shedding, one may see
that respectively 98"5 and 96*9 per cent, developed to normal blastulse.
On keeping, the ova in the tank water began at first to deteriorate
more rapidly than those in the pure sea water, but between the 24th
and 27th hours, those in pure sea water suddenly began to fall away,
and after 30J hours, only 025 per cent of the ova remained to undergo
normal development. The ova kept in tank water, on the other
hand, postponed their rapid degeneration till the 27th to the 35Jth
hours, or more especially till the 33rd to the 35 Jth hours. In order to
show more strikingly the suddenness of the onset of this abnormal
development on keeping the ova, another column has been added to
each half of the table, giving the percentage diminution of normally
developing ova per hour. For instance, after six hours development in
tank water, 3*2 per cent, less of the ova developed normally, or, on an

2 D 2

-552

Dr. K M. Vernon. The Effect of Stateness of the

average, 0*5 per cent, per hour for each of the first six hours. This value
is put, for convenience, against the " after six hours " line in the table,,
though it should rightly be placed between the " directly " and " after
six hours " lines. The other values are arranged in the same way. We
see then, that of the ova developed in tank water, from 0*5 to 1*8 per
cent, per hour underwent abnormal development up to the 27th hour, but
that then the percentage rapidly increased, till from the 33rd to the
35Jth hours, it reached to 13*5 per cent. In the case of the ova kept
in pure sea water, the result was still more striking. Thus till the
24th hour only 1*5 per cent, or less per hour developed abnormally,,
but from the 27th to 30|th hour, no less than 18*9 per cent.

Time of
fertilisation.

Per cent, diminution per hour.

Tank water \ Tank water
(1,000,000 per litre). (49,000 per litre).

Pure water
,000 per litre) .

(71 -6)
0-7

(83 -9)

(86-0)

After 9 hours

0-7

0-2

„ 20 „

3-3

6-6

0 3

„ 24 „

5 9

1-1

13 '8

29 „

0-5

0-05

3 5

» 32i „

0-2

o-o

1-5

„ 46 „

0-06

0-2

In this next table a similar series of observations is recorded, but in
addition a third series of determinations was made, in which ova and
sperm were kept in about twenty times as great a volume of water as
was used in the other experiments. Thus it was thought that perhaps
the keeping together of very large numbers of ova and of spermatozoa
in small volumes of water might tend to increase the rapidity of their
deterioration. As far as this single result can show, however, just the
reverse is the case. Thus when only 49,000 ova per litre were kept
together, the maximum rate of deterioration was reached between the
9th and 20th hours, whilst when 1,000,000 per litre (about the usual
state of dilution) were kept, the maximum rate was not reached till
the 20th to 24th hours.

In pure sea water, with a dilution of 680,000 ova per litre, the
maximum rate was also between the 20th and 24th hours, but a fair'
number also degenerated between the 24th and 29th hours. In this
table it will be noticed that the actual percentages of blastulse have
been omitted, and only the percentage numbers of ova per hour under-
going abnormal development given. The numbers given in brackets
against the " fertilised directly " line indicate the percentages of normal
blastulse produced on immediate fertilisation.

Sexual Cells on the Development of Echinoids. 353

Time of
fertilisation.

Per cent,
diminution per
hour.

Time of
fertilisation.

Per cent, j
diminu-
tion
per hour.

Time of
fertilisation.

Per cent. |
diminu-
tion
per hour.

1

Tank
water.

Pure
water.

Tank
water.

m ; i

Tank
water. I

Directly. . . .

(99 -7)

(98 -8)

Directly ....

(100 -0)

Directly. . . .

(98 -2)

After 6 hrs.

0-7

0-4

After 9 hrs.

1-6

After 6 hrs.

0-7

n 10 „

0-6

0-4

» 24 „

1 '0

„ 11 „

nil

„ 22 „

5-8

1-3

„ 33 „

6-7

„ 24 „

nil

5> 31 ,,

2-2

8-6

4 45 „

0-8

„ 32 „

11-7

„ 35

0-8

0-2

„ 36 „

0-1

In this next table the results of four series of observations are in-
cluded. In the first two the relative effects of tank and of pure sea
water, were again compared. In this case the pure water had a much
better preservative effect, the ova undergoing their maximum deteriora-
tion some ten hours later than those .kept in tank water. In the
remaining two series of observations, the sexual products were kept in
tank water, the maximum rate of deterioration being between the 24th
and 33rd hours, and the 24th and 32nd hours respectively. .This last
experiment is in some ways the most striking one made, as 94*8 per
eent. of the ova developed to blastulse until the 24th hour, whilst by
the 32nd hour only 0*8 per cent, so developed.

As a whole, therefore, these observations show a fair amount of con-
stancy. The mean times of the period of maximum deterioration in
the various series are respectively 34 J, 28f , 22, 14 J, 22, 18, 26 J, 28J,
and 28 hours, or, on an average, 24|rd hours after the shedding of the
ova and sperm. The reason of this constancy may have been the
similarity of the conditions of experiment. Thus all the observations
were made in the latter half of March and the first half of April, and
throughout the temperature of the water only varied between 13*5°
and 15-3°.

The chief conclusion to be gathered from these experiments seems to
me to lie in the comparative suddenness of the onset of the increased
rate of deterioration of the sexual cells. Thus in all but two out of
the nine experiments, the rate of increase of abnormal development
remained at about 1 per cent, per hour till the 20th to the 27th hour, and
then became so rapid, that within about nine hours the capacity for
normal development had almost entirely disappeared. Thus rapid
increase is well shown by the graphic method in the accompanying
figure. Here the four most striking results obtained are reproduced,
the values obtained in each of the different experiments being dis-
tinguished by different signs.

354 Dr. H. M. Vernon. The Effect of Stateness of tJie

•jnoffjdd dpynj^fg ^vlujo^ p uo/gnuiuj/g %

The probable reason of this increase readily suggests itself. Thus
supposing that animals developing from ova which are very nearly, but
not quite stale enough to avoid normal fertilisation and development,
are for that reason less strong and vigorous than those arising from
fresh sex cells, it follows that the period during which the sex cells
remain normal ought to be as long as possible, but that, once these have
begun to deteriorate, they ought to absolutely lose their functional
capacity as rapidly as possible, in order that the number of enfeebled
organisms which Ave have supposed to arise may be as small
can be.

as

Serual Cells on the Development of Echinoids.

355

In all the experiments thus far described, both ova and sperm were
kept for similar periods before fertilisation. In order to determine
whether the onset of abnormal development depended more especially
on the staleness of the one element or of the other, experiments were
also made in which either stale ova were fertilised with fresh sperm,
or fresh ova with stale sperm. At the same time some of the
stale ova were fertilised with the stale sperm, so that properly com-
parative results were obtained. These are collected in the subjoined
table.

Time of
fertilisation.

Percentage of normal blastulse.

? stale.
<j stale.

$ stale,
cj fresh.

$ fresh.
$ stale.

After 9 hours

a 9 „

„ 20 „•

„ 22 „

„ 24 „

„ 24 „

u 24 „

„ 27 „

„ 29 ..,

ta 33 „

„ 34 „

Mean

8S-0
85-6
29-0
91 -0
45 -3
70-1
94-4
73-3
2-7
10-6

o-o

53-6

81

•0

97 -0

70

•o

65-2

4

'7

95 -5

95

•0

93 -0

87

•3

96-2

87

•6

19-7

95

0

81 -1

67

■3

68-6

94

•3

5-9

48

•1

73-6

23

•8

36-0

68

•6

66 -5

On comparing the three columns, one can see that there is no regu-
larity in the figures. In two cases the maximum number of normal
blastulse was obtained from stale ova and stale sperm, in four cases
from stale ova and fresh sperm, and in five cases from fresh ova and
stale sperm. In the last line of the table are given the mean percent-
ages of all the observations made. From these one may perhaps con-
clude that whilst on an average just as many blastulse are obtained
with stale ova as with stale spermatozoa, yet that when both the sex
cells are stale, the proportion is slightly less. Probably, however, one
is more justified in concluding that within certain limits it is a matter
of indifference whether one or other or both of the sex cells is stale.
Thus if the last three observations in the table, made between the 29th
and 34th hours, be omitted, the average percentages of blastulae
obtained in the remaining eight experiments are respectively 72*1,
73*5, and 77*0, i.e., nearly as great with both sexual cells stale as with
only one. These last three experiments would, however, seem dis-
tinctly to indicate that when the sexual cells have reached their period
of rapid deterioration, it is a distinctly more favourable condition if
only one and not both of the cells be stale.

356 Dr. H. M. Vernon. The Effect of Staleness of the

The conclusion we have arrived at is not altogether an expected one,
as far as one could form any expectation from indirect evidence. Thus,
as before mentioned, 0. and E. Hertwig found that the ova of certain
Echinoids, if kept ten to twenty hours, underwent fertilisation in very
much larger numbers than freshly shed ova. The fresher and better
the condition of the sperm, however, the better the chance of cross
fertilisation. Again, in a former paper,* I showed that hybrid larvae
from the ova of Stnmgylocentrotus lividus and the sperm of Spluerechinus
granulans could only be obtained in any number during the months of
July and August, when the sexual products of Strong ylocentrotus were
found to reach their minimum maturity. Those of Sphcerechinus were,
on the other hand, in a mature condition. One would therefore be
inclined to conclude that in the present experiments, stale ova ferti-
lised by fresh sperm would yield a larger proportion of blastulae than
fresh ova fertilised by stale sperm. It must be remembered, however,
that the conditions are essentially different. Thus in direct fertilisa-
tion, it is the natural property of every ovum, whether fresh or stale,
to undergo fertilisation by any spermatozoon which still preserves its
vitality ; whilst in cross fertilisation it is, as a rule, the natural pro-
perty of every ovum to resist such impregnation, and this resistance
is only overcome when the vitality is diminished by keeping the ovum
in water, or by other means.

We have thus far examined only how far the staleness of the sexual
cells affects the number of normally developing blastulae. Does^it
have any more permanent effect, and do the larvae developed from
stale products differ either in form or size from those developed from
fresh products 1 In the paper just mentioned, a few experiments upon
this subject were recorded, and these showed a very distinct effect to
be produced : but as they were not very numerous, no great stress was
laid upon them. They have since been repeated, and sufficient con-
firmation of them obtained. The method of experiment was, as usual,
to mix portions of the liquids containing the stale or fresh ova and
sperm, and then, after an hour, pour them into jars containing about
2J litres of sea water. These jars were kept in a tank of running sea
water at a practically constant temperature for eight days, and the
larvae were then killed by the addition of corrosive sublimate, and pre-
served in 80 per cent, alcohol. They were then mounted in glycerin,
and measured under the microscope in groups of fifty, by means of a
micrometer eye-piece, f In a complete experiment, five series of mea-
surements had to be made, viz. : (1) of the normal larvae obtained
from the fresh ova fertilised with fresh sperm, (2) those from stale ova
. and stale sperm, (3) from stale ova and fresh sperm obtained from
another freshly opened Echinoicl, (4) from fresh ova and stale sperm,
* ' Phil. Trans ,' B, 1898, p. 465.

t For fuller details of the method, vide ' Phil. Trans.,' B, 1895, p. 577.

Sexual Cells on the Development of Echinoids. 357

(5) and, lastly, those from the ova and sperm of the freshly opened
Echinoids. It is of course impossible to get an exact basis of com-
parison for the larvae obtained from one stale and one fresh sexual
product. The best possible is to take a mean between the size of the
original normal larvae and that of the larvae obtained from the fresh
sexual products used for fertilising the stale products. The larvae
obtained with both sexual products stale are of course accurately com-
parable with the original normal larvae.

In the accompanying table are given the mean percentage differences
in the size of the larvae obtained with fresh and stale products, from
the original normal larvae in the one case, and from the mean between
the original and the fresh normal larvae in the other two cases. The
actual body length measurements of the original normal and fresh
normal larvae are also given, these being in micrometer eye-piece scale
units.

Condition of sexual
cells.

Fertilisation made after

9 lirs.

24 hrs.

33 lirs.

45 hrs. 24 hrs.

34 hrs.

Fresli ? , stale $

Stale $ , fresh $

-0'2
+ 7'1
-2-8

+ 1:9
+ 3-7
-3 0

4 1-1

+ 10-9

+" 2-0

ij

-1-9 :| -3-6
+ 1-5 +3-9
-15-9 1 -5-2

nil
— 2 ' 7
-13 0

31-01

Body lengtli of normal
Body length of fresli

29-94
26-12

30 59

25-85

I

M& 30-72

30-34 1 30 99

1

In the first group of observations given in the table these series of
measurements were repeated after keeping the sexual products respec-
tively 9, 24, 33, and 45 hours. When both the sexual cells were stale,
it may be seen that the size of the larvae was only slightly, if at all,
affected, the average variation from the original normal larvae being
only 0*2 per cent. With fresh ova and stale sperm, on the other hand,
the larvae were considerably increased in size, even those obtained
with sperm forty-five hours stale being slightly larger than the mean
normal. With stale ova and fresh sperm, there was in three out of
the four experiments a distinct diminution in the size of the larvae,
this amounting to no less than 15 '9 per cent, in the case of the ova
kept forty-five hours. This experiment thus affords a most satisfactory
confirmation of the conclusion tentatively put forward in the above-
mentioned paper, i.e., it shows that whilst larvce obtained from stale ova
and stale sperm are of the same size as those obtained from fresh sexual
products, those from fresh ova and stale sperm are distinctly larger than the
normal, and those from stale ova and fresh sperm distinctly smaller.

358

Dr. H. M. Vernon. The Effect of Stateness of the

The next experiment in the table is not so satisfactory, as in the one
instance in which stale $ stale ft larvae were obtained, there was
a distinct diminution in size ; whilst in one of the two sets of fresh $
stale ft larvae, there was also a diminution, instead of the expected
increase. In the two experiments with stale ova and fresh sperm, the
diminution was in each case very considerable, so that the somewhat
abnormal results obtained in this series of experiments may perhaps be
put down to the fact that all the larvae, whatever the conditions under
which they were obtained, showed a tendency to undergo diminution in
size.

In all the experiments thus far described, the fresh normal larvae
were measured as well as the original normal larvae. In another series
of observations, however, and in all the observations described in the
former paper, only the original normal larvae were measured ; hence,
though the values obtained for stale $ stale ft larvae are just as accu-
rate as before, those for stale $ fresh ft and fresh $ stale ft larvae
are presumably less accurate, as they are also compared against the
original normal larvae, and not the mean of the original and fresh
larvae. All these observations are collected in the subjoined table. In

Condition of sexual
cells.

Fertilisation made after

Means of

all
observa-
tions.

24 hrs.

18 hrs.

9 hrs.

22 hrs.

9 hrs.

Fresh $ , stale $ . . . .
Stale %, fresh $

-1-8
+ 5-5
-0-4

+ 11-0
-17-6

+ 9-6

+ 2'0
+ 3'5
-11-3

-3-4
-9-5
-1-8

-0 74
+ 4-05
-6-90

Body length of normal

30 '22

30 -72

29-41

30-95

the first series, the one not recorded in the former paper, there is a.
slight diminution in the size of the stale $ stale ft larvae, but a con-
siderable increase in that of the fresh $ stale ft larvae. In the next
experiment no stale $ stale ft larvae were obtained, but the fresh $
stale ft larvae showed a very marked increase in size, and the stale $
fresh ft larvae showed the maximum decrease of 17*6 per cent. Most
of the rest of the observations conform more or less to the rule above
laid down, but there is one very marked exception, the fresh $ stale ft
larvae being in one case 9*5 per cent, smaller than the normal, instead
of larger, as one would expect.

Taking means of all the observations in both series, we find that as.
an average of eight observations, the stale $ stale ft larvae were
diminished 0*7 per cent, in size ; as an average of eleven observations,
the fresh ? stale ft larvae were increased 4 per cent., and as an average;

Sexual Cells on the Development of Uchinoids. 359*

of ten observations, the stale $ fresh J larvae were diminished 6 '9
per cent. These values, being means of a fairly large number of
observations, may be regarded as trustworthy within certain limits, and
will, I think, be held sufficient to justify the rule above laid down.

Confirmation of the conclusion that larvae derived from stale ova and
fresh sperm are smaller than the normal was obtained from quite another
source. Thus in one case, it was found that the hybrid larvae obtained
on crossing the twenty -four hours stale ova of Sphcerechinus granulans'
with fresh Strong ylocentrotus sperm were 5*4 per cent, smaller than those
from the cross of the fresh ova and sperm, whilst those from some of
the same stock of ova after keeping an additional nine hours, and then
crossing with fresh sperm, were 9*3 per cent, smaller. Again, in another
experiment, larvae from twenty-four hours stale Sjjhcerechinus ova crossed
with fresh Strong ylocentrotus sperm, were 3*8 per cent, smaller than those
from the fresh ova and sperm. None of the repeated attempts at
obtaining crosses with stale ova and stale sperm succeeded, and in the
only case in which, fresh ova were crossed by stale sperm, the cross of
fresh ova and fresh sperm failed, and so prevented any comparison
being made.

In a previous part of the paper it was shown that the proportion of
blastulae obtained with both sexual products stale was nearly as great
as that with only one of them stale, and a similar relationship was found
to extend to the proportion of larvae. Thus, excluding the series of
observations made after forty-five hours, in which only 0*4 per cent,
or less of the ova arrived at the larval stage, and including only those
series in which all three of the fertilisations were attempted, it was
found that on an average stale ova with stale sperm yielded 43*5 per
cent, of larvae, stale ova with fresh sperm 55*5 per cent., and fresh ova
with stale sperm 49 "0 per cent. In this case, therefore, as in that of
the blastulae, the stale ova with stale sperm yielded the least propor-
tion, and the stale ova with fresh sperm the greatest, but the extreme
limits of variation are only comparatively slight.

It may perhaps be asked whether this somewhat curious result as to
the effect of staleness of one or other of the sexual products on the
size of the larvae is at all likely to be a factor of any importance
under natural conditions. To me it seems that in at least one respect
its value may be considerable, viz., it may be a very potent cause of
variation. As has been already stated, it seems probable that the
condition of relative staleness of the sexual products at the time of
fertilisation may very frequently occur in several phyla of the animal
kingdom, and hence it is by no means improbable that the average
variability of each generation may be increased considerably by this
means. As has been shown elsewhere,* the mean probable error of
variation in the body length of these Strongylocentrotus larvae is 6*1 per
* ' Phil. Trans.,' B, 1895, p. 615.

'360 Effect of Stale/less of Sexual Cells on Development.

cent. If, then, with a particular sea-urchin a third of the ova did not
undergo fertilisation till some twenty hours stale, whilst another third
underwent fertilisation by twenty hours stale spermatozoa, and only the
remaining third underwent fertilisation at once by fresh spermatozoa,
then the probable error of variation of all of the larva? so arising would
roughly speaking be doubled. Of course this is an extreme instance,
which could never occur in the case of Echinoids, but might easily
occur in the case of, for instance, man. Whether the variations so
produced would be in any degree transmissible by inheritance, is quite
another point, but supposing it to be only the size of the offspring
which is thus influenced, it may be merely a question of varying
degrees of nutrition, and so be directly transmissible. In any case,
whether inheritable or not, variation of itself may be in many cases
of value, as it may give a better chance to natural selection and other
agencies of picking out those individuals more adapted to their en-
vironment, and rejecting those less adapted. Thus, if all the individuals
are nearly alike the evolutionary process must needs be exceedingly
slow.

Finally, these results are of interest in that they prove the inequality
of the sex cells. The diminution in the size of larvse obtained from
stale ova and fresh sperm may perhaps be looked upon as one of
diminished nutrition, the result of the staleness of the yolk, but it is
difficult to imagine why the staleness of the spermatozoon used to
fertilise an ovuni should produce a larva larger than the normal, unless
one holds that the part played by the sex cells differs in some essential
particular.

Summary.

The following are the chief conclusions arrived at in this paper : —

(1) If the ova and sperm of the Echinoid St'rongt/loeentrotus lividus be
kept for various times in sea water before fertilisation, then for about
the first twenty to twenty-seven hours the number of normal blastula?
formed diminishes only about 1 per cent, per hour. After this ab-
normal development sets in rapidly, so that generally after a further
nine hours or so, no blastulse at all are obtained. The rate of falling
off in the number of normal blastulse may increase to as much as 18'9
per cent, per hour.

(2) If ova not more than twenty-seven hours stale be fertilised
with equally stale sperm, practically as many blastuke are obtained
as when stale ova are fertilised by fresh sperm, or fresh ova by stale
sperm. After twenty-seven hours, however, the number of blastula?
obtained with both products stale falls off" more rapidly.

(3) Larvas obtained from stale ova and stale sperm are of practically
the same size as those from fresh sexual products, but those from fresh
ova and stale sperm are distinctly larger than the normal, whilst those
.from stale ova and fresh sperm are distinctly smaller.

Influence of the Temperature of Liquid Hydrogen on Seeds. 361

" On the Influence of the Temperature of Liquid Hydrogen on the
Gerininative Power of Seeds/' By Sir William Thiseltox-
Dykr, K.C.M.G., CLE., F.E.S., Director of the Boyal Botanic
Gardens, Kew. Eeceived September 28, 1899.

The ' Connotes Bendus ' for August 28 (p. 43-4) contains a communi-
cation from Professor Dewar to M. Henri Moissan, " relative a la
solidification de l'hydrogene." It concludes with the following sentence,
which may be easily overlooked : — " Des graines refroidies dans de
l'hydrogene liquide conservent tout-e la propriete de germer."

This is the first announcement of an interesting experiment in which
Professor Dewar did me the honour to ask me to assist him. He has
further suggested to me to put on record the facts, as far as they came
under my observation, and any physiological conclusions to which they
seem to point.

"With this suggestion I have no alternative but to comply. Botanists
will naturally expect some more detailed account than is contained in
the brief announcement which I have quoted. But as my share in the
research has been of the smallest, I should have much preferred that
Professor Dewar should have given the result of the whole investiga-
tion himself.

When Professor Dewar first suggested the experiment to me, he
pointed out that it would be a costly one, that it would only be possible
to operate on very small quantities of seeds, and that the number of
kinds must also be few.

The dozen seeds experimented upon by Messrs. Brown and Escombe,
which were submitted to the temperature of liquid air, were apparently
selected as belonging to different natural families, and also in some
degree as to their composition.* My choice was much more restricted.
I took two out of their list for the sake of comparison ; Barley and
vegetable marrow. I added wheat, which had more than once been made
the subject of experiment. This gave me two farinaceous seeds and one
oily one. I then took shape and bulk into account. Wheat and
barley are roughly ellipsoidal and medium in size. The vegetable
marrow is relatively large but flattened. I therefore added another
oily seed, mustard, which is small and spherical. I followed Messrs.
Browne and Escombe in taking a pea, which is also spherical in shape
but nitrogenous in composition. Finally I sought a very minute seed,
and pitched upon musk.

The list then ultimately stood : —

* < Boy. Soc. Proc.,' vol. 62, p. 161.

362

Sir W. Thiselton-Dyer. On the Influence of

Brassica alba.
Pisum sativum.
Cucurbita Pepo.
Mimulus moschatus.
Triticum sativum.
Horcleum vulgare.

The next point seemed to be to eliminate the source of error which
might arise from defective germinative power. I therefore communi-
cated the list to Messrs. Sutton and Sons, of Eeading, and asked their
assistance. With their invariable kindness in any scientific enquiry,
they willingly complied, and sent the samples required with the follow-
ing report : —

" We now have pleasure in sending a packet of each of the seeds
you name. They are all of last year's growth, and of good germi-
nation.

" For your information we append the germinations arrived at by
our tests made in March last of the various parcels from which these
samples are taken.

" We have no doubt that each grain of wheat is a germinating seed,
as specially fine seeds have been picked out.

" In the case of musk a good growth was obtained, but the germina-
tion was not counted.

" Germinations : —

Mustard

1 Bountiful ' peas . .
Vegetable marrow

Musk

Wheat

Barley

I forwarded the samples (which were small) to Professor Dewar, and
suggested that they should be each divided into two portions, one for
a control experiment under ordinary conditions, the other to be
returned to me after being subjected to cooling. Owing to some
misunderstanding, this was not done ; but, as will be seen in the
result, the omission proved immaterial. The seeds, it should be stated,
were simply air-dried : they were ordinary commercial samples, and
no attempt was made to further desiccate them.

I pointed out to Professor Dewar the advisability of exposing the
seeds to extreme changes of temperature as gradually as possible, a
precaution which Messrs. Brown and Escombe carefully observed.*
He promised " to consider what can be done to avoid any disaster
from this cause."

* Loc. cit., p. 161.

100 per cent.
100 „

96 „
Good.

96
100

the Temperature of Liquid Hydrogen on Seeds.

863

On July 21 he wrote to me: — "In spite of the weather I have
carried out my promise, and cooled some seeds in liquid hydrogen for
half an hour. I had to seal them up in a glass tube, cool first in
liquid air, and then transfer to the hydrogen. They have, therefore,
been cooled to - 250° C, or -252°C, while being in a vacuum
(seeing the air left had no appreciable tension). The seeds, in other
words, have been transferred to a condition resembling that of moving
through space. Another set of the seeds have been cooled only in
liquid air for comparison."

On July 22 he added, on returning the seeds : — " There can be no
doubt about the seeds being cooled, as they were in the hydrogen for
more than an hour. In fact I used nearly 600 c.c. of liquid hydro-
gen."

The seeds came to me in the small packets of tinfoil in which they
had been placed in the tube. On opening these it was observed that the
seeds were as fresh and bright as before being subjected to the treat-
ment. There was not the slightest discoloration observable in the
green tint of the peas. This practically disposed of the only anxiety
which Professor Dewar felt as to the success of the experiment, and
expressed to me on July 25 : —

i; My own impression is that unless the sudden vacuum caused by the
liquid hydrogen cooling has produced physical rupture of the seeds,
they will germinate as usual. If they survive this awful strain, then
I believe no increase of the time of cooling could produce any effect
other than results from one hour's exposure to such severe cold."

The seeds were sown in a cool greenhouse, without heat, on July 27.
On August 1 they had all germinated. In the case of the mustard,
136 young plants were produced from 155 seeds ; the remainder had,
however, germinated, but the seedlings had damped off. One of the
packets of wheat, for some reason, germinated slightly more slowly
than the rest.

On August 5, I received a further packet of the seeds (the musk
excepted) indiscriminately mixed. Professor Dewar wrote the same
date : — " I have sent you seeds to-day which, if the treatment with
cold can kill, ought to be dead. They have been immersed in liquid
hydrogen for upwards of six hours, and no attempt was made to
graduate the cooling. They were placed in the vacuum vessel into
which the liquid hydrogen could drop from the apparatus, and had to
take their chance. The seeds have been soaked in liquid hydrogen,
and in this respect differ from the last that were cooled in a vacuum
from being sealed in a glass tube."

In this instance again- the seeds did not show the smallest visible
trace of the ordeal to which they had been subjected. They were
sorted out and immediately sown, under the same conditions as before.
By August 9 the seeds had all germinated without exception. I com-

364

Sir W. Thiselton-Dyer. On the Influence of

municated the result to Professor Dewar, and he informed me,
August 15 : — " The temperature Fahrenheit to which the seeds were
cooled was - 453° F. below melting ice."

These are the details of the experiment. As it is not likely to be
often repeated, I have thought it worth while to place them on record
as precisely as possible.

The first question that suggests itself is, what evidence we have
for believing that the seeds have actually been brought to the almost
inconceivable temperature with which they were surrounded. That
they were so brought, Professor Dewar himself has not a shadow of
doubt. That substances at widely extreme temperatures can remain
in juxtaposition at least for some time, and still maintain them, is
illustrated by a striking experiment shown by Professor Dewar at the
Eoyal Institution on April 1, 1898. Liquid air poured into a large
silver basin heated to redness, remained apparently as quiescent at
this high temperature as in cooler vessels, and maintained a spheroidal
condition. This is well understood. But the fact remains that liquid
air with a temperature of about - 190° C. was contained in a vessel
which had a temperature of 800° C., the difference in temperature
between the two being 1000° C.

If we turn for a moment to the effect of heat on living structure,
we know that a temperature of 75° C. is fatal to all protoplasm,
because at that temperature its proteids are coagulated. Yet there is
good evidence for the fact, that seeds have been exposed for pro-
longed periods to a temperature above 100° C, and yet have sub-
sequently germinated. It may be taken as absolutely certain, that in
this case that temperature never reached the embryo, but must have
been intercepted by the imperfect conducting power of the seed-coats.
Cohn again has found that the spores of Bacillus subtilis survive pro-
longed boiling,* and a similar observation applies.

It is probable that plant structures are deficient in thermal trans-
parency, and they are notoriously indifferent conductors. Nevertheless
it is difficult to believe that in the case of such small bodies as seeds,
their being brought to the temperature with which they are surrounded
can be more than a question of time.

That the thermal opacity of at least the seed-coats, may be really
considerable is not, however, impossible, even at low temperatures.
The following remarks by Professor Dewar -have an obvious bearing
on this point : —

" Pictet, after an elaborate investigation, concluded that below a.
certain temperature, all substances had practically the same thermal
transparency, and that a non-conducting body became ineffective at
low temperatures in shielding a vessel from the influx of heat. Ex-
periments, however, prove, that such is not the case, the transference
* See Tine's < Physiology of Plants,' p. 283.

tlie Temperature of Liquid Hydrogen on Seeds.

365

of heat observed by Pictet appearing to be due, not so much to the
materials themselves, as to the air contained in their interstices.
Good exhaustion in the ordinary vacuum vessels used in low tempera-
ture work, reduces the influx of heat to one-fifth of what is conveyed
when the annular space of such double-walled vessels is filled with
air/'*

It is to be noticed that in Professor Dewar's first experiment, the
seeds were practically in a vacuum. It is obvious from what has been
quoted above, that this would help them to retain their heat. Any
hesitation in accepting the results of the experiment on this ground
is however swept away by the second experiment in which the seeds
with absolutely no protection at all, were actually soaked in liquid
hydrogen for six hours. The extremity of caution can hardly resist
the conclusion that they must have been brought to the same tem-
perature.

Professor Dewar finds "that silica, charcoal, lampblack, and oxide
of bismuth, all increase the insulation to four, five and six times that
of the empty vacuum space." It might possibly be worth while to
try how far a packing of small air-dry seeds would compare, say with
charcoal. And this would in some degree be a measure of the thermal
transparency of seed structures.

Professor Dewar suggested to me that I should supplement this
statement by some remarks on the physiological bearing of the ex-
periment. This has already been discussed by Messrs. Brown and
Escombe, and there is perhaps little of moment to add to their con-
clusions.

The real interest of the whole investigation obviously lies in the
question how far it modifies our conceptions of the nature and pro-
perties of living matter. Protoplasm, whatever its source, has physical
properties and an ultimate chemical composition which are practically
uniform. This uniformity, however, overlies a potential diversity
which is not to be measured. Such diversity cannot be accounted for
by any purely physical conceptions, as physical conceptions are
understood.

We not merely know the ultimate constitution of protoplasm, but
we also know a good deal about its proximate constitution. Yet the
properties of living protoplasm are very far removed from the mere
sum of those of its constituents, and no light can be derived with
respect to them in this direction. And what we know about the con-
stituent bodies themselves is at present not a little obscure. They
belong, as it were, almost to the fringe of possible chemistry and
almost elude the methods of chemical research. But they, complex as
they are, are not themselves protoplasm. Their cumbrous molecules
are built up and broken down by ordinary chemical processes. They

* " On Liquid ^ir as an Analytic Agent," ' Roy. Inst.,' Apr. 1. 1898, pp. 7 and 8.
VOL. LXV. 2 E

366 Sir W. Thiselton-Dyer. On the Influence of

are not in themselves, in any intelligible sense, living though essential
to the exhibition of vital phenomena.

There our analysis of living matter by physical methods for the
present stops. But we are justified in pushing, at any rate, semi-
physical conceptions as far as we dare. We conceive, therefore, the
physical constituent molecules of protoplasm as aggregated into larger
molecules which, as they are unlike anything we know as purely
physical, we call physiological.* Of the properties of such molecules
we have some faint conceptions. The first is their instability. They
are kinetic ; " living substance is continually breaking down into
simpler bodies, with a setting free of energy ; on the other hand living
substance is continually building itself up, embodying energy into
itself, and so replenishing its store of energy." f This kinetic condi-
tion is essentially life ; when it ceases, we have hitherto believed that
the constituents of protoplasm come under the sway of purely inorganic
conditions.

If we pause for a moment to attempt a quasi-mechanical explana-
tion of the more developed phenomena of living organisms, such for
example as are included under heredity, we are led to suppose that
the physiological molecules may themselves be grouped into larger
aggregates. And each stage of aggregation introduces us into a new
order of phenomena. All that we can say is, that beyond the first
stage the properties which are characteristic of higher molecular aggre-
gates, are ultra-physical, taking physical in its ordinary signification.
That does not imply, however, that physical conditions are ever in
abeyance. Each stage of aggregation is conditioned by every one that
precedes it. In this sense life rests au fond on a physical basis.

A continuous kinetic condition appears to be one distinctive pro-
perty of physiological molecules. This not merely manifests itself in
continuous chemical activity, but under appropriate conditions in actual
visible motion. And it is to be remarked of the former, that though
chemical in kind, it is undoubtedly ultra-chemical as chemistry is under-
stood in the laboratory. A further characteristic of the physiological
molecule is that it possesses the power of breaking up chemical combi-
nations and reuniting their constituents in a way which absolutely
eludes the methods available to the chemist, and entirely outstrides
the pace at which he can proceed. There is the same kind of difference
between the two methods as there is between arithmetic and the cal-
culus in the solution of a mathematical problem.

The question then is, how far the effects of Professor Dewar's experi-
ments and of those who have preceded him in the same field, require
us to modify our conception of the physiological molecule. Are we

* Identical with Foster's " scmacula," ' Text-book of Physiology,' Part I, 5th ed.,
p. 6.

t Foster, loc. ext., p. 41.

the Temperature of Liquid Hydrogen on Seeds. 367

obliged to admit with Professor C. de Candolle and Messrs. Brown and
Escombe that it may descend to a purely static condition 1

This is really bound up with another question. The kinetic condi-
tion depends on the constant liberation of energy by chemical change.
Of this the most important is that due to oxidation. But we now
know that this is not the only source of energy in living matter, or in
all cases the indispensable one. The late Dr. Eomanes showed that
neither a high vacuum nor subsequuent exposure for twelve months to
absolutely indifferent gases, such as hydrogen or nitrogen, or even
poisonous ones such as hydrogen sulphide, had any effect on the ger-
minative power of seeds. Professor Pfeffer has, however, informed me
in conversation that an injurious effect is ultimately produced.

Vital processes have their optimum point as regards temperature.
Their superior limit for the reason already pointed out is tolerably
sharp; but the inferior is by no means equally so. According to
Boussingault the decomposition of carbon dioxide by green plants
may take place nearly at 0° C* Below the optimum there is then some
evidence of a "slowing down." While some processes reach their
limit, can we assume that all do ?

The question would be peremptorily answered for us by those
who assert that all chemical action is in abeyance at such temperatures
as I am discussing. Photographic action still takes place at the
temperature of liquid air, though this may be due to phosphorescence.
But a jet of hydrogen will burn in. it.

Messrs. Brown and Escombe sum up the two methods of explaining
what has been called " dormant vitality " with sufficient accuracy in
their paper. According to the one view, metabolic and its resultant
kinetic activity is " slowed down " indefinitely. In such a case as now
described, it might be said that this takes place along an asymptotic
curve, continually approaching but never becoming equal to zero.

According to the other, protoplasm passes absolutely from the
kinetic to the static condition. Its locked-up energy becomes purely
potential, and Professor C. de Candolle has not hesitated under these
circumstances to compare it to an explosive.

It has been pointed out that such a conclusion is absolutely in con-
flict with Mr. Herbert Spencer's well-known definition of life. But it
appears to me that that definition was only intended to apply to
higher stages of the aggregation of living matter than that of the
physiological molecule on which I have endeavoured to fix the discus-
sion. The question seems to me to be simply whether it is admissible
to regard that as capable of being brought to an absolutely static
condition.

Conceive two such molecules, one known to be living, but static,
and the other dead, and both to be maintained in a condition in which
* Sachs, ' Text-book,' 2nd ed., p. 729.

368

Mr. W. Edmunds.

they are not immediately susceptible to chemical change. What is the
criterion of life 1 There is none. It seems to me then that the question
I have propounded does not admit of any positive answer in the present
state of our knowledge.

A problem, perhaps somewhat scholastic, which once vexed the souls
of biologists was : — Whether life was the cause of organisation or
organisation of life. What is to be our answer if our starting point is
no more than a possible " explosive" 1

" Effects of Thyroid Feeding on Monkeys." By Walter Edmunds.
Communicated by Professor J. Kose Bradford, F.E.S.
Eeeeived September 28, 1899.

(From the Laboratory of the Brown Institution.)

The experiments were made altogether on nine monkeys ; from
them, however, it would be safer to exclude three : two because patho-
logical conditions were found in their lungs after death, and one
because the animal died in so short a time (seven days) after the com-
mencement of the treatment.

There remain then six experiments for consideration.

The thyroid preparations administered were either a powder made
according to the directions of Mr. Edmund White, or thyrocolloid
prepared by the method of Dr. Hutchison.

The doses given were large and corresponded to from one-half to
three sheep's thyroids (both lobes) to a monkey daily.

Marked effects were produced by this treatment : the eyes became
abnormally prominent ; the monkeys lost flesh, became weak, and
eventually died.

The symptoms produced may be tabulated as follows : (1) Proptosis,
(2) dilatation of pupils, (3) widening of palpebral fissures, (4) erection
of hairs on head, (5) hair falling out in patches from various parts of
the body, (6) paralysis of one or more limbs, (7) emaciation and mus-
cular weakness, (8) death from asthenia.

With respect to these symptoms, proptosis, dilatation of pupil,
widening of the palpebral fissure and erection of the hairs on the head
are effects which are produced by the stimulation of the cervical sym-
pathetic.

As the chief object of these experiments was to determine as to
exophthalmos, sketches were made of the monkeys before commencing
treatment and also after ; eye changes occurred in all the six monkeys,
and of four of them sketches exist showing these changes ; in two it
was not practicable to obtain the second sketches, but there was no
doubt about the eye changes in these also,

Effects of Thyroid Feeding on Monkeys. 369

Exophthalmos, after the administration of thyroid preparations, has
been observed before. Beclere, in a case of myxcedema in man in which
excessive doses of thyroid were given by mistake, noticed amongst
other symptoms a certain degree of exophthalmos. Ballet and
Enriquet, in one of six dogs which were fed on thyroids, found dis-
tinct exophthalmos. Cunningham also obtained this symptom in
rabbits and monkeys.

As the eye changes can be produced by stimulation of the cervical
sympathetic, it seems reasonable to infer that thyroid extract acts in
the same manner, and also that the increased secretion of the enlarged
and altered thyroid of Graves's disease is in part at least the cause of
the exophthalmos, that occurs in that affection.

The falling out of the hair in patches and its loosening generally
were common symptoms, and were seen in all the monkeys.

As to paralysis, in two of the experiments the hind limbs became
paralysed ; in one of these cases *the treatment was stopped to see if
any improvement followed, but it did not, and the animal died in a few
days. In another experiment both arms became paralysed; this
passed off in two days without the treatment being stopped.

These paralyses have also been noticed before. Beclere, in the patient
referred to above, found one day hemiplegia with aphasia and hemi-
anaesthesia, all of which disappeared in a few hours. In thyroidless
monkeys, Horsley has twice seen a complete hemiplegia, following an
attack of clonic spasms, and passing off in a day or so. In one of my
thyroidless monkeys a temporary paresis of one arm occurred.

The symptoms of loss of flesh and muscular weakness occurred in
all the experiments, and in all too the animal died, at periods varying
from forty to 116 days from the commencement of the treatment.

In two cases the feeding was stopped, but only when the symptoms
were well marked ; there was no improvement, and the animals died.

The main conclusion arrived *at is, that in monkeys thyroid feeding,
if large doses are given, produces exophthalmos.

REFERENCES.

Ballet and Enriquet : ' La Medecine Moderne.' December 26, 1895. No. 104.
Beclere : ' Gazette des Hopitaux.' October 16, 1894.

Cunningham : ' Journal of Experimental Medicine.' March, 1898. Yol. 3.
Horsley : ' Clinical Society's Report on Myxcedema.' Yol. 21. 1888.
Hutchison :< Journal of Physiology.' Yol. 23. (1898).
Edmund White : ' Pharmaceutical Journal.' September 2, 1893.

2 E 2

370

Dr. F. C. Penrose.

" On the Orientation of Greek Temples, being the Eesults of some
Observations taken in Greece and Sicily in the month of
May, 1898." By F. C. Peneose, M.A., F.E.S., F.E.I.B.A., &c.
Eeceived May 5 — Eead June 15, 1899.

The Cabeiron near Thebes.

My observations taken in 1892 were confirmed, and consequently
the doubt which a correspondent at Athens had thrown upon the
orientation of that temple* may be dismissed.

Girgenti.

For the orientations of the temples at Girgenti,! I had relied on
some observations taken in 1885 by means of the sun's shadow and a
plumb line, checked by data, given by various authorities. This year,
in the case of three of the temples, I was able to remeasure them by
solar observation with a theodolite, and also to obtain correctly the
altitude of the eastern horizon, as seen from each of the temples, with
the following results, viz. : —

The orientation angle of the Temple of Juno should be 262° 36'
instead of 264°, and the sun's altitude when surmounting the eastern
horizon should be 1° 45' instead of 0° 35'. The elements as recomputed
are as follows, viz. : —

Girgenti. Latitude 37° 18' 36".

Name of
temple.

Orienta-
tion
angle.

Stellar
elements.

Solar
elements.

Name
of
star.

Temple attri-
buted to
Juno Luci.n a

262° 36'

A, amplitude of star
or sun

B, corresponding alti-
tude

E, depression of suti
when star heliacal

E, E.A

Gr, approximate date. .

+ 10° 33' E.

3° 30'

+ 10° 15'
6h 14m

23° 56'
490 B.C.

+ 7° 24' E.

1 45
6 49
11° 30'

Ap. 6

v , >

a Arietis rising.

This amended date of the temple's foundation falls within the
Hellenic occupation of the site, and very nearly at the culminating epoch

* See ' Phil. Trans,' A, vol. 190, 1897, p. 46.
f Same vol., p. 55.

On the Orientation of Greek Temples.

371

of the city's prosperity. If the sun's depression had been taken at
11°, the date would have been about 440 B.C. and if at 12° about 540.
The depression at 11° 30' seems to accord best with the practice used
in temples of comparatively late foundation.*

The Temple of Hercules.

I found that the orientation angle of this temple agreed very closely
with that previously given, but that the eastern horizon was higher,
namely, 2~ instead of 0°35'; but if the solar depression be changed
from U° to 11° 54', the result of the calculation will be the same, and
the stellar elements and the date will remain unaltered.

The Olympieium.

I found that the orientation angle of the temple of Jupiter should
be increased from 257° 35' to 258° 44', whilst the altitude of the
eastern horizon has to be increased from 0° 35' to 1° 55'. This will
neutralize the effect of the alteration of the amplitude, and the stellar
elements and the date (430 B.C.) will remain unaltered.

I obtained one additional example from an ancient site in Greece,
namely, the Temple of Neptune at Calauria in the Isle of Poros ; the
scene of the last days of Demosthenes, who had chosen it for his
place of exile on account of its commanding a view of his much loved
Athens from the lofty ridge on which the temple was built. The
elements of the orientation are as below.

Calauria. Latitude 37° 31' 30".

Name of
temple.

Orienta-
tion
angle.

Stellar
elements.

Solar
elements.

Name
of

star.

Temple of
Neptune

247= 5'

A, amplitude of star
or sun

B, corresponding alti-
tude

E, depression of sun
when star heliacal

Gr, approximate date. .

+ 26° 10' E

+ 3° 0'

+ 22° 24'
6h 57m

7h 23m
960 B.C.

+ 24°53'E.
0

+ 19° 30'
8h lm
10°

8h 27ra
July 27

do

HQ

a
o

s

J

In the list of temples given in the same volume, viz. : Vol. 190,
p. 65, were ten examples of comparatively late date, and of most of

* See p. 65 of rol. cited.

372 Dr. F. C. Penrose.

Athens. Latitude 37° 58' 20".

Name of
temple.

Orienta-
tion
angle.

Stellar
elements.

Solar
elements.

Name
of

star.

Theseum ....
1

283° 6'

A, amplitude of star
or sun

B, corresponding alti-
tude

E, depression of sun
when star heliacal

F, E.A

G-, approximate date

-1° 7'E.

5° 30'

+ 2° 30'
5h 42m

llh 19m
470 e.c.

-10° 46'E.
5° 6'
-5° 17'

fjh 12m

17° 30'

12u 49m
Oct. 6

1

bb

3

J

New Erech-
theuni

265° 9'

A, amplitude of star
or sun

B, corresponding alti-

tude

E, depression of sun
when star heliacal

F, E.A

G-, approximate date

+ 6° 30' E.

4° 0'
+ 10° 35'

23h 58m
445 B.C.

+ 7° 20' E.

3° 25'

+ 7° 34'
7h 26m
12°

lh llm
April 9

bb

.s

'S3
'£

Kg

©
"u

8

J

New Erech-
tlieum con-
tinued

A, amplitude of star
or sun

B, corresponding alti-

tude

E, depression of sun
when star heliacal

E, E.A

G-, approximate date

+ 1° 53' W.

3° 0'

+ 3° 20'
5h 55m

21h 3m
450 B.C.

+ 7° 30' E. "

3° 25'

+ 7° 34'
7h 50m
14° 6'

10h 49m
Sept. 2

bb

"-+=>
©

r-tf

bD

a

J

New temple
of Bacchus
adjoining the
earlier temple

255° 49'

A, amplitude of star

or sun

B, corresponding alti-

tude

E, depression of snn
when star heliacal

Gr, approximate date

+ 11° 8'E.

3° 50'

+ 11° 8'
6h 15m

0h 4ra
340 B.C.

+ 14° ll'E. "

3° 10'

+ 13° 7'
8h 20m
17° 22'

2h 8m
April 23

1

bb

•S

i

©

J

New temple
of Jupiter
Olympius

270°

A, amplitude of star

or sun

B, corresponding alti-

tude

E, depression of sun

when star heliacal

F, E.A

G-, approximate date

-1° 13' W.

3°

+ 0° 52'
5h 48m

llh 33m
174 B.C.

0°E.
4° 31'
+ 2° 46'

ifh 5m
11°

0h 25m
March 27

On the Orientation of Greek Temples. 373

Ephesus. Latitude 37° 56' 30".

Name of
temple.

Orienta-
tion
angle.

Stellar
elements.

Solar
elements.

Wame
of

star.

Temple of
Diana as re-
built after
the fire

284° 35'

A, amplitude of star

or sun

B, corresponding alti-

tude

D, hour angles ....

E, depression of sun

when star heliacal

F, R,A

Gr, approximate date

-2° 32' E.

6°

i 1 0 AO'

+ 1

5h 35m

llh 25™
355 B.C.

-11° O'E.

4° 55'

to oe'
— 0 OO

15° 30'

12h 51m
Oct. 6

1

■oh

.9

o

*Ph

In this case the orientation follows the star, but with so considerable an increase of
amplitude that it would have been available for many centuries as a warning-
star.

which the years of their foundation are at least approximately known
historically, and which were shown to be conformable to the general
rule, but required a deeper depression of the sun, than was sufficient
for the distinct vision of the heliacal star. Of five of these examples
I had not given the elements, viz. : — The Theseum, the later Erechtheum,
the later Temples of Bacchus and Jupiter at Athens, and the great
temple at Ephesus, as rebuilt after the fire. As the cases of some of
these are very interesting in an archaeological point of view, I here
supply the elements in the same form as before.

With respect to the temple of Theseus at Athens, it appears to be
possible, proceeding from the approximate date given by the orienta-
tion work, to arrive at a much closer determination of the probable
exact year of the foundation (or perhaps dedication) of the temple,
and at the same time to confirm the traditional name of the temple,
which has been much disputed. What I have called the traditional
view is, that the temple was built under the influence of Cimon, who
during his supremacy at Athens, in the year 469 or 468 B.C., brought
from the Isle of Scyros the supposed relics of Theseus. The bones of
the hero were then interred at Athens with great solemnity, and a
temple or heroum was built over the grave and an annual celebration
was appointed, under the name Thesea ; which lasted two or three
days and commenced on the seventh day of the month named Pyane-
psion, a month which on the whole agrees with our October, but (owing
to the practice at Athens of commencing the year with the new moon
which occurred on or next after the summer solstice), when we com-
pare the two calendars, and reckon the days of the month together,
we find that they only agree together at intervals of 19 years — the

374

On the Orientation of Greek Temples.

Metonic cycle. It is therefore necessary to consult lunar tables to
see in what year or years 7 Pyanepsion would agree with October 6,
and the nearer to 470 B.C. this can be, the better it would also agree
with the orientation date.

The first Attic month was Hecatombaion having thirty days, in

general agreement with July.
Then Metageitnion having twenty -nine days, in general agreement

with August.

Then Boedromion having thirty days, in general agreement with
September.

After which Pyanepsion.

The first three months with seven days of Pyanepsion numbers
ninety-six days. Twenty-nine days of July with August, September
and six of October also give ninety-six days. It follows that when the
first of Hecatombaion agrees with the third of July, the seventh of
Pyanepsion will represent the sixth of October.

This would have happened in 466 B.C. and not again until 447.
The year 466, about three years subsequent to the recovery of the
Theseian relics, would have given time for the building of the Naos
of the temple, if not for its final completion ; so that it may well have
been the year of its dedication. On that year the astronomical com-
bination would have been exact (not but that on any year of the cycle
the rising sun would have nearly answered the purpose of the
Thesea celebration, though not quite so perfectly). Of the two years
above named which would have satisfied this condition, the earlier
seems preferable, firstly on architectural grounds (derived chiefly from
the greater spread of the capital of the column, compared with that of
the Parthenon, which was commenced about the later of the two
dates), and secondly, there is no record of Pericles, who was at that
time the guiding spirit in Athens, having built a great temple in the
lower city. If the earlier date be correct, the conclusion appears in-
evitable both from its combination with the Thesian relics in 469, and
from its connection with the Thesea year after year, that the temple
has been rightly named by tradition.

The orientation of the new Erechtheum corresponds with two of the
principal Attic festivals, and also offers a suggestion of the exact year
of its foundation or dedication. This is not drawn from the vernal
sunrise, of which I have given elements in combination with a star ;
but from the autumnal return of the sun to the same declination,
heralded by a Pegasi, which took place, touching the northern edge of
the eastern incolumnium on September 2, and parallel with the axis
on September 7. In the year 447, when the 2nd of September would
have agreed with the 2nd of Boedromion, and the 7th September
with the 7th of the same month; the former date would have

Determination of the Earth's Horizontal Magnetic Force. 375

been that of the great feast of the Niceteria in honour of Minerva's
contest with Neptune for the protectorate of Athens, and the other,
the annual celebration of the Marathon victory.

As respects the year 447, which is one year earlier than the supposed
commencement of the Parthenon, it seems appropriate because at that
time Pericles would have been supreme, and it is likely that a temple
of such sanctity as the Erechtheum would have called for his earliest
attention. It is true that the temple remained long unfinished, but
of the causes of this delay we are ignorant. The connection however
of the orientation with the feasts above mentioned would have been
exactly the same if the date had been 428.

The apparent discrepance between the orientation date (as respect
the day of the month) in the case of the Temples of Jupiter Olympius,
and that which is supposed by Mommsen to have been the day of
the celebration of the great feast to the supreme god (namely
Munychion 19, the tenth month of the year, which in a general way
corresponded with April), whereas the orientation dates give for the
earlier temple March 30-31, and for the later March 27, is explained
by the possibilities of the Metonic cycle, for when the Attic year
began as it would in its course on July 11, the 19th Munychion
would agree with March 30, or if July 8 with the 27th of March.

11 Collimator Magnets and the Determination of the Earth's Hori-
zontal Magnetic Force." By C. Chree, Sc.D., LL.D., F.Pl.S.,
Superintendent of the Kew Observatory. Communicated by
the Kew Observatory Committee of the Eoyal Society. Be-
ceived May 31 — Eead June 15, 1899.

Contents.

Sect.

1 . Introductory.

2. Temperature coefficients.

3. Induction coefficient.

4. Moment of inertia.

5. Coefficient P.

6. Tables of mean and extreme values of magnetic constants.

7. Dimensions of magnets, Table III.

8 — 14. Discussion of results of Tables I and II.
15—18. Relations between different magnetic conslants.

19 — 24. Probable errors in determinations of horizontal force due to errors in
Talues of magnetic constants.

25 — 28. Criticism of formulae for reducing horizontal force observations, from
mathematical standpoint.
29. Physical sources of uncertainty, variations of temperature.

30 -31. „ „ variations of horizontal force and declina-

tion.

VOL. LXV. 2 F

376 Dr. C. Chree. Collimator Magnets and the

Sect.

32. Uncertainties in determination of moment of inertia.
33 — 34. torsion.
35 — 37. „ ,, temperature coefficients.

38. „ ,, induction coefficient.

39 — 44. Asymmetry in magnets.

45 — 47. Law of action between magnets.

4S — 50. Concluding remarks.

§ 1. The present paper deals with magnets employed in measuring
declination and horizontal force. More than 100 collimator magnets
of English make have been examined at Kew Observatory, and the
record of the results forms probably a unique mine of information. So
far as I know, the only use hitherto made of this has been in the com-
pilation of a statistical paper on the mean and extreme values of the
temperature and induction coefficients by the late Mr. G. M. Whipple.*

The examination of a collimator magnet at Kew Observatory con-
sists mainly in the determination of certain " constants." These are
the "temperature coefficients" q and q', the "induction coefficient" /x,
and the " moment of inertia " K.

The values found for these " constants " are utilised in the construc-
tion of tables, intended for reducing the observations of horizontal force.
After the tables are constructed one or two observations are made.
Their primary object is to ensure that the application of the tables
leads to satisfactory results, and that there are no instrumental defects ;
but incidentally they afford the means of determining the magnetic
moment, m, of the collimator magnet, and also the value of a " constant,"
P, appearing in the expression

2m7?A--3(l+P?'-2)

for the couple exerted by the collimator magnet on an auxiliary magnet,
moment m", at distance r. The investigations described in the present
paper have been prosecuted at intervals during the last five years, as
the pressure of other work allowed. Their object has been twofold, 1°
to find out whether any relationships exist between the several con-
stants, and 2° to ascertain where our present knowledge wants
amplification, and where the present tests are least satisfactory.

I shall first explain the real significance of the " constants," and
describe briefly the method of determining them.

§ 2. Temperature Coefficients. — It is assumed that the magnetic moment,
m , at a temperature of t° C, the magnet being free from external force,
is connected with the moment m at 0° C. by the relation

m'/ra = 1-qt-q'P (1),

where q and g/ are absolute constants for the magnet concerned. If (I)
* 'Boy. Soc. Proc.,' vol. 26, pp. 218-222, 1877.

Determination of the Earth's Horizontal Magnetic Force. 377

be taken for granted, q and q can be determined by observing m' at any
three convenient temperatures. These temperatures, as the experiment
is conducted at Kew Observatory, lie usually within a degree or two of
0°, 18°, and 36° C. Magnets, however, destined for Arctic work are
exposed to a temperature below 0° C, while magnets destined for
tropical regions are exposed to a temperature over 40° C.

The changes of temperature are made rapidly by introducing hotter
or colder water into a wooden box containing the collimator magnet.
Changes in the moment of this magnet, accompanying observed changes
in the temperature of the water, are deduced from the variations in
azimuth of an auxiliary magnet, freely suspended at a fixed distance
from the deflecting magnet. In a single experiment, the cycle of
temperature " hot," " mean," " cold " is repeated three times, and the
mean of the three observations at each temperature is used in the final
calculation. It is customary to have two completely independent ex-
periments on different days, and to take the arithmetic mean of the
values deduced for q and q' on the two occasions. The exact times are
noted at which the" several readings are taken, and suitable corrections
are applied from the readings of the magnetic curves for variations in
the horizontal force and declination.

§ 3. Induction Coefficient. — This is denoted by ft, and really means
the temporary change in the magnetic moment of the collimator magnet
due to unit change in the field (parallel to the magnet's length), it
being assumed that the relation between temporary moment and
strength of field is linear.

The experiment* consists in observing the angles through which an
auxiliary magnet is deflected out of the magnetic meridian by the
collimator magnet, the latter being vertical, with its north pole alter-
nately up and down. The vertical plane through the centres of the
deflecting and deflected magnets is perpendicular to the latter's axis,
but the centres of the magnets are not in the same horizontal plane.

The change in the inducing field being double the intensity of the
vertical force at Kew is nearly 0*9 C.G.S. unit. As a rule, only one
complete experiment is made, but this involves inverting and reinvert-
ing the magnet several times. Originally fx was measured in British
units, so that conversion into C.G.S. units was necessary in many cases.

§ 4. Moment of Inertia. — This means the moment of inertia of the
magnet and all its appendages, when at a temperature of 0° C, about
the suspending fibre. A collimator magnet is a hollow steel cylinder,
about 9J cm. long and 1 cm. in external diameter, with screws cut on
its inner surface at both ends. The appendages consist of two small
cells, one holding a lens the other a glass scale, screwed into the

* A special apparatus is employed whose description would occupy undue space.
The method is practically that described on pp. 151-3 of Lamont's 1 Handbuch des
Erdmagnetismus.'

2 F 2

378

Dr. C. Chree. Collimator Magnets and the

ends of the magnet, and of a brass stirrup arrangement which carries
the magnet and affords the means of supporting, parallel to it, an
auxiliary solid brass cylinder. This brass cylinder is a regular geo-
metrical object, whose moment of inertia can be calculated from its
weight, length, and diameter. The actual inertia experiment consists
in observing the times of vibration of the magnet, under the earth's
horizontal force, when the auxiliary bar is in the stirrup, and when it
is removed. To reduce the possible effects of variation in force or
temperature, four complete series of vibrations are made, the first and
fourth without, the second and third with the auxiliary cylinder.
Allowance is made for the departure of the mean temperature from 0° C.
It is customary to make two independent determinations of the moment
of inertia, usually on different days, and in the event of serious dis-
crepancy a third experiment is made. In all the older experiments
conversion from British to C.G.S. units was necessary.

§ 5. Coefficient P. — The meaning of this has been already explained
generally. I need only add that in the deflection experiment the axes
of the two magnets are perpendicular, and the centr