Full text of "Induced cell-reproduction and cancer; the isolation of the chemical causes of normal and of augmented, asymmetrical human cell-division"

See other formats

(KJ

FIG. 1. — A typical field seen by the in-vitro method of staining. The leuco-
cytes are staining gradually.

INDUCED CELL-REPRO-
DUCTION AND CANCER

THE ISOLATION OF THE CHEMICAL CAUSES OF NORMAL AND
OF AUGMENTED, ASYMMETRICAL, HUMAN CELL-DIVISION

BY HUGH CAMPBELL ROSS

M.R.C.S. (ENG.), L.R.C.P. (LOND.)''

SURGEON, ROYAL NAVY (EMERGENCY LIST) ; DIRECTOR OP SPECIAL RESEARCHES AT THE

ROYAL SOUTHERN HOSPITAL, LIVERPOOL; AND HONORARY CLINICAL PATHOLOGIST

TO THE ROYAL LIVERPOOL COUNTRY HOSPITAL FOR CHILDREN

1 if

BEING THE RESULTS OF RESEARCHES CARRIED
OUT BY THE AUTHOR WITH THE ASSISTANCE OF

JOHN WESTRAY CROPPER

M.B., M.Sc. (LIV.), M.R.C.S. (ENG.), L.R.C.P. (LOND.)

ASSISTANT TO THE RESEARCH DEPARTMENT OF THE ROYAL SOUTHERN HOSPITAL, LIVERPOOL

WITH 129 ILLUSTRATIONS

PHILADELPHIA
P. BLAKISTON'S SON & CO.

1012 WALNUT STREET
1911

Printed ly

The Maple Press

York, Pa.

INSCRIBED TO

SIR WILLIAM P. HARTLEY, J. P.

OF LIVERPOOL
AND

JOHN HOWARD McFADDEN, ESQ.

OF PHILADELPHIA

MITOTIC FIGURE INDUCED IN A LARGE LYMPHOCYTE.

This illustration was obtained after the book had gone to press and does not appear in
the list of illustrations. It is included because it so clearly demonstrates the mitosis of the
lymphocyte which was unstained. The division was induced by means of bensamidine, one
of the several compounds containing the amidine grouping (N = C — N) the presence of which
appears to be necessary in a substance before it can cause cell-division. This point was de-
termined after the book had gone to press.

PREFACE

THE objects of this book are to describe in detail the
results obtained by a new method of experimentation
with individual living human cells, their importance in
the elucidation of the phenomena of healing, and in the
causation of cancer and other growths.

The old methods of examining dead tissues and
cells have been useful in the past, but I venture to
think that those who undertake the study of living
human cells, and especially blood-cells, by the in-vitro
methods of staining, which will be hereafter described,
will realise that they supersede all others. This method
enables us to observe cells in their proper shapes, and
an entirely new impression is obtained concerning
the functions of their constituent elements and of the
modes by which they divide and reproduce themselves.
The fact that the divisions of reproduction can be
induced on a microscope slide by means of the natural
chemical agencies which cause their proliferation within
the body has in itself opened a fresh vista of research
which has not only taught us the cause of prolifera-
tion of ' cells in healing, but has also suggested that
the cause of malignancy, which appears to be re-
lated to that normal process, is beginning to become
cleared up.

The methods which I shall describe are entirely

X PREFACE

new; some of the details have already been published
in the scientific journals, but the greater part of what is
herein set forth has hitherto been unknown. The new
methods have revealed many interesting facts which in
my opinion may be far-reaching in their influence in
the advancement of pathology.

It may not be out of place if I give a brief history
of the circumstances which led to the adoption of this
in-vitro method of microscopical investigation and of
the researches which have been made by means of it.
There can be no doubt that accidents have on more
than one occasion been responsible for valuable indica-
tions which have led to fruitful lines of work, and had
it not been for some of these accidents the results
attained would have been considerably less advanced
than they are now. I do not think that these re-
searches would have been started at all had it not been
for the firing of a gun. In the summer of 1905, when
I was a surgeon in the navy, my cabin being my
laboratory, I was interested in bacteriology, and
was endeavouring to grow organisms from the blood
of patients. To do this I had to invent an electric
incubator, it being impossible on board a battle-
ship to use one which was heated either by gas or oil,
the former not being available and the latter not
allowed. There was nothing for it, therefore, but to
invent an incubator which could be made on board,
and I look back upon this piece of apparatus with
interest. It was not so reliable as those which can
now be bought, but it worked fairly well. It had an
automatic thermostat, by means of which the lamp was

PREFACE XI

switched out at a given temperature, and was switched
on again when the temperature fell. Sparking gave
trouble, but I "blew out" the spark by a condenser.
It wras made by a "torpedo instructor," and was so
firmly bolted on to the steel bulkhead in my cabin that
apparently nothing would (or could) shake it down.

One afternoon, when my ship was in the Mediter-
ranean, we had, as I thought, finished heavy gun-firing.
I had placed some blood on to some nutrient agar
(sloped) in culture-tubes, which were in my incubator,
being kept at the blood temperature of 37° C. I was
working at the cabin table with the microscope and my
small stock of bacteriological apparatus. Suddenly,
without any warning, a "young gentleman" fired a
12-inch gun from the after-babette on the deck above;
for the captain had permitted the midshipmen to fire
a "round" after the main gunnery practice was over.
I extricated myself from the debris of microscope,
apparatus, pictures, etc., on the deck of my cabin, for
nearly everything was smashed. My incubator, firmly
fixed, as I have explained, on to the bulkhead, I did not
open, expecting that everything inside was shattered,
and it was not until the next day that I investigated its
contents. My surprise may be imagined when I found
that the culture-tubes were unharmed, but that, owing-
to the dislocation of the automatic thermostat, the
temperature inside the apparatus wras standing at 606 C.
On close examination of the culture-tubes, I noticed
that the red cells, which had been resting on the surface
of the jelly, were now diffusing as a cloud through the
jelly itself. The matter wras further investigated, and it

Xll PREFACE

formed the subject of my paper in The British Medical
Journal of May 5, 1906, on "The Diffusion of Red
Blood-corpuscles through Solid Nutrient Agar." The
reason why the diffusion had not been previously
observed was that one does not usually endeavour to
obtain cultures from the blood at 60° C., nor should
I have done so had it not been for the zeal of the
"young gentleman."

In January, 1906, before my paper was published, I
was demonstrating this remarkable diffusion of red cells
through the agar to my brother, Professor Ronald Ross,
at Liverpool, by placing some blood under a cover-glass
on a film of agar jelly spread on a slide. He was
impressed by the way in which the cells became spread
out between the cover-glass and the surface of the agar
film, and he suggested that it would be a good means
for blood-examination by the microscope, for the
corpuscles became admirably arranged and spread out in
such a way that each could be critically examined.
Then he remarked — a remark which has led to all the
researches described in this book, and to the discovery
that mitotic divisions in human cells are induced by
chemical agents— "I wonder what would happen if we
were to mix some stain with the jelly and then place
the living cells on it under a cover-glass." This sugges-
tion was promptly put into operation, and fortunately
(because it was the best which could have been chosen)
the first stain experimented with happened to be poly-
chrome methylene blue, with wrhich I obtained results
which determined me to adopt this method of examina-
tion to the exclusion of all others.

PREFACE Xlll

In July, 1906, I left the navy and proceeded to
Egypt, having received an appointment there in the
Public Health Department. Sir Horace Pinching,
the Director-General, permitted me to continue the
researches, and it was during^ the ensuing year that
the phenomena of achromasia and liquefaction of the
cytoplasm of leucocytes were investigated together with
some of the laws concerning the diffusion of substances
into cells. In January, 1907, I accidentally discovered
the excitation of amoeboid movements caused by
atropine, for, as will be described in the chapter relative
to this phenomenon, I was in reality trying to poison
the cells with the alkaloid.

In October, 1907, Sir Horace Pinching, my chief,
having retired, I w^as treated in such a manner by
Mr. W. P. Graham, the new Director- General of the
Public Health Department, that I was forced to leave
the Egyptian Government Service in December, 1907.
Mr. Graham objected to my doing scientific work dur-
ing my spare time, and also prevented my continuing
the mosquito campaigns which I had started, as he
apparently did not believe in them. This treatment
stopped may researches for the time being, and was the
cause of considerable delay in accomplishment of this
work. I was enabled, however, to complete my investi-
gations into the cause of achromasia by Dr. Marc
Armond- Ruffer, C. M. G., who came to my rescue and
temporarily gave me an appointment in the Quarantine
Department at Suakim in February, 1908. Here I
was also able to devise the technics for "measuring the
lives of leucocytes."

XIV PREFACE

In July, 1908, I obtained the appointment of Patho-
logist to the Royal Southern Hospital at Liverpool,
which I held for eight months, during which I was
able to investigate further the laws of the diffusion of
substances into living cells and to devise the technics
for the determination of the "coefficient of diffusion."
In the meantime I published the results of work done
while I was in Egypt. This could not be done before,
because Mr. Graham would not permit me to publish
scientific work. Two papers appeared in The Journal
of Physiology in September, 1908, four in The Lancet
in January and February, 1909, and that on the " Co-
efficient of Diffusion" in the Proceedings of the Royal
Society in April, 1909.

In August, 1908, I was demonstrating the excita-
tion of amoeboid movements caused by atropine to
Dr. Macalister, when he suggested to me that possibly
there might be some alkaloid-like excitant in the blood
of cancer patients; and this important suggestion was
the starting-point of the investigation of cancer by this
in-vitro method. Dr. Macalister and I read a paper
before the Royal Society of Medicine in November,
1909, on the researches which immediately followed his
suggestion.

In March, 1909, Professor Harvey Gibson also sug-
gested the important point, based on an observation
made by Professor Farmer, that nuclein might have
some influence on cell-division. I must acknowledge
the great assistance which I have received from Pro-
fessor Harvey Gibson on many occasions, as well as
the loan of a well-equipped laboratory in the Hartley

PREFACE XV

Botanical Department at the University of Liverpool.

It was when experimenting with a mixture of
polychrome stain, extract of haemal gland, and atropine
that I saw mitotic figures in lymphocytes for the first
time in May, 1909; and the. ensuing months were
occupied in the investigation of the cytology of these
cells and of the means whereby they might be induced
to reproduce themselves. It was not until October,
1909, however, that I was able to induce divisions in
polymorphonuclear leucocytes. In December, 1909, I
discovered almost accidentally that extracts of dead
tissues, if they were allowed to decompose by the
action of putrefactive bacteria, would, by themselves,
induce the division and multiplication of lymphocytes,
and this was followed by the investigation of the
action of "globin" in January, 1910. In February, 1910,
while investigating the epithelial cells present in some
vaginal secretion, it occurred to me to try to induce
divisions in them, and this experiment has been suc-
cessful in the case of one or two cells.

In April and May, 1910, when working with my
assistant, Dr. Cropper, I saw divisions induced by
kreatin and xanthin, the "extractives" contained in the
remains of dead tissues; and we then also investigated
the augmenting action on cell-division of the alkaloids
choline, cadaverine, etc., produced by the decompo-
sition of putrefaction. These points led me to elaborate
the theory regarding the cause of cancer which is
described in the latter part of this book.

The dates of the treatment of the two cases of
cancer are given in the description of them which

XVI PREFACE

has been written by Dr. Macalister. The crucial
experiment to try to determine whether my theory
regarding the cause of cancer was correct or not was
made in the first week of August, 1910. The treatment
of all the cases has been carried out under the immediate
supervision of Dr. Macalister.

In much of the latter portion of the experimental
work I have derived assistance from Dr. Cropper. He
has accomplished nearly all the investigations con-
cerned in counting the number of granules contained
in eosinophile leucocytes, and has given most valuable
assistance in the isolation of the active "auxetics"
from the remains of dead tissues, and in the investiga-
tion of the inhibitory action of blood-serum, some of
which he did entirely himself at my suggestion.

The Research Department at the Royal Southern
Hospital at Liverpool was started by Dr. Macalister
in April, 1909, in order that this \vork might continue;
Sir William Hartley, J. P., generously supplied funds
to last for one year in the first instance. In November,
1909, he extended his support for a period of three
years, and he also supplied me with the photomicro-
graphic camera which I had invented. Some of the
expenses attached to these researches, however, have
also been defrayed by Mrs. George Holt, Mr. and the
Misses Paton, and some of their friends. I would
like to take this opportunity of recording my personal
gratitude to Sir William Hartley and these ladies and
gentlemen, without whose assistance these researches
could not have been accomplished.

I wish also to record the manner in which I was

PREFACE XV11

enabled to obtain the assistance of Dr. Cropper. In
October, 1909, I happened to be discussing with Mr.
Sharpies, an energetic supporter of these researches,
the difficulty of my finding time to undertake the
investigation of some of the by.-issues revealed by the
new method — issues which might prove to be of
importance. Our conversation was overheard by a
gentleman, who I afterwards ascertained was Mr.
J. H. McFadden, of Philadelphia, whose acquaintance
I had only just made, and to whom I was practically
a stranger. Mr. McFadden immediately became in-
terested, and placed a large sum of money at my
disposal in order that I might obtain other assistance
to further these researches for a period of two years.
In March, 1910, Mr. McFadden further instructed me
to the effect that if I conscientiously thought that
further funds could usefully be spent in the advance-
ment of these researches I was to incur that expendi-
ture. In fact, he has not only supplied me with the
assistance of Dr. Cropper, but he has also been the
means of equipping a laboratory for him, kindly lent
by the Liverpool School of Tropical Medicine, but
also in defraying the serious expenditure connected
with the manufacture of the substance we call "globin"
from crystalline haemoglobin. Mr. McFadden has also
enabled me to take the large number of photomicro-
graphs which record the phenomena seen under the
microscope; ajnd, lastly, he has borne the entire cost
of the publication of this volume and the reproduction
of the photomicrographs which illustrate it.

I fear that I shall never be able to thank Mr.

XV111 PREFACE

McFadden sufficiently for his great generosity, which
I appreciate very greatly. It was extended to me at
a time when I was practically a total stranger to
him, together with the intimation that even if these
researches did not result in the advancement of knowl-
edge regarding cancer he would not consider that his
assistance had been misplaced or wasted.

In conducting prolonged researches of this nature
it is most gratifying to realise that one has such a
staunch supporter; and there can be no question that
if the results obtained lead to practical benefit, this will
be largely owing to Sir William Hartley, who enabled
the researches to be started, and to Mr. McFadden,
who enabled them to be brought so rapidly to the point
which has now been reached.

When this Research Department of the Royal
Southern Hospital was started, a Committee was
formed. It numbers amongst its members Professors
Sherrington, Herdman, Ronald Ross, Reynolds Green,
and Harvey Gibson, to all of whom I have frequently
appealed for advice on technical points; and when any
information has been required concerning the surgical
aspects of the healing process, or of cancer, I have
consulted Mr. Robert Jones, who is now Chairman of
the Committee, and also Dr. Alexander. I wish to
thank all these gentlemen most sincerely for their
kindness. I have also frequently received materials
from Mr. Jeans and Mr. Bickersteth, of the Royal
Infirmary, and many other members of the medical
profession in Liverpool have supplied specimens. The
beautiful sections of the growth with which the crucial

PREFACE XIX

experiments were made were kindly cut for me by
Dr. Moore Alexander. Professor Benjamin Moore
advised us in our researches to isolate the active auxetics
from the extracts of dead tissues, and Dr. H. E. Roaf
very kindly supplied us with the crystalline haemoglobin
from which globin was obtained for the first ex-
periments with that substance.

I cannot close this Preface without signifying my
thanks to Dr. Macalister. He grasped that this in-
vitro method would bring fruitful results, and it was
he who instituted this research into the cause of
cancer. Dr. Macalister's advice and constant encour-
agement, apart from the actual experimental work
which he has done and the clinical observations which
he has made, have been invaluable.

CONTENTS

CHAPTER I

PAGE

THE SCOPE OF THE NEW METHOD . 1

CHAFTER II

THE GENERAL PRINCIPLES OF THE METHOD — THE APPARA-
TUS REQUIRED — THE SPECIAL PHOTOMICROGRAPHIC
APPARATUS — THE REVOLVING APPARATUS . 15

CHAPTER III

THE PREPARATION OF THE JELLY-FILM 36

CHAPTER IV

CELLULAR STAINING, DEATH, AND ACHROMASIA 41

CHAPTER V

THE DIFFUSION OF SUBSTANCES INTO LIVING CELLS— THE

" COEFFICIENT OF DIFFUSION" 61

CHAPTER VI

THE PRACTICAL DETERMINATION OF THE " COEFFICIENT OF
DIFFUSION" OF CELLS, AND ITS APPLICATION TO THIS

"IN- VITRO" Method of Research 81

xxi

XX11 CONTENTS

CHAPTER VII

PAGE

DIFFUSION OF SUBSTANCES INTO CELLS TO EXCESS — DIFFU-
SION- VACUOLES OR " RED SPOTS" THE PROOF THAT THE

BLOOD-PLATELET is A LIVING CELL 103

CHAPTER VIII

THE EXCITATION OF AMOEBOID MOVEMENTS IN WHITE

BLOOD-CORPUSCLES CAUSED BY ALKALOIDS 130

CHAPTER IX

THE ADOPTION OF THE "IN- VITRO" METHOD FOR CANCER
RESEARCH — THE EXCITATION OF LEUCOCYTES CAUSED
BY CANCER PLASMA — FACTS KNOWN ABOUT CANCER —
THE AGE-INCIDENCE; VITALITY; DEATH; METASTASIS;
CHRONIC IRRITATION — THE POSSIBLE CAUSES OF CELL-
PROLIFERATION DISCUSSED 157

CHAPTER X

EXPERIMENTS WITH NUCLEIN — THE LOWERING OF THE CO-
EFFICIENT OF DIFFUSION CAUSED BY EXTRACTS OF DEAD
H.EMAL GLAND — DIVISIONS INDUCED IN LYMPHOCYTES
FOR THE FIRST TIME — REVELATIONS CONCERNING THESE
DIVISIONS — THE ROLES PLAYED BY THE ALTMANN'S GRA-
NULES, NUCLEI, AND NUCLEOLI IN THE CELL-DIVISION 172

CHAPTER XI

THE DIVISION OF LYMPHOCYTES INDUCED BY THE ANILINE
DYE — THE AUGMENTING ACTION OF ATROPINE AND
EXTRACT OF H<EMAL GLAND — "AUXETICS" — THE CYCLE
OF CELL-DIVISION — THE POSSIBILITIES OF THE INDUCED
CELL-DIVISION BEING DUE TO "DEATH-STRUGGLES"—
ASYMMETRICAL AND REDUCED DIVISIONS . 225

CONTENTS XX111

CHAPTER XII

PAGE

THE ''EXPERIMENTAL TEN MINUTES" — DIVISION INDUCED

IN THE SO-CALLED PoLYNUCLEAR LEUCOCYTES METHOD

FOR COUNTING THE NUMBER OF GRANULES IN EOSINO-
PHILE LEUCOCYTES, AND THE REDUCTION OF THIS NUM-
BER IN THE CELLS OF CANCER PATIENTS 249

CHAPTER XIII

THE AUXETIC ACTION OF CANCER-SERUM — THE INDUCED
DIVISIONS OF GRANULAR RED CELLS — THE AUXETIC
ACTION OF "THE REMAINS OF DEAD TISSUES," AND ITS
AUGMENTATION BY ATROPINE AND THE PRODUCTS OF
PUTREFACTION — THE ISOLATION OF THE AUXETICS
KREATIN AND XANTHIN — DISCOVERY OF CAUSES OF THE
CELL-PROLIFERATION OF HEALING 292

CHAPTER XIV

THE AUXETIC ACTION OF GLOBIN .

CHAPTER XV

THE PROOF THAT THE REMAINS OF DEAD TISSUES AND GLO-
BIN CONTAIN THE CAUSES OF THE CELL-PROLIFERATION

OF HEALING, AND OTHER CELL-REPRODUCTION — EXPERI-
MENTATION "IN VIVO" CONFIRMS "iN-VITRo" OBSERVA-
TIONS— THE CAUSE OF BENIGN TUMOURS 333

CHAPTER XVI

THE AUGMENTED DIVISIONS INDUCED BY PUTREFACTION
OF THE EXTRACTS is DUE TO THE ALKALOIDS OF PUTRE-
FACTION— A THEORY THAT CARCINOMA AND LYMPHA-
DENOMA MAY BE CAUSED BY THE COMBINATION OF THE '
AUXETICS OF CELL-PROLIFERATION AND CHOLINE AND
CADAVERINE — AN EXPLANATION OF THE AGE-INCIDENCE,
METASTASIS, AND THE OTHER FACTS KNOWN CONCERNING
CANCER — THE NECESSITY FOR A CRUCLA.L EXPERIMENT
TO PROVE THE THEORY . 348

XXIV CONTENTS

CHAPTER XVII

PAGE

THE INHIBITORY ACTION OF BLOOD-SERUM IN PREVENTING
THE ACTION OF AUXETICS IN CAUSING CELL-DIVISION—
MEASUREMENT OF THIS ACTION — THE TREATMENT OF
SOME CASES OF CANCER BY DEFIBRINATED BLOOD-
DESCRIPTION OF THE CASES — THE TREATMENT OF A

MALIGNANT ULCER BY MEANS OF GLOBIN — THE CRUCIAL

EXPERIMENT — CONCLUSION 374

APPENDIX I

TABLES DESCRIBING THE ENUMERATION OF THE NUMBER

OF GRANULES CONTAINED IN EOSINOPHILE LEUCOCYTES 401

APPENDIX II

METHOD FOR ESTIMATING THE NUMBER OL LIVING AND DEAD
LEUCOCYTES CONTAINED IN A GIVEN SAMPLE OF BLOOD,
AND MEASURING THE LIVES OF LEUCOCYTES 406

APPENDIX III

A "HANGING-DROP" PREPARATION WITH THE JELLY METHOD 419

APPENDIX IV

A CONTRIBUTION TO THE "THEORY OF IMMUNITY" 420

INDEX . 421

LIST OF ILLUSTRATIONS

ALL the photomicrographs which illustrate this book were takea
with the apparatus described in Chapter II. The objective used for
those taken with the higher magnification was a 2-mm. apochromatic
lens (Zeiss) N. A. 1 '30. The objectives employed for taking photo-
graphs of a lower magnification were Zeiss D and Zeiss A. The eye-
piece used in all of them was a "high-power projection eye-piece'^
(Watson).

In the actual preparations, as observed through the microscope, a
stereoscopic view of the dividing cells can be obtained, which facilitates,
the demonstration of the different phases. Unfortunately, this stereo-
scopic effect cannot be seen in the prints, although an examination of
them with a hand magnifying-glass will remedy the deficiency to some
extent.

The photograghs have all been produced without any alteration of
the original negatives.

FIG. PAGE,

1. A typical field seen by the in-vitro method of staining. The leuco-

cytes are staining gradually Frontispiece.

2. The photomicrographic apparatus. The microscope is ready to

be used for direct observation. The gas-burner can just be seen
at the lower end of the wooden plank. (N. B. — The sheets of
white paper have been placed in this position in this and the next
photograph in order to " show up " the apparatus.) 23-

3. The apparatus ready for photography. The mirror is swung aside,

and the eye-piece attached to the camera is inserted into the
microscope 25-

4. The photomicrographic apparatus. Showing positions of water-

cooling tank and Nernst burner. The microscope mirror is in
position for direct observation 29-

5. The photomicrographic apparatus. The microscope mirror is '

swung aside for photography 31

6. The granules of the leucocyte are gradually becoming stained.

The red cells are unstained. Low power 45-

7. The leucocyte's granules are stained. Its nucleus is unstained.

The pseudopodia are extruded in response to atropine, which is
diffusing into the cell as well as the stain 45-

XXV

XXVI LIST OF ILLUSTRATIONS

FIG. PAGE.

8. The same field as 7. The leucocyte is retracting its pseudopodia. 47

9. The same field as 7 and 8. The retraction of pseudopodia is nearly

complete. The lobes of the nucleus of the leucocyte are turning

a faint blue colour 47

10. A leucocyte excited by atropine. Its granules are deeply stained,

and its nucleus is also beginning to stain a blue colour. Low
power 49

11. A leucocyte which has just been killed by the staining of its nucleus.

Its granules are also deeply stained 49

12. The leucocyte has just died owing to the staining of its nucleus.

The cell-wall is beginning to bulge because the cytoplasm is
liquefying 53

13. The onset of achromasia. The same field as 12. The stain is

beginning to fade from the nucleus. The bulging of the cell-
wall has become general 53

14. Achromasia. The same field as 13. The stain has gone from the

nucleus, although the granules are still stained. Note that the
red cell is disappearing 57

15. Achromasia. The same field as 14. Many of the cell-granules

have lost their stain. The cell-wall is nearly invisible. The
red cell has disappeared 57

16. A stained leucocyte. The ordinary vacuoles (colourless patches

amongst the cell granules) are well shown. The cell has just
died 105

17. Diffusion-vacuoles in a leucocyte 105

18. A dead leucocyte in which diffusion-vacuoles are beginning to

appear 109

19. A diffusion-vacuole in a lymphocyte. Low power 109

20. A diffusion-vacuole in a granular red cell 115

21. A clump of normal blood-platelets. They are resting on a jelly

which will just stain their granules 115

22. Diffusion-vacuoles in blood-platelets. The cells are resting on the

same jelly-film as those in 21, but they had been subjected to
the action of morphine hydrochloride 119

23. Diffusion-vacuoles in blood-platelets. The jelly-film had the

same index of diffusion as that employed in 21 119

24. A specimen of blood which had been mixed with morphia solution.

Note the extreme vacuolation of the leucocyte. A blood-
platelet is also vacuolated. The same jelly as in 21 121

25. Patches resembling archoplasm induced in a leucocyte by sub-

jecting the blood to an extract of a dead tissue. The jelly-film

on which the cells are resting is similar to that employed in 21 . 121

26. An extruded pseudopodium becoming detached from a leucocyte

which is excited by atropine. No stain 125

27. Amosboid movements excited in a blood-platelet by the action of

atropine . 125

LIST OF ILLUSTRATIONS XXV11

FIG. PAGE.

28. Amoeboid movements excited in a leucocyte by the action of

atropine. Low power 135

29. Exaggerated amceboid movements in leucocytes which have their

granules stained. The movements were excited by atropine
sulphate 135

30. Excited leucocytes extruding their pseudopodia between red cells. 137

31. Excitation of amceboid movements in a4ymphocyte by the action

of atropine. No stain 137

32. Excitation of amoeboid movements in a lymphocyte which has its

granules stained 143

33. Extreme excitation of amceboid movements in a lymphocyte.

No stain 143

34. Excitation of two leucocytes by the action of choline. Low

power. No stain . 151

35. Excitation of a lymphocyte by the action of choline. No stain . 151

36. Excitation of amceboid movements in a leucocyte by the action of

cadaverine. No stain .153

37. A leucocyte excited by morphine. The cell's granules are stained . 153

38. Leucocytes excited by pyridine. No stain 173

39. A lymphocyte which has absorbed stain and atropine discarding

its granules (flagellation) 173

40. A resting lymphocyte. Note the deeply stained masses of gran-

ules in the cytoplasm, which is bulged out in places. The large
transparent nucleus and the stained ring-shaped nucleolus can
also be seen . 189

41. A resting lymphocyte. The Altmann's granules in the cytoplasm

are stained 189

42. A resting lymphocyte. The cytoplasm, the granules, the nucleus,

and the nucleolus can be distinguished 191

43. The earliest stage of mitosis. The nucleolus has divided into two

rings 191

44. Early mitosis in a lymphocyte. Looking down through the spindle

(polar aspect) . The nucleolus has divided into two centrosomes,
each of which is ring-shaped. The spindle is surrounded by a
belt of chromatin granules 193

45. Mitosis in a lymphocyte. Profile aspect. The two ring-shaped

centrosomes can just be seen towards the poles. The granules
are becoming formed into chromosomes 193

46. Foreshortened appearance of a mitotic figure in a lymphocyte.

The position of one nucleolus-centrosome at the pole of the figure "
is well shown 195

47. Profile aspect of mitosis in a lymphocyte. The relative positions

of the centrosomes and chromosomes can be seen ...... 195

48. Profile aspect of mitosis. The belt of chromatin is formed round

the waist of the cell . .197

XXV111 LIST OF ILLUSTRATIONS

FIG. PAGE.

49. One resting and one dividing lymphocyte. In the latter the

chromosomes are beginning to divide. The centrosomes appear

as dots of chromatin 197

50. Polar aspect. The belt of chromatin granules is dividing into

chromosomes 199

51. Polar aspect. The chromosomes are becoming semicircular . . 199

52. Polar aspect. An "aster" stage of mitosis in a lymphocyte . . 201

53. Polar aspect. Some of the chromosomes are semicircular-shaped;

some are dots of chromatin 201*

54. Polar aspect. One centrosome can be seen at the pole of the

"aster" figure 203

55. Polar aspect. Sixteen chromosomes could be counted in this cell . 203

56. Profile aspect, of mitosis in a lymphocyte 205

57. Profile aspect of mitosis in a lymphocyte 205

58. Profile aspect. The chromosomes can be seen at the waist of the

spindle 207

59. Profile aspect. A figure frequently seen 207

€0. Profile aspect of mitosis 209

61. Oblique aspect of mitosis in a lymphocyte 209

62. Polar aspect of mitosis in a large lymphocyte from a patient

suffering from carcinoma. There are sixteen chromosomes . . 211

63. Polar aspect. The chromosomes were V-shaped with their apices

inward to be attached to the nucleus-spindle, which can dimly

be made out 211

64. Polar aspect of mitosis in a large lymphocyte from a cancer

patient. The chromosomes are dividing 213

65. Profile aspect of mitosis 213

66. Profile aspect. The figure is fully formed. One nucleolus-ceiitro-

some is ring-shaped; the other is a dot of chromatin 215

67. Profile aspect. The sixteen chromosomes could be counted . . 215

68. The cell has become constricted in its centre 217

69. Profile aspect. Complete division is about to occur. The chro-

mosomes are being reconverted into granules, but the mitotic
figure is not quite finished at the dividing-point 217

70. Profile aspect. The spindle and chromosomes have divided, but

the cell-wall has not yet separated 219

71. Completion of mitosis in a lymphocyte 219

72. Asymmetrical mitosis in a lymphocyte induced by azur stain aug-

mented by atropine 233

73. Asymmetrical mitosis induced by azur stain augmented by

atropine 233

74. An early stage of delayed mitosis induced by a jelly with a low

index of diffusion. The number of chromosomes is more than
sixteen 241

75. Thirty-two chromosomes could be counted in this cell. Early

mitosis delayed 241

LIST OF ILLUSTRATIONS XXIX

FIG. PAGE.

76. A resting polymorphonuclear leucocyte. Its granules are stained

but not its nucleus. The cell was alive 253

77. A basophile leucocyte in the act of cell-division. The granules

of the cell are in the centre. The lobes of the nucleus are at the
poles of the cell which is dividing into three 253

78. An eosinophile leucocyte in the earliest stage of division. The

granules were arranged in lines radiating outwards from the
centre of the cell. The lobes of the nucleus were at the poles . . 259

79. Early stage of division of a neutrophile leucocyte 259

80. A dividing leucocyte 261

81. A dividing leucocyte 261

8lA. Division of a leucocyte. The linear arrangement of the granules

could be well seen 263

82. A dividing leucocyte • . 263

83. A dividing leucocyte 265

84. A dividing leucocyte 265

85. A dividing leucocyte 267

86. A dividing leucocyte 267

87. An eosinophile leucocyte with its granules stained . 275

88. A field containing a neutrophile, an eosinophile, and a basophile

leucocyte. The upper cell is the neutrophile and the lower one
the basophile cell. All the cells are ruptured, but their granules
are stained 279

89. A basophile leucocyte whose stained granules have been turned

black by heat 279

90. One of the negatives of a ruptured eosinophile leucocyte (negative

No. 52) 285

91. One of the negatives of a ruptured eosinophile leucocyte (negative

No. 54) 285

92. Counting the granules. The image of the ruptured cell depicted

on negative No. 52 is projected on to a sheet of white paper
pinned on to a screen 287

93. Counting the granules of negative No. 54 287

94. A dividing red cell from a cancer patient 295

95. A dividing red cell from a cancer patient. The granules seem to

be arranged in an indefinite figure 295

96. Very early stage of mitosis in a lymphocyte induced by decom-

posed extract of suprarenal gland. No stain 301

97. Mitosis of a lymphocyte induced by decomposed suprarenal

extract. No stain 301

98. Mitosis induced in a lymphocyte by decomposed extract. No

stain 303

99. Asymmetrical division induced by decomposed extract. No

stain or atropine is present 303

100. Mitosis induced by fresh extract of suprarenal gland. No stain

or augmentor present 307

XXX LIST OF ILLUSTRATIONS

FIG. PAGE

101. Mitosis induced by fresh suprarenal extract. Xo stain is present . 307

102. A dividing polymorphonuclear leucocyte induced by suprarenal

extract alone. Xo stain 311

103. Mitosis induced in a lymphocyte by suprarenal extract which had

purposely been allowed to become putrid. Xo stain . . . .311

104. Mitosis induced in a lymphocyte by suprarenal extract which had

purposely been allowed to become putrid. Xo stain .... 313

105. Asymmetrical mitosis induced by suprarenal extract augmented

by atropine. Xo stain 313

106. Mitosis induced in a lymphocyte by kreatin. No stain or extract . 317

107. Division in a leucocyte induced by kreatin. Xo stain or extract . 317

108. Mitosis in a lymphocyte induced by globin augmented by atropine.

No stain, extract, or kreatin 327

109. Asymmetrical mitosis induced by globin augmented by atropine.

No stain, extract, or kreatin 327

110. Mitosis induced in a lymphocyte by means of decomposed globin

solution. Xo stain, extract, kreatin, atropine 329

111. To show the way in which globin is " dotted " over the sufrace of an

ulcer 343

112. To show the scab formed by the application of globin to an ulcer. 343

113. Mitosis induced by a mixture of kreatin and choline. No stain,

extract, or atropine 353

114. Asymmetrical mitosis induced in a lymphocyte by a mixture of

suprarenal extract and globin, augmented by choline. No
stain or atropine 353

115. Mitosis in a lymphocyte induced by globin and choline. XTo stain

or other auxetic 355

116. Mitosis induced in a lymphocyte by suprarenal extract and choline.

No stain or other auxetic 355

117. Mitosis induced in an epithelial cell by a mixture of stain and

extract 357

118. Early mitosis in an epithelial cell from the vagina induced by stain

and extract 357

119. Section from the case of scirrhus of the breast. Low power . . 385

120. The same as 119. High power 385

121. To show the way in which globin was " dotted " on to a portion of

the malignant ulcer 387

122. Section of a portion of the ulcer after treatment. Low power . .391

123. The same as 122. High power 391

124. Section of the treated portion of the ulcer after the application of

globin augmented by choline, showing reinfiltration. Low
power 393

125. The same as 124. High power 393

Induced Cell- Reproduction
and Cancer

CHAPTER I

THE SCOPE OF THE NEW METHOD

THE study of the individual living human cell and of
the effects of chemical reagents upon it marks what
may almost be regarded as a new scientific departure.
Although much has been written concerning the
passage of substances into cells, mainly the outcome
of experiments not made actually upon the individual
cells themselves, and certainly not while they w^ere alive,
little practical work has been done with reference to
the behaviour of individual cells while substances are
being made to pass into them. This has been owing
to the lack of satisfactory methods, and because the
laws which govern the diffusion of substances into the
individual living cells have not been recognised. These
laws are of the greatest importance, and must be
thoroughly understood if in-vitro experimentation is
to prove serviceable or successful, and later on a
section will be devoted to this subject. In the mean-
time the elemental fact must be simply stated that
living cells are examined by placing them, under a
cover-glass, on to the surface of a film of jelly, which
may contain dissolved in it any substance we may wish

2 THE SCOPE OF THE NEW METHOD

to experiment with, and which has, while in a molten
condition, been poured on to a microscope slide and
allowed to set there. The jelly may, for instance,
contain an aniline dye; and by watching the way in
which the living cells absorb the stain from the jelly,
and by experimentation with it, many of the laws of
the diffusion of substances into living cells have been
ascertained; and by the application of these laws we
can now add other substances to the jelly and make
them also diffuse into the living cells, and watch the
results by means of the microscope. The cells are
pressed into the jelly by the cover-glass, and therefore
they can absorb only what is in the jelly (there is
nothing else for them to take), provided that the
conditions have been correctly arranged for the pass-
age of the substances from the jelly into the cells. It
is essential to note that one class of cells differs from
another with reference to the rate at which they absorb
materials from the media in which they are placed, so
that the composition of any given jelly must be cor-
rectly arranged for experimentation with any particular
class of cell with which it may be desired to work.

The word "cell" in this book refers to the living
cell unless otherwise specified. Cells must always be
freshly removed from the body when they are placed
on the jelly. It occasionally happens that the cells
may have just died or be dying when they are ex-
amined, as when mitotic divisions are being induced
by azur stain, as will presently be described; but,
generally speaking, after the cells in a specimen are
dead the specimen is thrown away. It is obvious

ADVANTAGES OF THE IN- VITRO METHOD 3

that "specimens" of living cells cannot be kept. All
attempts to "fix" the jelly films (on which the cells
are resting) at the end of the experiments have so far
failed, so it is impossible to retain the specimens for
future examination or for purposes of collection; and
consequently when dead, or when finished with, speci-
mens have to be discarded. This circumstance has
led me, at the suggestion of Professor Sherrington, to
devise a rapid method of recording the actual experi-
mental facts observed by means of photomicrography;
and although the photographs, many of them taken with
the highest powers of the microscope, are not com-
parable by any means to what is seen with the eye,
we at least have the satisfaction of knowing that a
truthful image is recorded which cannot be influenced
in the way that drawings, however carefully made, are
apt to be. The photomicrograph is therefore the best
substitute for microscope "specimens" which we have
to offer.

In the past very little has been learned from the
study of individual living cells either in physiology or
in pathology. Presumably this has been due to the
fact that it has been difficult to stain cells satisfactorily
when they are alive; for, since the discovery of the
aniline dyes, stains have been used in nearly all micro-
scopical work. It is true that a good deal of work
has been done in the way of attempting to stain unfixed
cells by mixing them with solutions of methylene blue
and neutral red; but the results have not been very
satisfactory, and no doubt the advances made in the
study of dead cells by means of differential staining

4 THE SCOPE OF THE NEW METHOD

with dyes dissolved in alcohol have done something to
retard in-vitro methods of investigation, because dyes
dissolved in alcohol cannot, of course, be used to stain
living cells. As a matter of fact, with this new "jelly"
method it is simpler to stain certain living cells than it
is to stain dead ones by the old methods, and better
pictures are obtained although less skill is required.
No matter how rapidly a cell or tissue is killed, the
fact remains that it is dead, and the means usually
taken to prepare it for examination by placing it in
preservative solutions or in others necessary for fixing
and staining it — not to speak of the processes of em-
bedding and freezing and the subsequent cutting with
razors and so forth — can only add to the fallacious
results of its examination. So far as blood-cells are
concerned, the study of their morphology and cytology
has hitherto been almost entirely based on the exami-
nation of dead specimens, with the result that some
erroneous impressions, both as to form and func-
tion, have become generally accepted. For instance,
let an experienced worker with the older methods
look for a "hyaline leucocyte" with the new one, and
he will marvel at his credulity. The hyaline leuco-
cyte is a dead lymphocyte which has become achro-
matic. By the new method we see cells stained while
they are alive, and admirably spread out on the jellies,
so that they can readily be examined by the highest
powers. One can now cause any soluble substance to
diffuse into them at any rate one pleases, and with the
help of this knowledge one can, by specific chemical
agents, cause leucocytes and other cells to divide on

IT REVEALS FALLACIES 5

the microscope slide. By this study of vital activity
new lessons have been learned concerning the real
functions of the morphological elements of the cells.
For instance, owing to the fact that the older methods
merely showed pictures of dead cells and the arrange-
ment of their component parts after they are dead,
controversies have arisen regarding the functions of
the cellular elements. Unfortunately, theories regard-
ing these functions have sometimes become accepted
as facts. The "lobes of the nuclei" of leucocytes are
generally recognized as being analogous to the nuclei
of other cells, in spite of the fact that the act of cell-
division has never been seen in leucocytes.1 The very t
designation of the cells — "polymorphonuclear" — is
even based on this theory; but in reality the "lobes
of the nuclei" are the centrosomes. We hear it said
even now that the blood-platelet is a precipitate,
although a single glance at a specimen in vitro, espe-
cially if an alkaloid is present in the jelly, demonstrates
beyond denial that a blood-platelet is a living creature
arid a highly amoeboid cell.

The new method reveals new points in every direc-
tion which are difficult to reconcile with the old
theories based upon the examination of dead specimens,
some of them so firmly rooted that people may be slow
to discard them.

Infinite interest and variety awaits the investigator
of cells by this new method. He is dealing with living

1 Throughout this book the word "leucocyte" refers to the polymorpho-
nuclear cell; the mononuclear cell from the peripheral circulation is called a
lymphocyte.

6 THE SCOPE OF THE NEW METHOD

creatures which are amenable and can be excited or
made to divide almost at will. It is remarkable to
think that one can order samples of one's own or some
other person's white blood-corpuscles to reproduce
themselves at a given time, and that if they are properly
treated they will do so with obedient regularity.
Instead of the diagrammatic representations of karyo-
kinesis, from which every student learns his impressions
of cell-division, one is now able to appreciate mitosis
in its reality and to watch it through its various phases.
This is a very striking fact, but its interest grows when
we consider another very important lesson derived from
it, insomuch that, as will be seen later, there is strong
evidence that white corpuscles will multiply only in
response to a specific chemical agent. We now believe
that it is essential for a leucocyte to absorb an "aux-
etic" (<lv£ ^TI/COS, an exciter of reproduction) before it will
make any attempt to proliferate, and we have also
evidence that it is more than probable that other
human cells, and possibly all of them, proliferate in
response to a similar agency. It will be realized, there-
fore, that this method of study of the cell and of the
influences of chemical agencies upon it has opened
up a new field of work, not only in pathology, but in
physiology also.

The proliferation of cells is the main theme of this
book. By this in-vitro method it has not only been learnt
that cells will divide in response to certain chemical
agents, but that these agents exist in the remains of
all dead tissues. Two of the substances which are
directly responsible for cell-reproduction within the

IT ELUCIDATES CELL-DIVISION 7

body have been isolated in crystalline form : they are
the extractives, kreatin and xanthin, and individual
cells divide in response to them according to the
amount of each substance absorbed by the cell.

It will be shown that cell-proliferation depends upon
cell-death, and this affords an explanation of the cause
and origin of benign tumours. "Development" is a
basis of physiology; and since the multiplication of
human cells is due to chemical agents, as is shown by
this method, one cannot but suppose that the facts
learnt may lead to the explanation of points connected
with the growth of the embryo.

One of the foundations of pathology is the phenome-
non of "healing," which is caused primarily by the
proliferation of certain cells. The causes of this
proliferation have been ascertained for the first time
by this method, and the ultimate chapters of this
book will describe proofs that these causes are now
known. If the finger is cut, or if disease gains a
footing in any part of the body, an attempt is made
by the tissue-cells to proliferate and to heal the injury;
but up to now no one has known why this proliferation
took place or how it was caused. This mystery is now
elucidated: The knowledge that cell-proliferation in
the body is due to chemical exciters, of reproduction
(auxetics) is, we think, the beginning of an innovation
which must lead to developments of practical value.

The effect of any given substance, so long as it is
soluble, can be tested on many individual human cells
and the results watched. I fear that we, personally,
have only been able so far to try the effects of auxetics,

8 THE SCOPE OF THE NEW METHOD

alkaloids, and a few other substances ; but a whole field
of investigation of the actions of substances on indi-
vidual cells remains to be carried out, and this is now
possible by this "jelly" method of in-vitro staining.

The action of chemical substances on living cells
is closely associated with the diffusion of substances
into these cells (a subject to which a section of this
book will be devoted), and this diffusion is governed
by the "coefficient of diffusion" of the cells them-
selves, a phenomenon which has been so far entirely
studied by this in-vitro method. Up to the present,
however, we have only had time to ascertain the com-
parative rates of diffusion of substances into some of
the classes of human cells and into a few species of
bacteria. The determination of the coefficients of
diffusion of all the rest of the cells of the whole ani-
mal and vegetable kingdoms remains as a "legacy"
for those who will undertake the work.

Methods will be described by which the lengths of
the lives of leucocytes can be measured after they
have been removed from the body. By this means
the comparative effects of different poisons on the
cells can be tested, and the small amount of work
done in this direction will be summarised. We think
that there are possibilities that farther investigation
of the actions of specific poisons, such as bacterial
toxins, will lead to fruitful results; in fact, one of us
(C. J. M.) has already shown by this method that
chorea and rheumatism are less closely related than
is generally supposed.1

1 British MedicalJournal, August 23, 1909.

SHOWS THE CELL-STRUCTURE 9

In the last chapter experimental evidence is given
to prove that blood-serum has an inhibitory action on
cell-division; and it will also be seen that it is pos-
sible to measure this inhibitory action. Since the
cell-proliferation of healing is * caused by chemical
substances contained in the soluble remains of dead
tissues, and since, as will be shown, bacteria decom-
pose these solutions, a field of research is opened for
the investigation of this decomposing action by various
pathogenic bacteria; for in decomposing the sources
of the causes of healing they greatly modify that pro-
cess, and the healing process must play an important
part in immunity against disease. Further, bacteria
may have an action on the substances contained in
blood-serum which restrain cell-division. We fear that
we have hitherto been able to do little towards the
investigation of this factor in the problem of immunity,
which is now mentioned for the first time.

These are only a few of the fields which have been
pried into by experimentation with this new method.
It has been impossible for us to investigate all the
paths of research which have been opened up, and
prospective workers may be assured, from our own
personal experience, that research with stained living
cells will amply repay the time and patience expended
on it.

For the examination of the arrangements of the
cells in living tissues we have not, so far, been able
to make this in-vitro method so useful as is the older
method of examining sections of dead tissues, but we
think that improvements may be possible. For blood-

10 THE SCOPE OF THE NEW METHOD

examination, on the other hand, it takes one into a
different realm compared with the older methods.
Examined by the older methods, a cell appeared
usually as a flattened, stained diagram; by the new
one it appears as a sphere. The difference is com-
parable to that which exists between an old Japanese
print in which there is no perspective and a perfect photo-
graph seen through a stereoscope. By the older meth-
ods, for instance, the nucleus of a lymphocyte appears
as a flattened, homogeneously stained mass, or perhaps
the stained chromatin resembles a "spireme" within
the nucleus; by the new method it is seen at a glance
that the nucleus in the living cell is a round, trans-
parent ball, studded on its outside by minute chromatin
granules. There is no doubt that the observation of
the living cell is a new study. In almost every slide
one sees something of interest which has not been
seen before. Living cells seem to have small points
of individuality which can only be seen when they are
stained alive.

Take for example the phenomena of cell-division.
The rnitotic divisions, although the same in general
principles (unless of course we take steps to induce
asymmetrical divisions by an alkaloid) are almost
always slightly different, depending to some extent
upon the stage of division reached, and upon the
attitude in which the cell happens to be presented to
the observer.

By this means of cytological study we may frankly
say that we cannot tell what revelations may turn up
at any time. This book will record a few of them, but

APPLICATION TO CANCER RESEARCH 11

there are doubtless many more to come. The feeling of
astonishment may be imagined when one of us for the
first time — and the cells have been discovered for more
than a century — saw most of the polynuclear leucocytes
in the specimen in the act of divisron. It was expected,
it is true; but the way in which these cells divide was
by no means expected.

We have carefully searched the literature relating to
our subject, without discovering points which have
helped us. Most of the literature is devoted to de-
scriptions of morphology which are not of much
assistance in this kind of experimental work. There
is no literature dealing with the effects of chemical
substances on stained, individual, living human cells,
and if a point is to be unravelled we have found it
better to make experiments for its solution rather
than to depend upon any literature dealing with the
observation of dead cells.

The new investigator will have to begin at the
beginning, which is not far off, and he will have to
do so with an open mind.

The foregoing points indicate briefly the scope of
this book descriptive of the new methods, and of the
paths of research which have been opened by them.
But we shall also describe in detail the main path
which we have followed — namely, the adoption of the
methods for the elucidation of the cause of cancer.
It must be obvious that since we can now induce
proliferation in human cells, and since the proliferation
of certain human cells is the fundamental condition
which characterises cancer (for that is what it is), we

12 THE SCOPE OF THE NEW METHOD

can, by investigating the chemical cause of prolifera-
tion, throw considerable light on the cause of cancer.
Cancer is essentially a growth caused by excessive
cell-proliferation, and the new methods are the only
ones which have given us the power to induce an
individual cell to reproduce itself.

As will be seen later, we can say more than this,
for we can induce by certain specific chemical agents
those remarkable asymmetrical mitotic divisions in
human cells which are characteristic of many of the
divisions which occur during malignant proliferation.
The latter part of this book will therefore relate to
Cancer Research.

Before closing this chapter, two other points must
be mentioned. The usual cytological phraseology has
been found to be difficult to apply to many of the facts
seen by the new methods. For instance, the word
"nucleus" has a very vague meaning, and yet every one
uses it. It arose, we believe, from the examination of
cells with the lower powers of the microscope, which
are commonly employed in the study of "pathological
specimens." The nucleus of a cell, studied from this
aspect, is merely a deeply stained body within the cell;
but in reality the nucleus is composed of several dif-
ferent parts, each of which has a separate function
during cell-division. The body which appears as the
nucleus in some cells has a very different function to
that which appears as the nucleus in others. For
instance, the body which appears as the nucleus of a
lymphocyte under low magnification forms the spindle;
whereas what are usually described as the nuclei of_

CYTOLOGICAL DEFINITIONS 13

leucocytes are their centrosomes. The so-called nuclei
of leucocytes ought, we think, in reality, always to be
called the centrosomes, and the word "nucleus" deleted
from their morphology. We have done our best to
retain the usual cytological terms in the senses in which
they are usually employed; but we must ask some
indulgence when referring to those cells in which
divisions have been seen for the first time, and in which
these divisions differ very materially from those which
occur in other types of cells. Again, we use the defini-
tion "amoeboid" for the exaggerated movements
exhibited by cells under the influence of alkaloids,
but it must be understood that these movements
differ from the blunt and sedate amoeboid move-
ments which are commonly seen—that is to say,
they are far more exaggerated and are absolutely
characteristic.

We think that, from the persistent examination of
dead structures, cytology has been rather led away
into a maze from which it will be difficult to extricate
it; and it is possible that pathology may have to be
modified in some of its points now that we know
a great deal more regarding the causes of the prolifera-
tion of cells. *

The last point to which attention must be directed
is, that one ought to be careful how attempts are made
to demonstrate new facts observed by this method to
other people. If the specimen is actually under the
microscope, and other people are present, then, of
course, a few persons can see the new fact. But these
living cells never last long, and many has the occasion

14 THE SCOPE OF THE NEW METHOD

been that a few persons have seen, say a beautiful
mitotic figure, when suddenly a later arrival at the
microscope says that he can see nothing, and on exam-
ination it has been found that the figure has completely
vanished owing to the onset of achromasia. If other
people wish to see any experiment, two or three should
await beside the microscope; but they may have to wait
a long time before a typical specimen is found, for, as
has been pointed out, cells rarely present exactly the
same appearances every time. It is of common occur-
rence that on one day perfect specimens continually
present themselves, but on the next every cell appears
to be distorted, or always in the wrong position. For
this reason we have found it better to take photomicro-
graphs and convert them into lantern slides rather than
attempt demonstrations to many people.

It is right to mention that this method requires the
expenditure of patience and time on the part of the
investigator. One cannot attain good results in a few
minutes, but if some time is devoted to it the value of
this in-vitro method will be appreciated.

CHAPTER II

THE GENERAL PRINCIPLES OF THE METHOD— THE

APPARATUS REQUIRED THE SPECIAL PHOTOMICRO-

GRAPHIC APPARATUS— THE REVOLVING APPARATUS.

THIS method by which cells are observed in vitro is
very simple. They are placed on a film of agar jelly,
which holds in solution any material with which we
may wrish to experiment. To prepare the film, a drop
of molten jelly is poured on to a slide, which is then
laid on a level surface until the jelly sets firmly. A
drop of the citrate solution in which, say, blood-cells
are suspended is then placed upon a cover-glass, which
is inverted and allowed to fall flat on the film. It
might be thought that the weight of the cover-glass
would be sufficient to kill the cells; but they sink into
the jelly to' some extent, and so become protected.
Before this happens, however, they spread out centri-
fugally from the centre to the periphery of the cover-
glass, and if a drop of blood be examined in this way
on stain-containing jelly they may be seen by the
naked eye rushing in every direction towards the edges
of the cover-glass. When this movement has ceased,
if the slide is held up between the observer and the

16 THE GENERAL PRINCIPLES OF THE METHOD

window it will be seen that the surface of the jelly
over which the cells have passed is studded with
corpuscles.

If the jelly has been properly made the slide may
be handled freely. It may be tilted to any angle, and
even turned upside down without the cover-glass sliding
off or the jelly becoming displaced. This is a fortunate
fact, because it enables the microscope to be placed
at any convenient angle for examination of the slide
or for purposes of photography. If the specimen is
quickly focused under the microscope while the
spreading-out process of the cells is going on, using
a ^-inch objective and, say, a No. 4 eye-piece, the
picture presented is a very remarkable one. The cells
will be seen rushing along in a direction from the
centre of the cover-glass towards its margin; they
tumble over each other, leucocytes and red cells,
lymphocytes and blood-platelets, bumping into each
other and apparently all striving to reach some imagin-
ary goal. Gradually the flowT becomes slower and
slower, the cells cease to "barge" into each other so
fiercely, they squeeze past one another, and it will be
realized what a marvellous power blood-corpuscles
have of accommodating their shapes to almost any
requirements.

Leucocytes and red cells all behave in the same
way. They allowr themselves to be squeezed through
gaps between other cells, which appear to be so small
that if it had not actually been seen one never would
believe it. As the flow becomes slower it will be
seen that suddenly a passing leucocyte goes "ashore";

PREPARATION OF THE SPECIMEN 17

its course is arrested because it has adhered to the
jelly, or between the jelly and the cover-glass. Some-
times the rest may be only momentary, when the cell
may be seen to revolve on its own axis for a few mo-
ments, and then pass on again in the slowing stream.
Leucocyte after leucocyte afterwards becomes arrested
in this way; they apparently stop first because they
are larger and more "sticky." Then the red cells
gradually stop, until at last the field is dotted with
living blood-corpuscles, which may happen to become
arranged in groups or rest singly side by side.

The specimen may now be moved about by means
of the mechanical stage, when it will be seen that all
the cells in the film of blood under the cover-glass
have become arranged in a manner very suitable for
examination. The frontispiece of this book is a
photomicrograph of a typical field presented by this
method.

The living cells all come to rest in a short time, and
each one has its own share of jelly-surface, from which
it has no alternative but to absorb any substances
which have been previously dissolved in the jelly.
Having focused a field, therefore, which contains an
example of the cell writh which one wishes to experi-
ment, it is only necessary to wait until that cell has
sufficiently absorbed the contents of the jelly for it to
respond to the agent which has been dissolved in it.

"Artefacts" do not exist; the surface of the jelly
is the same all over. One has no control over the
attitude which a cell may adopt, no matter what
part of the jelly-surface it may come to rest upon, nor

18 THE GENERAL PRINCIPLES OF THE METHOD

over the other cells which form its immediate sur-
roundings. The cells are always placed on the jelly
in identically the same way as has just been described,
and therefore the only way in which one can in-
tentionally affect the individual cells is either by
deliberately (1) mixing some other substance with the
jelly before it is set on the slide, or (2) by keeping the
slide at various temperatures. It sometimes happens
that unintentionally the cells may become distorted
by the presence in their neighbourhood of some for-
eign substance which has been accidentally mixed with
them in the citrate solution in which they have been
suspended prior to being placed upon the film; but
such a foreign body may easily be recognized.

The apparatus required for these researches is not
very elaborate. Many of the earlier experiments were
made in a cabin in a battleship, where there is not
much room for scientific apparatus, but we simply
enumerate them here for the benefit of those who may
desire to commence the study of in-vitro methods for
the first time. They consist of:

1. Microscope slides.

2. Cover-glasses. These should be very thin and
| of an inch in diameter. A few larger ones, say f of
an inch, may occasionally be needed. A silk hand-
kerchief is required to polish the cover-glasses, which
should be very clean and kept in alcohol.

3. Capillary glass tubes. These are constantly in
use, and it is well to begin with a stock of 100 of them.
They should be about 4 inches long, having an internal

GENERAL APPARATUS 19

diameter of 2 millimetres, and should be kept in water
which has been sterilised.

4. A watch-maker's file for removing the sealed
ends of the capillary tubes.

5. Hair-lip pins are most convenient for pricking
the finger or the ear to obtain the blood.

6. Two or three needles in handles for teasing out
tissues, etc.

7. Pipettes; several 1-cc. pipettes, graduated in lOths
and lOOths; a graduated 10-cc. pipette, and one or
two ungraduated of 5-cc., 3-cc. and 2-cc.-capacity.

8. Two beakers. These are used for boiling water
in. The jellies are melted and made liquid by im-
mersing the test-tubes containing them in water which
is boiling in the beakers.

9. Tripod stand and gauze cover.

10. A Bunsen burner or good spirit-lamp.

11. A 100-cc. graduated measure.

12. Two small flasks.

13. Some glass funnels and filter paper.

14. A selection of test-tubes.

15. A centrifuge.

16. An ordinary chemical Centigrade thermometer
for recording the room temperature.

17. A good incubator, which should maintain a
temperature of 37° Centigrade, i.e. the temperature of
the blood. Hearson's is a very good one, but any of
the ordinary water-jacketed types will do. An auto-
matic thermostat is a convenience.

18. The microscope is the most important part of
the outfit, and it should be a good one.

20 THE GENERAL PRINCIPLES OF THE METHOD

This work consists largely of cytology, requiring
accurate observation as to details, and the highest
powers of magnification. Any good microscope stand
will do, but we think that the English tripod one
is the best, especially if the special photomicrographic
apparatus is adopted, in which case it is almost es-
sential. The larger and heavier the stand the better.
It must have a mechanical stage, which should be built
with the instrument; not an "attachable" one. The
lenses must give good definition. Two objectives
only are necessary — a sixth-inch, and an immersion
twelfth. We use equivalents of these in a Zeiss D,
and a Zeiss 2-mm. apochromatic lens, which is com-
pensated for the long-draw-tube of 250 mm., and which
has a numerical aperture of 1 . 30. There is no doubt
that an apochromatic objective for this work is vastly
superior to an ordinary twelfth-inch lens, especially if
photography is to be used.

The eye-pieces we employ are the No. 4 and No. 8
Zeiss compensated ones, and these, or their equiva-
lents, will be found most useful.

The light should always be artificial; daylight is
not suitable for this method. We have found that the
inverted incandescent gas-burner gives the best light
for ordinary work, or if electricity is preferred, the
1 -ampere Nernst lamp is most suitable. If neither
gas nor electricity are available, the spirit-lamps which
give a light by heating an inverted mantle have proved
most suitable in our hands. No matter which light is
used, it is better always to use the same, in order that
contrasts may be detected readily.

SPECIAL APPARATUS 21

It is well to remember that with this method one
cannot afford to waste much time in manipulating the
adjustments of the microscope. The cells, under some
conditions, die quickly, and we therefore have to search
the specimen very rapidly before "achromasia" occurs,
when all the cells vanish, as will be presently described.
It is better, therefore, to have everything ready before
the specimen is prepared.

The microscope should be fitted with a nose-piece,
so that the objective can be changed quickly. When
using the immersion lens, great care must be exercised
in placing the drop of cedar oil on to the cover-glass,
for the cells and jelly-films are easily destroyed if it is
accidentally touched with the solid oiler. There is
neither time nor necessity to reverse the mirror from
concave to plane when the objective is being changed
from a dry to an immersion one. When searching
through the specimens of living cells, rapidity of
focusing will be found to be of more value than too
much attention to accurate microscopy, which is
difficult, if not impossible, to adhere to with this
method. The focusing of the substage condenser on
the specimen cannot be very accurate. Most micro-
scopes are adjusted for slides of a certain thickness,
but we have to place a comparatively thick film of
jelly on top of the slide, and hence the objective
is always farther away from the condenser than" it
ought to be.

The photomicrographic apparatus (figs. 2-5) in-
vented for this method has been designed so that a
photograph can be obtained quickly of any field in

22 THE GENERAL PRINCIPLES OF THE METHOD

a specimen without disturbing either the microscope
or the specimen. Having obtained the negative, the
camera is removed in a moment and the examination
of the particular cell or specimen under observation
can be immediately proceeded with in the usual way.
The old forms of cameras which necessitated the moving
of the microscope or the specimen are not useful for
recording specimens of living cells. An instrument
is required capable of being immediately connected
with the microscope as it stands, so that two or three
records of the same cell may be taken before it dies or
becomes achromatic and vanishes. It is necessary to
use a powerful light, and the light itself will kill the
cells if they are exposed to it for very long. For this
reason we employ a powerful light for the photography
and another for the eye work, but each of them fixed
and capable of being used independently of one another.
The inverted gas-burner above referred to, being placed
at a distance of two feet above the mirror, gives a soft,
indirect illuminant for ordinary work, the other being
a powerful electric Nernst burner, which is placed
behind (that is, underneath) the mirror. When a
photograph is to be taken the mirror is swung aside,
and the light from the Nernst lamp replaces that from
the gas one.

The microscope is fixed on the bench and tilted
at an angle of about 45° from the vertical. All
the microscopes which we use are bolted perma-
nently on to the bench, and they can only be
moved with the aid of a screwdriver. The instru-
ments are not placed vertically, but are tilted at

SPECIAL APPARATUS

Oi'

OQ -^- .„

3 -g

•S°l "

-°T3 &

O C 93

*" v'£

>1 ;-, +i ^

^ S a S3
S3 ^-S5
°~ rt

.22 o> § a

QJ'S^ &

S-+. a^

o r: _.

o « fl a;

03 <K _S

/*\ ^H (11 "-1— t

a
o^o

tn K BJ

9> « O • Ei

£43J3 ^

«HHS
S 1^

3d«fc
o o -^

'MSs

t> B

i -a
•5^*

j|.g

LJ CC M J5

H-S-c^

^T)
-u "Q. fl
O ^ cS

• a; /-<

«.fa S.22

r"-OT3^3

O **

tn O

O >_«

SPECIAL APPARATUS

bC o
G 02
3 P

o ca

.G °

-o~c

(D 43

SPECIAL APPARATUS 27

an angle, because this is most convenient for comfortable
use. This last point is most important, for one may
have to spend hours searching through films with this
method, and it is most wearying to have to work in an
uncomfortable position.

Behind the mirror, and standing a little way back
from it, there is a Nelson's aplanatic condenser (Wat-
son) with iris diaphragm, and immediately behind
this again is fixed a rectangular all-glass water-tank.
This small tank has an outlet pipe above, and an inlet
pipe below, connected by means of rubber tubes with
a sink and a cold-water supply respectively. The water
is kept circulating through this tank when the apparatus
is in use. Lastly, behind the tank is the burner of
a 1-ampere Nernst lamp.

Above the microscope and set at an angle corre-
sponding to its tilt a rigid wooden board is arranged,
being fixed to the ceiling above and, by means of a pair
of legs on either side of the microscope to the bench
below. The board, which is about ten inches in width,
by seven-eighths of an inch thick, has a slot cut into it
in which a box camera can easily slide up and down and
be capable of being fixed at any point by means of a
screw clamp. The camera is fitted with a shutter
(instantaneous and time exposures) the aperture of
which is connected with a "high-power projection eye-
piece" (Watson) by means of a flexible velvet collar.

The Nernst burner, the cooling tank, the two
condensers, and lastly the camera must all be very
carefully centred to the microscope, and immovably
fixed so that the whole apparatus may always be ready

28 THE GENERAL PRINCIPLES OF THE METHOD

for use, the Nernst lamp being kept lighted as well as
the gas one during any experimentation in order that
a photograph can be taken at a moment's notice. Of
course, so long as the mirror is in its usual position no
light reaches the specimen from the Nernst lamp;
swing the mirror out of its position, and the light is
instantly changed from that of the gas-burner to the
powerful one from the Nernst burner. The distances
between the Nernst lamp, aplanatic condenser, and the
substage condenser, are of great importance. It must
be determined at the outset by trials which distances
give the best results. The presence of the water-tank
renders it difficult to make a rule.

When a cell or other object comes under observa-
tion which it is desirable to photograph, the working
eye-piece is removed from the microscope draw-tube;
the camera is allowed to slide down the beam until its
shutter is about an inch from the mouth of the draw-
tube, when it is clamped to fix its position. The
projection eye-piece, which is already attached to the
camera-shutter by means of the flexible velvet collar,
is inserted into the microscope draw-tube. The mirror
is now swung on its gimbals out of the focal axis,
thus allowing the light from the 300-candle-power
Nernst burner to replace that of the gas-burner; and
the former, after being cooled by transmission through
the intervening water-trough, is projected directly
through the two condensers. The image of the field
of the specimen will then be seen on the ground-
glass screen at the back of the camera, where it can
be rapidly focused.

SPECIAL PHOTOMICROGRAPHY

29-

FIG. 4. — The photomicrographic
water-cooling tank and Nernst burner,
for direct observation.

apparatus. Showing positions of
The microscope mirror is in position

SPECIAL PHOTOMICROGRAPHY

Fia. 5. — The photomicrographic apparatus. The microscope mirror is
swung aside for photography.

SPECIAL PHOTOMICROGRAPHY 33

If preferred, focusing may be done with a lens;
but in the case of a specimen of blood, the edges of
the red cells afford a good indication of its accuracy, for
they seem just to disappear when the accurate focus
is obtained. When they are out of focus the edges of
the cells stand out in high relief. Having obtained the
focus — and stress must be laid on this point — the cell
or other object is deliberately thrown out of focus to the
extent of about erirth of a millimetre1 by screwing down
the fine adjustment so as to bring the objective nearer
the object. The reason for this is that the cells are
resting on a jelly under a cover-glass which is all the
time slowly sinking into the jelly, and, of course,
carrying the cells with it. The latter, therefore, are
sinking out of focus all the time. By deliberately
"over-focusing," when the exposure is actually made
the focus will become accurate, and the sinking of
the cover-glass compensated for.

The length of the exposure varies with the objective
used and the candle-power of the light, which in its
turn varies with the voltage. It is best to find the
length of the exposure by experiment, but we give
about twenty seconds with the apochromatic objective,
using "backed" Imperial plates. The water cooling
tank cuts out light, but it is very necessary to use it
in order to delay death of the cells and the onset of
achromasia, both of which are accelerated by heat
rays. The tank cuts off some of the heat rays, but
allows the passage of the actinic ones. Many specimens

1 The fine adjustments of most microscopes are graduated to allow of this
measurement.

/ *

34 THE GENERAL PRINCIPLES OF THE METHOD

were lost owing to achromasia before the cooling tank
was employed.

The photograph having thus been quickly taken,
the mirror may again be swung into position, the
camera pushed out of the way, and, having inserted
the working eye-piece, the examination of the specimen
may be proceeded with, or other fields explored. We
have taken a negative in fifty seconds with this appa-
ratus, and as many as five negatives have been taken
from different fields in a single specimen; but such
speed is not often necessary.

All the photographs which illustrate this book have
been taken with the apparatus just described. It never
gives trouble, and has proved most useful in supplying
a means of recording the "specimens." It used to be
most annoying to see unique mitotic figures or other
interesting specimens slowly vanish before one's eyes
without being able to record them satisfactorily. In
fact, the best mitotic figure I have ever seen in a
lymphocyte was induced before we possessed a camera ;
and although thousands of figures have been seen
since then, I have never seen a picture comparable
to it. It wras seen by Professor Harvey Gibson as
well as by myself.

There is one other useful piece of apparatus which
requires mentioning, viz. the "revolving apparatus."
This is a simple clock-work contrivance which keeps
a long test-tube revolving on its long axis. The
test-tube is placed horizontally. The object of the
appliance is to keep the blood-cells constantly moving
in the "citrate solution," or other medium in which

THE REVOLVING APPARATUS 35

they may be suspended, while samples are under
examination. If the capillary tubes containing the
specimen are laid for some time on the table, the
corpuscles will sink to the most dependent part of
the citrate solution, and will ultimately adhere to the
glass. By placing the tubes in the "revolving appa-
ratus" this is effectively prevented. It is a good thing
to have in the laboratory, for it delays loss of vitality
in the cells; but it is not essential.

CHAPTER III

THE PREPARATION OF THE JELLY FILM

AGAR, the substance used for making the film on
which the cells are examined, is obtained from sea-
weed. It is very cheap, and may be bought in strips
or as a powder. We have used Merck's powdered
agar, which is quite neutral and pure. It is insoluble
in cold water, but immediately soluble in boiling water.
This solution, therefore, on cooling, sets as a jelly.
It is necessary to have a stock of jelly constantly in
hand, and a 2-per-cent preparation is used throughout.
This will melt when its boiling-point is approached,
but will not set again until the temperature has fallen
almost to 40° C. Unlike gelatine, this jelly may be
boiled over and over again, and it will always set at
its usual temperature.

The jelly is made in 2-per-cent strength for the
reason that it will stand diluting with its own volume
of water or other solution, and will still set as a
jelly — that is, a 1-per-cent solution of agar jelly will
set on a slide in the form of a film as it cools. By
using the 2-per-cent preparation we are enabled
to add an equal volume of any solution we please,

SALTS ARE NECESSARY 37

so that the result is that the 1-per-cent jelly may
contain quite a variety of substances, and if some
human cells are placed on its surface we may try
the effect on those cells of any of those materials
which have been added in solution to the 2-per-cent
agar. We are thus able to investigate, by a method
which is simplicity itself, the effects of drugs or
chemical' substances upon the individual human cell.

Before we begin to discuss this subject, however,
we must be certain that the cells are alive when they
are being subjected to the drug. It is, of course,
well known that when, say, a drop of blood is re-
moved from the finger the leucocytes are alive; but
it is necessary to be certain that they are not killed
immediately they are placed on the jelly-film. As
will be discussed at greater length later on, we can
always ascertain whether white blood-cells are alive
or not by mixing a certain quantity of an alkaloid
with the jelly; for alkaloids excite amoeboid move-
ments, and it is obvious that these movements cannot
occur in a dead cell. Since alkaloids have supplied
the means of determining this point, we have also
been able to ascertain how to make the jelly so that
it will keep the cells alive as long as possible; for it
is clear that a jelly which will allow cells to remain
excited for the longest period with a given quantity of
alkaloid must be the best jelly for keeping the cells
alive when made without the alkaloid. The presence
of a combination of certain salts is essential.

Suppose a drop of blood is placed on to a film
of jelly which contains only agar and water and no

38 THE PREPARATION OF THE JELLY FILM

salts. The red cells will haemolyse immediately. The
white cells are worth watching. As soon as they
come to rest, or even before that, the polynuclear
leucocytes seem to swell up, the granules exhibit
"furious" Brownian movements, and in a few moments
the cell totters and then bursts. Water kills blood-
cells instantly if there are no salts present. Let the
experiment be repeated, but, instead of using merely
agar and water, now make the jelly with sodium
chloride in the strength of "normal saline solution."
It can be made thus: Melt a few cubic centimetres
of 2-per-cent agar jelly and place 1 cc. in a test-
tube. Prepare a solution of 1 . 8-per-cent sodium
chloride in water. To the 1 cc. of molten 2-per-
cent agar jelly add 1 cc. of the sodium-chloride
solution. The test-tube will now hold 2 cc. of a
1-per-cent agar jelly containing 0.9-per-cent sodium
chloride, i.e. "normal saline solution." The whole
is melted again, and a drop poured on a slide. If
some blood is now examined on this jelly, it will be
seen that the red cells do not "lake" immediately.
The leucocytes, however, again die very quickly, as is
seen by their swelling up, the onset of "dancing"
movements of the granules, and by rapid bursting,
although the rupture will not be quite so rapid as
when only water was present.

Now let the experiment be repeated a third time,
but instead of adding a solution which contains only
sodium chloride, let it contain in addition some sodium
citrate, thus: To 1 cc. of 2-per-cent agar jelly add
1 cc. of a solution containing 1 . 8-per-cent sodium

STOCK SOLUTION OF AGAR 39

chloride, and 2-per-cent sodium citrate.1 (When this
jelly is spread on the slide it will contain 1-per-cent
agar, 0.9-per-cent sodium chloride, and 1-per-cent
sodium citrate.) The picture presented by blood spread
on such a film is very different from those in the last
two experiments. The red cells are not crenated, but
are beautifully spread out. The leucocytes are not
dead, but alive and amoeboid ; no Brownian movements
of the granules can be seen, and the cells do not burst;
on the contrary, they will live now for an hour or more.
It may therefore be said that for the examination of
living blood-cells (and it has been found that it is also
the case for all cells yet tried) the jelly must always
contain a certain amount of the salts sodium citrate
and sodium chloride. "Normal saline" is not enough
by itself. Cells die immediately when they are resting
on a surface which contains only sodium chloride.

These three experiments will prove instructive for
the beginner with this jelly method, for they demon-
strate how the jelly is prepared. It must be observed
that for the purposes of these researches the supply
of jelly is always kept as a 2-per-cent solution of agar.
When, however, it is placed as a film on the slide, it is
always diluted with an equal volume of some other
solution, so that the film invariably contains 1 per cent
only of agar. It is in the diluting solution (always
added in an equal volume) that the salts, and any other
substances to be experimented with, are contained, and,
obviously, before being added to the agar they must
be of twice the required strength so as to be reduced

1 Potassium oxalate may be substituted for sodium citrate.

40 THE PREPARATION OF THE JELLY FILM

to the proper one in the resultant jelly with which the
film is made. It is imperative to explain the way in
which the jelly is made even at the risk of being
verbiose. Bear in mind, therefore, that two solutions are
required — namely, No. 1, a stock 2-per-cent solution
of agar, and No. 2, a solution which contains the
other substances the effects of which are to be tried on
the cells. Solutions Nos. 1 and 2 are always mixed
together in equal parts and then boiled up to form
No. 3, from which the jelly-film on the slide is prepared.
No. 1 is always the same. No. 2 may contain a variety
of substances, but no matter how much of any sub-
stance No. 2 may contain, No. 3 will always have half
that amount. For example, if one wishes a cell to rest
on a jelly containing 1 per cent of morphine one must
have 2 per cent of morphine in No. 2, so that when the
two solutions are mixed in equal parts the combination,
that is No. 3, will contain 1 per cent of morphine.

A word is necessary as to the effects on cells of the
agar itself. It appears to be innocuous. We have tried
it in strengths double and even four times as great as
that contained in the stock solution, without apparently
producing any deleterious effect upon the vitality of
the cells experimented with.

CHAPTER IV

CELLULAR STAINING, DEATH, AND ACHROMASIA

BY far the larger number of cells examined in these
researches have been blood-cells taken from the finger.
The white blood-corpuscles have offered a very interest-
ing study, and since they respond to chemical agents in
a way very similar to those observed in several other
varieties of cells, and, since they are very easily obtained
and can be very carefully watched, it is convenient to
describe what wre have seen with them. These cells
play an important role in the phenomenon of healing,
and ultimately go to form some of the fixed tissue-
cells, especially after an injury has been sustained.

For the examination of blood-cells in in vitro it is
best first to mix the sample of blood (which should be
drawn freshly from the finger) with an equal volume of
"citrate solution."1 The citration of the blood not
only prevents it coagulating, but it also keeps the cells
alive sometimes for as long as seven days.

The way in which wre citrate the blood is as follows:
One end of a capillary tube, such as has been described
(Chapter II.), is dipped into the citrate solution, some of

1 Three-per-cent sodium citrate, and 1-per-cent sodium chloride.

42 CELLULAR STAINING, DEATH, ACHROMASIA

which runs up into it. The amount drawn into the
tube can, if necessary, be controlled by keeping the
finger on the other end. It has been found most service-
able to allow the solution to fill the tube to the extent
of about half an inch, and any excess can always be
removed by tapping the lower end of the tube upon the
table, which causes some of it to run out. Having got
a sufficient quantity of citrated solution into the tube, it
is run down to one end of it, and a mark is made at the
upper limit (or meniscus) with a grease pencil. The
fluid is now run along the tube by depressing its other
end until its lower meniscus stands at a level of the
mark, and a second mark is then made at the upper
meniscus, after which the tube is again placed vertically
so that its contents runs down to its original position.
The finger having been pricked, a drop of blood is
squeezed out and at once allowed to run into and mix
with the citrated solution in the tube, the greatest care
being taken that no air-bubble intervenes between the
fluids. The blood should be allowed to run in until
the upper meniscus of the mixed fluids reaches the
upper mark. Thorough mixing of the blood with the
citrated solution is ensured by rocking the tube in such
a way that its contents runs from end to end. The
mixture in the capillary tube will now consist of equal
proportions of blood and citrate solution, and of this a
drop is tapped out on to a cover-glass, which is then in-
verted and allowed to fall on the agar film in the usual
way. When tissues are to be examined, a small portipn
of the growth or normal structure is either teased
out or scraped into a little of the citrate solution in

THE STAINING OF A CELL 43

a watch-glass. A drop of the cell-containing mixture
is then placed on to the cover-glass and similarly placed
upon the jelly.

The citrate solution simply acts as the vehicle
in which the cells are kept in a living condition before
being placed upon the jelly, and furthermore, by dilut-
ing the blood, . it reduces the actual number of cells
which come to rest in any field of the film. If no
citrate solution is used they are apt to become huddled
or crowded together owing to their great numbers,
and the leucocytes may become completely hemmed
in by erythrocytes so that a clear observation of the
whole cell cannot be obtained.

We must now pass on to the study of some of
the phenomena connected with the staining of the
cells, which have been the means of elucidating many
cytological details which have led to the correct
appreciation of the effects of chemical substances on
cells. In this chapter, however, I do not propose to
discuss very deeply the actual laws by which the
staining of the cells is controlled; that will be reserved
for discussion when I come to speak of the diffusion
of the substances, including stain, into the cells. In
the meantime I shall simply describe what happens
to the cell as it absorbs the stain (say Unna's
polychrome methylene blue, Grubler) ; how the stain
causes the gradual death of the cell (the staining of
the nucleus invariably kills it) and how death is
followed by achromasia. The amount of stain which
is put into any given jelly is not added in a hap-
hazard way, the actual amount necessary to cause

44 CELLULAR STAINING, DEATH, ACHROMASIA

leucocytes to stain deeply in a given time being
a very definite one, as will be described in the next
chapter; but in the meantime we must assume that
a jelly has been correctly prepared containing, besides
the proper proportions of sodium citrate and sodium
chloride to keep the cells alive, the proportion of stain
requisite to enable us to observe its gradual passage
into the leucocytes as they absorb it.

The agar jelly, of course, will be coloured purple
owing to the stain it holds in solution, but it will be
quite transparent and will allow sufficient light to
penetrate it so that the cells may be clearly observed.

Having without delay placed the film, with the blood-
cells upon it, under the microscope, at first the cells
will be quite unstained, but the white corpuscles
may easily be recognized owing to their granulation
and size. Let a polymorphonuclear leucocyte be
watched. Gradually its granules become tinted a
faint red colour (fig. 6) and about the same time
amoeboid movements may begin. If certain propor-
tions of alkaloid have been added to the jelly, these
amoeboid movements will be very marked (fig. 7).
The staining of the granules becomes deeper and
deeper, always maintaining the same bright scarlet
colour. In spite of the deepening coloration of the
granules, amoeboid movements will continue, showing
that the cells are alive and that their vitality is appar-
ently unaffected by the staining of their granules.
It is not only the polynuclear leucocytes that behave
in this way, but the mononuclear, or lymphocyte,
cells as well.

A STAINED NUCLEUS MEANS DEATH

FIG. 6. — The granules of the leucocyte are gradually becoming stained. The
red cells are unstained. Low power.

FIG. 7. — The leucocyte's granules are stained. Its nucleus is unstained.
The pseudopodia are extruded in response to atropine, which is diffusing
into the cell as well as the stain.

A STAINED NUCLEUS MEANS DEATH 47

FIG. 8. — The same field as 7. The leucocyte is retracting its pseudopodia.

FIG. 9. — The same field as 7 and 8. The retraction of pseudopodia is
nearly complete. The lobes of the nucleus of the leucocyte are turning a
faint blue colour.

A STAINED NUCLEUS MEANS DEATH

FIG. 10.— A leucocyte excited by atropine. Its granules are deeply stained,
and its nucleus is also beginning to stain a blue colour. Low power.

FIG. 11. — A leucocyte which has just been killed by the staining of its
nucleus. Its granules are also deeply stained.

A STAINED NUCLEUS MEANS DEATH 51

After a short time the extrusion of pseudopodia
ceases, and it will then be noted that general retraction
(figs. 8, 9) of pseudopodia begins to occur. In the
meantime the lobes of the nuclei of the polynuclear
cells begin to turn a faint blue colour (fig. 10). If
two or more leucocytes happen to be in the same
field, it will be seen that they all behave in a like
manner, for the stain affects them all equally. In a
few moments all the amoeboid movements cease, for
death is about to occur, and then, sometimes quite
suddenly, the nuclei turn bright scarlet (fig. 11) and
the death of the cell takes place.

We have never yet seen a cell show any amoeboid
movements when its nucleus has stained scarlet. By
mixing some blood with a citrated solution of stain
one can cause first the granules and then the nuclei
of the leucocytes to stain, the difference depending
on the length of time the mixture has been made.
If now cells with only their granules stained are placed
on to a jelly-film which contains an alkaloid, we can
excite these cells, showing that they are alive. But
if their nuclei are stained scarlet, no excitation or move-
ments of any sort can be produced, and there can be
little doubt, therefore, that the staining of the nu-
cleus kills the cells. White blood-corpuscles do not
seem to mind the staining of their granules; but the
staining of their nuclei invariably causes their death.
This is a rule to which we have never yet seen an
exception in any cell which we have examined. A
stained nucleus is incompatible with life.1

1 We have tried several stains, but this rule holds good with them all.

52 CELLULAR STAINING, DEATH, ACHROMASIA

When the lobes of the nuclei stain scarlet, the
chromatin network within them shows up well. The
blue coloration which precedes the scarlet one is due,
I think, to the staining of the nuclear wall. The
polychrome dye contains two stains, a red and a blue
one, and the nuclear wall seems to have an affinity
for the blue one, while the chromatin combines with
the red. The staining of the nucleus, therefore, is a
sign that the cell has died, and one now sees a circular
dead cell (in reality it is a spherical cell which has
become flattened out) with its granules stained scarlet,
and in their midst there is the polylobed nucleus, also
stained scarlet. Let the specimen be watched still
further. Gradually the cell-wall is seen to bulge out
in places (fig. 12), apparently away from the granules.
After a few moments this bulging becomes general
(fig. 13), and the cell presents a clear halo of cell-wall
and cytoplasm outside the limit of the mass of granules
in its centre. This is due to the gradual liquefaction of
the cytoplasm which occurs at death, beginning at the
periphery and progressing slowly towards the nucleus.
Sometimes a few stained granules appear to migrate by
the "dancing" Brownian movement into the liquid
cytoplasm which has bulged out the cell- wall. Under
suitable conditions the Brownian movement becomes
general, showing that all the cytoplasm has liquefied—
a certain sign of death.

No matter whether the cytoplasm has completely
liquefied or not, however, one of two things is bound
to happen after a short time. The granules and nucleus
may remain stained for half an hour or so, especially

ACHROMASIA

FIG. 12. — Tne leucocyte tias just died owing to the staining of its nucleus.
The cell-wall is beginning to bulge because the cytoplasm is liquefying.

FIG. 13. — The onset of achromasia. The same field as 12. The stain
is beginning to fade from the nucleus. The bulging of the cell-wall has
become general.

ACHROMASIA 55

if a temperature of about 30° C. is maintained, which
may prevent the bulging of the cell-wall; but after
that time the cell will either burst and become
achromatic, or become achromatic without bursting
(fig. 14). In either case achromasia, or loss of stain
from the cell, invariably occurs. If the temperature
is low (say that of the room) the cell will probably
burst and its granules will be scattered about on the
surface of the jelly. Now, when a cell bursts on a jelly
which contains salts — such as the one with which we
are supposed to be experimenting — there is another
rule to which there is no exception, namely, that the
cell's nucleus loses its stain instantly. In a flash all
coloration has gone from it. But the granules may
remain stained for half an hour or more; and then
they also gradually lose their stain (fig. 15), and appear
slowly to vanish from the scene. The phenomenon
of achromasia always overtakes the cells sooner or
later.

If the cell does not burst, the stain disappears, but
its disappearance is much slower. This is a pretty
phenomenon to watch; but it requires a warm room or
warm stage. Suppose we are watching a cell which
is dead, having its nucleus and granules stained bright
scarlet. The stain gets a deeper colour, and one
wonders how deep a shade it will attain to. Suddenly
the staining seems to stop, and the depth of colour .may
remain the same for a quarter of an hour or so. Then,
almost imperceptibly at first, the colour becomes paler,
and with an accelerating speed the colour fades away
from the lobes of the nucleus, until that structure

56 CELLULAR STAINING, DEATH, ACHROMASIA

remains as unstained as it was when the cell first came
to rest on the jelly-film. After a few minutes the
granules slowly lose their stain also, until nothing seems
to remain. Ultimately the red cells disappear too (figs.
14, 15), and the field, which a short time previously was
dotted with red cells and stained leucocytes, now
becomes a blank, and a new-comer looking at the
specimen would hardly believe that there had ever
been any cells under view. The picture afforded by
the successive occurrence of the staining, death, and
onset of achromasia in the cells is well worth seeing.
First the slow diffusion of the stain into the cells,
staining first their granules and then their nuclei;
the gradual retraction of pseudopodia as the nuclei
stain, and then the bright scarlet coloration of the
nucleus itself as death occurs... After a pause the
gradual fading of the stain, first from the nucleus and
then from the granules, until at last nothing remains
visible of the leucocyte in the place it filled among
the neighbouring red cells. The whole phenomenon
reminds one of a lantern dissolving view — the onset
of staining, its climax, and then its disappearance.
Achromasia invariably occurs after a time — nothing
which we know of will prevent it; but heat greatly
accelerates its onset, and a ruptured cell always becomes
achromatic before a whole one.

I do not propose here to give the details of ex-
periments which I made some years ago, to try to
investigate the nature of this phenomenon of achro-
masia; they will be found in a paper, ''On the Cause
of Achromasia," in The Lancet of January 23, 1909.

FIG. 14. — Achromasia. The same field as 13. The stain has gone from
the nucleus, although the granules are still stained. Note that the red cell
is disappearing.

Jbiu. 10. — ^ciii-uaiasia. 'ine same held as 14. Many of the cell-granules
have lost their stain. The cell-wall is nearly invisible. The red cell has
disappeared.

CAUSE OF ACHROMASIA 59

I shall now simply state the conclusions which were
arrived at then, subsequent experimentation not having
altered my opinion in any way.

Achromasia seems to be part of the general dis-
organization which occurs in a cell after death. I have
never seen the phenomenon in a living cell, and one
cannot excite an achromatic leucocyte or lymphocyte.
It is by no means necessary for a cell to be stained
before it can become achromatic; on the contrary, one
frequently sees dead cells which refuse to stain, although
their living neighbours will stain well under suitable
conditions. The rapidity of onset of achromasia de-
pends upon the temperature and the presence and
amount of salts. It also appears to depend to some
extent on the completion of the liquefaction of the
cytoplasm. The more advanced the liquefaction, which,
of course, only occurs after death,1 the more readily
does achromasia take place. Heat and salts accelerate
it greatly. If there are no salts present, even the
nuclei of ruptured cells do not become achromatic for a
long time. These stained nuclei may sometimes be
seen floating about free from cytoplasm, granules, or
cell- wall. I believe that achromasia is due to the
chromatin passing out of the dead and liquefied cell by
osmosis. If the chromatin is stained, the stain will dis-
appear with the chromatin ; if the cell is unstained, it is,
of course, impossible to stain its chromatin if the latter
has already passed out by osmosis. I believe that this
is what has happened in the cell which is commonly

1 See paper in Journal of Physiology, "On the Death of Leucocytes,"
vol. 37, No. 4, 1908.

60 CELLULAR STAINING, DEATH, ACHROAIASIA

known as the hyaline cell. Achromatic lympho-
cytes resemble them strongly. A dead lymphocyte,
from which the granules have disappeared, will not
stain, and the cell resting on the jelly looks like a
phantom. We have never been able to excite such
a cell. If a specimen of fresh blood is placed carefully
on a jelly-film, one does not usually see any such cells,
for all the cells will stain; but after a while the film
may contain many examples of achromatic cells which
appear to be exactly like what are known as hyaline
leucocytes by the older methods. Achromasia is a
certain sign of death, and the recognition of its very
characteristic appearances is of the utmost importance
in this form of research. It should be borne in mind
that a cell with a stained nucleus is dead, and so is
a cell which is achromatic.

CHAPTER V

THE DIFFUSION OF SUBSTANCES INTO LIVING CELLS—
THE "COEFFICIENT OF DIFFUSION"

THE following points regarding the diffusion of sub-
stances into cells have been determined by experimen-
tation with this method. A cell does not respond to the
action of a chemical substance unless the substance has
diffused into it. It is of the utmost importance, there-
fore, that the lawTs wThich are concerned in this diffusion
should be understood, both for the practical application
of this in-vitro method in the study of the actions of
substances on individual cells and also, I think, because
it throws light upon the way in which drugs produce
their effects upon the various systems and organs of the
animal body. Very little has hitherto been known
concerning the diffusion of substances into individual
living cells; and although we do not claim to have
advanced the knowledge of the subject to a very great
extent as far as its scientific basis is concerned, wre can
safely say that wre are nowT in a position to cause sub-
stances to diffuse into individual cells according to our
will. We do not know7 all the scientific facts, it is true,
for more \vork will be necessary before these can be

62 DIFFUSION OF SUBSTANCES INTO LIVING CELLS

determined, but we do know the more important laws
which are sufficient for practical purposes.

The diffusion of substances into living cells is purely
a physical process. A cell does not seem able to exert
any vital control or power of selection whatever over
the diffusion of substances into its cytoplasm. In every
cell with which we have experimented this has been the
case, and if proof is needed it is afforded by the fact
that we can at will cause cells to be excited, to repro-
duce themselves, or to die, by employing the knowledge
of the laws, which shall presently be described, enabling
us to make substances diffuse into cells at any speed we
please.

Owing to a variety of means at our disposal cer-
tain cells can be made to die in two minutes or in
two hours, whichever one likes, merely by accelerating
or delaying the diffusion of the agents into them; and
it is clear that if a cell could control this diffusion it
would at least make some effort to do so in order to
save its own life. They cannot do so, however, and
always die with clockwork-like regularity at the end of
various given periods of time, which are determined by
controlling the diffusion of certain chemical substances
into the cell's cytoplasm. As will be shown later, the
rate of diffusion of substances into living cells can be
calculated by means of a simple equation; and since
excitation, reproduction, and death can each in succes-
sion be induced in vitro by causing the diffusion of
substances into cells, it follows that excitation, repro-
duction, and death may also be induced according to
the rules which can be plotted as a simple equation.

A PHYSICAL PHENOMENON 63

Cells, as living entities, cannot refuse to absorb
substances, and it is also a rule that they cannot "pick
and choose" what they absorb. For instance, a cell
cannot take from a solution which surrounds it a pro-
tein and refuse an alkaloid. If it is surrounded by both
these substances it has to take both. On theoretical
grounds, I believe that a solution could be prepared
(although we have not yet been able to assure our-
selves that such is the case) from which a given cell
would be able to absorb nothing; but such a contin-
gency, as far as can be seen, would be impossible in the
body. A cell does not appear to " feed " in the ordinary
sense of the term — that is to say, it cannot seek after
food. It has to take what is there according to certain
laws, even if it dies in consequence; but it cannot
"help itself" in any sense of the term. The life of a
cell depends upon substances in its surroundings, even
its reproduction depends upon them, and the associa-
tion between them and its life depends on the diffusion
of these substances into the cell itself, which diffusion
is in its turn undoubtedly dependent on physical laws
over wrhich the cells themselves can individually exert
no control. A cell cannot take a cake and leave a bun,
so to speak: it has to take a bit of cake and a bit of
bun wrhether it likes them both or not — a law which
has been amply confirmed by work extending over a
period of five years.

One is open to criticism in this matter; for the
objection may be raised that it is well known that
some cells of the body are affected by some agents
or drugs, while others apparently are not. This

64 DIFFUSION OF SUBSTANCES INTO LIVING CELLS

seems to be the case, judging from results; but it
does not prove that all the other cells of the body
have not also absorbed some of the drug in question
as well, and that they may have been affected too,
but have shown no signs of this effect. Strychnine,
for example, stimulates certain cells of the nervous
system, as is shown by the twitchings produced by
it. Strychnine also causes amoeboid movements in
leucocytes — a fact which is easily demonstrated micro-
scopically; but this excitation of amoeboid movement
within the body gives rise to no symptoms, and of
course passes unnoticed. It is important to remember,
therefore, that because a certain drug affects certain
cells and gives rise to symptoms related to these cells,
it does not follow that only those cells have absorbed
the drug. All have absorbed a share, but not neces-
sarily to the same extent, as will be shown directly.

A cell, therefore, cannot control the diffusion of
substances into itself; but after it has actually absorbed
them the protoplasm of different classes of cells seems
to treat a substance differently, and the cells may* by
this peculiarity of their protoplasm, be able to make
use of it, or, on the other hand, they may leave it
unchanged, or thirdly, they may have to die from
its effects. We shall presently describe how cells
can be made to absorb aniline dye which contains
two substances — one which causes the cell to re-
produce itself, the other a poison which kills it. As
both substances diffuse into the cell together, and
as the cell cannot control this diffusion, it will respond
to both. It will reproduce itself by cell-division in

"IN VITRO" VERSUS "IN vivo" 65

response to one element, and it will die in the act
of mitosis from the effects of the poisonous one.
This experiment, which will be described at length
later, proves these two points, about which I wish
to be emphatic; viz. that a cell cannot control the
diffusion of substances into itself, nor can it choose
from its surroundings any one substance and leave
another. Even at the expense of its life, a cell is
bound to absorb from its surroundings any substance
which may be present; and this absorption depends
entirely upon certain chemical and physical factors.

Before proceeding to describe these laws and
factors, other points must be mentioned. We are
dealing with in-vitro experimentation; and we have
no proof that the diffusion of substances into cells
in vitro is identical with this diffusion into cells
in vivo. There is, however, strong presumptive evi-
dence that similar conditions prevail. As a matter
of fact, apart from the mere phenomenon of diffusion,
this possible distinction between the facts learnt from
in-vitro experimentation and what actually occurs
in vivo must always be borne in mind in researches
of this nature. The force of this point will become
apparent later on when we come to deal with induced
cell-division; for although one can induce the diffusion
of substances into cells or cell-division at will on a
microscope slide, it will be seen that these phenomena
in the body occur under very different conditions,
which must be taken into consideration in forming
deductions from in-vitro experiments. In the final
chapters, however, it will be shown that the results

66 DIFFUSION OF SUBSTANCES INTO LIVING CELLS

obtained by in-vitro experimentation have been con-
firmed in some instances by experimentation in the
living body, and hence one may, I think, reasonably
infer what goes on in vivo from what is observed
in vitro, and that these experiments into individual
cells may be undertaken with confidence.

Before continuing this subject another matter con-
nected with it must also be stated. In the previous
chapter it was mentioned that cells observed in vitro
must be resting in a solution or on a jelly which con-
tains certain salts the presence of which are necessary
for keeping them alive. In the body one of these salts,
sodium chloride, is actually present; but there is no
sodium citrate, a solution of which has proved to be
the best one for leucocytes and other cells to live in.
Obviously in the body there must be some salt or salts
for which the sodium citrate is a substitute or equiva-
lent. One of the roles played by sodium citrate in
in-vitro experimentation is its property of preventing
coagulation of the blood, which seems to be an im-
portant one, for related to this is the curious fact that
leucocytes will live longer in citrated plasma than in
undiluted serum, a point which will be alluded to in
the description of the method of measuring the lives of
leucocytes. Sodium citrate, however, is detrimental to
leucocytes, and there is no solution known which will
keep leucocytes or other human cells alive for more
than a few days. If there was we should now be in
a position to cultivate families of human blood-cells in
test-tubes. At present, by means of sodium citrate,
one can only make leucocytes "exist" for some hours

GAUGING THE DIFFUSION 67

while we experiment with them; and it must be borne
in mind that since sodium citrate is detrimental, leuco-
cytes or other cells placed in it gradually lose vitality
all the time, and that they are under experimental
conditions.

The laws of diffusion — or rather what we know of
them — are simple in their experimental application;
but they are difficult to describe.

There are two methods by which it may be known
when substances have diffused into a cell. If the
diffusing substance consists of a colouring matter which
will combine with or otherwise colour the molecules of
protoplasm within the cell, one can see the extent of the
diffusion by watching the progress of the coloration.
The other method consists in the use of a substance
which has a specific action on the cell and causes it to
give a definite response which will tell us wrhen the
substance has diffused in. Of the twro methods, the
former is obviously the better, for by seeing the gradual
staining of the morphological elements of the cell one
can more accurately gauge the extent of the diffusion
than one can by measuring roughly the degree of a
response such as excitation of amoeboid movements or
even cell-division. It is, of course, possible to employ
a combination of both methods, by which much can be
learnt; in fact, in this book I shall describe what has
been observed, in the first place, by using colouring
substances only, afterwrards a combination of stain and
other substances, and lastly by experimenting with
other substances by themselves.

Suppose the jelly on which a given cell is resting

68 DIFFUSION OF SUBSTANCES INTO LIVING CELLS

contains a certain quantity of an aniline dye, such as
Unna's polychrome methylene blue. This dye com-
bines with the cell-granules and stains them red, and
the rate of the diffusion of the dye can be estimated
by observing the depth of coloration of the granules
and the time occupied before the nucleus stains. The
first granules to stain, of course, are those which are
nearest to the jelly, for the cell is pressed against it by
the cover-glass. With a given quantity of dye, the
depth of coloration and the rapidity of the extent of
staining will take a certain length of time. No matter
how often this experiment is repeated, provided the
arrangement of the jelly is always the same, with the
same type of cell, the result is always the same; but
if a fresh jelly is prepared, with double the quantity of
stain, the depth of coloration will be double, and the
same extent of staining will be reached doubly as
quickly as with the first jelly. If the concentration
of the dye is trebled or quadrupled, etc., the depth of
coloration and the rapidity of the given extent of
staining are also trebled, quadrupled, etc., as the case
may be.

Hence we arrive at the first law, which is, that the
diffusion of a substance into a cell varies directly with
the concentration of the substance in the solution in which
the cell is resting. The more concentrated the sub-
stance, the more it will diffuse into the cell, apparently
in arithmetical proportion. In a given time, ceteris
paribus, a 2-per-cent solution of a substance will have
double the effect on a cell as compared with a 1 -per-
cent solution.

THE FACTORS CONCERNED 69

Briefly, therefore, we may say that the diffusion is
proportional to the amount of substance diffusing, or
we may plot it thus:

diff = S

Obviously, the diffusion of a substance into a cell
takes time. If there is only sufficient dye to combine
with a certain amount of protoplasm, the combination
will occur in a certain time, and then the diffusion will
cease, for all the dye will be used up; but if there is
a sufficiency of stain for it to go on diffusing indefi-
nitely into the cell until it kills it by staining the
nucleus, then the diffusion will go on for a longer
time — in fact, it will go on diffusing minute after
minute until death occurs. Hence WTC may say that
the longer the time which we observe the diffusion, the
greater will that diffusion be, unless the substance is
all used up — a contingency which in reality cannot
occur in practical experimentation, but it may occur
in the body. It must be remembered that once the
experimental jelly- film is made it cannot be altered,
whereas in the body there can be no doubt that the
solutions are being continually modified during meta-
bolism.

With a given concentration of dye or other sub-
stances in the jelly, therefore, the greater the time
during which the cell is resting on the jelly, the more
of that substance will diffuse into the cell, also in
direct arithmetical proportion. Each minute will see
an equal amount of substance diffusing, provided the
supply of that substance is constant, and that other
conditions remain the same during the time.

70 DIFFUSION OF SUBSTANCES INTO LIVING CELLS

Conversely, in a given time, the greater the con-
centration of the substance diffusing, the more of that
substance will pass into the cell, as was shown in the
first law. We have now considered two factors, there-
fore, viz. that diffusion is equal to the concentration of
the substance and the time, or thus:

diff = S+T

The next factor to be considered is heat. In vivo,
of course, variations of temperature are not very great,
but with in-vitro experimentation the temperature must
be carefully considered, for we may keep the slide with
its jelly-film with which we are working at a variety of
temperatures, ranging from that of the room in which
one works to that of the blood. Heat increases the
diffusion of substances into cells in a marked degree,
and this increase is also in arithmetical proportion.
Each degree of temperature means a definite increase in
the diffusion, and therefore the diffusion can be regu-
lated to a nicety by keeping the slide on which the cells
are resting at a definite temperature. Of course if
extremes of heat are used death will occur; but within
reasonable limits, which are compatible with life, one
can employ heat to great advantage in these experi-
ments. H eat therefore must be coupled with concentra-
tion and time as a factor which increases diffusion; and
our equation now stands thus:

diff = S + T + H

There is one other factor which increases the diffusion
of substances into cells more than any of the three other

THE FACTORS CONCERNED 71

factors already mentioned. Alkalies and alkaline salts
greatly increase the diffusion of other substances into liv-
ing cells. By means of a strong alkali one can cause a
substance like stain to diffuse into a cell so rapidly as to
induce death and staining of the-nucleus almost instantly.
And this marked increase of diffusion caused by alkalies
also takes place in an arithmetical progression; that is
to say, if the jelly or solution contains 2 per cent of an
alkali, another substance present will diffuse into the
cell twice as rapidly as it would if the jelly or solution
only contained 1 per cent of the same alkali. Our
equation must therefore contain a symbol for alkali
also:

All the above four factors — namely, the concentration
of the substance diffusing, the time, the heat, and the
alkalies — increase the diffusion. Neutral salts, however,
decrease it. The more of a salt one adds to the jelly,
the less of any other substance, ceteris paribus, will
diffuse into the cell in a given time. And this retarding
effect of a neutral salt also varies exactly with the
amount of the salt present. We may therefore add
salts to our equation with a minus sign before them,
thus:

diff = S+T + H+A-salts

Acids, of course, delay the diffusion of other sub-
stances, for they neutralise alkalies; and the amount of
retarding effect due to an acid is in exact proportion to
its neutralising effect on any alkali present. But apart
from this neutralising action of acids, they also actually

72 DIFFUSION OF SUBSTANCES INTO LIVING CELLS

retard diffusion themselves according to their strength;
that is to say, that if a certain amount of diffusion of a
substance will occur from a neutral jelly, the addition of
an acid will delay that diffusion in direct proportion to
the amount of acid present. As a matter of fact, acids
play only a very small part in these researches, for it
has been our endeavor to copy the conditions found
in the body as much as possible, and cells do not
normally come into contact with acids to a great extent.
For this reason, as will be shown later, we actually take
steps to eliminate the consideration of acids from our
experiments, in order to simplify matters.

The foregoing, then, are the factors which increase
or decrease the diffusion of substances into living cells.
We have no right, of course, to assert that all alkalies
increase and all salts retard the diffusion of substances
into cells, for we have not tried them all; but as far as
we have experimented they seem to obey a general
rule. As has already been stated, one can only touch
on the main principles of this subject of the passage of
substances into individual cells, about which little was
known before this jelly method of in-vitro staining
was invented.

Up to the present I have used the expression "cell"

in its widest sense. Cells exist as individuals, and as

individuals in classes. One may say that polynuclear

'neutrophile leucocytes are a class of cell, and that

erythrocytes are another class of cell.

The diffusion of substances into cells is generally
the same in individuals of a class, but it presents great
differences in the various classes. For instance, if a

TAKES PLACE AT VARIOUS RATES 73

jelly is suitably prepared to stain the nuclei of leuco-
cytes in a given time, it will stain the nuclei of all
the leucocytes in that time, and it will always do so.
There will, of course, be a few exceptions among
individual cells which have died or which have become
achromatic, but generally speaking all the cells obey
the rules of their class. In some classes of cells, how-
ever, such as those of the epidermis, we have not yet
succeeded in causing anything to diffuse into them at
all; and in some of the larger cells, such as some
epithelial cells, only a few types will absorb sub-
stances in vitro; yet if some of the cells of a class in a
specimen will absorb a substance at a certain rate, the
others of the same class, which are not achromatic,
will also absorb the substance at the same rate. It
must therefore be grasped that the individual cells of
a class will absorb substances in the same way as each
other, and the diffusion into them will be influenced
by the usual factors in the same way in each cell of the
class; but substances diffuse into the cells of different
classes at different rates.

Now we come to an extremely important factor
which has not been mentioned before, and which is the
last one to be taken into consideration. It is the
"coefficient of diffusion."

We may prepare a jelly containing a certain con-
centration of stain, alkali, and salts which will allow
a certain amount of diffusion of the stain into a certain
class of cells at a certain temperature in a certain
number of minutes. Another class of cells may then
be tried on a film made from the same jelly under

74 DIFFUSION OF SUBSTANCES INTO LIVING CELLS

the same conditions, when it may be found that now
no staining, or less staining, may take place. If one
adds more stain or more alkali, or more heat, or allows
more time, this second class of cells may then stain.
Hence we may say that the second class of cell has
a higher "coefficient of diffusion" than the first, for
it requires more of one or more factors which increase
diffusion to cause a certain extent of diffusion into
it than did the first class of cells. Different classes
of cells may therefore each have different coefficients of
diffusion, but in spite of this fact the diffusion of
substances into all classes of them depends on the
factors already expressed by the equation:

diff = S+T+H + A-salts.

That is to say, that the factors given in the equation
increase or decrease the diffusion of substances into
all cells; but some classes of cells require more or
less of them to cause the same amount of diffusion
than do others.

It is obvious, therefore, that we must always find
the coefficient of diffusion of a class of cells before we
can attempt to make substances diffuse into them; and
we find the coefficient of diffusion by ascertaining the
number of the factors expressed in the equation, and
the amount of each of them required to cause a certain
extent of staining of the cell. By means of the equa-
tion we can set down algebraically the number of
factors and the amount of each of them required to pro-
duce this certain extent of staining; and then they are

MODE OF DETERMINATION 75

all added up to make a grand total figure — which repre-
sents the "coefficient of diffusion," or, to express it
briefly, the "c/" of the cell.

The coefficient of diffusion of a cell is determined
by adding up the total amounts of the factors required
to cause a certain extent of staining of the cell. The
extent of staining which we always use as a standard is
the staining of the nucleus. Now, the "moment" of
the staining of the nucleus of a cell can be recognized
through the microscope, and it has an additional
importance, insomuch as it is coincident with and
signifies the death of the cell. In reality, therefore, the
determination of the coefficient of diffusion of a cell, as
well as supplying the rate of diffusion of substances into
it, also tells us how much of the stain, together with
the other associated factors, are required to make it (the
stain) diffuse into the cell so as to cause the cell's death
in vitro. In other words, it tells us the amount of a
standard dye required to be in the immediate surround-
ings of a cell, so that it may diffuse into it to such an
extent as to cause its death by combining with the
chromatin within the nucleus.

In order to determine the coefficient of diffusion of
a cell, howrever, it is necessary to count up, not only the
number of the factors required to cause staining of the
nucleus, but also the amount of each factor required.
To do this it is necessary to measure each factor. One
could, of course, measure the chemical factors, such as
alkalies, salts, etc., in grammes, the heat in degrees of
temperature, and the time in seconds; but this would
necessitate a complicated total figure involving grammes,

76 DIFFUSION OF SUBSTANCES INTO LIVING CELLS

degrees, and seconds. It has been found preferable to
measure these factors in special units which can, if
necessary, be resolved into their proper ones of
grammes, degrees, and seconds.

For instance, in order to remember the rate of stain-
ing of a class of cells it would be most inconvenient
to have to make a note of a statement such as this:
To stain the nuclei in twenty minutes, it is necessary to
keep the cells at 20° C. on a film made from a jelly con-
taining 0.5 cc. of Unna's stain, 0. 16 gramme of sodium
chloride, 0 . 03 gramme of sodium citrate, and 0 . 3 cc. of a
5-per-cent solution of sodium bicarbonate. It is much
simpler to say that the jelly contains so many "units"
of stain, salts and alkali, heat and time. One may go
farther and express these units as a simple equation,
thus:

Stain. Alkali. Heat. Time. Slats.

cf = (5s+3a

A letter by itself means one unit of the factor; a
number before a letter means that there is that number
of units of the factor: c means a unit of sodium citrate,
3c would mean three units of it, and so on.

It will be grasped that it is better to make "one
unit" of any factor a standard quantity, and these
quantities have been chosen with a special object. As
has been previously explained, the coefficient of diffusion
of a cell is the total number of units of the factors
required to cause staining of the nucleus. Some of the
factors increase the diffusion into the cell, and others
decrease it. A unit of a factor which increases diffu-
sion is so chosen that the increase it causes is equal to

STANDARDISATION OF FACTORS 77

that of one unit of any other factor which also increases
diffusion. Likewise a unit of any factor which retards
diffusion is also equal to a unit of any other factor
which does the same thing. But further still, a unit
which increases the diffusion of a substance into a cell
is so chosen that the increase which it causes can be
exactly neutralised by a unit of a factor which retards
diffusion. The units are all equal in value, so to speak.
Some increase diffusion, and some decrease it. Any
number of units of factors which decrease diffusion
retard exactly the increase of diffusion due to the
same number of units of factors which cause increase of
diffusion.

By the first law, if we double the quantity of the
dye in the jelly, we double the rapidity of its diffusion
into the cells. A convenient quantity was chosen,
namely, 0.1 cc., and this contained in 10 cc. of jelly
constitutes one unit of polychrome dye.1 Let us
suppose that this quantity (one unit) causes staining
of the nucleus of a given cell in a certain time. If
now another unit is tried, the cell will stain in half the
time it did before.

The alkali, sodium bicarbonate, increases the dif-
fusion of other substances into cells, and therefore it
greatly increases the rapidity of the staining by poly-
chrome methylene blue. Now, since all units must be
equal in value, it was ascertained experimentally that
0.1 cc. of a 5-per-cent solution of sodium bicarbonate
exactly doubled the rapidity of diffusion of one unit of

1 Unna's polychrome methylene blue (Griibler) is only supplied^in
solution, which is standardised. It cannot be made in a powder.

78 DIFFUSION OF SUBSTANCES INTO LIVING CELLS

polychrome dye. Hence the unit of alkali is 0 . 1 cc.
of a 5-per-cent solution of sodium bicarbonate.

Time is a factor. One unit of time is ten minutes;
and since time increases diffusion in arithmetical pro-
portion, therefore in twenty minutes (two units) the
diffusion of one or more units of the dye or other
substance will be doubled.

The unit of heat is 5° C. ; unity is 10° C., because
one cannot conveniently work at a temperature below
this point; 20° C. is three units, etc.

Salts delay diffusion. The two commonly employed
are sodium citrate and sodium chloride. Their units
respectively are 0 . 03 gramme and 0 . 08 gramme. One
unit of sodium citrate or sodium chloride will prevent
the increase of diffusion due to one unit of alkali, heat,
or time; an extra unit of stain will neutralise a unit
of salt.

Hence the units of all the factors are so measured
experimentally that they are as nearly as possible equal
in value. The delay in diffusion caused by a unit of
a substance which decreases diffusion is equal to the
acceleration occasioned by one which increases diffusion.
It will therefore be realised that the units can be
substituted for each other. A unit of alkali will double
the diffusion of the dye, but so will another unit of the
dye itself. Again, the unit of time is ten minutes;
since time increases the diffusion by arithmetical pro-
gression, another ten minutes of time is equal to a unit
of alkali or another unit of dye. Salts delay diffusion;
we have found out how much of a salt, such as sodium
citrate, is required to decrease this diffusion in equal

EXPRESSED AS AN EQUATION 79

proportion to the increase caused by one unit of alkali,
dye, or time. All the units are equal to each other
as regards the increase or decrease of diffusion, and
therefore they are interchangeable. Hence we may
simplify our equation by adding together all the units of
all the factors and making a grand total of them ; thus :

=13-3,

or, simpler still:

c/ = 10.

This method of determining the coefficent of
diffusion is intended principally to assist experimenta-
tion with these in-vitro technics. The act of its
determination gives up the comparative rate of the
diffusion of other substances into the cells under
observation, and tells us how to prepare jellies for
further experimentation with these substances. For
practical purposes, the equation and the measurements
of the units of the several factors (wThich are used
continually, not only in the initial determination of the
coefficient of diffusion, but in all subsequent experi-
mentation) have been devised with a view to the
simplification of the practical methods to be described
in the next chapter, wrhere full details for the prepara-
tion of the jellies, etc., will be stated. These laws of
diffusion were ascertained in the first instance by me
with the jelly method as described in the paper in the
Journal of Physiology already referred to, and they
soon led to the method of determining the coefficient
of diffusion by the same method wrhich was published
in a paper in the Proceedings of the Royal Society

80 DIFFUSION OF SUBSTANCES INTO LIVING CELLS

(B. Vol. 81) ; and the description of the methods and
laws given herein are in reality an elaboration of the
original ones given in the papers mentioned. Much
work has been done, however, since those papers were
written, including induced cell-division by a variety
of chemical substances, and all of it has been carried
out with those specifications as bases. The point is
mentioned in order to show that the method is fairly
reliable. New technics of this nature, where one is
dealing with a series of factors, all of which are
variables, are prone to become involved in their
application. The "moment" of the staining of the
nucleus cannot be a very accurate one, and the more
factors and units one deals with, the more do small
errors assert themselves.

It is a simple matter to note the effects on a cell
of two or three units of a dye and a unit or two of
alkali. But when one deals with complicated equa-
tions involving twenty or thirty variable units, each
of which modifies the action of its neighbour, it
sometimes follows that complicated situations arise.
For instance, the units of the two salts are satisfactory
when small quantities of them are used; but with
larger quantities it will be found that they are not
quite so accurate. For practical purposes, however,
the units given have been found to be sufficiently
useful ; but if this method was to be employed to
determine the more scientific data of the action of
the several physical factors in increasing and decreasing
diffusion, I am prepared to admit that some units will
require modification.

CHAPTER VI

THE PRACTICAL DETERMINATION OF THE "COEFFICIENT
OF DIFFUSION OF CELLS," AND ITS APPLICATION
TO THIS IN- VITRO METHOD OF RESEARCH

IN the foregoing chapter I endeavoured to give an
outline of the principles of diffusion of substance into
individual cells, and the method of the determination
of the coefficient of diffusion. In the present chapter
I shall describe, in detail, how those principles are
applied experimentally, and how one can find out
the coefficient of diffusion of a given class of cells.
The preparation of the jellies from which the films
are made constitutes the most important part of the
procedure. The chemical substances which are to be
made to diffuse into the cells are contained in the
jelly together with the other chemical factors, which
increase or decrease diffusion. The factor heat is
measured by keeping the slide on which the jelly-film
is set at a certain temperature, and the length of time
the slide is kept at this temperature determines the
amount of the factor time. The coefficient of diffusion
of a cell, as already pointed out, is arrived at by

82 "COEFFICIENT OF DIFFUSION OF CELLS"

ascertaining experimentally the lowest sum of units
of the factors— both chemical in the jelly and physical
as applied to the slide — which will just cause the
cell's nucleus to stain. In the original paper, already
referred to, which specified this method and the co-
efficients of diffusion, the following definitions were
given :

When a film of agar jelly contains stain and
other substances, its Index of Diffusion (/#) may be
defined as the sum of its constituents, which delay
diffusion subtracted from the sum of its constituents
which accelerate diffusion added to the quantity of
stain contained in the jelly.

The Coefficient of Diffusion (c/) of a cell is that
Index of Diffusion plus the time and temperature
required to cause staining of the nucleus, or staining
of the cytoplasm in unnucleated cells (e.g. red cor-
puscles), when the specimen is prepared by a standard
method.

It should be noted that the index of diffusion
refers to the composition of the jelly, and that the
coefficient of diffusion refers to the rate at which
the cell absorbs substances from the jelly.

The standard method of placing the cells on the
jelly-film and the general principles of preparing the
film have already been described. The cells are mixed
with a little "citrate solution" (3-per-cent sodium
citrate and 1-per-cent sodium chloride), which acts
as a vehicle to keep them alive, and in which they
are placed on the cover-glass. Since this citrate solu-
tion spreads to the periphery of the cover-glass, it does

"COEFFICIENT JELLY" 83

not materially influence the diffusion of the stain from
the jelly into the cells. When experimenting with
blood-cells the blood is mixed with an equal volume
of the solution. In the case of other cells the mix-
ture is made as may be convenient. In some instances,
when the cells are naturally suspended in a fluid-
such as pleuritic fluid — it is unnecessary to use any
citrate solution at all, and the cells may be placed,
suspended in their own fluid, straight on to the cover-
glass.

The general principle of preparing the jelly-film,
as given in Chapter III., may be recalled. A 2-per-
cent solution of agar in water forms a jelly basis for
these experiments. This jelly is kept stored in large
test-tubes, so that small quantities of it may be used
without having to melt it in bulk every time some
is wanted; and it should be filtered when it is made
in a manner similar to that employed for the prepara-
tion of "nutrient agar," although, of course, it has
no "nutrient" ingredients added to it.

As already mentioned, the 2-per-cent solution of
agar has such a consistency that it can, when melted,
be diluted with an equal volume of a liquid and yet
will set as a firm jelly on a slide when it cools.

Experimentation with this method is essentially a
process by which one contrasts the effects of one sub-
stance on cells compared with those of others; hence it
is important that all the conditions must be the same
in each experiment, except the actual difference in the
amount of the substance which has to pass into the
cells. To this end the jelly-basis is always the same in

84 COEFFICIENT OF DIFFUSION OF CELLS

every way, and each film is always made from a tube
containing 10 cc. of this jelly. The substances which
are to be tried on the cells are added to the jelly in the
form of the Solution 2 (see Chapter III.), which in its
turn is added to the Solution 1. The combination
(Solution 3) is always in the quantity of 10 cc., and the
film is prepared from this.

It has already been shown that the jelly-film must
always contain certain quantities of the salts sodium
citrate and sodium chloride, or the cells will not live on
it. These salts are therefore added to the jelly-basis or
Solution 1. They are added to it in bulk, so that any
portion of it contains them, and, in consequence, it is in
a condition to cause cells to live on it for as long as
possible.

The jelly is prepared as follows :

In a beaker of water stand several of the large test-
tubes which contain the stock 2-per-cent agar jelly.
The amount required will be at least 50 cc. The
water in the beaker should be heated until it boils,
when the jelly in the test-tube will be melted.

1 gramme of sodium citrate and then 0.8 gramme
of sodium chloride should be weighed out accurately.
The two salts are then placed in a flask, which should
be of such a size that it also can be steeped in the
beaker of boiling water; 49 cc. of the molten 2-per-
cent agar solution from the test-tube are now measured
out and poured into the flask. The salts slowly dis-
solve in the molten agar, and, while this solution is
taking place, the flask should be steeped in the boiling
water in order to keep the jelly molten.

"COEFFICIENT JELLY" 85

It is important that the sodium citrate should be
neutral. Sodium citrate is inclined to become alkaline
when exposed for long to the air, owing to the deposit
of sodium carbonate. The jelly in the flask, therefore,
must be tested and neutralised to litmus with citric
acid.

Previous to melting the jelly solution, a solution
containing 8 . 3 per cent of citric acid should have been
prepared; and now 1 cc. of that solution is added to
the 49 cc. of the molten agar solution in the flask.
This renders the whole of the jelly acid, the reason for
which will be given directly.

A series of ten clean test-tubes should be ready, and
with a pipette 5 cc. of the acid jelly with its salts in
solution is measured into each test-tube. Each of the
ten test-tubes now contains 5 cc. of the jelly: total
50 cc. in all. The 10 test- tubes are placed in a stand
until the jelly is set, and a plug of wool is placed in
each; otherwise moulds may growT on the jelly, as
it contains salts.

Every one of the 10 test-tubes contains 5 cc. of
a 2-per-cent agar jelly, which is acid, because it contains
in solution 0.0083 gramme of citric acid. It also
contains 0 . 1 gramme of sodium citrate and 0 . 8 gramme
of sodium chloride ; and these tubes of jelly are known
for convenience as tubes of "coefficient jelly."

To any one of these tubes we may add 5 more cc.
of any solution or solutions; and if the whole is boiled
and mixed by shaking, any portion of the 10 cc. of
jelly mixture now contained in the test-tube will
set on a slide as a firm jelly-film when it cools.

86 "COEFFICIENT OF DIFFUSION OF CELLS"

Since it is essential that all jellies must be alike
in all respects except in the actual quantities of the
chemical substances which are to be tested on the cells,
it follows that every jelly-film is always made from
10 cc. of jelly. A film is never made direct from a tube
of 5 cc. of "coefficient jelly" unless it previously has
had added to it an equal quantity (5 cc.) of some
solution. If this rule is followed, every jelly-film
wTill be identical in that the strength of the agar will
be the same, and the initial strength of the salts and
acid will be the same; but since the second 5 cc. may
be composed of any solution, each 10 cc. of jelly may
also contain a variety of other substances.

It is in the extra 5 cc. of solution or solutions that
the chemical substances, with which one wishes to
experiment on the cells, and any chemical factors
additional to those already contained in the "coefficient
jelly" which are required to increase or decrease dif-
fusion, are added to that "coefficient jelly."

The chemical factors, therefore, such as alkalies
and salts, which increase or decrease diffusion assist
to constitute the second 5 cc. of jelly which is always
added to the 5 cc. of "coefficient jelly." Now, one
could, of course, weigh out the right number of units
of each factor for every experiment, but it is much
simpler to add them from standard solutions. These
standard solutions should be kept ready to hand in
flasks, on the labels of which should appear the exact
amount of each which constitutes one unit.

The 'same may be said of the chemical substances
the action of which one wishes to try on the cells.

UNITS OF THE FACTORS 87

For instance, in the determination of the coefficient
of diffusion, the stain, as well as the alkali, is kept
in standard solution, and is added to the 5 cc. of
"coefficient jelly"; but it is most important to re-
member that no matter how many units of each
factor or substance may be contained in the 5 cc. of
solution added to the 5 cc. of "coefficient jelly," the
former solution must never be more nor less than 5 cc.
Therefore, every jelly-film on the slide is always made
from 10 cc. of jelly, which in its turn is composed of 5 cc.
of "coefficient jelly" and 5 cc. of another solution bearing
the units of the chemical factors. No matter how
many units of no matter how many factors the second
5 cc. of solution contains, it is always added in the
quantity of 5 cc. — no more and no less. Hence, the
Solution 3, from which the film is prepared,will invari-
ably consist of 10 cc. Solution 2 may contain one unit
of one factor, or it may contain any number of units
of any of the factors.

If all the units of the contained factors exactly
amount to 5 cc., all well and good; but if they do
not do so, the balance must be made up to 5 cc. with
water. By this means there will always be 10 cc. in
the tube of jelly used for an experiment, but it may
contain a great variety of units of the chemical factors
which increase or decrease diffusion.

The standard solutions of the several factors must
be prepared with due regard to this rule. They must
not be too dilute or their total may exceed 5 cc. The
following list (abridged from the original paper on
the "Coefficient of Diffusion") gives not only the actual

88 "COEFFICIENT OF DIFFUSION OF CELLS"

units of the several factors, but also convenient standard
solutions of them. It is useful to keep this list ready
to hand in the laboratory.

Alkali (sodium bicarbonate) increases diffusion.—
Unit 0.005 gramme. Standard solution 5 per cent,
unity being 0.1 cc. It is convenient to remember
that this solution is neutralised by a 4. 175-per-cent
solution of citric acid, and that 1 unit of alkali is
neutralised by 0 . 1 cc. of such a solution. Since the
agar at the outset is acid to the extent of 0 . 083 gramme
to 50 cc., a tube of 10 cc., made up as described, must
contain 0 . 0083 gramme of acid. This is exactly neutral-
ised by 0.2 cc. of the standard alkali solution; that is,
the agar at the outset, before any stain or other factor
is added, delays diffusion to the extent of 2 units.
Or, the addition of 2 units of sodium bicarbonate
will render the agar neutral.

Sodium Citrate delays diffusion. — Unit 0 . 03 gramme.
Standard solution 10 per cent, 0.3 cc. being unity.
Since 50 cc. of agar contains 1 gramme at the outset,
the 10 cc. of jelly may be said to contain about 3 units.

Sodium Chloride delays diffusion. — Unit 0 . 08 gramme.
Standard solution 10 per cent, unity being 0.8 cc. The
10 cc. of jelly contains this from the outset.

Heat hastens diffusion. — Each unit 5° C. ; 10° C. is
unity, 15° C. is 2 units, 20° C. 3 units, etc. For
practical purposes I call 37° C. 7 units.

Time increases diffusion. — Ten minutes is 1 unit,
twentv minutes 2 units, and so on.

REASON FOR ACIDITY OF JELLY 89

Stain, Unna's polychrome methylene blue (Grubler),
behaves as if it increased diffusion. — Unit 0.1 cc.

The reason why the "coefficient jelly" is made acid
at the outset is this. Alkalies increase diffusion;
acids delay it. Acids neutralise alkalies, and between
the two there is a neutral point. If the jelly is
neutral at the outset, we might have to add acid in
the case of a cell having a very low coefficient of
diffusion. Again, we may frequently have cases of
cells which stain on a neutral jelly. Our equation,
therefore, would have to include these three factors
— alkalies, acids, and a neutral point — which would
be very complicated, as the neutral point would
introduce zero. The object throughout has been to
make the determination of coefficient of diffusion of
cells, and the estimation of diffusion of substances
into them, as simple as possible in their practical
application, and in order to do this the "coefficient
jelly" is rendered acid at the outset and one deals
only with the one factor — alkali. The original 50 cc.
of jelly contains 0 . 083 gramme of citric acid ; therefore
each tube of 5 cc. of "coefficient jelly" contains 0.0083
gramme of citric acid; and each tube will ultimately
be made up to 10 cc., which, of course, will also
contain this amount of citric acid, unless it is
neutralised by alkali. This 0.083 gramme of citric
acid represents 2 units of acid, and it is neutralised
by 2 units of alkali. If we wrant to try a jelly which
is acid to the extent of 2 units, wTe simply add no
alkali. We are not likely to want a jelly which is

90 "COEFFICIENT OF DIFFUSION OF CELLS"

more acid than this, for we have never yet seen
any cell stain on a jelly which is acid beyond the
extent of 1 unit. We must remember all along,
however, that the jelly at the outset is acid to the
extent of 2 units, and then go ahead with alkali. If
we add to Solution No. 2 10 units of alkali, we say
that the jelly contains 10 units; but in reality it is
only alkaline to the extent of 8 units, for two of
them have been utilised in neutralising the original
2 units of acid. Of course, the neutralisation of
the acid increases the content of sodium citrate to
a slight extent, but it is so small that it can be
neglected. As a matter of fact, by saying that the
content of sodium citrate is usually 3 units, which is
in excess of the reality, we compensate for the extra
salt produced by neutralisation of the acid. The
neutral point we ignore. If a jelly contains only
2 units of alkali, it is in reality neutral; but wre
need not trouble about that. There is no neutral
point in the equation, nor is there a symbol for acid;
yet the neutral point and acid both exist in the
equation, for the symbol "2a" means neutrality;
and the symbol "a" means 1 unit of acid, whereas
the absence of the symbol "a" means 2 units of acid.
To recapitulate: Acids and the neutral point are
omitted from the equation, but the jelly is acid at
the outset, and we deal only with alkali. If a jelly
contains only 2 units of alkali, that jelly is neutral.
If a jelly contains 15 units of alkali, it really is only
alkaline to the extent of 13 units. The jelly basis
with which we work is known as "coefficient jelly."

METHOD OF DETERMINATION 91

It is kept in quantities of 5 cc. in test-tubes ready
to hand. Each "coefficient jelly" contains sufficient
salts for cells to live on it; it is acid to the extent
of 2 units; and another 5 cc. of some solution must
be added to it before it is poured on to a slide to make
the "jelly-film." The film is always made from 10 cc.
of jelly.

In experimenting with a certain class of cells,
one must in the first instance always estimate their
coefficient of diffusion. The cells are mixed with
"citrate solution" and kept ready at the room
temperature, preferably in the revolving apparatus
(see Chapter II.).

In order to determine the coefficient of diffusion
of these cells a tube of "coefficient jelly" is taken and
a few units of stain are added to it, together with
2 or 3 units of alkali solution. The content of the
tube is then completed up to 10 cc. with water. The
tube is steeped in the beaker of boiling water until the
"coefficient jelly" all melts, when the stain and alkali
become mixed with it, as will be presently described.
A film is prepared from it on a slide, and a drop of the
citrate solution, with the cells in suspension, placed on
to it under a cover-glass. The specimen is kept at a
certain temperature — representing a certain number of
units of heat — until a certain number of minutes — repre-
senting a certain number of units of time — have elapsed,
and then the specimen is examined under the micro-
scope. If the nuclei of the cells are not yet stained, a
few more minutes (e. g. another unit of time — ten min-
utes) are allowed. If, then, the nuclei are not stained,

92 "COEFFICIENT OF DIFFUSION OF CELLS"

a fresh film is made from the same jelly, and it is kept
at a higher temperature — or so many more units of
heat — and again examined. If again it is found that the
nuclei are not stained at the end of, ,say, half an hour,
a fresh jelly is made, but with more units of alkali or
stain, or both, added to a fresh Solution 2, which is
added to a fresh tube of "coefficient jelly." If the
nuclei still again remain unstained, one must try more
units of time and more units of heat again. Thus we
can go on trying fresh jellies, each of which contains
more units of alkali, or of stain; and we try each jelly
for a few minutes, first at a low temperature and then
with a few more units of heat, until at last we find that
the nuclei of the cells are just beginning to stain. The
number of units of stain, alkali, heat, time, etc., of each
film is noted on a piece of paper, and therefore there is
no difficulty in knowing exactly how many units the
jelly contained which was instrumental in staining the
nuclei of the cells. The units of this jelly are then
written out in the form of an equation, and those
which retard diffusion — i.e. the units of the salts-
are subtracted from those which increase diffusion,
the difference being the number which is the co-
efficient of diffusion of the class of cells experimented
with.

Examples. — We wish to find the cf of the neut.ro-
phile polynuclear leucocyte. A small quantity of
citrate solution is drawn up into a capillary tube, as
already described, and, the finger having been pricked
and a small bead of blood squeezed out, an equal
volume of blood is added to the citrate solution in

EXAMPLES 93

the capillary tube, which is placed in the "revolving
apparatus."

Take a test-tube of 5 cc. of "coefficient jelly," which
of course, being cold, is set in the bottom of the tube.
Add to it 0.4 cc. (4 units) of Unna's stain; 0.2 cc.
(2 units) of alkali solution. Then the tube must have
added to it 4.4 cc. of water, to make its total contents
up to 10 cc. The colorless "coefficient jelly" will be
set at the bottom of the tube, and above this will
be 5 cc. of the mixture of stain, alkali, and water.
The test-tube is then steeped in boiling water, when
the jelly melts, and, as it does so, the stain, alkali, and
water pervade the whole of its contents of 10 cc. In
reality this 10 cc. of molten jelly is neutral, for the
2 units of alkali have just neutralized the original
acidity of the "coefficient jelly." When all is melted
and mixed, the tube is taken out of the boiling water,
and the contents are actually boiled, until they froth up
in the tube, by holding the end of the tube in the flame
for a few minutes. A drop of the boiling, stained
mixture is then run on to the slide. Here it will set
firmly in about three minutes, and if it is held up to
the light the jelly-film will be found to be quite trans-
lucent. A clean cover-glass is prepared, and a drop of
the citrated blood is tapped out of the capillary tube on
to it. The size of the drop is immaterial. The cover-
glass is taken up bet\veen the finger and thumb, in-
verted so that the drop of fluid is undermost, and it
is allowed to fall flat on to the agar-film on the slide.
The blood spreads over the film under the cover-glass,
and the slide is then placed in the 37° C. incubator

94 "COEFFICIENT OF DIFFUSION OF CELLS"

(7 units of heat) for 10 minutes (1 unit of time). The
index of diffusion of the jelly is this:

/ar=(4* + 2o) — (3c + n),

where s = unit of stain, a = unit of alkali, c = unit of
sodium citrate, and n = unit of sodium chloride.

At the expiry of the ten minutes the specimen is
examined, when it will be seen that the lymphocytes
are quite unstained; but the granules of the polynu-
clear leucocytes are just beginning to colour red. To
find the cf of these cells, however, it is stipulated that
their nuclei should just stain. The specimen is there-
fore replaced into the incubator for a further ten
minutes. Now it will be found that the nuclei of
the eosinophile leucocytes are just staining. Hence,
although this jelly has not yet given us the coefficient
of diffusion of the neutrophile leucocyte, it has de-
termined that of the eosinophile cell, which may be
set down as follows:

Eosinophile leucocytes

where h = unit of heat and / = unit of time.

In order to stain the nuclei of the neutrophile cells,
we could place the same specimen for another ten
minutes in the incubator; but it is not a very safe
thing to do, for the cells by this time may be dead.
It is better to make a fresh film from another jelly
which contains more units of a factor which increases
diffusion. We may add more stain or more alkali.
Let us try another unit of each, thus: To a fresh tube

EXAMPLES 95

of 5 cc. of "coefficient jelly" add 5 units (0.5 cc.) of
stain, 3 units of alkali (0.3 cc. of a 5-per-cent sodium
bicarbonate solution) , and 4.2 cc. of water to make
it up to the 10 cc. of jelly. The film is prepared as
before, and it is incubated at 37° C. for 10 minutes.
On examination, it will be seen that the nuclei of the
neutrophile cells are just turning scarlet. Hence this
jelly at 37° C. in 10 minutes has the right Index of
Diffusion for the coefficient of diffusion of neutrophile
leucocytes. The equation may be thus set down:

Neutrophile leucocytes

cf = (5s

The lymphocytes have a cf of 14 (2 units higher
than that of the neutrophile cells). We may cause
their nuclei1 to stain in 10 minutes at 37° C. by using
a jelly similar to the last one, but by either increasing
the amount of alkali by 2 units, or by increasing the
concentration of the stain by 2 units, or by increasing
the alkali by 1 unit and the stain by 1 unit; or by
increasing the time by 2 units. Let us try a jelly
which contains 2 more units of stain, for now the
chromatin of the cells will stain deeply and show up
well. The jelly is made thus: To a tube of 5 cc.
of "coefficient jelly" add 7 units (0.7 cc. of stain), 3
units of alkali, and, since we now have more stain,
only 4 cc. of water is needed to make the contents
of the tube up to 10 cc. The whole mixture is boiled
and a drop of it spread on a slide in the usual manner.
After the blood has been mounted on it the slide is

1 See Chapter XII.

96 "COEFFICIENT OF DIFFUSION OF CELLS"

incubated at 37° C. for 10 minutes, when it will be
seen that the nuclei of the lymphocytes have turned
scarlet.

Lymphocytes

In this specimen the neutrophile leukocytes will
have burst, for the jelly has an index of diffusion too
high for them by 2 units — their cf being 12. For
the same reason the eosinophile cells will also be achro-
matic, and the same may be said of the basophile cell,
although it is very difficult to stain the nuclei of these
cells. Their of, however, is about the same as that of
the neutrophile leucocyte.

The simple equation has other advantages. It can
be inverted, so to speak, and the units of the different
factors can replace each other to some extent; for since
the units of the several factors are equal to one another
as regards their power of increasing or decreasing the
diffusion, one can interchange them at will. We can
make two jellies, for instance, one of which contains
5 units of stain and 2 of alkali; and another which
contains 2 units of stain and 5 of alkali; and provided
the other factors are the same in the films made from
each tube, the result obtained by each jelly will be
identical. The equations will both give the same
total :

=9.
cf = (2s + 5a + 4/z, + 2<) - (3c + n) =9.

Any of the factors may thus be interchanged.

We may delay this diffusion by adding more units

REVERSING THE EQUATION 97

of salts. The (3c+n), however, is the usual content of
salts which is always present in the "coefficient jelly,"
but more salts may be added in the shape of solutions
to the 5 cc. which also contains the stain and alkali.
Whatever is added must be put down in the equation.
The only substance not in the equation is agar,
which, as already noted, does not affect the cells, and
which is always present in the same strength in every
experiment.

Since the units of the factors are equal and inter-
changeable, and since their sum is equivalent to the
coefficient of diffusion, the numeral which constitutes
the coefficient of diffusion in the equation can be inter-
changed with an equivalent number of units of one
or more of any of the factors. We may reverse the
equation, therefore, and, provided we already know the
coefficient of diffusion of the cell experimented with,
we can, by this reversal, determine in a moment the exact
quantity of any factor required to obtain staining of the
nucleus. That is to say, that if the coefficient of diffusion
is known, and if all the other factors, except one, are
given quantities, then we can determine the required
quantity of the one unknown factor simply by reversing
the equation; always remembering the well-known
algebraic law that in bringing one factor from one side
of the equation to the other, we must change the signs.
For instance, suppose a strain of Spirochceta refringens
has a coefficient of diffusion of 20, and one wishes to
stain a sample of them : Let us suppose there is a jelly
to hand which contains 6 units of stain, 8 units of alkali,
and the usual content of salts in the "coefficient jelly"

98 "COEFFICIENT OF DIFFUSION OF CELLS"

from which it was prepared. The total contents of the
tube has already been made up to 10 cc. as usual. The
specimen is prepared and incubated at 37° C. Then the
question must be asked, How long must the specimen
remain in the incubator before the spirochsetes will be
stained ? We could, of course, keep taking the speci-
men out and looking at it, but every time we did this
we should lower the temperature and spoil the experi-
ment. It is much simpler to plot the equation. The
coefficient of diffusion of the spirochsetes is a known
quantity, i.e. 20; the time is now the unknown factor.
We therefore exchange the places of the symbols cf
and t, thus:

=3.
t= 3, or 3 units of time, i.e. half an hour.

Likewise, since the units of all the factors are equal,
we may interchange any of them. Another example
may be given. A certain strain of Amoeba coli from
a "culture" has a cf of 13. We want to stain the
nuclei of these parasites in 10 minutes with a jelly
which contains 7 units of alkali. But we want to
stain them at the room temperature of 20° C. The
jelly contains its usual content of sodium citrate and
sodium chloride — i.e. 3 units of the former and 1 of
the latter. How much Unna's stain must we add to
the tube of jelly to obtain the required result ?

The number of units of s is the quantity required,
hence :

PRECAUTIONS 99

six units of stain will be required, or 0.6 cc. of Unna's
dye.

The foregoing examples show how the coefficient of
diffusion is determined, and how, after it has been
ascertained, one can, by means of the equation, find
out other factors, which may be unknown quantities.
It follows that by this procedure other substances can
be made to diffuse into the cells. This method of
calculation has been used throughout these researches,
and it will be seen that further examples will be given
in the future chapters of this book.

The factors most often employed are alkali and heat.
Salts are not usually varied a greal deal, although their
amounts can be altered if necessary by adding more of
them to the second solution.

The determination of the units of any other sub-
stance is carried out on the principle that all units must
be equal. Let us take a substance like urea, for in-
stance. It delays the diffusion of other substances,
such as Unna's stain. All that has to be done is to
find out how much urea must be contained in the
10 cc. of jelly to neutralize the increasing action of a
unit of alkali. Having found out the unit of the fresh
substance, that unit is added to the equation in the
usual wray. If it increases diffusion it is placed in
the bracket with the alkali and heat; if it delays diffu-
sion it is bracketed with the salts.

Lastly, having obtained the coefficient of diffusion
of a class of cells by measuring the rate of diffusion
of the stain, the stain may be omitted and any other
substance substituted for it. If more than one sub-

100 " COEFFICIENT OF DIFFUSION OF CELLS"

stance is made to diffuse into the cells, they may each
affect the diffusion of the other; for they may them-
selves be alkalies, acids, or salts. In this case the unit
of each must be found, and their action on the diffusion
of other substances into the cell allowed for.

It is necessary to point out that this method is
reliable only within certain limits, and precautions must
be taken to prevent errors due to employing excessive
amounts of substances, heat, and time, and those due
to possible contingencies arising when dealing with cells
from tissues.

The following list of precautions has been copied
from the paper in the Proceedings of the Royal Society:

Precautions. — As regards Life and Death: In a
previous paper1 it has been shown that the staining
of the nuclei of leucocytes, when examined by this
method, is a sign of death, and that the nuclei of dead
cells will stain, ceteris paribus, before those of living
cells. Consequently all the experiments given in the
present paper have been made with fresh normal cells,
and in the case of micro-organisms, with cultures not
more than forty-eight hours old. It may also be men-
tioned that the liquefaction of the cytoplasm which
occurs after death materially alters the conditions of
staining of leucocytes, and that the cf of living cells
falls gradually after the blood has been shed.

As regards Excess of Alkali, causing rapid death
and liquefaction of the cytoplasm, with consequent
prevention of staining (achromasia) : The addition of

1 "On the Death of Leucocytes" (H. C. Ross, Journal of Physiology,
vol. xxxvii., p. 327,1908).

PRECAUTIONS 101

excess of alkali may cause death, staining of the nuclei,
liquefaction, and the loss of stain on the part of the
cells.1 This may occur before a preparation can be
focused, in which case the cells appear unstained
and will refuse to stain, no matter how much stain or
alkali are tried. Therefore it is better to begin with a
low index of diffusion and to try tube after tube, each
containing a little more alkali, until staining is ob-
tained. Further, the amount of sodium bicarbonate
should not exceed twenty units, because, as has already
been pointed out in a former paper,1 if added to excess
it may act as a neutral salt and delay diffusion.

As regards Excess of Deficiency or Heat: A tem-
perature above 40° C. may allow the cells to diffuse
through the agar.2 A temperature below 15 degrees
has not been experimented with, because, even at a
temperature of 20° C. it requires a minimum of 3 units
of stain to cause staining of the nuclei of leucocytes
in spite of the addition of a large amount of alkali,
for the alkali is not sufficient, per se, to cause the
cells to absorb sufficient stain to colour the nuclei
unless the stain is concentrated.

As regards Excess of Time: A period of more
than half an hour has not been employed for fear
of death and liquefaction of the cytoplasm, for the
cells may die and become achromatic before there
has been time for sufficient stain to diffuse into them
to cause staining of the nuclei, in which case, of course,
the cells will never stain.

As regards Excess of Stain: More than 10 units
of stain may cause precipitation of the agar as the
film cools on the slide, and the precipitate carries

1 "On the Death of Leucocytes" (H. C. Ross, Journal of Physiology,
vol. xxxvii., p. 327, 1908).

"The diffusions of Red Blood Corpuscles through Solid Nutrient Agar"
(H. C. Ross, British MedicalJournal, May 5, 1906).

102 "COEFFICIENT OF DIFFUSION OF CELLS"

some of the stain down with it, vitiating the results,
for it has been shown that agar is not very soluble
in cold stain.1

As regards Examination: The observation of cells
floating near a bubble under the cover-glass should
be avoided. The fact that blood-cells in such a
situation will stain before others has already been
noted.1 I consider this to be due to these cells floating
in a small quantity of alkaline citrated plasma collected
round the bubble.

Consequently the experiments have all been made
within the compass of the above restrictions. It may
also be advised that when unnucleated cells contain
granules in their cytoplasm the staining of the gran-
ules gives a more constant rate than the staining of the
cytoplasm. By this means it is seen that the cf of the
blood-platelet is identical with that of the polymorpho-
nuclear cells.

1 "On the Death of Leucocytes" (H. C. Ross, Journal of Physiology,
vol. xxxvii., p. 327, 1908).

CHAPTER VII

DIFFUSION OF SUBSTANCE INTO CELLS TO EXCESS-
DIFFUSION- VACUOLES OR "RED SPOTS" —THE PROOF
THAT THE BLOOD-PLATELET IS A LIVING CELL

IN this chapter I shall discuss the effects of the diffusion
of substances into a cell, when that diffusion occurs to
excess. A cell's protoplasm can utilize only a certain
amount of the dissolved substance or substances which
diffuse into it from the immediate neighbourhood of
the cell. One can, however, push this diffusion by the
agency of one or both of the factors — heat and alkali
—which increase diffusion, and if wre do this some of
the neighbouring liquid itself passes into the cell and
remains suspended as minute droplets in the cytoplasm.
These droplets have been called " diffusion- vacuoles."
When they were first seen, five years ago, the cells were
resting on stained jelly/and since the "diffusion- vacuoles"
were stained they were therefore called "red spots."

Diffusion-vacuoles must not be confounded with
the ordinary vacuoles (fig. 16) which appear as colour-
less, non-granular patches in leucocytes. Many theories
have been advanced regarding these latter vacuoles,
but although we have so often seen them, we have no
explanation to offer as to their nature. They are

103

104 DIFFUSION- VACUOLES

certainly not composed of liquid; they are not cavities;
and, so far as we have observed, they play no part in
cell-division. When the cytoplasm liquefies at death
they disappear, and when a cell divides they seem to
migrate into the cytoplasm, remaining outside the
chromosomes and centrosomes.

The diffusion-vacuole is quite another kind of body
(fig. 17). It is never seen in a normal cell which has
been freshly removed from the tissues. "Red spots"
alwrays appear gradually (fig. 18), beginning as minute
coloured points in the cytoplasm, which gradually
become larger until — in the case of leucocytes — they
may become as large as a lobe of the nucleus. Two
or more may coalesce to form one large diffusion-
vacuole; and their appearance depends entirely on the
laws of diffusion; in fact, they may be produced in
leucocytes at will by arranging the plus factors, heat
and alkali, in the equation in such a way that they
promote the diffusion of a substance to excess.

Diffusion-vacuoles appear only in living protoplasm.
After death the cytoplasm liquefies and the cell
becomes disorganized, when diffusion-vacuoles cannot
appear in it. The actual passage of a substance, say,
stain, through a living cell's cytoplasm occupies a
certain amount of time, which can be shortened by
Jieat or alkalies and lengthened by salts. If heat and
alkali are present, but the salts are absent, the stain
diffuses into the cell so quickly that death may ensue
in a few moments, because the nucleus becomes stained..
Indeed, one may thus cause death in a few seconds;
and death is accompanied by liquefaction of the cyto-

THEIR NATURE

105

FIG. 16. — A stained leucocyte. The ordinary vacuoles (colourless patches
amongst the cell-granules) are well shown. The cell has just died.

FIG. 17. — Diffusion-vacuoles in a leucocyte.

THEIR NATURE 107

plasm, which, when it is alive, appears to be in the form
of a jelly. Now, it is obvious that if the cytoplasm
liquefies in a few seconds, diffusion-vacuoles cannot ap-
pear, for it is unlikely that a liquid like a solution of stain
cannot remain suspended in droplets in another liquid
like liquefied cytoplasm. On the other hand, if the
cytoplasm is alive and jelly-like, any excess of stain which
diffuses into it will become suspended in it as a "red
spot." Hence, if death is caused extremely rapidly, no
matter to what excess the diffusion is increased, diffusion-
vacuoles will not appear, and, owing to the excess,
liquid passes into the cell. If this excess is great, the
dead cell will be seen to burst (it appears even to
explode sometimes, especially if there are no salts to
delay the diffusion), and the cell-granules are scattered
about the field of the microscope. It is a well-known
fact that if water is mixed with blood, the leucocytes
will burst, the reason being the same, for the water
passes into the killed and liquefying cytoplasm, and the
intracellular tension is so great that rupture of the cell-
wall takes place. There are no salts to delay the
diffusion of the water, which now occurs to such excess
that it causes the cell to rupture.

In order to demonstrate the diffusion-vacuoles,
therefore, it is necessary to delay death, which can
be done by placing salts in the jelly- film such as are
present in the "coefficient jelly." Diffusion is then
increased by alkali or heat until maximum diffusion,
short of causing death, is obtained; for it must be
remembered that all artificial substances will kill
human cells, and the more they diffuse into them the

108 DIFFUSION- VACUOLES

more rapidly will the cells die. If now half a unit
more of alkali is added to the jelly, or two or three
more degrees of temperature are tried, diffusion-
vacuoles will gradually make their appearance in all
the cells.

For instance, "red spots" are readily produced
in leucocytes. Any jelly which has the correct index
of diffusion for a coefficient of diffusion of 12 will
cause them to appear if another drop of alkali is
added to the jelly. The diffusion of the stain must
be excessive; but not so excessive as to cause death
in a few seconds. It is necessary to hit off those
amounts of alkali and heat which will cause liquid
to pass into the cells, but which will not unduly
hasten death by staining the nucleus too rapidly.
If this is done accurately, these remarkable diffusion
vacuoles suddenly begin to appear. If the diffusion
is still further increased, the cells will burst and
become achromatic instantly. The appearance of
the diffusion-vacuole may be regarded as the safety-
point of diffusion; and it is a signal that no more
alkali or heat may be tried, or the cells may burst.

It is interesting to watch the fate of these vacuoles.
Since the substance is diffusing into the cells to excess,
this diffusion must cause death in a short time,
.even though the cells do not burst. Before this
occurs, however, the diffusion steadily increases, and
the "red spots" get larger. When death takes place
the cytoplasm liquefies slowly, beginning at the.
periphery and progressing more and more towards
the nucleus. As the cytoplasm liquefies, more and

DISPERSAL OF THE VACUOLES

109

FIG. 18. — A dead leucocyte in which diffusion- vacuoles are beginning to

appear.

FIG. 19. — A ditiusion-vacuole in a lymphocyte. Low power.

DISPERSAL OF THE VACUOLES 111

more of the cell-granules show the remarkable danc-
ing Brownian movements, and the liquefying cyto-
plasm gradually involves the diffusion-vacuoles, one
by one. When the liquefying cytoplasm, which
immediately surrounds a vacuole, becomes of the
same consistency as the liquid within the vacuole,
the latter, which in reality is like a bubble of liquid
suspended in a liquefying jelly, suddenly bursts and
disperses, leaving a cup-shaped cavity in that portion
of the more central cytoplasm which has not yet
become liquefied. One by one all the vacuoles
disperse, and either immediately before or after
their dispersal general achromasia of the cell ensues,
for achromasia also depends on the liquefaction of
the cytoplasm. Vacuoles have never been seen to
"disperse" in a living cell; it is necessary for the
cytoplasm to liquefy for this to happen, and lique-
faction occurs only at death. Diffusion-vacuoles will
frequently be seen when experimenting with this
in-vitro method, large numbers of them sometimes
making their appearance in a single cell; but they
will all disappear after a short time. I have seen
them in all varieties of blood-cells (figs. 19, 20).

The colour of the diffusion-vacuole depends on the
colour of the solution or jelly in \vhich the cell is
resting. If the jelly contains a red dye, such as
polychrome blue, the vacuoles will be red; ordinary
methylene blue causes them to appear blue. If no
stain is present the vacuoles will be colorless; if
stain is present the coloration of a vacuole is always
deeper than the colour of the surrounding jelly. We

DIFFUSION- VACUOLES

believe the reason for this is, that when the droplet of
liquid becomes suspended in the jelly-like cytoplasm
it forms a cavity in it, and the walls of the cavity
actually become stained. This is readily seen when
the vacuoles disperse, for portions of the stained wall
of the cavity can be demonstrated. When cytoplasm
is wounded (the formation of a vacuole in it really con-
stitutes a wound of it) the cytoplasm stains deeply with
an anilin dye, and this appears to be the reason why the
"red spots" seem to be so deeply coloured. Moreover,
being spherical droplets, they are highly refractile.

We have never seen diffusion-vacuoles in normal
cells immediately after they have been removed from
the body; it is always necessary to induce them.
There is an exception to this rule, however, in the
cells of some malignant growths, especially cancer of
the breast, in which we have frequently seen large
"red spots." We think that it is possible that the
injured cytoplasm associated with these spots may
produce the deeply staining patches which have been
described as "archoplasm" in these cells when they
are fixed and stained by the older methods. In a
former paper we also suggested that archoplasm might
be derived from chromatin which has diffused through
the cytoplasm to some extent, and we still think that
this may be possible, but it is also probable that archo-
plasm is derived from the fixation of injured cytoplasm
connected with a dift'usion-vacuole. We have never
seen anything like the commonly described archoplasm
in a normal living leucocyte, and it certainty does not
play any role in their cell-division.

CAUSED BY A LOWERED COEFFICIENT 113

There appears to be little doubt that archoplasm
does not exist normally in a living cell; it can be
produced in them, however, by lowering their co-
efficient of diffusion by keeping them for some hours
in extracts of dead tissues — aYid this is, we believe, the
reason why it appears so frequently in living cancer-
cells.

It is obvious that substances will, ceteris paribus,
diffuse more readily into a cell if it has a coefficient
of diffusion lower than its normal one, and, for the
same reason, vacuoles can more easily be induced in
it. For instance, no diffusion-vacuoles will appear in
fresh leucocytes when they are resting on a jelly which
will not cause the maximum diffusion of stain into
them; but if we lower their coefficient of diffusion,
and again place them on the same jelly, not only may
the maximum amount of stain now pass in, but it may
pass in to excess, and diffusion-vacuoles will appear.
This fact has led to the determination of a point of
scientific interest which has been controversial for more
than half a century. It has proved that the blood-
platelet is a living cell1; for diffusion-vacuoles will
not appear in the normal blood-platelets, but if their
coefficient of diffusion is lowered by causing gradual
death, the lowering of the coefficient of diffusion so
occasioned will now permit "red spots" to be induced
in them.

Our experiments up to the present have revealed
the fact that the coefficients of diffusion of all cells
fall gradually as their vitality fails, provided this loss
of vitality is gradual. The coefficient of diffusion of

'A paper by myself on "V. Ph. Vacuolation of Blood-platelets " was
published in The Proceedings of the Royal Society, B., vol. Ixxxi, 1909.

114 DIFFUSION- VACUOLES

leucocytes may fall by as much as one unit if the cells
have been shed for more than twenty-four hours and
kept in citrate solution at the room-temperature.
There are certain substances also which expedite this
loss of vitality and its accompanying lowering of the
coefficient of diffusion, especially certain alkaloids and
extracts of dead tissues; and it was in the course of
experimentation with the alkaloid morphine hydro-
chloride that diffusion-vacuoles were seen in the
blood-platelets for the first time.

The events which led to the discovery of diffusion-
vacuoles in blood-platelets are worthy of mention.

Soon after this method of in-vitro staining was
suggested by Professor Ronald Ross about five years
ago, either he or I saw the "red spots" for the first
time in leucocytes. The laws of diffusion which I
have described were not then known, and only minute
vacuoles had been seen in the cells, for alkalies had
not been employed. These spots only appeared as
minute red points in the cytoplasm, and in appear-
ance they certainly resembled the centrosomes of
plants and other cells; for it must be remembered
that hitherto leucocytes have never been seen to divide,
and no one knew what their centrosomes were like.
In the preliminary note in The Lancet1 by Professor
Ross and Messrs. Moore and Walker, in which this
in-vitro method was first described, these "red spots"
were mentioned, and it was suggested that, from their
appearance, they might possibly be centrosomes. Now,
it is well known that the nature of the blood-platelet

lThe Lancet, July 27, 1907.

NATURE OF BLOOD-PLATELETS

115

FIG. 20. — A diffusion-vacuole in a granular red cell.

FIG. 21. — A clump of normal blood-platelets. They are resting on a jelly
which will just stain their granules.

NATURE OF BLOOD-PLATELETS 117

(fig. 21) has been a matter of great controversy for many
years ; some say that they are normal constituents of
the blood, but are precipitates of the plasma. Others
say that they are extruded nuclei of red cells, and
again it has been suggested that they are derived from
leucocytes. Lastly, some people say, even to this day,
that they arise from all three sources. In view of this
controversy, Professor Ross and his collaborators, con-
sidering the "red spots" in leucocytes to be centro-
somes, suggested that if anybody could find them in the
blood-platelets it would, of course, settle once and for
all the real cellular nature of these bodies.

A short time after this, while I was working to
determine the laws of diffusion by this method, I
appreciated that "red spots" were not centrosomes at
all, but were diffusion-vacuoles — a fact which I pub-
lished in The Journal of Physiology* and a fact which
was afterwards confirmed when divisions were induced
in leucocytes and the real centrosomes demonstrated.

This knowledge rendered Professor Ross's sugges-
tion of less importance, for since the spots are not
centrosomes, the discovery of them in the platelets
would not prove that these bodies found in the blood
were cells capable of reproduction. But when I was
experimenting with morphia on blood-cells I acci-
dentally discovered the "red spots" in all the blood-
platelets (figs. 22, 23).

Now, in spite of the fact that these spots are
not centrosomes, their appearance in the blood-plate-
lets still proves that these minute bodies are living

1 Journal of Physiology, vol. xxxvii, No. 4.

118 DIFFUSION- VACUOLES

cells; because these diffusion- vacuoles will appear only
in living cytoplasm.

Moreover, vacuoles will never appear in normal
blood-platelets — they are never seen in fresh blood-
films. It is necessary to keep the blood in an equal
volume of citrated solution of morphia (a 1-per-cent
solution of morphine hydrochloride in citrate solution)
for four hours at 37° C. A drop of the mixture is then
examined on a film of jelly in the usual wray, and the
film of jelly should have the correct index of diffusion,
and be kept at the right temperature and time for
the coefficient of diffusion of leucocytes, i.e. 12. The
diffusion-vacuoles will then appear in all the blood-
platelets. This experiment is a very easy one, and
certain in its results.

The action of the morphia is the same on the
blood-platelets as it is on leucocytes and lympho-
cytes. It lowers the coefficient of diffusion to a
marked degree, and it appears to do this by causing
gradual death. Morphia, in the 1-per-cent solution,
is a slow poison for leucocytes, for it will kill most
of them at 37° C. in about six hours. After incuba-
tion for four hours, however, when the cells are placed
on the jelly, the cells are still alive, but their coeffi-
cient of diffusion is so lowered by the poison that
the jelly, instead of merely causing maximum diffusion,
now causes diffusion to excess, and the leucocytes
and lymphocytes become intensely vacuolated (fig.
24). Further, the blood-platelets will now exhibit
"red spots."

In addition to lowering the coefficient of diffusion

ARCH OPL ASM

119

FIG. 22. — Diffusion-vacuoles in blood-platelets. The cells are resting
on the same jelly-film as those in 21, but they had been subjected to the
action of morphine hydrochloride.

FIG. 23. — Diffusion-vacuoles in blood-platelets. The jelly-film had the
same index of diffusion as that employed in 21.

ARCHOPLASM

FIG. 24. — A specimen of blood which had been mixed with morphia
solution. Note the extreme vacuolation of the leucocyte. A blood-platelet
is also vacuolated. The same jelly as in 21.

» •^^•I^BMBm^C^HBHaE ••

FIG. 25. — Patches resembling archoplasm induced in a leucocyte by
subjecting the blood to an extract of a dead tissue. The jelly-film on which
the cells are resting is similar to that employed in 21.

ARCHOPLASM 123

by causing gradual death, morphia undoubtedly has
a profound effect on the cellular cytoplasm, for other
alkaloids, so far as they have been tried, do not cause
vacuolation of the platelets. On the other hand,
we have occasionally seen a vacuolated blood-platelet
from a specimen of blood which has been mixed for
about twelve hours with a citrated solution (100 per
cent of suprarenal extract). It has already been
mentioned that extracts of dead tissues lower the
coefficient of diffusion, and in producing vacuolation
they also produce archoplasm in leucocytes (fig. 24).
Possibly, as mentioned before, the archoplasm which
is so frequently seen in cancer cells is derived from
the vacuolation caused by the action of the remains
of dead tissues on the cells. If leucocytes which have
been subjected to morphia and have been placed on
jelly as above described are watched for some time,
patches which might be described as archoplasm may
often be seen in them as a result of the dispersal of
many of the diffusion-vacuoles induced by the alkaloid.
We cannot, of course, state definitely that these patches
are identical writh what is known as archoplasm, and
we have never seen anything which could be described
as it in normal leucocytes examined by this method;
but that induced in them by extracts and morphia is
nearer the usual interpretation of archoplasm as seen
in fixed specimens than anything we have seen.

Since the blood-platelets can be made to become
vacuolated by lowering their coefficient of diffusion
by the action of the poison morphia; and since all the
blood-platelets in a specimen thus respond to it, it

124 DIFFUSION- VACUOLES

is clear that the blood-platelet is a living cell and
not a precipitate. As far as we know, no precipitate
has a coefficient of diffusion, and even if such a
thing were possible, one certainly could not lower it
by causing approaching death with morphia.

Blood-platelets unquestionably are living cells; and
they can actually be seen to be produced by leuco-
cytes when they are examined on a jelly-film by
this method. They are all the same class of cell, ap-
parently produced in the same manner. If a speci-
men of fresh blood is spread on a jelly which contains
stain and an alkaloid such as atropine sulphate, as
will be described in the next chapter, the leucocytes
and the lymphocytes become excited and extrude long
pseudopodia. Sometimes these pseudopodia become
detached from the cells (fig. 26), when the fragment
appears to be identical with a blood-platelet. They
may contain a few granules derived from the leucocytes.
Moreover, the blood-platelet is also highly amoeboid
under this excitation; and their amoeboid movements
can easily be seen by this method. Deetjen, several
years ago, asserted that blood-platelets showed amoeboid
movements, although, of course, he did not employ
alkaloids to excite them. By this method, however,
not only can they be readily seen to show movements,
but they have also actually been photographed in the
act (fig. 27). We have also succeeded in obtaining
a negative of a blood-platelet apparently being produced
by a leucocyte (fig. 26) . As a matter of fact, the plate-
lets stained by this method have such a remarkable
resemblance to leucocytes that in the very earliest

NUCLEATED RED CELLS 125

FIG. 26 — An extruded pseudopodium becoming detached from a leucocyte
which is excited by atropine. No stain.

G. 2 1. — Amoeboid movements excited in a blood-platelet by the action

of atropine.

NUCLEATED RED CELLS 127

experiments it became apparent that these bodies were
associated with those cells. We have never succeeded,
however, in making a platelet reproduce itself, even with
the most powerful exciter of cell-division. Blood-plate-
lets frequently become clumped into masses, especially
if the jelly contains an extract of a tissue; we therefore
think that this clumping may have some function in the
phenomenon of healing.

At this juncture it may be useful to dispose of an
old theory, that the blood-platelet is the "extruded
nucleus" of a red cell. In the first place, no diffusion-
vacuoles have ever been seen within the nucleus of any
cell, and the platelets, therefore, can hardly be nuclei.
This suggestion of the nuclear origin of the platelet
would never have occurred, I think, if the originators
of it had used in-vitro staining. Red cells are never
seen to extrude their nuclei by this method as they
sometimes seem to be in the act of doing when they
are spread out and fixed on a slide by the old methods-
It is difficult to imagine that any cell could extrude its
nucleus bodily, and a glance at the stained nucleus of
an unfixed nucleated red cell will dismiss such a fallacy
in very short time. The nucleus of a living red cell
seems to consist merely of a mass of chromatin granules,
which appear to be identical with those of the "red
cell with basic granules." The granules ultimately
seem to disappear altogether, for in normal blood one
sees about 1 per cent, of these granular cells, which
sometimes have only one or two granules, whereas in
anaemia the number of granules is much greater in
most of the cells. Presumably, when the granules

128 DIFFUSION- VACUOLES

have disappeared altogether the cell resembles the
ordinary erythrocyte. The nucleated red cell has a
coefficient of diffusion of about 11, and so has the
granular cell. An ordinary erythrocyte's coefficient of
diffusion seems to be much higher, however; but since
it has no nucleus or granules to stain, it is difficult to
determine it.

To stain the stroma of an ordinary red cell it re-
quires a jelly with an index of diffusion of nearly 20.
Like other blood-cells, the coefficient of diffusion of
red cells falls the longer the blood has been shed,
until, with a jelly suitable for staining the nuclei of
leucocytes, the stroma (or perhaps it is the haemoglobin
itself) of red cells will stain in many instances. It
is presumed that this more rapid staining of the stroma
or haemoglobin of red cells which have been shed some
time is due, as in other cells, to the lowering of the
coefficient of diffusion, for extracts of dead tissues and
morphia also have this effect on them.

The stroma or haemoglobin — whichever it may be—
stains more readily in nucleated and granular red cells
than in the others. "Red spots" will fairly often be
seen in nucleated red cells and in granular ones; but
they have only been seen three times in ordinary
erythrocytes.

Apart from their scientific interest, however, diffu-
sion-vacuoles are not of great importance, we think,
except that their appearance, as noted above, is a
signal that maximum diffusion is being occasioned.

I have now described what we know concerning the
diffusion of substance into living cells. It is a complex

FACTORS ACT ON THE CELLS 129

subject, which will require careful elucidation if the
actual physical laws on which it is based are to be
found out, and I venture to think that this method will
supply a means by which these laws can be determined ;
a large amount of careful experimentation will be
necessary, however, with a large variety of substances.
The chemical factors, such as alkalies and salts, will
have to be tried in greater variety; after which it
seems to me probable that one will be able to settle
whether the increase and decrease in diffusion which
they cause is due to their atomic weight or the osmotic
pressure, or what. One point, however, should be
clearly appreciated, which is this, that these chemical
factors which increase or retard the diffusion of other
substances, act not on the substance diffusing into the
cell, but on the cell itself. For instance, as will be
shown later on, alkalies, by increasing the diffusion
of kreatin or xanthin, increase the rapidity of cell-
division induced by these extractives. But the alkalies
have no effect on either kreatin or xanthin. The way
they increase diffusion into the cell is by causing the
cell to absorb substances more readily. And so with
acids, salts, and other chemical factors.

Lastly, these simple laws of diffusion must be taken
into consideration throughout researches with this
method, for no results will be obtained if they are
forgotten. The equation has been found to be of
more use when stain is employed. Later on, when one
is experimenting with single substances and no stain,
the arrangement of the jellies is more simple, and the
equation is not used so much.

CHAPTER VIII

THE EXCITATION OF AMCEBOID MOVEMENTS IN WHITE
BLOOD-CORPUSCLES CAUSED BY ALKALOIDS

SOON after this in-vitro method of staining was invented,
it occurred to me that it might be employed for
measuring the lives of human leucocytes after their
removal from the body. Much work had been done
by others in the way of determining the effects of
virulent disease-germs on men and animals, and soon
after the discovery of "opsonins" by Wright and
Douglas, many researches were made to find out how
individual human cells defended the body by attacking
pathogenic organisms; but little was known about
the effects of virulent germs and their poisons on the
protecting leucocytes. Hence, if one could measure
the lives of leucocytes, it would be a simple matter
to mix them with the toxins produced by bacteria,
and determine the virulence of these toxins by seeing
how long it took for them to kill leucocytes.

In order to measure the length of time that
leucocytes will live in a given sample of blood removed
from the body, it is obvious that the first thing to

130

LIVING AND DEAD CELLS 131

be done is to be able to distinguish accurately between
a living and a dead leucocyte; it is impossible to say
how long a cell will live if there is no means of telling
when it is dead. It may appear strange, but it is a
fact, that it took two years to find out the difference
in the appearance between a living and a dead leucocyte.
During this two years many of the points regarding
the diffusion of substances into cells, vacuolation, and
achromasia were found out; but although many efforts
were made experimentally to try to perfect a method
of measuring the lives of leucocytes, this difficulty, that
one could not accurately distinguish between living and
dead cells, always stood in the way. When the point
was discovered, it may almost be said that it was by
accident, and even then its value as a method of
measuring the lives of the cells was not appreciated
for some time.

It was known, of course, that leucocytes lived for
some hours after their removal from the circulation, for
they sometimes showed amceboid movements; but
in order to measure the lengths of their lives it was
necessary to be able to say at any given moment that
so many leucocytes in a given sample of blood were
alive, and that so many were dead. The cells were
always examined on jelly which contained stain; some-
times they showyed movements and sometimes they- did
not; but the absence of movements was no evidence
that death had taken place. Many experiments were
made, and at last it was resolved deliberately to kill
some cells by a virulent poison, and to see whether the
cells so killed appeared in any way different from those

132 THE EXCITATION OF AMCEBOID MOVEMENTS

not so treated. The poison was mixed up with the
stained jelly, and that jelly was alkaline, in order to
cause the diffusion of the stain and of the poison into
the cells. The poison chosen in the first instance was
hydrocyanic acid, and then nitrobenzol was used; but
after subjection to them, the cells presented very little
difference from others not so treated and known to be
alive. At last atropine sulphate was tried, with a very
astonishing and unexpected result, for every leucocyte,
far from being killed outright, became excited to great
activity. Some time afterwards it was realised that

•

this excitation by atropine was very constant, and that
if a cell was placed on a suitable jelly which contained
atropine, it wTould, if alive, respond with absolute
certainty by exhibiting excited amoeboid movements.
Thus the means of measuring the lives of leucocytes
was determined, and it became a simple matter, by
examining the leucocytes in a given sample of blood-
over a series of intervals — to discover how long they
lived under varying conditions, for one was enabled
at once to say whether the leucocytes were living or
dead, the living ones showing exaggerated movements,
the dead ones remaining immobile.

This method of measuring the lives of leucocytes,
and the details connected with it, will be found in the
Appendix. It was originally intended to use it for
ascertaining the effect of toxins on leucocytes, and we
think that for this it will have a useful application.
Owing to the fact that the excitation by the alkaloid
led to other work, we have not yet had time to in-
vestigate the actions of toxins very far.

KINETIC JELLY 133

Apart, however, from being able to measure the lives
of leucocytes, it is very necessary in this in-vitro work
to be able to tell at once when the cells with which
we are experimenting are alive, for it is essential that
one should deal only with living cells; hence wre will
now give the formula for the preparation of a suitable
jelly which will excite amceboid movements in living
leucocytes. This jelly has been called for convenience
"kinetic jelly," for it will always excite living leucocytes
to activity. It is as well always to have a tube of it
ready to hand, in order that at any time a film may
be prepared, so that we may be able to make certain
that the cells in a sample of blood are alive. It is
prepared as follows: To a tube of 5 cc. of "coefficient
jelly" add five units (0.5 cc.) of Unna's stain, six units
of alkali solution (0.6 cc. of 5-per-cent sodium bicar-
bonate), and 0.7 cc. of a 1-per-cent solution of atropine
sulphate. The content of the tube is made up to the
total of 10 cc., with 3.3 cc. of water. The mixture
should be melted and boiled until it froths up in the
tube, and a drop of the stained jelly poured on to a
slide and allowed to set. A drop of fresh citrated
blood is then placed on a cover-glass, which is inverted
on to the film in the usual manner. It should be
examined at the room temperature, which may be said
to be about 18 or 20° C.

When the cells come to rest on the jelly they will,
of course, be unstained. Slowly their granules begin
to turn red. A field which contains a few leucocytes
should be watched. In about fifteen minutes it will
suddenly be noticed that around the circumference of

134 THE EXCITATION OF AMCEBOID MOVEMENTS

first one leucocyte and then in the others small bodies
like minute beads appear. These beads seem to come
from underneath the cell. The beads get larger, and
quickly develop into long snake-like processes of cyto-
plasm, which are extruded from the cell. In a few
moments every leucocyte in the specimen will appa-
rently be putting out these long "feelers" until the
cells may almost be said to look like tarantulas (fig. 28) .
There are usually two or three of these long pseudo-
podia extruded from each cell. At first the pseudo-
podia are composed of clear cytoplasm (fig. 29), but
later on a few granules from the cell are seen to move
into them. Leucocytes seem to endeavour to push
their pseudopodia into the crevices between the neigh-
bouring red cells if they can (fig. 30), although we
have no reason to give for this propensity. These
excited movements differ from ordinary amoeboid
movements in that they are far more exaggerated.
The picture of a field containing several excited
leucocytes is a striking one, for these movements are
very different from the ordinary sluggish amoeboid
movements seen when the cells are merely kept on
a warm stage. Moreover, it must be remembered
that we are using the room temperature and no warm
stage.

The excited movements are due to the action of the
atropine. All the time, however, the stain is diffusing
into the cells as well as the alkaloid, and as time pro-
gresses the stain will reach the nuclei which now begin
to turn a faint blue colour. Now, it has already been
pointed out that the staining of the nucleus of a cell

KINETIC JELLY

135

FIG. 28. — Amoeboid movements excited in a leucocyte by the action of
atropine. Low power.

FIG. 29. — Exaggerated amoeboid movements in leucocytes which have their
granules stained. The movements were excited by atropine sulphate.

KINETIC JELLY

137

so.— Excited leucocytes extruding their pseudopodia between red cells.

FlG. 31.— Excitation of amoeboid movements in a lymphocyte by the action
of atropine. No stain.

KINETIC JELLY 139

will kill it, and therefore all the leucocytes in the
specimen slowly begin to retract their pseudopodia,
for leucocytes endeavour to resume their spherical shape
before death. The long snake-like processes can be
seen to shrink back gradually into the cells (figs. 7,8),
until in most cases they are completely retracted
(fig. 9). Occasionally, however, a constriction appears
in a pseudopodium where it arises from the cell-wall
(fig. 26), and separation has actually been seen to take
place; the separated portion, often containing a few
cell-granules, will now resemble a blood-platelet. Soon
after the pseudopodia have been retracted the cell dies,
and either bursts or becomes achromatic.

If the jelly has been properly prepared the whole
phenomenon of excitation of amoeboid movements will
be over in about twenty-five minutes or so. The action
of this kinetic jelly is instructive, for it affords
another example of the diffusion of substances into the
cells, and of the accuracy of the equation used in its
preparation in relation to this diffusion. The way of
making the jelly has been described, and it must be
remembered that it contains 0.7 cc. of a 1-per-cent
solution of atropine sulphate. Now, this is a salt, and
it delays diffusion; hence its unit must be ascertained
before the correct equation can be made for this jelly.
The unit of atropine sulphate (as found by experiment)
is .013 gramme, and therefore since the jelly contains
0 . 7 cc. of a 1-per-cent solution, it must contain 0 . 5 of
1 unit, which may now be added to the equation among
the other salts wrhich are minus factors. We may now

140 THE EXCITATION OF AMCEBOID MOVEMENTS

state the formula for the index of diffusion of this
jelly:

where z = the unit of atropine sulphate.

The specimen is kept at the room temperature, or
3 units of heat; and the object of the jelly is to excite
amceboid movements in fifteen minutes (or 1 . 5 unit
of time) in neutrophile polynuclear leucocytes, which
have a cf of 12. This jelly, of course, is arranged for
the coefficient of diffusion of leucocytes, and it may thus
be set down :

Now, if these equations are carefully considered, it
should be noticed that they are apparently wrong: the
coefficient of diffusion of neutrophile leucocytes is 12,
not 11.

This brings us to another rule. It is obvious that
if the jelly was prepared for the exact coefficient of
diffusion of leucocytes, we would not obtain excitation
of those cells in the given time — we would only obtain
staining of their nuclei, and staining of the nuclei means
that the cells would be dead. This would mean that
we should defeat our object, for dead cells with their
nuclei stained will certainly not respond to the atro-
pine. * The determination of the coefficient of diffusion
of nucleated cells involves death," because the stain-
ing of the nucleus is the moment by which the cf is
obtained.

But this difficulty can be overcome by subtracting

KINETIC JELLY 141

one digit from the coefficient of diffusion, and making
the jelly accordingly. Hence the equation given above
in reality is correct, for the coefficient of diffusion of
leucocytes is 12, and subtracting one digit from it makes
11, as given in the equation. With the jelly arranged
for 11, the nuclei are not yet stained, and death will not
occur for another unit of time. On the other hand, the
diffusion has already been sufficient for the atropine to
excite the cells, and when the given fifteen minutes of
time has elapsed, the cells will be seen, not dead, but
in the height of their excitation.

Thus the rule is that, having ascertained the co-
efficient of diffusion of a cell, if we wish that cell
to be alive at the expiry of the given time, subtract
one digit from its coefficient of diffusion, and make
the jelly accordingly.

This rule is an important one in this work, for
we shall, of course, frequently have to observe cells
in the act of excitation, which is an easy matter if
their coefficient of diffusion is known, as it only
remains to subtract one unit from any of the factors
which increase diffusion, and we get the right result.1

All forms of the polynuclear leucocyte respond
to atropine by exhibiting excitation of amoeboid
movements. In making them respond, however, the
different coefficients of diffusion of each class - of cell
must be duly regarded. The eosinophile cell has a
coefficient lower by one unit than the neutrophile;
and if it is required to excite it especially, the jelly-

1 It will doubtless be realised that subtracting one unit of a factor which
increases diffusion, is similar in effect to subtracting one digit from the cf.

142 THE EXCITATION OF AMCEBOID MOVEMENTS

film must also have an index lower by one unit than
that for the neutrophile corpuscle. The lymphocyte,
or mononuclear corpuscle, also becomes excited to a
marked degree by absorption of atropine (figs. 31-3) ;
indeed they extrude longer pseudopodia than any of
the other classes of blood-cells, a fact which is more
interesting, because it is generally supposed that the
lymphocyte is not a very amoeboid cell, a supposition
which is erroneous. To induce amoeboid movements
in the mononuclear cells, however, it is best to treat
them as though they had a coefficient of diffusion
lower than that of the leucocytes by about one unit,
as these cells seem to die before the nucleus becomes
stained. It was pointed out in the original specifi-
cation that the staining of the nucleus indicated the
point at which the coefficient of diffusion is determined.
It has already been mentioned that this term nu-
cleus is rather vague, and, as will be shown later,
death is occasioned in the lymphocyte by staining
of the nucleolus, which frequently becomes coloured
before the nuclear wall. For general purposes,
however, the original specification stands good.

The foregoing experiment, by which one can
excite amoeboid movements in leucocytes which have
their granules stained, proves that the staining of
their Altmann's granules is not very harmful to cells.
The granules can, and do, become deeply stained,
and all the while the cells will continue to extrude
and retract pseudopodia in response to the alkaloid.
This point is a very important one when we come
to study induced cell-division, for it affords a clue as

DIFFUSION OF TWO AGENTS

143

FIG. 32. — Excitation of amoeboid movements in a lymphocyte which has
its granules stained.

FIG. 33. — Extreme excitation of amoeboid movements in a lymphocyte.

No stain.

DIFFUSION OF TWO AGENTS 145

to how the chemical exciters of reproduction act in
the causation of mitosis.

Another point is learnt, however, by experimenta-
tion with this combination of. stain and atropine, for
here we have two chemical agents, an anilin dye
and an alkaloid, both diffusing into the cells side by
side and each producing its effect on the cell-proto-
plasm. One excites the cell, the other kills it, and
each carries out its function in direct proportion to
its own concentration; for if the content of the stain
in the jelly is reduced, the cells become less stained,
and death is delayed; but if the alkaloid alone is
reduced, the staining is as usual, but there is less
excitation. At the same time, it must be remembered
that the alkaloid is a salt, and, like other salts, as it
diffuses itself into the cell, it delays the diffusion of
the stain.

The diffusion of a combination of substances into a
cell, therefore, is not a simple matter, for it represents
an equation of variables, although those variables, if
applied in the same manner, always have the same
effect with mathematical precision.

The excitation of amoeboid movements in white
corpuscles is due entirely to the atropine. Using a
jelly-film which is alkaline,1 and which contains stain
but no atropine, no amoeboid movements will occur,
and the cells retain their spherical shape. If the jelly
is neutral, however, occasionally sluggish movements
occur, even at the room temperature. At a tempera-
ture of 30 to 37° C. sluggish movements may occur

1 The alkalinity oHhese jellies is not sufficient to precipitate the alkaloids.

146 THE EXCITATION OF AMCEBOID MOVEMENTS

even in the presence of alkali. But in any of these
instances the movements are not comparable to the
deliberate extrusions caused by atropine, which are very
striking in character, and if once seen will always be
remembered.

We can, of course, make kinetic jelly suitable for
the temperature of the blood (it is merely necessary
to reduce the content of alkali in the jelly by 3 units,
because we increase the temperature by 3 units), and
still the excitation occurs, although (and this is a
remarkable circumstance) the excited movements are
not so marked at the temperature of the blood as they are
at that of the room. Many persons who have seen the
action of kinetic jelly have disparaged it, saying that
they have often seen marked amoeboid movements in
leucocytes; but when questioned, the fact is always
elicited that they have employed the warm stage. It
is the deliberate and constant exaggerated movements
which invariably occur in all living leucocytes at low
temperatures which constitutes the striking effect of
atropine sulphate upon them. Let a control experi-
ment be made with a jelly which contains no atropine—
and no stain either if one wishes to — and the difference
is immediately apparent. Excited by the alkaloid, the
cells with their stained granules, extruding their long,
snake-like pseudopodia in all directions, as if they were
searching for something (which, as far as can be found
out, they are not), form a very pretty picture, which,
when seen through the microscope, will be a revelation
to those who have only worked with films of dead cells.

Atropine sulphate is not the only substance which

NOT DEATH-STRUGGLES 147

causes this excitation. We have tried several alkaloids,
and all of them have had this effect. It does not matter
what the alkaloid is, nor whether it is a salt or an
alkaloid; the result is the same. In fact, we think
that it is probable that this power of exciting amoeboid
movements is a property of alkaloids generally. It is
true that we have not yet tried all known alkaloids,
but we have experimented with many, and we think
that they probably all have this effect. Moreover, the
parent substances of alkaloids, such as pyridine and
quinoline, also excite wrhite blood-corpuscles.

Some alkaloids cause more excitation than others;
atropine has so far proved the most effectual, morphine
the least. To man atropine is very poisonous; mor-
phine is not so poisonous, weight for weight. To a
man's leucocyte, however, it is curious to note that
morphine is the more poisonous, and atropine not
nearly so dangerous. By means of this jelly method
we can try the effects of alkaloids and substances in
various strengths on leucocytes and other cells, and if
the jelly contains atropine, by noting the extent of the
excitation one can find out the dose of an alkaloid
which will cause maximum excitation and the dose
which will cause death in a given time. Generally
speaking, it requires three times as much of a given
alkaloid to cause death as it does for it to cause
maximum excitation.

This latter point is an important one, for it has
been suggested to us that the excitation by alkaloids
is in the nature of a death-struggle. It is clear,
however, that if it was, the excitation would steadily

148 THE EXCITATION OF AMCEBOID MOVEMENTS

increase as more alkaloid was absorbed; but such is
not the case. Moreover, this excitation is not caused
by poisons, such as nitrobenzol and prussic acid. The
possibility of the excitation being due to a death-
struggle is also precluded by the fact that if no
stain is employed the excited movements may be
watched for an hour. Death-struggles, as seen in
higher animals, do not usually last very long, and
always commence immediately before dissolution. The
excitation appears to be a specific one caused by
alkaloids, although we have seen a similar form of
excited movements, but not to the same extent,
caused by arsenic.1

As already mentioned, we have ascertained the
amounts of other alkaloids which cause maximum
excitation of leucocytes, and in finding out these
"doses" we have always used a similar jelly containing
no stain, and the temperature employed has been that
of the room in every case. The jelly was alkaline, as
it contained 5 units of alkali solution, and the alkaloids
were each used in a 1-per-cent solution, thus: To
5 cc. of coefficient jelly, 5 units of alkali solution, and
the amount — whatever it is — of alkaloid solution were
added, and the balance, up to the usual total of 10 cc.,
was made with water. The jelly was then boiled, and
a film prepared from it in the usual way, fresh citrated
blood being used in each case.

The following is a list of alkaloids which we have
tried on leucocytes, and the amount of each of them

1 The effects of oxygen have been tried on leucocytes by bubbling the gas
through the jelly; but its action seemed to be negative.

STRENGTHS OF ALKALOIDS 149

which, when mixed with the jelly, produces maximum
excitation. Treble this amount, and death will gener-
ally occur without excitation, although leucocytes will
stand even ten times the dose- of codeine and bruceine
without dying.

To produce maximum excitation in twenty minutes:

Alkaloid. Amount of 1-per-cent solution of it

contained in 10 cc. of jelly.

Bruceine 1 cc.

Morphine 0.2 "

Pilocarpine Nitrate 0.5 "

Cocaine Hydrochloride .... 2 "

Strychnine 1 "

Atrophine Sulphate 0.7 "

Aconitine 0.5 "

Codeine . . 3 "

Atropine is undoubtedly the most active of the
vegetable alkaloids; but, as will be shown later
choline (figs. 34, 35), and cadaverine (fig. 36), two
of the animal alkaloids produced by putrefaction, are
nearly as effective. The action of morphine in exciting
exaggerated movements is very poor (fig. 37), but still
it does have this effect. The dose may be doubled
with cocaine, and the excited movements continue.
Strychnine is not so effective an excitor for leucocytes
as atropine. Codeine acts more effectively than
aconitine. Pyridine is fairly effective (fig. 28).

The excitation of leucocytes by alkaloids is a very
remarkable thing, for it seems to be a functionless
procedure on the part of the cells. The alkaloids do
not appear to cause the cells to migrate at all; they
remain in their original position, and merely extrude

150 THE EXCITATION OF AMOEBOID MOVEMENTS

and retract their pseudopodia aimlessly. Quinine
hydrochloride excites them fairly markedly ; and it
must be noted that the statement has been made by
other authors that quinine stops diapedesis. We have
made a hanging drop preparation1 with a jelly-film
in such a way that the cells are not pinned down by
the cover-glass, but still absorb atropine, and they
therefore were in a position to move about if they
wished to. They remained in their original positions,
however, and seemed to be content to extrude and
retract their "feelers."

Experiments were made to see if this excitation
was due to any chemotactic influence of the alkaloids.
Two jellies were made, one of which contained atropine
and the other none; and they were so mixed on a
slide that there was atropine in one part of the film
and not in another. Some citrated blood was placed
over the line of demarcation to see if the cells neces-
sarily extruded their pseudopodia in the direction of
the concentrated alkaloid. They did not do so, how-
ever, for, provided a cell absorbed the alkaloid suffi-
ciently, the extrusions were made in all directions as
usual.

In order to try to find out whether this excitation

«/

increased the ingestion of bacteria by leucocytes, a
sample of fresh blood was mixed with a volume of
citrate solution and atropine, which contained bacteria
in suspension. Having been incubated for some
minutes, the cells were spread on jelly; but when

1 This method will be found in the Appendix, where also another method
of preparing kinetic jelly will be found.

EXCITATION AND PHAGOCYTOSIS

151

FIG. 34. — Excitation of two leucocytes by the action of choline. Low power.

No stain.

FIG. 35. — Excitation of a lymphocyte by the action of choline. No stain.

EXCITATION AND PHAGOCYTOSIS 153

FIG. 36. — Excitation of amoeboid movements in a leucocyte by the action
of cadaverine. No stain.

FIG. 37.— A leucocyte excited by morphine. The cell's granules are stained.

EXCITATION AND PHAGOCYTOSIS 155

the number of bacteria ingested were compared with
those phagocytosed in control experiments where no
alkaloid was used, it was seen that the excited cells
did not ingest more germs ^than usual. Excitation,
therefore, does not increase phagocytosis; and we have
noticed that if a mixture of living leucocytes and
germs are mixed and spread on jelly which contains
atropine, the cells do not purposely extrude their
pseudopodia in the direction of any bacteria which
may be near them. On the contrary, if a pseudo-
podium happens to strike against a bacterium, the
latter is usually pushed out of the way.

Whether leucocytes are excited or not, we have
never seen a cell actually ingest bacteria. We have
often seen cells with bacteria inside them, but we
have never seen the actual act of ingestion, nor have
we any explanation to offer as to how it occurs.
Moreover, we have often seen leucocytes with red
cells apparently inside them, although how they came
to be absorbed we do not know. It is possible that
the laws of diffusion may play some part in the
actual act of phagocytosis.

Another point in connection with phagocytosis may
be mentioned. In the making up of fixed films,
germs and other substances may be crushed into
leucocytes. By the examination of living cells this
cannot happen. We have seen fixed specimens which
showed phagocytes apparently crammed with germs;
but on looking at another sample of the same cells
alive a very different impression was obtained. We
have mentioned this point in view of the possibility

156 THE EXCITATION OF AMCEBOID MOVEMENTS

of fallacy arising in the technique of the "opsonic
index," if it is carelessly carried out, because in that
technique fixed films are usually employed.

The possibility of foreign substances being crushed
into cells during the preparation of fixed films is also
the reason, we think, for the common, fallacious
supposition — which has already been mentioned — that
the blood-platelets are the extruded nuclei of red
cells, for in the preparation of fixed films platelets are
crushed into red cells, to which they often adhere, and
after fixation they appear as if they were emerging
from them; an artefact never seen with the jelly
method.

In concluding this chapter it should be mentioned
that Professor Osier, many years ago, pointed out
that certain alkaloids excited amoeboid movements
in leucocytes, although this fact was not known to
me when the effects of atropine mixed with the jelly-
film were first tried.

As will be shown later, alkaloids have a far more
important action on cells than merely exciting amoe-
boid movements, for they greatly augment the action
of the exciters of reproduction.

CHAPTER IX

THE ADOPTION OF THE IN-VITRO METHOD FOR CANCER

RESEARCH THE EXCITATION OF LEUCOCYTES

CAUSED BY CANCER PLASMA FACTS KNOWN

ABOUT CANCER THE AGE INCIDENCE; VITALITY;

DEATH; METASTASES; CHRONIC IRRITATION — THE
POSSIBLE CAUSES OF CELL-PROLIFERATION DIS-
CUSSED.

IN August, 1908, on an occasion when the excitation
of leucocytes by atropine was being demonstrated to
one of us (C. J. M.) he remarked that he had often
enough thought that patients dying from cancer ex-
hibited symptoms resembling those of poisoning by
some alkaloids, and he suggested that an investigation
might be made by means of the in-vitro method to
try to find out whether there existed in the blood
of cancer patients any substance of an alkaloidal
character which might be responsible for these
symptoms. This suggestion, based on bed-side obser-
vation, taken in conjunction writh the fact that a group
of chemical agents existed which were capable of
exciting human cells, warranted a research in this
direction, for if such a substance existed in the blood
of these patients it was felt that either it might have

157

158 APPLICATION TO CANCER RESEARCH

some bearing upon the cause of the disease or that
it might be an effect of it.

A cancer consists essentially of cells of the body
which have multiplied irregularly and too rapidly, and
it was quite reasonable to think that this form of
excessive proliferation might be the result of some
abnormal excitation.

It should be remembered that in August, 1908,
the actual cause of cell-division was quite unknown,
and multiplication of individual human cells in direct
response to a chemical agent had, of course, never
been seen. It was realized that the problem of the
nature of cancerous growths could only be solved by
the discovery of the cause of cell-division. The cells
of the body are continually multiplying by cell-division,
and the correct appreciation, not only of the nature of
new growths, but also of the problem of healing, and
in reality most of the problems of pathology, depend
upon the cause of cell-reproduction.

The most commonly accepted theory regarding the
cause of cell-proliferation was that cells divided owing
to some inherent vital propensity — that is to say, that
they multiplied because it was their "duty" to do so.
As a matter of fact, however, nothing was known as
to the immediate cause of individual cell-reproduction.

So much work had been done with reference to
cancer, and in spite of it so little was known concerning
the cause of that disease, that we felt justified in follow-
ing any clue, however slender it might appear at first
sight. It was true that the excitation by alkaloids had
so far only resulted in the production of exaggerated

EXCITANT IN CANCER PLASMA

movements in white blood-cells; but still it was an
excitation, and for all we know, although they had not
yet been seen, the excitation might produce other
results as well. This clue, however, arising from micro-
scopical experimentation and from a clinical observation,
has proved to be of great importance, and has led by a
singular chain of events to the knowledge that cell-
division in the body results from the presence of specific
agents, the action of which becomes remarkably aug-
mented if the cells are in a condition of excitation
resulting from the presence of an alkaloid.

This cancer research became instituted in this way,
and the first step undertaken in connection with it
was to test the blood of cancer patients experimentally
in order to find out whether it, or other of the body
fluids, contained any substance which would, like the
alkaloid, excite exaggerated movements in leucocytes.

Ten cases of well-marked carcinoma were examined
in the following way: A certain quantity of the
patient's blood was mixed in a capillary tube with
an equal volume of citrate solution. The tube was
then centrifugalised and the corpuscles removed. To
the remaining plasma a certain quantity of fresh blood
taken from a healthy person (usually one of ourselves)
was added and thoroughly mixed. The sealed tube
containing the mixture was then placed in the revolving
apparatus in the incubator and kept at 37° C. for half
an hour, at the end of which time a drop of it was
examined at 20° C. upon a slide under a cover-glass.
Blood plasmata taken from fifty healthy people, or from
people suffering from diseases other than cancer, were

160 APPLICATION TO CANCER RESEARCH

similarly tested, and it was found that the leucocytes
of healthy people bathed in the plasmata of cancer
patients undoubtedly showed amoeboid movements
which were exaggerated and different in character from
those which were observed in the corpuscles suspended
in the plasmata of normal persons, or of persons
suffering from a number of other diseases. The differ-
ence was one of degree, however, for leucocytes fre-
quently under these conditions showed some amoeboid
movements; but we were quite satisfied that there wTas
a distinct difference, although the test could not be
considered a very delicate one.1

This series of experiments made us consider that
there probably is some agent in the body fluids
of cancer patients which causes excitation of cells,
and one of us was charged with the task of further
confirming the correctness of this observation, and
of finding out what the substance is and how it is
produced. It was appreciated that this substance
could only be present in the blood in small concen-
tration, and that to isolate it from serum would prove a
very difficult task.

As a preliminary to this part of the research, it
was considered advisable to review the known facts
concerning cancer to see whether they harmonised
with the possibility of the disease being associated
with an excitation of cells by chemical agents. After-
wards we proceeded, by means of the new jelly method,
to try the effects of different substances either taken

1 A paper by Dr. Macalister and myself describing these experiments was
read before the Royal Society of Medicine in November, 1908.

AGE- INCIDENCE 161

from growths, or which we knew were associated
with growths, on individual living healthy cells. By
this means it was hoped that we might find some
exciting substance from cancerous growths which
might in the first place cause normal individual cells
to undergo a change and become similar to those
cells taken from the growths themselves. In the
event of such a substance being found, it would, of
course, then be necessary to try to prove the argu-
ment by experimentation with the substance in the
body itself. In other words: believing that cancer
might be due to a chemical agent, we proposed to try
to find that agent, and to test its effect, in the first
instance on individual cells under the microscope, and
lastly to test its action on groups of cells in the tissues
of the body.

Malignant disease may be separated into two
main divisions — carcinoma and sarcoma. The former
attacks gland-tissues and epithelial cells; the latter is
a disease of connective tissues. These researches are
almost entirely concerned with carcinoma, and the
term "cancer" in this book refers to that disease.
There is, we think, a close association between these
two forms of malignant disease, although there is
a line of demarcation in the age incidence and in
some of their morphological and clinical characteristics
which separates them. Cancer — that is to say,
carcinoma — attacks people over the age of forty,
although there are occasional exceptions to this rule;
but sarcoma may occur at any age from infancy onwards.
At the outset we turned our attention exclusively to the

162 APPLICATION TO CANCER RESEARCH

consideration of carcinoma, for we considered that if we
succeeded in throwing any light on the causation of that
disease, it would be time enough for us to investigate
sarcoma. Cancer is much more common than sarcoma ;
but it has to be remembered throughout that, from
the similarity of the cardinal symptoms of the two
diseases, there is probably an intimate association
between the causes of both. The connective tissues
can become malignant at any age. The epithelial
tissues are usually attacked after the age of forty.
This age-incidence of carcinoma is most striking, and
it necessarily constitutes a fundamental fact with
which all our thoughts regarding the cause of
cancer must ultimately harmonize. It is a disease of
senescence; it attacks people when they are robust and
apparently in a state of highest vitality, just when they
are in the prime of life, or having just passed it. We
have to remember in this connection that the expres-
sion "prime of life" in its physiological sense may be
taken to refer to middle life — that is, somewhere about
the age of 35; and we may further understand that
before that age a man is being built up, whereas after-
wards he enters upon the downward trend and steadily
progresses towards physiological death, which may be
taken to occur about the seventieth year. We may
therefore consider that the climax of his physiological
life is reached at 35.

The age incidence of cancer is unique; there is no
other disease which has this limitation in its age
averages. Exceptions do occur, it is true, but the
number of cases occurring during senescence, when

VITALITY 163

man has passed the climax of his age, is so enormous
that the possibility of fallacy due to "the error of
random sampling" is reduced almost to zero. It is a
salient feature of the disease which cannot be disputed,
and we may regard is as an axiom that cancer attacks
people when they are trending downwards from their
physiological prime. The question is, therefore, What
happens in the tissues during this senescence which
renders them liable to the onset of cancer ? At the
time when these researches were first applied to the
investigation of cancer, this question could only be
answered in a speculative manner; but it was appre-
ciated that the conditions present after the prime of
life which predisposed to the disease might merely
depend on something in the nature of the oversetting
of a physiological balance.

Vitality seems to be worthy of consideration as a
factor connected with the onset of malignancy. Very
old persons do not appear to be so liable to cancer
as those between the ages of 40 and 55 — a circum-
stance which may possibly be due to a loss of vitality,
for it has already been mentioned that cancer is a
disease of the robust. Premature ageing, on the other
hand, seems to favour the onset of cancer; but in
conditions of decrepitude there is more freedom from it.
Tottering persons, such as are seen in asylums -and
institutions, do not so frequently develop carcinoma;
but people who are sufferers during their senescence
from the atrophic form of osteo-arthritis or from gout
are common victims. Let the reader visit a home for
incurables, and he wrill there learn that many of the

164 APPLICATION TO CANCER RESEARCH

cases of cancer arising in the institution are also afflicted
with rheumatoid arthritis. The setiology of cancer is
a large subject, and for full information regarding
what is known of it reference may be made to an
excellent volume by Mr. W. T. Gibson, on The
Etiology and Nature of Cancerous and other Growths.

This book enumerates in detail the trades and pro-
fessions the members of which are especially prone to
cancer, and it furnishes a valuable aid to pathological
cancer research. Therein it is shown that chronic
alcoholism is a predisposing factor. Syphilis also is
undoubtedly a predisposing cause of cancer, provided
the disease is not too severe. We have been reminded
of this point by Mr. Fernet,1 whose experience of
syphilitic patients has left him convinced of an associa-
tion between the two diseases.

The conditions of decrepitude and chronic enfeeble-
ment — to which reference has been made as ones which
render persons less liable to malignancy — affect not
only the general vitality of the body, but also pre-
sumably the vitality of the individual cells.

Cancer is a disease which is general throughout the
world as far as we can find out, but climatic conditions
appear to influence its incidence to some extent. Sir
William MacGregor has told one of us that as far as
he can remember he has never seen a case among the
Esquimos, an observation which is interesting in con-
nection with the association of cancer and some
putrefactive products, which will be discussed in the

1 "The Intramuscular Syphilitic Treatment," by George Fernet, Transac-
tions of the American Medical Association, June, 1909.

CHRONIC IRRITATION 165

later chapters of this book, for, generally speaking,
putrefaction of organic substances must be reduced to
a minimum in the ice-bound regions of the far North.

Death is the ultimate result of cancer in most
instances, unless the progress of the disease is success-
fully interrupted by surgery, and this is a fact which
must be carefully considered. Cancer consists of a
growth, of human cells. Why should such a growth
kill the person it afflicts. Benign growths do not
necessarily cause death ? It may be, of course, that
the original cause of the disease increases with the
growth, and that it is this cause which is instrumental
in killing the victim. We speak of death from cancer
as resulting from the vague condisions described as
** exhaustion and cachexia," but why these conditions
result from cancer was not even within the realms of
speculation.

Cancer is a disease which seems to aggravate itself.
Once the disease is started in the circumscribed area
—for it always begins in one spot — it will go on steadily
if it is left to itself. Moreover, one of the features of
a malignant growth is that it produces metastases. Why
should malignant growths and not benign ones produce
metastases ? It is usually considered that metastases
are due to embolism, and that the transplanted cells con-
tinue to multiply in their new surroundings ; but, again,
why should these emboli only continue to mutiply in
malignant tumours ? Benign growths, like the malignant
ones, are supplied with vessels and lymphatics, and
there seems to be just as much reason why portions of
both forms of growth should be swept away to form

166 APPLICATION TO CANCER RESEARCH

metastases in other parts of the body. Still, the fact re-
mains that metastases occur only in malignant disease.

The foregoing points formed our axioms. Whatever
experiments we undertook had to harmonise with them
all in their consideration. There was one other factor,
however, which has already been mentioned; the mys-
tery of the cause of cell-division in the body, and a
well-known predisposing factor in the causation of
cancer which is intimately associated with it, namely,
chronic irritation.

The body consists almost entirely of living cells;
individual living creatures, each of which is capable of
separate existence for a short time, but which in con-
junction with one another form the tissues which in
their turn have special functions. Each cell is merely
an individual in a multitude; a unit in an organ. Cells
not only have functions to perform for their own
individual welfare, but they also act collectively for the
general welfare of the body.

Since cancer consists of a tumour composed of cells,
we may attack the problem of its causation from two
aspects — the investigation of the individual cells, and
the investigation of collective masses of them. At the
outset, the first aspect is obviously the one to receive
consideration; and since cancer consists of a growth of
cells which have multiplied too often and have so
formed a tumour, the first question to be asked is,
What makes this excessive multiplication take place ?

Before this question can be approached, however,
another question must be answered, namely, What
makes any multiplication of cells take place ?

WHAT MAKES CELLS DIVIDE? 167

The multitudes of cells which form our bodies have
been evolved from a single pair of cells. The maternal
ovum is a single cell, and always remains as such until
it is fertilised by the paternal spermatozoon, which in
its turn is also a single cell. The conjugation of the two
at once causes cell-division to take place within the egg.
Multiplication occurs, and where there was one cell
there are now two; and each daughter cell divides and
divides until generation after generation of new cells
are produced, and this cell-reproduction ultimately
leads to the formation of the new individual. The
basis of the formation of new beings is the reproduction
of cells by their division in response to the conjugation
of an original pair of cells. We had therefore to ask
ourselves why this conjugation should cause cell-
division; but unfortunately the answer was unknown.

Throughout our lives, although we cannot actually
feel it, the cells in our bodies are continually repro-
ducing themselves by division by mitosis, and individual
cell-death is also constantly taking place. It is true
that some cells, such as some cells of the nervous
system, prpbably live throughout the length of our
lives, but myriads of other cells are constantly dividing
to help to build up the tissues. "Birth" and death
are continually going on among the individual units
of ourselves. When a tissue is sectioned and examined
microscopically it will frequently be seen that some of
the cells are in the act of division by mitosis; but when
we asked what makes the division occur, and what
makes cells multiply to build up the tissues, we could
only say once more that the reason was quite unknown.

168 APPLICATION TO CANCER RESEARCH

When a child grows to form a man he grows by the
multiplication of his cells, but we did not know what
causes him to grow, or what makes his cells to attain
this object.

Again, if we injure or wound ourselves in any part
of the body, the tissues always make an attempt to
repair the damage. No matter to what extent the
injury may occur, attempted healing always takes
place. The phenomenon of healing is due to the
proliferation of white blood-cells, which multiply by
cell-division to repair the tissues which are damaged.
Not only do leucocytes and lymphocytes proliferate
when a tissue is damaged, but other cells also multiply.
For instance, epithelial cells will also proliferate to heal
a damaged site. The cell-proliferation of healing forms
one of the bases of pathology, and therefore of medicine
also; yet it had to be admitted that nothing whatever
was known as to why this cell-proliferation occurs when
any part of our bodies is damaged. The process of
healing is occurring in our bodies throughout our lives,
and this sudden multiplication of cells must be con-
stantly before the consideration of medical men; but
although this multiplication by reproduction is an
established fact, one never hears the question asked,
Why do cells immediately divide to reproduce them-
selves when a tissue is damaged ? If the question was
asked, however, the answer would have had to be,
"We do not know."

For the cell-proliferation of healing to occur it is.
not necessary for the skin to be actually broken. On
the contrary, extensive cell-proliferation of healing may

PROLIFERATION OF HEALING 169

occur as the result of a bruise or disease; and chronic
irritation, which is an invariable predisposing factor in
cancer, may give rise to exuberant multiplication of
the cells in the neighbourhood of the irritated part.
A common instance of the proliferation due to chronic
irritation is shown in the case of a "corn." An ill-
fitting boot irritates a certain portion of the foot by
pressing unduly on a certain portion of the skin. The
skin becomes hardened, and a small tumour may even
be formed on the irritated spot. This hardening is
due to excessive proliferation of the living cells in
and immediately underneath the skin. A wart is
an example of the proliferation due to irritation;
but although this irritation leads to proliferation, we
do not know exactly wrhy the cells proliferate in
response to it. If we think the problem out care-
fully, wre can picture a group of living cells multiplying
by division, and then try to grasp how irritation of
that group by mechanical pressure can possibly make
the individual cells reproduce themselves; for this is
what they do. At the time when these cancer re-
searches wrere started we could not grasp this point.
It seemed incredible that a cell would reproduce itself
because its cell-wall was "tickled" or pressed upon.1
Why should a living cell undergo the complex phe-
nomenon of mitosis for a reason of this nature ? . Be-
sides, living cells are very delicate, and we know that
they will not stand much handling or pressure without
dying. No; it was necessary to find some better
explanation of the cause of cell-division than mere
mechanical irritation, and we appreciated that irritation

1 The pressure of a cover-glass does not cause cell-division.

170 APPLICATION TO CANCER RESEARCH

in reality causes localized cell-death, and the cause
of the proliferation due to irritation would in all
probability be found to be due to the same cause or
causes which make our cells multiply in order to heal
up a cut or a sore.

Now, there can be no doubt that cancer occurs in
sites where there has been previously some form of
chronic irritation, and cancer is another name for
malignant proliferation of cells. Since such irritation
is probably directly associated with the cause of the
cell-proliferation of healing, we made our first endea-
vours to try to find out this cause of cell-reproduc-
tion in the body, for it was considered that if that
could be found a step in the right direction would
be made. Some attempts at healing are always going
on in the parts that are subject to chronic irritation,
and we may safely say that the cell-proliferation of
healing is going on in these parts. Cancer super-
venes on old ulcers and sores, which, of course, are
healing sites. In the breast and uterus, two of the
commonest places for cancer, the cell-proliferation of
healing occurs every month during the ages of actual
sexual function, and at the climacteric a large involu-
tion takes place, accompanied by destruction of tissue.
The irritation which causes cancer of the lip is usually
the pressure of a tobacco-pipe; X-ray cancer usually
follows the ulceration and damage due to burns; and
there are many other examples. The cause of the
cell-proliferation of healing, therefore, constituted our
first investigation in the path of cancer research.
V / Leucocytes and lymphocytes are the cells which pro-

PROLIFERATION OF HEALING 171

liferate to a great extent when healing occurs anywhere,/
and these w^hite blood-corpuscles formed the objects
of our observation in the first instance. What made
these cells divide wTe did not know, how they divided
was also unknowrn; but we knew that amoeboid move-
ments could be excited in them by means of alkaloids.
It is an astonishing thing that when any injury
occurs, in no matter what part of the body, those
neighbouring cells which have not been damaged wTill
immediately reproduce themselves. If the damage is
persistent, and healing becomes very chronic in persons
over the age of 40, cells may reproduce themselves
in a malignant manner, and then they go on dividing
and dividing, producing a cancerous growth which
ultimately kills the person the part of whose body
they are, and whose damaged tissue it was their en-
deavour to heal. The first thing to do, undoubtedly,
was to try to find the cause of the cell- proliferation
of healing.

CHAPTER X

EXPERIMENTS WITH NUCLEIN THE LOWERING OF THE

COEFFICIENT OF DIFFUSION CAUSED BY EXTRACTS

OF DEAD HAEMAL GLAND DIVISIONS INDUCED IN

LYMPHOCYTES FOR THE FIRST TIME— REVELATIONS

CONCERNING THESE DIVISIONS THE ROLES PLAYED

BY THE ALTMANN'S GRANULES, NUCLEI, AND
NUCLEOLI IN THEIR CELL-DIVISION

BEFORE proceeding to discuss the problem of the
causation of cell-division, it is necessary to state that
another piece of information was at our disposal which
we believed to be intimately associated with cell-division,
although the fact was not appreciated when the point
was first noticed. During experimentation with a
mixture of stain and alkaloid on blood-cells it was
noticed that with a citrated mixture of Unna's stain
and the alkaloid atrbpine the lymphocytes sometimes
extrude granules (fig. 39) from their cell- walls. These
granules remain attached to the cell by means of a
streamer, apparently derived from the cell-wall itself.
The extrusion appears to be a deliberate one on the
part of the cell, and the granule ultimately becomes
separated from it altogether. This extrusion, or " flagel-
lation" as we erroneously called it, has been confirmed

172

FLAGELLATION

173

FIG. 38. — Leucocytes excited by pyridine. No stain.

BBHBHHHBBI

FIG. 39. — A lymphocyte which has absorbed stam and atropine discarding,
its granules (flagellation).

"FLAGELLATION" 175

by L'Engle, of Philadelphia, who has also seen it occur
in poly nuclear leucocytes. When we examined fresh
blood-cells mixed with the pla,sma of cancer patients,
we again noticed that the lymphocytes extruded
granules in some cases apparently in response to
something in the cancer plasma, a point which Dr.
Macalister and I published in The British Medical
Journal on January 16, 1909. Dr. Buchanan, how-
ever, has informed us — and this is a most interesting
point — that he had previously seen similar extrusions
take place in cases of leukaemia, a fact which he
mentions in his book; and a fact which we shall
recall later on. We, however, had never seen these
extrusions occur unless alkaloid or cancer plasma had
been mixed w^ith the cells.

As already mentioned, the commonly accepted
explanation regarding the cause of the reproduction of
cells by individual cell-division is not very satisfactory.
One of the characteristics of living matter is that it
is capable of reproducing itself, and the theory as
to its causation in animal cells was that they, being
living creatures, reproduce themselves because it is an
intrinsic function of the protoplasm — that is to say, that
it is a vital propensity on the part of every cell to
divide automatically, so to speak, and to continue to
do so until it dies. This explanation, however, does
not harmonise with certain known facts concerning cell-
proliferation. For instance, physiologically cell-division
is influenced by conditions outside the cell. The limi-
tation of the size of an organ must be controlled by
some governing factor which influences not only the

176 DIVISIONS INDUCED IN LYMPHOCYTES

proliferation of individual cells, but that of multitudes
of them. It is very difficult to believe that the develop-
ment of an animal from the ovum can be entirely an
automatic function of the protoplasm of individual cells,
unless that function is so controlled that the cells act
together in masses. Moreover, the phenomenon of
healing which has been mentioned presents features
which tend to dispose of the "automatic theory"— a
theory which does not explain why cells immediately
reproduce themselves at a much quicker rate than
normally when a tissue is damaged. Leucocytes, for
instance, will not divide when they are removed from
the body, nor have they ever been seen in the act of
division when examined from the peripheral circulation.
Yet when these cells are shed into a damaged tissue
they proliferate immediately.

Jacques Loeb was, we believe, the first to show that
cell-division in the ova of star-fish can be accelerated
by certain chemical reagents; and further observations
were made in this line of work by B. Moore, H. E. Roaf,
and E. Whitley, who proved that the regularity and
rapidity of growth of the cells of the fertilised ova of
echinoderms could be greatly influenced by certain
alterations in the alkalinity of the water in which they
normally divide. B. Moore has also shown that the
alkalinity of the blood-plasma in cancer is increased—
a point which is of great importance, especially when
we remember that alkalies increase the diffusion of
substances into living cells.

O. and R. Hertwig and Galleoti have described
how, when mitosis occurs in some of the cells of

THEORIES OF CYTOGENY 177

lower animals which have been subjected to certain
alkaloids, such as quinine, nicotine, and cocaine, and
also to antipyrene, the mitotic figures may be of the
asymmetrical type, and that in the case of certain epi-
thelial cells of salamanders the mitotic divisions which
occur in the presence of these substances closely re-
semble the asymmetrical divisions seen in human
cancer-cells. These points took us back once more
to our own knowledge that alkaloids excited leuco-
cytes; but we have never seen divisions, asymmetrical
or otherwise, actually induced in leucocytes by any
alkaloid or other substance.

Farmer, Moore, and Walker had closely studied
the cytology of cancer-cells. They had frequently
seen cells in the act of division in their stained speci-
mens, and they described the asymmetrical "maiotic"
mitoses by which cancer-cells frequently appear to
divide. By the expression "maiotic division" a "re-
duced" division is meant; that is to say, that a cell
divides with a reduced number of chromosomes, and
instead of having its customary somatic number, that
number may be reduced to one-half. In man the
somatic number of chromosomes is thirty-two, and
cancer-cells sometimes divide with sixteen chromo-
somes. Farmer, Moore, and Walker also describe
other characteristics of the several maiotic phases of
mitosis, and they specify two forms of maiotic di-
vision— namely, the first change in a cell's life-
history from its somatic division to the maiotic, which
they call the first (heterotype) maiotic division, and the
succeeding maiotic divisions of its life-history, which

178 DIVISIONS INDUCED IN LYMPHOCYTES

are called the homotype maiotic divisions. These
authors, however, believe that it is not only cancer-
cells which divide by maiotic divisions, but that certain
other tissue-cells also normally proliferate by maiotic
reproduction, especially some cells of the testis, and
the "wandering" cells of the body.

In March, 1909 we discussed the problem of the
causation of cell-division with Professor Harvey Gibson,
who suggested that we might try the effect of nuclein
on cells; and he founded this idea on the well-known
fact that in the sexual generation of the normal alter-
nations of generations of plants the nuclei have only
half the number of chromosomes which are present in
the nuclei of the asexual generation, and that what is
normal in the plant appears to resemble what is patho-
logical in the human being's cancer-cells. It is thus
suggestive that a cancerous growth might be looked
upon as consisting of abnormally induced "gameto-
phytic" or sexual tissue. Professor Gibson, with this
in his imind, suggested that it might be possible by
some means to induce the nuclei deficient in nuclein
to absorb more, and so get back to the normal somatic
condition. Farmer and others have shown that it is
possible to induce such changes in the tissue of ferns,
and for many months one of us (C. J. M.), acting on this
knowledge, treated some cancer patients with nuclein,
which was made by Professor Reynolds Green, but
without proof that it conferred benefit. We, however,
determined to experiment with it on individual cells.1

1 Quoted from a paper, "A Report on Cancer Research," by Dr. Macalister
and myself, in The British MedicalJournal, October 23, 1909.

EXPERIMENTS WITH NUCLEIN 179

From the foregoing facts, believing that it was
reasonable to suppose that chemical agents might
influence human cell-division, we resolved to try the
new in-vitro method. Bearing in mind that the cell-
proliferation of healing appeared to be associated with
the proliferation of cancer, our first step was to try the
effect of nuclein on leucocytes. A saturated solution
of it was made in "citrate solution," and this was mixed
with an equal volume of fresh blood. It was found
that the nuclein seemed to lower the coefficient of
diffusion of the cells very markedly compared with a
control experiment in which no nuclein was employed.
Some nuclein was then mixed up with jelly which
contained stain and which had the right index of
diffusion to stain leucocytes deeply, without killing them,
in twenty minutes. But nuclein did not excite amoe-
boid movements in the cells.

In the next place some juice was squeezed from a
malignant growth and citrated, and the citrated mixture
was in its turn mixed with some fresh normal blood.
It was found that this juice, like the nuclein, lowered
the coefficient of diffusion of the leucocytes, but in
addition it excited amoeboid movements in them.

The lowering of the coefficient of diffusion due to
nuclein was striking, because not only does the juice of
a growth do the same thing, but the cells of cancer
patients usually have a lowered coefficient.

We were not satisfied, however, with this experi-
ment with nuclein, because the preparation of it which
we had obtained was very insoluble in neutral solution,
and it was impossible to employ it in any more concen-

180 DIVISIONS INDUCED IN LYMPHOCYTES

trated form because more powerful solvents damaged
or killed the cells. In place of this nuclein, therefore,
extracts of some dead tissues were made, which we
believed would contain the dead chromatin of cells, and
it is said that chromatin contains nuclein. Moreover,
it was thought advisable to keep as closely as possible
to chemical substances which might be produced in the
body, and the insoluble nuclein which we had used had
been extracted by an elaborate process with hydro-
chloric acid.

To obtain this extract containing — as we believed—
the chromatin of cells, we adopted a principle based on
our observations of the phenomenon of achromasia. It
may be recalled that achromasia is believed to be due
to the chromatin of cells passing out of them, by
dialysis, after their death. Achromasia will readily
occur if cells are allowed to die in a solution which
contains salt; and its onset after death is accelerated
by heat. We therefore made an extract of a tissue by
chopping it up in "citrate solution" and keeping it for
twenty-four hours at 60° C. The first tissue chosen
was lymphatic gland — the reasons for this being the
knowledge that cancer frequently spreads through the
lymphatic channels and glands, that lymphocytes are
always seen in large numbers in growths, that lymphatic
glands contain large numbers of lymphocytes, and
especially because lymphocytes proliferate to a large
extent when a tissue is chronically damaged.

The small prevertebral (haemal) glands of lambs
provided the lymphocytes whose chromatin we hoped
to extract. These glands are composed almost entirely

"PLIMMER'S BODIES" 181

of lymphocytes. In the first instance a dilute extract
was made in citrate solution, kept at 60° C. for twenty-
four hours, and then filtered. Some fresh blood was
mixed in a capillary tube with an equal volume of the
filtrate and incubated at 37° C. for three hours. A drop
of the mixture was then examined on the stained jelly
which excites amoeboid movement in leucocytes (kinetic
jelly). It was at once seen that the coefficient of
diffusion of the leucocytes and lymphocytes had fallen
remarkably — a greater fall than had ever been seen
except that produced by morphine. The nuclei actually
stained on this jelly in about fifteen minutes; and this
jelly will never stain the nuclei of normal leucocytes—
for they burst before that happens. It was also noticed,
however, that the cells contained oval vesicles within
their cytoplasm which closely resembled "Plimmer's
bodies." After a while these bodies became identical
with diffusion- vacuoles of large size, and when they
burst some of them resembled archoplasm. It may
be noted that other authors have suggested that
"Plimmer's bodies" and archoplasm are identical. We
think that these vesicles induced in leucocytes by the
extract are diffusion-vacuoles due to the lowered
coefficient of diffusion.

The next series of experiments was made to observe
the effects of this extract of haemal gland on leuco-
cytes when the cells are spread on jelly which contains
the extract. The jelly-films also contained the correct
amount of Unna's stain to stain the granules of the
cells, so that, if the extract had any action on the
individual cells under these conditions, they would be

182 DIVISION INDUCED IN LYMPHOCYTES

observed nicely stained and yet alive while this action
was taking place. At first a dilute extract was used,
as before, and the films in some instances were incu-
bated for a short time, while others were suitably
prepared for the room temperature. In one or two
cases the lymphocytes seemed to contain some rod-
shaped bodies in the cytoplasm. These rods stained
a bright scarlet, similar to the staining of chromatin,
and nothing had been seen like them before. They
certainly were not bacteria, for we have often seen
ingested bacteria which have quite a different ap-
pearance; besides, they wrere only seen in the lympho-
cytes, which we have never seen to ingest bacteria.
With great hesitation we thought they might be
chromosomes.

Before proceeding farther it is necessary to explain
that at the time when these experiments were made
the appearance presented by the chromosomes of
lymphocytes were unknown; in fact, it was not known
whether these cells from the peripheral circulation
divided by true mitosis or not. One of us had
examined leucocytes by the in-vitro method for four
years, and had never seen anything, previous to these
last experiments, wThich appeared in any way connected
with division of the white blood-cells. It was appre-
ciated that, with the new method, a possibility existed
that cell-division in white blood-cells might some day
be seen; but to observe what appeared to be chromo-
somes in lymphocytes, after we had tried only one or
two groups of substances, seemed to be too good to be
realised. It was necessary to be very careful, however,,

LYMPHOCYTES MADE TO DIVIDE 183

before we came to any conclusion as to the nature of
the bodies which had been observed in the cells.

The first striking point noticed about the red-
staining rods was that they were not within the nuclei,
but were in the cytoplasm outside the nucleus. This
did not seem to be right, if tHte rods were chromosomes.
It is usually considered that the phenomena of mitosis
goes on within the nucleus as it does in plant-cells.
Hitherto, mitosis in human cells, or animal cells gener-
ally, had been seen only in cells which had been killed
and fixed with heat or chemical agents at a time when
they happened to be in the act of cell-division. From
observation by the older method, it was usually under-
stood that during mitosis the nuclear wall vanishes,
and the chromatin within the nucleus forms into
chromosomes, which then migrate into the cytoplasm.
We were prepared to believe that the older methods
might be fallacious owing to distortion caused by the
killing and fixing of the cells, and to the fact that
cells w^ere only caught in the act of mitosis, not
observed undergoing the whole phenomenon from start
to finish. If our observations were correct, the rods
in the lymphocytes were in the cytoplasm right enough,
but the nuclear wall was still there internal to the
chromosomes.

The experimentation was then improved. Instead
of the dilute extract being used, a concentrated one
was made consisting of 50 grammes of ha3mal gland
chopped up in 50 cc. (i.e. 100 per cent) of citrate
solution, kept at 60° C. for twenty-four hours and then
filtered as before. A jelly-film was made thus: To a

184 DIVISIONS INDUCED IN LYMPHOCYTES

tube of 5 cc. of "coefficient jelly," 0.5 cc. (5 units) of
stain and 0 . 8 cc. of alkali solution (8 units) were added,
together with 3 cc. of the 100-per-cent extract, and the
content of the tube was made up to a total of 10 cc.
by 0.7 cc. of water. The jelly was boiled and a film
made from it in the usual way, fresh citrated blood
being placed on it. The object was to see whether this
jelly wrould cause the rod-shaped bodies again to appear
in the lymphocytes, for we believed that it was the
extract which caused their appearance. It was neces-
sary, therefore, to raise the index of diffusion of the jelly
as high as possible short of killing the cells, in order
to cause maximum diffusion of the contents of the jelly
into the lymphocytes. The coefficient of diffusion of
these cells is 14, and we added one more unit of alkali
to the jelly in order to cause the extract to diffuse to
the utmost into the cells. This is the equation :

cf=(5s + 8a + l.5x + 7h + t} - (Qc + l.Sn) = 15.

where x is the 3 cc. of extract which is alkaline to the
extent of about 1 . 5 units, and contains 3 per cent
(3 units) of sodium citrate and 1 per cent (0.5 unit)
of sodium chloride.

Several fields of the specimen were first looked at
and the ordinary resting condition of the lymphocytes
noted. The slide was then placed in the 37° C. incubator
for eight minutes. The same fields (containing the
same lymphocytes) wrere then again examined, and
pictures were seen which had never been seen before,
for nearly every lymphocyte in the specimen was

PHENOMENA OF MITOSIS 185

unquestionably in the act of reproducing itself by /
mitosis.

If any doubt existed as to whether the rod-shaped
bodies which had been seen in the cytoplasm of the
cells were really chromosomes, that doubt was now set
at rest. The cells were certainly not reproducing them-
selves when they were first placed on the jelly-film;
but after they had absorbed the contents of the jelly
during the eight-minutes' incubation at 37° C., they
gradually went through the process of cell-division by
mitosis, and on the removal of the slide from the
incubator they were found in the act of reproduction
with their chromosomes and centrosomes stained bright
scarlet.1

These mitotic divisions, induced for the first time
in living human cells, revealed the fact that the phe-
nomenon of mitosis in lymphocytes differed in many
details from the commonly accepted ideas regarding
karyokinesis which have been adopted from the study
(with the older fixation methods) of dead cells other
than lymphocytes. The nucleus does not vanish; it
forms the spindle. The chromosomes are not derived
from within the nucleus, but are formed from the
normal Altmann's granules which exist in the cyto-
plasm. The centrosomes are not mere "dots" at the
poles of the spindle, but are derived from the nucleolus
which has divided into two.

Fresh films were made, and bloods taken from other

1 That division had been seen in lymphocytes with this jelly, and some of
the facts which led up to this discovery were published by us in The British
MedicalJournal, October 23, 1909.

186

DIVISIONS INDUCED IN LYMPHOCYTES

persons were tried, and before long hundreds of mitotic
figures were induced in lymphocytes, some of which
closely resembled the karyokinesis, as described in
the diagrams and drawings in well-known books on
Cytology. The Altmann's granules, however, always
form the chromosomes,1 the nuclear wall forms the
spindle, and the nucleolus forms the centrosomes.
Thus:

1 The chromosomes of lymphocytes do not always appear as definite
"rods," but may look as if they were composed of masses of granules. See
photos.

THE MITOSIS OF LYMPHOCYTES 187

As will be shown in the succeeding chapters, one
can now induce mitosis in lymphocytes whenever one
pleases, and we have seen all stages of their cell-division.
It must be remembered that to induce all these stages
occupied many months of work, and involved the em-
ployment of many varieties of the jelly-films. I shall
now describe these divisions in detail, because we have
since been able to induce divisions in other human
cells, and therefore there is reason to believe that the
phenomenon of mitosis in other varieties of cells is
similar, if not identical, with that of lymphocytes,
especially as regards the Altmann's granules forming
the chromosomes and the nuclear wall forming the
spindle, both of which are important cytological
points.

The normal lymphocyte (figs. 40-2) occurs in a
great variety of sizes in the body. In the blood
one usually sees the smaller sizes, but in the glands
(and not only in the lymphatic glands) the cell may
reach large proportions. As will be shown later, it
is quite a different class of cell, cytologically, from
the so-called polymorphonuclear leucocyte, and it
must spend only a portion of its life in the peripheral
circulation. The lymphocyte has a large round or
kidney-shaped nucleus, writhin which there are one or
two nucleoli. In the living cell the nucleus appears
to be a transparent membrane (fig. 40) which stains
a faint blue with Unna's polychrome dye,1 and it is
tucked in at its poles to be attached to the nucleolus.
Outside the nucleus, and studded on its surface, a

1 Chromatin stains scarlet.

188 DIVISIONS INDUCED IN LYMPHOCYTES

large number of chromatin granules (figs. 41-2) are
found which really are in the clear cytoplasm, and
they are frequently extruded with the cytoplasm into
the pseudopodia, especially if amoeboid movements are
excited by atropine. When a lymphocyte "flagellates,"
these granules are thrust out through the cell-wall
and become separated. When the cells are on jelly
which makes them divide, amceboid movements cease,
and then the procedure is as follows: The nucleolus,
which is shaped like a minute ring, and stains as if
it was composed of chromatin, splits either into two
rings1 (figs. 43-4), or into two dots of chromatin
which form the centrosomes. They then separate
and emerge at opposite poles of the cell out through
the mass of granules in the cytoplasm (figs. 45—7),
and in doing this they seem to pull out the nucleus
into the form of a spindle. The chromatin granules
of the cytoplasm in the meantime are gradually
collected into masses round the waist of the spindle
(fig. 44), and ultimately they form a belt of
chromatin round it on its outside (figs. 48-9). In
a specimen in which one can see down through the
spindle it will be observed that this belt divides into
a number of chromosomes (figs. 50-1), which become
semilunar-shaped with their points inwards (figs. 52-5).
Each chromosome is in contact with its neighbours
at its points (figs. 62-3). Each one of them then
divides into two (fig. 64). One half of every chro-
mosome travels towards one nucleolus-centrosome,

JOne of us (J. W. C.) has recently seen a ring-shaped centrosome in the
act of division. It appeared hour-glass shaped.

CHROMOSOMES OUTSIDE NUCLEUS

189

FIG. 4(J. — A resting lympnocyte. i\ ote the deeply stained masses of gran-
ules in the cytoplasm, which is bulged out in places. The large transparent
nucleus and the stained ring-shaped nucleolus can also be seen.

FIG. 41. — A resting lymphocyte. The Altmann's granules in the cyto-
plasm are stained.

CHROMOSOMES OUTSIDE NUCLEUS 191

•^^^••^•^^••••^•••••••ftr ^

FIG. 42. — A resting lymphocyte. The cytoplasm, the granules, the nucleus,
and the nucleolus can be distinguished.

Fig. 43. — The earliest stage of mitosis. The nucleolus has divided into two

CHROMOSOMES OUTSIDE NUCLEUS

193

FIG. 44. — Early mitosis in a lymphocyte. Looking down through the
spindle (polar aspect). The nucleolus has divided into two centrosomes,
each of which is ring-shaped. The spindle is surrounded by a belt of chro-
matin granules.

FIG. 45. — Mitosis in a lymphocyte. Profile aspect. The two ring-
shaped centrosomes can just be seen towards the poles. The granules are
becoming formed into chromosomes.

CHROMOSOMES OUTSIDE NUCLEUS

195

FIG. 46. — Foreshortened appearance of a mitotic figure in a lymphocyte.
The position of one nucleolus-centrosome at the pole of the figure is well
shown.

FIG. 47. — Profile aspect of mitosis in a lymphocyte. The relative positions
of the centrosomes andi-chromosomes can be seen.

CHROMOSOMES OUTSIDE NUCLEUS

197

FIG. 48. — Profile aspect

f mitosis. The belt of chromatin is formed round
the waist of the cell.

FIG.. 49. — One resting and one dividing lymphocyte. In the latter the
chromosomes are beginning to divide. The centrosomes appear as dots
of chromatin.

CHROMOSOMES OUTSIDE NUCLEUS 199

r i

L A

jTIG 50.— Polar aspect. The belt of chromatm granules is dividing into

pIG_ 51. — Polar aspect. The chromosomes are becoming semicircular.

CHROMOSOMES OUTSIDE NUCLEUS

201

FIG. 52. — Polar aspect. An "aster" stage of mitosis in a lymphocyte.

FIG. 53. — Polar aspect. Some ot tne chromosomes are semicircular-shaped;
some are dots of chromatin.

CHROMOSOMES OUTSIDE NUCLEUS 203

FIG. 54. — Polar aspect. One centrosome can be seen at the pole of the

"aster" figure.

FIG. 55. — Polar aspect. Sixteen chromosomes could be counted in this cell.

CHROMOSOMES OUTSIDE NUCLEUS 20.5

JTIG 56.— Profile aspect of mitosis in a lymphocyte.

FIG. 57. — Profile aspect of mitosis in a lymphocyte.

CHROMOSOMES OUTSIDE NUCLEUS 207

FIG. 58. — -Profile aspect. The chromosomes can be seen at the waist of the

spindle.

FIG. 59. — Profile aspect. A figure frequently seen.

CHROMOSOMES OUTSIDE NUCLEUS 209

FIG. 60. — Profile aspect of mitosis.

FIG. 61. — Oblique aspect of mitosis in a lymphocyte.

CHROMOSOMES OUTSIDE NUCLEUS

FIG. 62. — Polar aspect of mitosis in a large lymphocyte from a patient
suffering from carcinoma. There are sixteen chromosomes.

FIG. 63. — Polar aspect. The chromosomes were V-shaped with their
apices inwards to be attached to the nucleus-spindle, which can dimly be
made out.

CHROMOSOMES OUTSIDE NUCLEUS

213

FIG. 64. — Polar aspect of mitosis in a large lymphocyte from a cancer
patient. The chromosomes are dividing.

FIG. 65. — Profile aspect of mitosis.

CHROMOSOMES OUTSIDE NUCLEUS 215

FIG. 66. — Profile aspect. The figure is fully formed. One nucleolus-
centrosome is ring-shaped ; the other is a dot of chroma tin.

FIG. 67. — Profile aspect. The sixteen chromosomes could be counted.

CHROMOSOMES OUTSIDE NUCLEUS

217

FIG. 68. — The cell has become constricted in its centre.

FIG. 69. — Prohle aspect. Complete division is about to occur. The
chromosomes are being reconverted into granules, but the mitotic figure is
not quite finished at the dividing-point.

CHROMOSOMES OUTSIDE NUCLEUS

219

FIG. 70. — Profile aspect. The spindle and chromosomes have divided, but
the cell wall-has not yet separated.

FIG. 71. — Completion of mitosis in a lymphocyte.

CHROMOSOMES OUTSIDE NUCLEUS

and the other half towards the other centrosome
(figs. 56-61, 65-7). The spindle divides in the centre
(figs. 68-70); and lastly the cell itself divides (fig. 71).
In each daughter cell the chromosomes return to
their granular condition and pervade the whole
cytoplasm. The single centrosome (for there is now
one only in each daughter cell) again becomes tucked
into the centre of the transparent nucleus — which
consists of one half of the original spindle, and thus
the cycle of mitosis is completed. Doubtless each
chromosome granule divides during some part of the
cycle, but owing to their minute size we have not
been able to see their division.

The cells, of course, do not usually divide in
definite stages such as the aster and diaster, although
sometimes a cell will be found which presents one
of them. Sometimes one sees that the chromosomes
may be dividing in one part of the cell, while some
chromosomes in another part are being reconverted
into granules of chromatin. The way in which a
cell is lying on the jelly must be taken into con-
sideration in the determination of the stage of mitosis.
One rarely finds a perfect figure as described in
diagrammatic drawings of other types of cell, for the
cells frequently appear foreshortened owing to the
oblique manner in which they happen to come to
rest under the cover-glass. The position of the ring-
shaped nucleolus-centrosomes is of prime importance
in the determination of the stage of the initotic
figure.

In observing any stage of mitosis, however, it will

DIVISIONS INDUCED IN LYMPHOCYTES

be seen at once — a point on which we must lay stress
—that the chromosomes are outside the nucleus
are formed by the conglomeration of the Altmann's
granules in the cytoplasm. As will be shown later
on, this is also the rule in leucocytes and some epithelial
cells as wrell as in lymphocytes.

The phenomenon of mitosis, then, as seen in these
cells when they are stained alive, differs very materially
from the usual descriptions of it as seen in cells which
have been killed, fixed, cut into sections or otherwise
manipulated, and stained. The old idea was — although
divisions had not been seen actually in lymphocytes—
that the chromosomes were formed out of some
chromatin which is within the nucleus, and that, inside
this again, a spindle, which does not exist in the rest-
ing stage, is formed. The nuclear wall was described
as vanishing during mitosis according to most con-
ceptions. But, as I have described, mitosis is a much
simpler phenomenon.

The misconception has been due, I think, to several
factors. In the first place most cytological research has
been carried out with plant-cells, and animal cytology
has arisen from it. In the second place, cells up to now
have only been caught in the act of mitosis; their cycle
of cell-division has not been followed from the resting
stage to completed division in one individual cell. The
morphological elements of a resting cell have been
studied, and then those in one killed in the act of
division, and the part played by each element has been
deduced from its new position — not watched through-
out. Lastly, owing to manipulation, the so-called

ALTMANN'S GRANULES 223

Altmann's granules of some cells have been crushed
into the nucleus, in which case they look as if they
formed part of it, or were inside it — a fallacy which has
given rise to great controversy regarding the nature of
these granules, to the statement that they do not exist
in some cells, e.g. lymphocytes and cancer cells, and to
failure of appreciation of the fact that the chromosomes
are formed out of them. Let mitosis be induced in
a living cell and no second glance will be required
to realise the real sequence of events.

The mitosis of plant-cells seems to go on within
the nuclear wall, but this is not the case in the animal
cells which we have seen. The granules in the cyto-
plasm of Altmann 's granules are larger in some classes
of cells than in others. For instance, they are much
larger in eosinophile leucocytes than in lymphocytes.
When they are large their position is obvious, but when
they are small, as in lymphocytes and cancer-cells,
during the killing of the cell as it is fixed the small
granules — which are composed of chromatin — adhere
to and are merged into the nucleus. No matter how
we try to fix a specimen, death takes time, and the
liquefying cytoplasm bulges out the cell-wall. The
rapidity of death depends upon the diffusion of the
fixative into the cell, and this diffusion takes time.
Hence when a cell is stained and fixed, it appears as
if its nucleus is a mass of chromatin — which it
is not — and its halo of cytoplasm, which has bulged
out of the cell- wall, is now apparently devoid of
Altmann's granules. This is a pitfall into which we fell
ourselves, for, although we had seen the granules of

224 DIVISIONS INDUCED IN LYMPHOCYTES

lymphocytes and cancer-cells forming the chromo-
somes, we thought that these granules were composed
of chromatin, but that Altmann's granules as ex-
emplified in polynuclear leucocytes were of quite a
different nature. As will be shown later, the granules
of leucocytes also form the chromosomes in the same
way as those of lymphocytes and cancer-cells. Professor
Lorraine Smith suggested that the apparent absence of
granules in lymphocytes and cancer-cells might be due
to the fixative, and he is right. The cytoplasm of every
living lymphocyte is full of minute granules which stain
like chromatin with aniline dyes, and these granules
clump together to form the chromosomes during cell-
division — a point about which there can be no question
whatever.

In some other cells, such as some large cells of the
liver, we have seen large granules in the cytoplasm (as
well as fat globules), which will not stain. What their
function is we do not know, for we have not been able
to induce divisions in these cells. The granules of
lymphocytes we shall henceforth style "chromosome-
granules," the nucleolus as the "nucleolus-centrosome,"
and the nucleus as the "nucleus-spindle."

CHAPTER XI

THE DIVISION OF LYMPHOCYTES INDUCED BY THE ANILINE

DYE THE AUGMENTING ACTION OF ATROPINE AND

EXTRACT OF HAEMAL GLAND " AUXETICS " -— THE

CYCLE OF CELL-DIVISION THE POSSIBILITIES OF

THE INDUCED CELL-DIVISION BEING DUE TO
" DEATH-STRUGGLES " —ASYMMETRICAL AND RE-
DUCED DIVISIONS

THE fact that mitotic figures could be made to appear
in lymphocytes was very satisfactory, for it seemed to
us to be a step in the solution of the problem of the
cause of the multiplication of cells. It was true that
we had only seen them in lymphocytes; but still these
mitotic divisions had occurred in response to the action
of a chemical substance, and if these cells were capable
of dividing in response to it, it appeared reasonable to
suppose that other cells would do the same and that
it was possible that they would only divide when they
absorbed a chemical substance. We believed at first,
of course, that the substance which had induced the
divisions was the extract of the dead haemal gland;
but before many experiments had been made this
suggestion received a check. One day a cell was seen

225

226 THE CYCLE CELL-DIVISION

stained in an early stage of mitosis; its ring-shaped
nucleolus-centrosome was lying at a pole of the nucleus-
spindle in the cytoplasm, outside the mass of granules
which had not yet collected round the waist of the cell.
Now when this early figure was seen by me, I remem-
bered that I had seen something very similar to it before,
and on turning up a paper (British Medical Journal,
January 16, 1909), which described some work done
more than a year previously, it was found there men-
tioned that the nucleoli sometimes appeared outside
the nucleus in the cytoplasm. Now, this position of
the nucleolus-centrosome is the first step in mitosis,
and therefore it was grasped that this mitosis must have
been seen before, although the fact was not realised at
the time. Another far more important point was also
grasped, viz. that when the mitosis had been seen a year
previously, no extract of hsemal or any gland had been
either used or thought of.

The notes of the previous work were referred to,
and it was found that when the — as it turned out-
early stage of mitosis had been seen, the cells had been
resting on a jelly which contained only Unna's stain
and atropine. It was clear, therefore, that either one
or both these substances would induce divisions in
lymphocytes and our hopes were rather damped, for
both these substances, unlike the extract of haemal
gland, are entirely artificial, and could not possibly be
concerned in the cell-proliferation of healing.

Each of the ingredients of the jelly — described in
the last chapter — which induced well-marked mitosis
in lymphocytes was now tried separately. Jellies were

MITOSIS DUE TO THE DYE 227

prepared which contained each of them in turn, and
jellies were prepared which contained only the salts
sodium citrate and sodium chloride. Many experi-
ments \vere made from each, and several different
strengths of the different substances were tried re-
peatedly on fresh lymphocytes. It was thus ascertained
that Unna 's polychrome methylene blue (Grubler) con-
tains some substance which will induce divisions in
lymphocytes. It requires a high concentration of this
stain for this purpose, and this was the reason why
advanced divisions had not been seen in the several
years' previous work with this dye. Unless the jelly
contained Unna's stain, no mitosis whatever would
occur. Repeatedly they were tried, but none of the
other ingredients by themselves could be made at that
time to cause lymphocytes to reproduce themselves.
The 100-per-cent extract of hsemal gland by itself
certainly did not do so, nor did the atropine; but
both the extract and atropine — and this was an im-
portant point — greatly augmented the action of the
stain in inducing mitosis. By itself at least 10 units
(1 cc.) of polychrome stain were required to induce
mitosis; but if a certain quantity of atropine or of
extract, or, better still, of both, was also added to the
jelly, one could cause advanced mitosis in lymphocytes
with only two or three units of stain. It was a
remarkable state of affairs that neither atropine nor
extract would induce divisions by themselves, but that
they augmented the action of the stain in doing so to
a very marked degree.

During all this experimentation, which occupied a

228 THE CYCLE OF CELL-DIVISION

considerable time, many points connected with the proc-
ess of inducing divisions were learnt. We had three
factors to deal with, viz. polychrome dye, atropine, and
extract consisting of the soluble remains of dead haemal
gland of 100 per cent. It was found that lympho-
cytes would not make any attempt whatever to divide
unless they absorbed some of the polychrome stain. As
the stain passed into the cells, it stained first their
chromosome-granules and their nucleolus-centrosomes.
Like polynuclear leucocytes, lymphocytes do not appear
to surfer much harm to their Jives while their granules
are stained, but as soon as their nucleolus-centrosomes
are reached by the dye death occurs. Mitosis takes
place about the time when the granules are staining,
and therefore the rapidity of the onset of mitosis depends
on the rapidity of the diffusion of the dye into the cells.
It is thus evident that the gradual diffusion of the
stain first causes mitosis and then death because it
kills the cells by combining with and staining the
nucleolus-centrosome. The rapidity of the diffusion
of the stain is increased by concentrating it, by the
presence of alkalies, or by heat. These factors also
hasten death and they likewise hasten cell-division.
With regard to the factor heat, however, we must add
the qualification that no lymphocyte will divide below
a temperature of 30° C. or above about 40° C., and for
this reason we have employed a temperature of 37° C.
throughout these experiments for inducing division.

Now, mitosis is a process which occupies a certain
amount of time. If the diffusion of the stain is
very slow, the time taken by the act of mitosis is

TIME TAKEN BY ACT OF MITOSIS 229

correspondingly slow. But as far as we can see, mitosis
cannot occur completely in less than about three
minutes. It can take a very long time in its accom-
plishment ; but it cannot be completed in less than three
minutes. Hence, if mitosis can take place slowly,
without the cell being killed by the stain, complete
mitosis can occur ; but if the nucleolus-centrosome
stains in less than a minute or so, death will occur
before the cell has had time to divide. This fact
governs the wrhole of this experimentation, for when
inducing cell-division with the aniline dye it must
be remembered that the mitosis has to occur after the
cell-granules have begun to stain, but before death is
occasioned by the staining of the nucleolus-centrosome.
We have the power of accelerating and delaying the
diffusion of the stain into the cells by adding or sub-
tracting alkali, or by increasing or decreasing the
concentration of the stain by rules which can be plotted
in an equation, and therefore by such an equation we can
ascertain the rate of cell-division as induced by the
chemical agent. But throughout it must be appre-
ciated that it is the stain which is inducing the
cell-division, and that if the stain is not sufficiently
concentrated no division will occur at all. On the
other hand, it must also be remembered that an excess
of stain will poison the cells too quickly. A cell must
absorb a certain amount of stain before it will divide,
and the absorption depends on the concentration of the
stain in the jelly and on the alkali. One may place
living blood-cells on a jelly wrhich contains the best
ingredients for inducing cell-division; but unless the

230 THE CYCLE OF CELL-DIVISION

alkali is correct according to the equation — that is to
say, unless the index of diffusion of the jelly is correct
for the coefficient of diffusion of the cells, the latter
will take no notice whatever of the mitosis inducing
agents in their surroundings. But make diffusion
factors of the jelly right and the cells will then re-
spond immediately, and as many as 90 per cent of
the lymphocytes in a specimen may be made to divide.
Not only does the rapidity of the onset of mitosis
depend on the physical laws of the diffusion of sub-
stances into cells, but the actual stage reached in a
given cycle of cell-division also depends on them; for
the completion of the mitotic cycle occupies a certain
amount of time, which varies inversely with the quantity
of the stain absorbed by the cell, and this absorption
depends on the coefficient of diffusion, heat, alkali, etc.
The following experiment illustrates this point. A
jelly-film was made which induced almost completed
divisions in lymphocytes in ten minutes. By making
several films and removing them, one at a time,
from the 37° C. incubator at each minute, it was
seen that mitosis began with the staining of the
granules at about the seventh minute, and that death
occurred at about the ninth. The experiment was
repeated, and at the seventh minute, immediately while
mitosis was occurring, the slide was quickly removed
from the 37° C. incubator to one which maintained 32° C.
The sudden lowering of the temperature delayed the
diffusion of the stain into the cells, and the interesting
point is that the mitosis ceased when the diffusion of
the stain was suddenly arrested, and the cells died

INFLUENCE OF VITALITY 231

slowly. Twenty minutes afterwards, when all the
chromatin was stained, it was seen that the mitosis
had been arrested in those early stages reached at the
seventh minute.

Thus it appears from this experimentation that not
only will a lymphocyte not reproduce itself in vitro
unless it absorbs a chemical "exciter of reproduction,"
but also the actual stage reached in its act of mitosis
varies directly with the quantity of that substance
which has diffused into the cell. It follows that,
in vitro, before a cell will reproduce itself completely
it must receive a definite quantity or dose of the
chemical substance.

In addition to the above factors, the divisions of the
cells depend upon their vitality. If some blood is
citrated and kept for two days, it is very difficult to
induce divisions in the lymphocytes. The longer cells
have been shed the slower they are to respond to the
division-inducing action of the stain, in spite of the fact
that their coefficient of diffusion has fallen. It is im-
possible to induce divisions in cells with auxetic jelly if
other cells from the same sample of blood will not show
excited movements on kinetic jelly.

The foregoing points showed that the reproduction
of lymphocytes in vitro depended entirely on the
aniline dye. The dye did not merely increase the cells'
propensity to divide; it actually caused the division.
Lymphocytes had never been seen to divide before, and
they certainly will not divide in vitro unless one takes
deliberate steps to make them do so. Mitosis is a
complex phenomenon which only occurs as an act of

THE CYLE OF CELL-DIVISIOX

cell-reproduction, and in vitro the only way to cause
it to take place was to force the cells to absorb the
chemical "exciter of reproduction" contained in the
aniline dye. It appeared reasonable to us to suppose
that there might be other "exciters of reproduction,"
not only for lymphocytes, but for other cells as well,
and therefore we proposed to call the substance in the
aniline dye which caused cell-division in lymphocytes
an "auxetic" (dwfjjrwos, an exciter of reproduction),
a convenient term suggested by Professor Harvey
Gibson, which might be applied to other substances
having a similar action if such were proved to exist.

The next steps were to investigate the "augment-
ing" actions of both atropine and the extract of haemal
gland. It has been pointed out how atropine, being
an alkaloid, greatly excites amoeboid movements in
lymphocytes and leucocytes, and it was soon seen that
atropine also greatly augments the action of the poly-
chrome dye in inducing mitotic figures in lymphocytes.
The best strength of atropine to be added to the jelly
which contains the stain is that which causes maximum
excitation of amoeboid movements. If this is done
lymphocytes can be caused to divide with the strength
of the stain reduced to one-fifth of the minimum
amount of it which will, by itself, induce mitotic
^figures. In other words, atropine will not by itself
induce divisions on the microscope slide, but it will
augment the "reproducing" action of polychrome stain
five-fold. Another point was also noticed, which was
very material to the main object of these researches,
in that stain, plus atropine, caused lymphocytes to

STAIN ACTS ON THE GRANULES 233

FIG. 72. — Asymmetrical mitosis in a lymphocyte induced by azur stain
augmented by atropine.

FIG. 73. — Asymmetrical mitosis induced by azur stain augmented by

atropine.

STAIN ACTS ON THE GRANULES 235

undergo curious one-sided mitoses in some instances
(figs. 72, 73).

We now investigated the "augmenting" action of
the extract of haemal gland. This was even more
powerful than that of atropine. So great was it that
one can employ a jelly which only contains three units
of polychrome stain — which will never induce divisions
by itself; and if 3 cc. of the 100-per-cent extract of
dead haemal gland is also contained in the jelly, complete
divisions can be induced in lymphocytes without the
cells actually being coloured by the stain at all. Yet
all attempts at this stage to cause the extract to induce
divisions by itself had failed.

Thus, by means of a mixture of a little stain, say , (\
4 units, 0 . 7 cc. of a 1-per-cent solution of atropine
sulphate, 3 cc. of the 100-per-cent extract of haemal
gland, 6 units of alkali solution, and 0.3 cc. of water
added to 5 cc. of coefficient jelly to make a total of
10 cc., one can induce advanced divisions in lympho-
cytes, without the cells staining at all in ten minutes.

We have already stated that mitosis occurs about
the time when the stain has diffused into the cells
sufficiently to stain the granules. But now with the
combination of stain and the augmenting substances
mitosis will occur without the stain colouring the
granules at all. In spite of this, however, stain is
essential. Hence we suggest the theory that the
stain induces divisions by acting on the chromosome

236 THE CYCLE OF CELL-DIVISION

granules; but that, since it is not necessary for it
actually to colour these granules, as shown by the last
experiment, it seems probable that the stain induces
divisions by virtue of some substance contained in it
which does not colour granules. It is not the stain
itself which induces divisions; it is some constituent
of it, and the action of that constituent is greatly
augmented by atropine and extract.

The next point is that when mitosis is induced on a
microscope slide with stain, death is premature. Even
if there is not sufficient stain to colour the nucleolus-
centrosomes, death rapidly follows. We believe that
this dye contains at least two constituents which can
be utilised differently by the cell's protoplasm — a
substance \vhich, by combining with the cell-gran-
ules, causes the cell to reproduce itself, and a poison
which kills it. Both diffuse into the cell together;
mitosis is induced and then the cell dies prematurely.
If the stain is sufficiently concentrated, the chromatin
after it is dead will combine with it, and the chromatin
then turns bright scarlet. From prolonged observation
of these induced divisions we think that the scarlet
coloration of the chromatin is a post-mortem effect.
The stain as it diffuses into the cell induces division as
it combines with the granules, which die and become
coloured one by one. All the time the stain is passing
farther into the cell, and later and later stages of mitosis
are being induced. Ultimately the nucleolus-centro-
some is reached and the cell dies; and thus it is seen
dead in the act of mitosis with its chromosomes and
centrosomes stained bright scarlet. If, on the other

THE "AZUR" PRINCIPLE 237

hand, the concentration of the stain is reduced and its
action augmented by atropine and extract, still the
poison, but in less strength, passes into the cell; and
although mitosis occurs to an advanced degree, never-
theless premature death occurs in spite of the fact that
there is not sufficient strength of colouring matter to
give rise to the post-mortem coloration of the chromo-
somes and centrosomes. Death is a gradual process-
presumably it is molecular as well as cellular, for the
post-mortem scarlet coloration occurs gradually; but
it is not until the nucleolus-centrosome is reached
that all mitosis ceases. One cannot excite amoeboid
movements in a cell which has its nucleolus stained.

Since Unna's polychrome methylene blue contained
the active principle which caused the cells to divide,
and the other two substances appeared merely to be
augmenters, we now turned our attention more es-
pecially to the dye. Polychrome methylene blue stains
chromatin scarlet and the nucleus-spindle a faint blue.
It is made by "poly chroming" methylene blue. Fresh
methylene blue stains chromatin blue, and it is not so
effective as the polychrome dye in inducing mitosis.
The "poly chroming" process consists of rendering
a solution of methylene blue alkaline with sodium
carbonate and naturing it for some time at a high
temperature. The methylene blue turns a purple
colour. This is due to decomposition — an oxidation
occurs with the production of a dye known as "azur."1
This azur dye can be obtained from dealers, and it
can be extracted from the polychrome dye by means of

1 Centralblatt fur Bakteriologie, bd. xxix., 1901, p. 765.

238 THE CYCLE OF CELL-DIVISION

chloroform. It was found that the constituent which
induces divisions in lymphocytes is almost entirely
confined to the azur dye. The more the azur was
extracted the less efficient the polychrome dye became,
and the azur is very potent although it does not stain
the chromatin as well as the polychrome dye.

A concentrated solution of azur dye was made
thus: In a burette 20 cc. of Unna's polychrome methy-
lene blue (Grubler) had added to it 20 cc. of chloroform,
and the mixture was allowed to stand for 12 hours.
The chloroform, which sinks to the bottom, carrying
some of the azur dye with it, was then run off into
a shallow dish, where it was allowed to evaporate.
20 cc. more of chloroform was then added to the
original 20 cc. of stain in the burette, and, after
12 hours, it, in its turn, was run off into the same
dish and also allowed to evaporate. This procedure
was repeated five times, and the dry azur dye was
so obtained. Lastly, a solution of this dye was made
by adding 5 cc. of water to the dish. This potent
dye is a fluorescent red one, which, when dry, shines
with a metallic lustre. A very powerful jelly for
causing mitotic divisions in lymphocytes was made by
substituting 0 . 4 cc. of this potent solution for the 0 . 2 cc.
of Unna's stain in the last equation. By means of this
<jelly all stages of divisions can be readily obtained, it
being only necessary to vary slightly the content of
alkali in producing early or late mitosis in the ten
minutes. It is better to keep at least two units of
polychrome stain in the jelly, in order to stain the
chromosomes more deeply.

REDUCTION DIVISIONS 239

In experimenting with this last jelly containing
azur dye, an important point was found out. By
delaying death as long as possible, by employing the
minimum amount of alkali wrhich will make the cells
undergo mitosis, we at last succeeded in keeping the
lymphocytes alive for twenty minutes, and yet mitosis
was being induced during the greater part of the time.
That is to say, mitosis was induced as early as pos-
sible, for, as will be shown in the next chapter, we
cannot, under the experimental conditions, keep up
the cell's vitality longer than twenty minutes, and it
is difficult to keep them alive to divide for a longer
time than even ten minutes; still their lives have been
prolonged for twrenty minutes.

The point revealed by this experiment was that
the so-called reduction division is not a special form
of mitosis in lymphocytes. The somatic number of
chromosomes in the body is thirty-two, but hitherto
in all the dividing lymphocytes in which it was pos-
sible to count the actual number of chromosomes their
number was either sixteen or thereabouts (figs. 62, 63,
67). In other words, the divisions which we induced
with the stain in ten minutes were of the reduced
variety, or what Farmer, Moore, and Walker called
"maiotic" divisions. By prolonging life, however, for
twenty minutes, and inducing the divisions slowly,
especially if only the early stages of mitosis were
induced in the time, it was found that now lympho-
cytes divided by somatic divisions with more than
sixteen and sometimes with a full number of thirty-
two chromosomes (figs. 74, 75), and the statement

240 THE CYCLE OF CELL-DIVISION

that the wandering cells of the body only divide by
reduced divisions of the "reproductive" type is thereby
disproved. In lymphocytes examined on a microscopic
slide the question of the number of chromosomes seems
to be entirely one of degree — it depends on the rapidity
of the division, which, in its turn, depends on the
quantity of the "auxetic" absorbed by the cell. By
increasing the alkali one can induce divisions very
quickly, provided of course there is not too much
alkali. We have seen lymphocytes divide with less
than sixteen chromosomes, and on one occasion, when
mitosis was very rapidly induced, the number was
reduced to eight only; but the number of chromo-
somes seems usually to remain in these round numbers,
namely, thirty-two, sixteen, eight, the last one being
very rare. If a division is induced in the usual way
with a jelly wThich will kill the cells in about ten
minutes, the number of chromosomes is nearly always
sixteen, but a slow division will be a somatic division.
There does seem, however, to be a difference in the
way the chromosomes split. We have seen them in
the act of splitting longitudinally, and also, and more
commonly, they split transversely; although Avhether
the longitudinal splitting is significant of a "first
(heterotype) maiotic" division or not we are not in
a position to state.

The asymmetrical mitoses induced when atropine
is present, especially if it is present to excess, are
interesting. The mitosis seems to be going on in
one side of the cell. We have not seen a completed
division in one of those asymmetrical mitoses, but we

FACTS SUMMARISED 241

FIG. 74. — An early stage of delayed mitosis induced by a jelly with a low
index of diffusion. The number of chromosomes is more than sixteen.

FIG. 75. — Thirty-two chromosomes could be counted in this cell. Early

mitosis delayed.

FACTS SUMMARISED 243

think that, from the appearance of the cells, they are
about to divide into more than two daughter cells by
some quite atypical arrangement of the chromosomes.
The point is a very important one, for asymmetrical
divisions are reported to be frequently seen in can-
cerous growths.

We may now summarise the facts learnt from the
mitotic divisions induced in lymphocytes by the aniline
dye. (1) Lymphocytes will not divide in vitro unless
they absorb the chemical agent. (2) The rapidity of
the onset of division depends on the rapidity of the
diffusion of the agent. (3) The time occupied by the
act of division depends on the amount of the agent
absorbed and the time occupied in the diffusion of the
substance into the cell. (4) If the diffusion is slow
the cells divide with the somatic number of thirty-
two chromosomes; but if it is rapid the number is
reduced. (5) A "reduction division" means that a
cell is very prolific, owing to its absorption of a large
quantity of the chemical agent. (6) The rapidity of
the absorption of the agent depends on the coefficient
of diffusion of the cell, the concentration of the agent
in the surrounding fluids, and on the presence and
strength of the factors which increase or decrease the
diffusion of the substances. (7) Lastly, it depends on
the vitality of the cells themselves. In fact, the
division of lymphocytes on a microscope slide depends
entirely on the presence or absence of a chemical
agent, and, if it is present, on its strength and on its
diffusion into the cell.

In our opinion, judging from the mitotic figures

244 THE CYCLE OF CELL-DIVISION

which have been induced in cells, mitosis should not
be described as the phenomenon of nuclear division.
It is part of the cycle of cell-division, and the whole
of the cell-protoplasm takes part in it. The Altmann 's
granules form the chromosomes, the nucleolus forms
the centrosomes, and the nucleus forms the spindle.
The protoplasm of the cytoplasm and cell-wall also
reproduces itself and divides during mitosis.

The active principle in the stain which causes
mitosis in lymphocytes is a constituent of the azur
dye. This dye also contains a substance which kills
the protoplasm, and having done this it will, if in
sufficient concentration, cause that protoplasm to stain
scarlet. Mitosis occurs by the action of the active
principle on the chromosome granules; cell death
occurs by the action of the poison on the centrosomes.
So far the active principle has proved to be inseparable
from the poison in the anilin dye.

Up to this stage in the researches the only sub-
stance we had found which would induce divisions in
lymphocytes was this anilin dye; but its action was
augmented by atropine and an extract of dead haemal
gland. Atropine augments its action five-fold if it is
absorbed in suitable strength, in which case it may
induce asymmetrical mitosis. Neither atropine, in no
matter what strength, nor extract of dead haemal gland
in the strength of 100 per cent will by themselves
induce mitotic figures in lymphocytes.

Great care must be exercised in the practice of
inducing the mitotic figures in lymphocytes. The
jelly must be accurately prepared, but it is better to

TECHNIQUE 245

allow it to be deficient in alkali by a unit or so at
first. A film is made with some fresh blood spread
on it and incubated for ten minutes. The temperature
of 37° C. must be accurate; many failures have resulted
owning to the neglect of regulation of the incubator
temperature. If the chromosome-granules of the
lymphocytes are unstained, a drop or two of alkali
solution is added to the jelly and a fresh film is tried.
Soon the right alkalinity will be obtained to induce
early mitosis in the ten minutes ; and now if more alkali
is very carefully added to the jelly and another film
made, later stages of mitosis will be induced. It is
instructive to proceed farther and once more add alkali,
when the cells will be killed too quickly, and only very
early stages will be seen in the ten minutes, for there
has not been time for late phases to occur before the
cells have died. If more alkali is again added, owing
to rapid death the cells will appear quite at rest, as if
there had been no agent to cause cell-division in the
jelly at all. But the granules and nucleoli will be
deeply stained, and the polymorphonuclear cells will
probably all be burst and achromatic.

When first we showed the mitotic figures to some of
our friends we received some adverse criticisms. It is
always possible to induce mitosis in lymphocytes, but it
is not always possible, at a few minutes' notice, to find
figures which resemble the diagrammatic drawings of
mitosis in the cells of plants and the lower animals as
given in the text-books on cytology. However, when
a convincing figure did appear, the nature of the
chromosomes, the spindle, and the centrosomes were

246 THE CYCLE OF CELL-DIVISION

immediately appreciated. We, of course, maintained
that the divisions were induced by a specific chemical
substance contained in the stain, pointing out that
lymphocytes had never been seen to divide before, and
that mitosis will onlv occur in them if they absorb a

«/ «/

certain quantity of the substance. But our friends
"one and all began to make excuse." Some said that
the divisions were in the nature of a death-struggle;
they pointed out — a fact which we admitted — that
death always was premature, and it usually occurred
during the act of mitosis. We explained the cause of
death, but still the suggestion of the "death-struggle"
was maintained by some in the absence of proof
against it.

Others suggested that the divisions were entirely
artificial and not at all like the natural method of cell-
proliferation, although they had never seen the latter.
We admitted the fact that at this stage of our re-
searches we could only induce divisions in lymphocytes,
and we could only do this with an entirely artificial
anilin dye; but still we found it difficult to appreciate
why a cell should go out of its way to divide by an
entirely abnormal process. We suggested that we
thought that if a cell was going to divide at all, it
would try to do so by the normal process to which
it was accustomed. But the suggestion that the mitotic
divisions were "freaks" remained to be disproved.

It appeared to us a remarkable thing that a cell
should try to reproduce itself by cell-division in a
death-struggle; it seemed such a futile thing for it
to do. Moreover, other stains — such as ordinary

TECHNIQUE 247

methylene blue — do not induce divisions like azur,
and yet they kill the cells by staining the chromatin
of the nucleolus-centrosome. Why should only the
latter dye cause cell-division; presumably both would
cause death-struggles ? Moreover, we have often killed
cells by prussic acid and nitro-benzol, but no division
occurred and nothing resembling a death-struggle.
Again, in connection with the experiment given in
this chapter in which it was ascertained that the stage
reached in a single act of mitosis varies directly with
the quantity of the chemical substance absorbed, it
appeared to us that if these mitotic figures induced in
lymphocytes were in the nature of death-struggles,
a cell once it had been started in its act of mitosis
would continue that act until it was complete. But
as the experiments showed, they did not do so, for
when the diffusion of the chemical agent was arrested
the mitosis ceased even in its early stages.

These suggestions were worthy of consideration,
and the only way to disprove them was to continue
the investigations. It appeared to us reasonable to
suppose that other cells, besides lymphocytes, would
possibly respond by dividing to the chemical auxetic,
and we also considered it possible that other chemical
agents existed which would induce divisions. Lym-
phocytes responded in such a constant manner, and
always required a definite quantity of the substance
that we thought it possible that there might be some-
thing similar to it in the body which would cause their
proliferation. The question was, What was this sub-
stance and where was it contained ? Was it associated

248 THE CYCLE OF CELL-DIVISION

with the cell-proliferation of healing ? Lymphocytes
proliferate in healing, especially in chronic healing, and
chronic healing is a forerunner of cancer. Still
leucocytes proliferate even more than lymphocytes in
healing, but we had never up to this point, nor had
any one else, ever seen a leucocyte divide. If the agent
in the azur dye was analogous to a chemical substance
in the body which caused the cell-proliferation of
healing, it ought, strictly speaking, to cause the multi-
plication of polymorphonuclear leucocytes as well as
lymphocytes. But so far it had not done so.

All these points were carefully considered at this
stage, and they urged us to make further researches.
Whether we were right or wrong in supposing that
there might be a chemical auxetic in the body which
caused proliferation of cells in a manner similar to
the agent contained in the azur .dye, we had made
one step in causing one class of human cells to divide
on the microscope stage by means of a chemical agent.

CHAPTER XII

THE "EXPERIMENTAL TEN MINUTES" —DIVISIONS IN-
DUCED IN THE SO-CALLED POLYMORPHONUCLEAR

LEUCOCYTES METHOD FOR COUNTING THE NUMBER

OF GRANULES CONTAINED IN EOSINOPHILE LEUCO-
CYTES, AND THE REDUCTION OF THIS NUMBER IN
THE CELLS OF CANCER PATIENTS.

THE reduced number of chromosomes exhibited in
lymphocytes when they are forced to divide in ten
minutes by increasing the diffusion of the stain into
them with alkali reminded us forcibly that the cells
were undergoing mitosis under the stress of experi-
mental conditions. The somatic number of chromo-
somes of human cells is 32; and since lymphocytes will,
if their divisions are delayed, divide by the somatic
number, it appeared to us that the normal time occu-
pied by the division of the cells in the body is probably
much longer than the ten minutes allowed to them
on the microscope slide. It was appreciated that it
would be better if we could delay the diffusion of the
stain into the cells to such an extent as always to
produce somatic divisions. But unfortunately herein
we met with a difficulty which has not yet been over-
come. As we have pointed out, death has been delayed

249

250 THE DIVISION OF LEUCOCYTES

for twenty minutes while mitosis has been induced, but
this experiment has only been followed by success on
very few occasions. For general practical purposes it
must be remembered that whatever is done in the way
of attempting to induce divisions in cells when they are
resting on a jelly-film, this must be done in ten minutes.
If the diffusion of the chemical agent is delayed beyond
this time, except in very few instances, the cells will
refuse to divide at all, simply because they die.

The reason for this is a question of vitality, which
brings us back to the disadvantages of in-vitro experi-
mentation. All cells of the body lose vitality gradually
after they have been shed. White blood-corpuscles will
live in citrate solution for two or three days at the room
temperature, but they lose vitality all the time. As
already pointed out, there is no known medium in
which blood-cells will live and thrive, and in the best
medium at our disposal they merely exist for this short
period. When cells are resting on a jelly-film, how-
ever, they are not even in the best available medium;
but at present we are bound to employ the jelly
method, for we have not succeeded in inducing divisions
in any other way. The reason for this is two-fold:
firstly, because substances can be made to diffuse into
individual cells more quickly if the cells are pressed
into the jelly which contains them; and, secondly, we
think that lymphocytes prefer to be at rest when they
divide, for we cannot induce divisions \vith the cells
floating in a solution, although we have tried to do
so many times in solutions which have contained the
necessary constituents.

EXPERIMENTAL TEN MINUTES 251

At present there is nothing for it but to induce
divisions with the cells spread out on jelly under a
cover-glass; and it must be remembered throughout
that these conditions are most detrimental to the cells.
Pressed in this way into the jelly by means of a cover-
glass, which to living cells must be proportionately of
enormous weight, leucocytes and lymphocytes will not
live more than about three-quarters of an hour. Al-
though they wrill exist for this time, and although
amoeboid movements may be excited in them during
greater part of it, it is obvious that the cells are in
reality dying slowly all the time. Since the ease with
which one can induce divisions in lymphocytes varies
directly with the vitality of the cells, it is clear that
wrhatever is done to induce mitosis must be done
quickly, and by practical experiment it has been found
best to observe the general rule that, when one attempts
to induce cells to divide on the microscope slide one
must so arrange the jelly-film that the cells will be in
the act of mitosis within ten minutes. This is a serious
disadvantage appertaining to in-vitro experimentation,
which cannot so far be overcome, and it is important
to remember it throughout. The cells are labouring
under abnormal difficulties which modify one's deduc-
tions from the facts seen ; and since this important point
will frequently have to be considered, it is convenient
to standardise these detrimental conditions and desig-
nate them the "experimental ten minutes."

Two corollaries depend on the "experimental ten
minutes." Since the induction of a division in a cell
depends on the diffusion into it of a certain quantity of

252 THE DIVISION OF LEUCOCYTES

a chemical agent, and since the quantity of the agent
required must increase with the rapidity at which we
wish mitosis to occur, it is obvious that a greater con-
centration of the chemical agent will be required to
induce a division in the "experimental ten minutes"
than would be required to make a cell reproduce itself
if it were resting in its normal surroundings, where it
might take a much longer time in its division.

The second corollary is that if the jelly on which
the cells are resting contains a saturated solution of a
given substance which is diffusing into the cells to the
utmost in ten minutes, and if that substance does not
induce divisions in the "experimental ten minutes," it
does not prove that that substance will not within the
body, with the cells in their natural surroundings, cause
them to proliferate.

It was a matter of concern to us that the azur
dye did not make the polymorphonuclear leucocytes
(fig. 76) divide. So far only lymphocytes responded.
If the contention was correct that the dye contains a
specific agent which was possibly analogous to some
similar agent in the body which causes proliferation
of lymphocytes, it appeared reasonable to expect that
some similar agent, if not an identical one, would also
cause divisions in leucocytes; for the latter cells always
proliferate together with, and to a greater extent than,
lymphocytes during the process of healing. So far,
however, we had not seen anything resembling a
division in a polymorphonuclear leucocyte. It must
be admitted that we had no idea as to what a leuco-

SPECULATIONS REGARDING THEM

253

76.— A resting polymorphonuclear leucocyte. Its granules are stained
but not its nucleus. The cell was alive.

FIG. 77. — A basophile leucocyte in the act of cell-division. The granules
of the cell are in the centre. The lobes of the nucleus are at the poles of
the cell which is dividing ito three.

SPECULATIONS REGARDING THEM 255

cyte would look like when it divided, for no one had
ever seen a division in a leucocyte. These peculiar
cells are large, and easily examined. They differ from
all other cells in that they contain a polylobed
nucleus, and it was very difficult to imagine how
mitosis would occur in such a cell. Speculations have
been made from time to time to the effect that these
cells divide by pluripolar mitosis. Each lobe of the
nucleus is said to undergo a mitosis of its own, and
that the chromatin within the lobe forms up into
chromosomes. This would mean that a cell with five
lobes to its nucleus would divide into ten cells. Such
a speculation makes no allowance for the Altmann's
granules, which attain a large size in these cells, or
for the filaments which unite the several lobes of the
nucleus. Since the cytological process of mitosis in
lymphocytes was so different to what was expected,
we were prepared to see the speculation disproved; but
in spite of this it must be admitted that when at last
the divisions of leucocytes were seen the arrangement
of their cytological elements came rather as a reve-
lation.

Jelly-films were made which contained greater
strengths of the azur dye, extracted from polychrome
stain in the way which we have described. The
possibility of divisions being induced in leucocytes
was considered to be an event which would be seen
before long; but when it was first seen it, like the
first mitosis in lymphocytes, was not recognized or
appreciated. The increased quantity of azur dye was
added to the jelly in reality to see what the effect of

256 THE DIVISION OF LEUCOCYTES

excess of it on lymphocytes might be. On one occa-
sion a "basophile" leucocyte was found lying on the
jelly with its granules arranged in rows, and forming
a sort of radiating pattern. Moreover, the granules
were in the centre of the cell, which is an unusual
position for them, and there were clear spaces outside
them which evidently contained the lobes of the nuclei,
although the latter were not stained, as it is very diffi-
cult to stain the nuclei of basophile leucocytes in vitro.
The condition had not been seen before, but it was
passed over, for at the time lymphocytes were being
sought for. Some days afterwards another basophile
cell was seen in a similar condition, and then it was
more carefully observed. The lobes of the nucleus of
this cell could just be made out, and they were external
to the granules. The cell- wall itself was indented in
three places, so that the leucocyte looked like the pro-
peller of a steamship. The granules were deeply stained
and turning black (fig. 89) , which sometimes occurs in
vitro in the stained granules of basophile cells ; and they
were again arranged in indefinite lines or rows. It
was this arrangement of the granules which specially
arrested attention. The Altmann's granules of lympho-
cytes form up into rows to form the chromosomes, and
it looked as if something of some similar nature was
.happening on this occasion in a leucocyte.

This curious condition of the basophile leucocyte
seemed to have occurred in response to the excess
of azur dye. Still more of it was therefore added to
some coefficient jelly which also contained atropine,
polychrome dye, and extract, with the idea of de-

MITOSIS OF LEUCOCYTES 257

liberately producing this condition of the basophile
leucocytes. As a matter of fact, the jelly contained
0.6 cc. of the azur dye, and its index of diffusion was
now arranged for the coefficient of diffusion of the
basophile leucocytes, which is the same as that of the
ordinary neutrophile leucocyte.

After removal from the incubator at the end of
ten minutes, it was seen that the lobes of the nuclei
of the neutrophile polymorphonuclear leucocytes were
just staining a faint blue colour, and — there could be
no question about it — nearly every leucocyte in the
specimen was in the act of division. Neutrophile,
basophile (fig. 77), and those eosinophile (fig. 78)
leucocytes which were not ruptured were undergoing
the act of reproduction on the jelly-film. They were
dead owing to the staining of the lobes of their nuclei,
but the lines of demarcation between the individual
daughter cells could be distinctly seen. The cytological
procedure by which these cells divide is identical in
all varieties of leucocyte. As in lymphocytes, the
Altmann's granules were formed into rows, and pre-
sumably they are analogous to chromosomes; the rows
of granules become arranged into indefinite lines
radiating outwards from the dividing-point, which is
in the centre of the cell. Running down through the
centre of the mass of granules, the filament which
unites the lobes of the nuclei evidently forms a basis,
analogous to the spindle of other cells, to which the
chromosomes are attached; and at the poles of this
filament, or spindle, the so-called lobes of the nuclei
appeared. It was then immediately appreciated that

258 THE DIVISION OF LEUCOCYTES

these bodies are in reality the centrosomes of the
cells.

If a leucocyte has two lobes to its nucleus it will
divide into two cells; if it has three lobes it will divide
into three cells, and so on. It will thus be seen that
when these cells proliferate each daughter cell will have
one centrosome until that centrosome itself divides and
assumes the appearance of being polylobed. Further,
a tissue made up of such daughter cells would be de-
scribed as consisting of "mononuclear cells." The
chromatin-staining lobes within the leucocytes are there-
fore not nuclei but centrosomes, and the so-called
Altmann's granules, which have been variously sur-
mised to be collections of food or secretion, are the
elements of the chromosomes themselves.1 As in
lymphocytes so in leucocytes, the chromosomes are
outside the nucleus. Divisions have been induced in
hundreds of leucocytes, and the procedure is always
the same in all of them (figs. 79-86).

Now, the increased quantity of the azur dye con-
tained in the jelly did not improve the mitosis induced
in the lymphocytes; in fact, it seemed too strong for

1 Professor Sherrington has a specimen of an eosinophile leucocyte of a
cat in which the individual granules are elongated and almost rod-shaped.
We have also seen elongated granules in these cells in human blood.

MITOSIS OF LEUCOCYTES

259

FIG. 78. — An eosmopliiie leucocyte m tue earliest stage ot division. The
granules are arranged in lines radiating outwards from the centre of the
cell. The lobes of the nucleus were at the poles.

FIG. 79. — Early stage of division of a neutrophile leucocyte.

261

FIG. 80. — A dividing leucocyte.

FIG. 81. — A dividing leucocyte.

MITOSIS OF LEUCOCYTES

263

FIG. 8lA. — Division of a leucocyte. The linear arrangement of the granules
could be well seen.

'lG. 82. — A dividing leucocyte.

MITOSIS OF LEUCOCYTES 265

FIG. 83. — A dividing leucocyte.

FIG. 84. — A dividing leucocyte.

MITOSIS OF LEUCOCYTES

267

FIG. 85. — A dividing leucocyte.

FIG. 86. — A dividing leucocyte.

MITOSIS OF LEUCOCYTES 269

them, as only early stages of mitosis were seen in
them. Hence it became apparent that the auxetic
constituent of the aniline dye which induces divisions
in lymphocytes also does the same thing with leucocytes ;
but it evidently requires more of it, ceteris paribus,
to induce a division in a leucocyte than in a lymphocyte.
The coefficient of diffusion of the lymphocyte is higher
than that of the leucocyte if the staining of the nucleus
is the moment by which it is determined; but so far
as inducing divisions is concerned the coefficient of
lymphocytes seems to be lower, for they require less
of the chemical agent than do leucocytes.

The divisions of reproduction had now been induced
in both leucocytes and lymphocytes by an artificial
chemical agent. These cells are the ones which pro-
liferate when a tissue is damaged, and it is by their
multiplication that the healing of an injury takes place,
and it must be borne in mind that cancer, with its in-
creased malignant proliferation, is intimately associated
with chronic healing. Judging by the divisions induced
in these white blood-corpuscles it appeared that their
reproduction takes place in a cycle which depends on
some chemical substance absorbed by them. The cycle
consists apparently of the division of the centrosomes,
division of the chromosomes, and the division of the
cell. Whether there is a "resting stage" in the strict
sense of the term, we are not in a position to state,
for we do not know how long a time is occupied in
the division of the centrosome. If a cell is absorbing
the agent which causes it to divide, presumably the
cycle of mitosis is going on in direct proportion to the

270 THE DIVISION OF LEUCOCYTES

amount of the agent absorbed. The division of
the centrosome seems to be part of this cycle, but
how long this part takes we do not know. Hence it
appears possible that what is commonly known as
the resting stage is in reality the time occupied by
the division of the centrosome.

The method of division of leucocytes and lympho-
cytes is so constant that we thought it was reasonable
to expect that the proliferation of healing would be
ultimately proved to take place by a similar process,
and that if so there must be produced in an injured
tissue some chemical substance very similar in its effects
to that contained in azur dye. Up to this time, however,
we had not succeeded in inducing divisions at all with
any substance which we could call a "natural" sub-
stance. There is nothing in the body that we know of
at all like the aniline dye. It was true that an extract
of dead haemal gland augmented the action of the ani-
line dye; but it would not induce divisions by itself.
Extracts of tissues other than haemal gland were tried,
made in the same strength — namely, 100 per cent — and
it was found that suprarenal glands of sheep augmented
the action of the stain in inducing divisions even better
than haemal gland, and several extracts, such as those of
muscle and liver, did the same, but to a lesser degree.
In spite of the augmenting action of all these extracts,
however, none of them alone in the strength tried would
induce divisions either in lymphocytes or leucocytes in
the experimental ten minutes.

This inability to cause cell-division by entirely
"natural" substances, such as the extracts named, was

THE EXTRACTS CONCENTRATED 271

believed, after mature consideration, to be due to the
fact that we had not tried extracts in sufficient strength
for cells to respond to them under the detrimental
circumstances of the "experimental ten minutes." It
has already been pointed out as a corollary to these
circumstances that if a cell refuses to respond to a given
substance by not dividing in the experimental ten
minutes, it does not prove that that substance does
not actually contain an active principle for inducing
cell-division. We therefore considered the advisability
of concentrating the extracts, and then trying them
again by themselves.

In the first instance this concentration process took
some little time. At this stage of our researches we
were unaware that the active augmenting principle
contained in the extracts was "thermostable" and
would resist boiling, and in consequence, at the outset,
we evaporated the extracts at the room tempera-
ture, which was a most tedious process. Moreover,
during this slow concentration of the extracts it was
necessary to test them from time to time to see if their
augmenting action was impaired at all with keeping.

In the meantime other points were considered. It is
well known that cancer-cells frequently are seen to be
dividing with a reduced number of chromosomes. As
we have already stated, we believe that a reduction in
the number of chromosomes is due to increased pro-
lificity in cells; and this being the case, it seemed
probable that there might be some increase in or
augmentation of the cause of proliferation of the cells
of cancerous growths. Further, if this is the case, the

272 THE DIVISION OF LEUCOCYTES

augmenting substance might appear in the peripheral
circulation. It has already been pointed out in
Chapter VIII. that cancer plasma excites amoeboid
movements in leucocytes, and that alkaloids also excite
these movements. Since the alkaloid atropine augments
the action of azur stain in inducing divisions, it was
thought possible that the exciter of amoeboid move-
ments found in cancer plasma might be in the nature
of an alkaloid possibly derived from the neighbourhood
of the growth. Now, atropine and stain together cause
white blood-cells to extrude granules of chromatin, a
phenomenon which we erroneously called "flagellation"
(see Chapter X.), and this extrusion had also been
observed in cells which have been subjected to cancerous
plasma.1 The suggestion followed that the cells might
be extruding their granules deliberately in response, not
only the artificial combination of stain and alkaloid,
but also to some possibly similar combination derived
from the malignant growth. Moreover, since in both
the artificial and the natural circumstances, cells appear
to divide with a reduced number of chromosomes, and
since the granules form the chromosomes, it was sur-
mised that the extrusion might be part of the process
of reduction. It must be remembered that our ex-
perimentation left us convinced that the divisions of
lymphocytes and leucocytes occur just as the stain is
combining with the chromosome-granules; and as the
extrusion of the granules — which has been seen by
others as well as ourselves — seems to be a deliberate

1 "The Flagellation of Lymphocytes in the Presence of Excitants both
Artificial and Cancerous," by H. C. Ross and C. J. Macalister, British
Medical Journal, January 16, 1909.

COUNTING GRANULES SUGGESTED 273

action on the part of the cells, we went so far as to
theorise that the cells might be discarding their granules
in order to prevent some of the combination of the
"auxetic" and their granules, and so delay their pro-
liferation to some extent.

This theory led to the suggestion that we should
try to count the number of granules contained in the
blood-cells of cancer patients, with a view to see if they
were reduced in number in that disease. The blood of
cancer patients seems to contain a body which excites
amceboid movements and the extrusion of granules,
and, therefore, the blood-cells themselves as well as the
cells composing the growth might also have a reduced
number of the granules wThich form their chromosomes.

We must admit that these suggestions were based
on slender grounds of evidence, and it was appreciated
that to count the number of granules of the leucocytes
of cancer patients would require considerable work,
especially as many control experiments would have to
be made, for we did not even know the normal number
of granules in healthy persons' cells. Still, it was very
necessary to try to find out whether the clue on
which we were engaged wTas in any wTay correct, and
it was realised that in order to make the counts it
would be necessary to examine a large number of
samples of blood-cells from many patients and from
normal and other persons — a procedure which had not
yet been done by this in-vilro method. Hence no
matter how far-fetched it appeared at first sight for us
to count the granules of blood-cells in cancer patients,
I thought that an endeavour to do so would be justified,
and I devised the following technique for doing so.

274 THE DIVISION OF LEUCOCYTES

It was seen from the outset that it would be quite
impossible to count the number of granules contained
in lymphocytes, and the same could be said of those
of the common neutrophile leucocyte (fig. 76). But it
is possible to do so in the so-called eosinophile cells
(fig. 87). These cells have large granules, which stain
a deep scarlet with the polychrome dye, and therefore
these cells were chosen for this series of experiments,
especially as they are fairly common (2 to 4 per cent) .

Three difficulties presented themselves in arranging
a technique for counting the number of eosinophile
granules :

1. To the novice the basophile cell is sometimes
very difficult to distinguish by in-vitro staining from
the eosinophile cell, and mistakes seriously modify the
results. If specimens of each class of cell are seen
lying side by side (fig. 88) there is no difficulty in
distinguishing them, the eosinophile cell being much
the larger, although there is very little difference be-
tween the size of their granules. But in spite of the
fact that the cells rarely are thus found lying side by
side, with a little experience they can be readily dis-
tinguished; the granules of the basophile cell are more
discrete, and the lobes of its nucleus will practically
never stain by this in-vitro method.

2. A living leucocyte is spherical in shape, and it
usually appears with its granules heaped one on top
of another, rendering it impossible to count them
accurately.

3. If one attempts to count through the microscope
a group of granules not arranged in any definite order,

COUNTING GRANULES 275

FIG. 87. — An eosmopmie leucocyte with its granules stained.

COUNTING GRANULES 277

one is apt to count the same granule more than once,
and it is easy to lose one's place — in which case it
becomes necessary to begin all over again. On looking
at the cells through the microscope, the granules appear
as though they might with care be counted, and it is
most inviting to attempt to do this and to rely on it;
but on testing this rough-and-ready method we have
found that it usually involves an error of nearly 50 per
cent. No estimate whatever can be made of the
number of granules contained in a cell by merely
looking at it through the microscope, no matter what
magnification is used.

Obviously the granules must be stained, and then
it is necessary: (1) to distinguish readily between an
eosinophile and a basophile leucocyte; (2) to kill the
cells, and then to burst them so as to cause their
stained granules to rest discretely side by side in one
plane and not on top of one another; (3) to magnify
the image of the ruptured cell in such a way that one
can "tick off" each granule with a pencil on paper
as it is counted, so as to avoid counting the same
granule twice over.

By the following procedure the staining, killing,
differentiation, and bursting can be readily accom-
plished. In order to magnify the image of the ruptured
cell so as to count its granules and to "tick them off,"
it is necessary to obtain a photomicrograph negative of
it, and then to project the photographed image on to
a paper screen with an optical lantern, when the image
of each granule can be marked off on the paper with
a pencil.

278 THE DIVISION OF LEUCOCYTES

It is necessary to employ the photomicrographic
apparatus which I have already described, and the
photographs must be taken with as little delay as
possible after the cells have been ruptured. Unfixed
cells may rapidly become achromatic after death, and,
in the case of a ruptured cell, the loss of stain may
occur with great rapidity.

The blood of the person to be examined is drawn
into a capillary tube and there mixed with an equal
volume of citrate solution. At the room tempera-
ture this solution will keep the cells alive for some
days; but when it is intended to count the granules of
the eosinophile leucocytes, it is better to examine the
blood as fresh as possible.

A jelly is prepared thus: To a tube containing
5 cc. of coefficient jelly add 4 units of TJnna's poly-
chrome stain, 7 units of the 5-per-cent alkali solution,
and, instead of making the contents of the tube up
to a total of 10 cc. with water, 3.9 cc. of a molten
2-per-cent solution of agar in water is used. The
last solution contains agar in order to make the jelly
exceptionally firm, so that the ultimate bursting of the
cells can be facilitated. The jelly is melted and boiled
and a drop of it run on to a slide, where it is allowed
to set. A drop of the citrated blood is then placed
on a cover-glass, which is inverted and allowed to fall
flat on the film in the usual way. The slide is then
placed in the 37° C. incubator for three minutes exactly.
When examined microscopically it should be seen that
the nuclei of the eosinophile leucocytes are just staining
scarlet, showing that death is occurring; the granules

COUNTING GRANULES

279

FIG. 88. — A Held containing a neutrophile, an eosinophile, and a baso-
phile leucocyte. The upper cell is the neutrophile and the lower one the
basophile cell. All the cells are ruptured, but their granules are stained.

FIG. 89. — A basophile leucocyte wnose stained granules have been turned

black by heat.

COUNTING GRANULES 281

of the cells should be deeply stained. If the nuclei
are not yet stained, a little more alkali must be added
to the jelly and a fresh specimen made. If the cells
are achromatic or disorganised, or if the nuclei of the
neutrophile cells are deeply stained, the jelly is too
alkaline, and a little acid solution must be added to it.
But if the coefficient jelly and other solutions are
correct, the nuclei of the eosinophile cells will just be
staining.

Using a |-inch or equivalent objective the specimen
is searched until a suitable eosinophile cell is found.
If a cell is distorted or hemmed in by red cells, it is
necessary to pass it over and find another.

If there is any doubt as to whether a cell is an
eosinophile or basophile one, the slide is removed from
the mechanical stage in such a way that on returning it
to the microscope the same field can be focused again.
The slide is then again incubated for three minutes, but
at 47° C. On examination of the cell, if it is an eosino-
phile leucocyte, its granules will still appear scarlet ; but
if it is a basophile cell, its stained chromosome granules
will have turned black1 (fig. 89). With a little experi-
ence of the method of staining, however, the difference
between the classes of cell can be detected without this
procedure of incubation at 47° C., which is apt to cause
premature rupture and achromasia.

The next step is to burst the cell. The photo-
micrographic apparatus being ready on its slide above
the observer's head, the immersion objective is " turned

1 These granules may turn black at 37° C. We have no explanation to
offer of this phenomenon.

282 THE DIVISION OF LEUCOCYTES

on" and focused, and the cell is brought into the
centre of the field. Watching it through the eye-
piece, keeping one hand on the fine adjustment,
the cover-glass, which of course is resting on the
jelly-film, is gently struck (tapped) with a glass rod
held in the other hand. At each tap the cells are
seen to be jerked out of the field, but, provided the
taps are not too forcible, the eosinophile cell can
easily be followed by using the mechanical stage.
It is usually necessary to strike the cover-glass two
or three times, and generally at the third blow the
eosinophile is seen to totter and then burst, scattering
its stained granules about on the surface of the jelly
in the field of the microscope. This is a trick, of
course, which was devised by one of us (J. W. C.),
and with a little practice rupture can nearly always
be assured.

When a cell ruptures on this jelly — which contains
salts — its nucleus loses its stain instantly, but at the
room temperature the granules do not usually become
achromatic for some little time. On the other hand,
in some instances they may become unstained in a
few moments, and for this reason, in order to secure
the photographic negative, speed is now required.
The ruptured cell is placed in the centre of the field
'with the mechanical stage; the working eye-piece is
removed from the microscope, the camera is allowed
to slide down the wooden slide, and its projecting
eye-piece, which is already attached to it by means of
a flexible velvet collar, is inserted into the draw-tube
of the microscope. By the simple movement of swing-

COUNTING GRANULES 283

ing the microscope mirror on its gimbals out of the
focal axis, the working 32-c.p. gas-light is changed
to the water-cooled ray of light from the 1-amp.
Nernst lamp. The image of the ruptured cell will
then be seen on the ground-glass screen at the back
of the camera, where it can be rapidly focused.

The special precautions regarding the focusing
with this method have already been described, but
it should be remembered that in order to be able to
count the number of the granules in the ruptured
cell it is most important to obtain as perfect a nega-
tive (figs. 90, 91) as possible.

If the photography has been accomplished quickly,
the camera may be pushed up out of the way, the
microscope mirror replaced, and the specimen may be
searched for more eosinophile leucocytes.

To count the number of granules contained in a
ruptured cell, the negative must — after it has been
developed and dried in the usual way — be placed in an
optical lantern, and the image of the ruptured leucocyte
projected on to a screen which has a sheet of white
paper pinned in front of it. One stands close in front
of the screen and counts the granules, each of which
will now appear about the size of a shilling-piece, and
the image of each granule can be "ticked off" with
a pencil on the paper (figs. 92, 93) . It is thus impossible
to count any granule twice over, and an accurate
enumeration can be made.

Such is the technique. By it there have been
counted 38,759 granules from 235 cells from 96 persons,

284 THE DIVISION OF LEUCOCYTES

22 of whom were suffering from undoubted1 cancer,2
and 47 of whom apparently were not. Of the latter,
which for convenience will be called the "control" cases,
some were "healthy" and others were suffering from
various diseases (hospital patients) . We did not count
the granules in each of the 235 negatives as the latter
were obtained, but the plates were developed and then
put away until a hundred or more had collected. The
name of the person from whom the blood had been
taken wras entered into a book with the age, sex,
disease (if any), and other details. The negatives were
numbered consecutively, and the numbers corresponded
with similar ones in the book against the names of the
persons from whom the cells had been derived. The
samples of blood were taken from persons, cancerous or
otherwise, as they came into hospital, and therefore,
without referring to the book, a number on a negative
gave no indication from whom the cell it depicted was
derived. With three exceptions, the samples of blood
were collected and photographed by one of us, who
kept the book in his laboratory. The counting wras
done by another, who had no idea to whom the
numbers on the negatives referred.

The only possible source of error is in the counting.
Some of the negatives were not quite perfect, and
' some of the granules appeared blurred ; hence there may
be a small error in some of the numbers, but it cannot
be very important judging by the uniformity of the
averages.

1 Determined either by such clinical manifestations as recurrence or
metastasis, or by pathological examinations. 2 Carcinoma.

THE NUMBER OF GRANULES

285

FIG. 90. — One of the negatives of a ruptured eosinophile leucocyte (negative

No. 52).

FIG. 91. — One of the negatives of a ruptured eosinophile leucocyte (negative

No. 54).

THE NUMBER OF GRANULES 287

FIG. 92. — Counting the granules. The image of the ruptured cell
depicted on 'negative No. 52 is projected on to a sheet of white paper pinned
on to a screen.

FIG. 93. — Counting the granules of negative No. 54.

THE NUMBER OF GRANULES 289

At first only two cells were taken from each person,
but since it was found that there was frequently a wide
variation in the number of granules contained in in-
dividual cells, this number was afterwards increased to
five. Averages were then struck, and the tables given
in Appendix I. give the number of leucocytes ex-
amined and the persons from whom they were derived,
together with the number of granules contained in the
largest and smallest cells from each person. From
these averages it will be seen that the sex makes
practically no difference in the average number of
granules contained in the cells; but more experiments
will be needed before the same can be said about age.
The averages can, in the first place, be divided into
two groups, male and female. Each of these groups
can be subdivided into two, viz. control persons
(healthy and diseases other than cancer), and cancer
persons. Neglecting fractions, the average number of
granules in the cells appear thus :

Number of Average number of granules in cells of

persons males females

47 controls ...... 168 ... 168

22 cancer 159 ... 161

Thus, between normal1 males and females there is no
difference, and between carcinoma males and females
there is very little difference; but the number of
granules in cancer-cells is well below the normal in
the averages of both males and females, and it would
appear from this that the number of granules con-
tained in the cells of cancer patients is actually reduced.

1 Control cases (normal, and diseases other than cancer).

290 THE DIVISION OF LEUCOCYTES

The table and its summary supply further details.
As was expected, the reduction is not very large, but
the striking point is that, in addition to the total
cancer averages being below the normal, a subdivision
into such groups as male and female demonstrates
that the reduction in cancer is again present in both
groups.

Every individual case of cancer in the category
does not, by any means, have a reduction in the
average number of granules contained in the cells, and
it will be seen that many of the individual controls
showed a reduction; but when one comes to deal
with comparatively large numbers, the reduction in
carcinoma is demonstrated. It must be remembered
that in everybody there is a great variation in the
actual number of granules contained in individual cells;
and when sampling say five cells from a person, one
may by chance hit upon five larger or five smaller
cells. Obviously, therefore, it is only by the observa-
tion of many cells from large numbers of persons that
one can reduce to a minimum the "error of random
sampling." We think, however, that the enumeration
of the granules contained in 235 cells, from 22 cancer
patients and 47 controls, diminishes this error to such
an extent that the results are fairly trustworthy. At
the same time, it must be remembered that in experi-
mentation of this nature the error of random sampling
can never be altogether eliminated, and therefore the
reliability of the averages depends entirely on the
extent of this error among the cells which have been
photographed.

THE NUMBER OF GRANULES 291

Among the control cases three cases of sarcoma
are included. In all of them there was no apparent
reduction, and the same can be said of another case
tested recently. But the number of cells is too small
to form any conclusion from, and more cases will be
required.

CHAPTER XIII

THE AUXETIC ACTION OF CANCER SERUM THE IN-
DUCED DIVISIONS OF GRANULAR RED CELLS—
THE AUXETIC ACTION OF "THE REMAINS OF
DEAD TISSUES," AND ITS AUGMENTATION BY
ATROPINE AND THE PRODUCTS OF PUTREFACTION
—THE ISOLATION OF THE AUXETICS KREATIN

AND XANTHIN DISCOVERY OF THE CAUSE OF

THE CELL-PROLIFERATION OF HEALING

COUNTING the granules of eosinophile leucocytes from
cancer patients, therefore, seemed to us to show that
the clue on which we were working was to some extent
correct. Judging from the comparison between the
number of granules contained in the cells of cancer
patients and those of other people, there appears to
be a reduction in cancer, and this reduction presumably
is due to the presence in the blood of some agent which
^causes more proliferation than normal. It has been
pointed out that increased prolificity owing to excessive
absorption of a chemical agent makes cells divide by a
reduced number of chromosomes as seen in carcinoma
cells ; and now apparently other cells, such as the eosino-
phile leucocytes, have in that disease a slightly reduced

292

AUXETIC IN CANCER PLASMA 293

number of chromosome granules. But this digression
from the main researches also taught us that other
facts were to be learnt from the comparison of samples
of peripheral blood from twenty- two cancer patients
and forty-seven " others." Never before had systematic
examination of blood from such groups of persons been
made by the in-vitro staining of their cells, and it was
soon noticed that in the samples of cancer blood the
actual number of eosinophile leucocytes was reduced;
in fact, four cases could hot be included in our category,
because, even after repeated examination of many
samples of their blood, no eosinophile leucocytes could
be found; and in all the other cases, with the exception
of three, there was an undoubted reduction in the
number of eosinophile cells. In the three exceptions
there appeared to be an eosinophilia.

In some cases of carcinoma, also, there was a large-
lymphocytosis, especially in the advanced cases. But
this is by no means an absolute rule, and, moreover, a
large-lymphocytosis was fairly common among the
control specimens.

But a still more important point was observed.
We have already shown how a mixture of azur dye
and atropine causes excitation of amoeboid movements
in leucocytes and lymphocytes, then the discard of
granules (flagellation), and lastly augmented cell-
division; also that an agent has been detected in
the plasma of carcinoma patients which induces the
first two — i.e. excitation of amoeboid movements
and the discard of granules. We have just shown
that the granules in certain leucocytes in cancer

294 THE CAUSE OF HEALING

patients are reduced in number. The inference is that
this reduction is made in response to the same agent
which causes the excitation and discard of granules.
The important point is that while engaged in these
blood examinations the fact became apparent that this
agent in cancer plasma (presumably it is the same
agent) will help to induce cell-division.

The large lymphocyte requires a considerable
quantity of stain, extract, or atropine before it will be
induced to divide in the "experimental ten minutes."
In the technique, described in the last chapter, for
counting the granules of eosinophile leucocytes the
jelly employed contains only 4 units of polychrome
dye, the efficiency of which for inducing divisions is
infinitesimal (the jelly containing no extract of dead
tissues or atropine). Yet in the examination of the
blood of three of the cases of carcinoma some of the
large lymphocytes showed well-marked stages of early
mitosis, whereas this result could not be obtained in
any of the controls. It is clear, therefore, that the
cells in these three cases were inclined to divide before
they were ever placed on the jelly, and the trifling
assistance which they received from the 4 units of the
polychrome dye caused them to show well-marked
mitotic figures (figs. 62, 64) , whereas the large lympho-
cytes in all the control specimens, made under exactly
the same conditions, remained at rest.

Moreover, in two other cancer patients (both cancer
of the stomach), owing to anaemia, many granular red
cells were seen in their blood. On its being examined
on jelly which contained azur dye, extract, and atropine,

CONCENTRATION OF EXTRACTS 295

FIG. 94. — A dividing red cell from a cancer patient.

;)">.— A dividing red cell from a cancer patient. The granules seem
to be arranged in an indefinite figure.

CONCENTRATION OF EXTRACTS 297

amitotic1 divisions (figs. 94, 95) were induced in these
granular red cells. The granular red cells of normal
and other persons have never hitherto been seen to
make any attempt to divide on auxetic jelly or any
jelly, and hence it appears that these cells from these
cancer patients are also more prone to divide than
those of other people.

Since it has been shown that the reproduction,
certainly of lymphocytes and leucocytes, and possibly
of other cells, depends (on the microscope slide) on
the quantity of an auxetic absorbed by them, it is
reasonable to suggest that the plasma of these cancer
patients contained some such agent which caused this
inclination to divide on the part of the large lymphocytes
and red cells. Presumably this is the same agent
which had been previously found to cause excitation
of amceboid movements, and the discard of granules
for the combination of stain and atropine will also
do this as well as cause augmented divisions.

It is interesting to note that it is only the red
cells which have granules which can be induced to
divide, for it bears out the theory that the auxetic
contains a specific agent which induces cell-division
by acting on cell-granules.

We may now return to the study of the extracts.
It may be remembered that we had only succeeded
in inducing divisions in lymphocytes and leucocytes
with the artificial azur dye. Extracts of several

1 We are uncertain whether some of the granular red cells were not
dividing mitotically (fig. 95), as their granules appeared to be arranged in an
indefinite figure.

298 THE CAUSE OF HEALING

dead tissues, especially that of suprarenal gland, in
the strength of 100 per cent, would augment the action
of the azur dye, but they would not in themselves
induce divisions or even the early stages of mitosis
in the experimental ten minutes. We had therefore
made arrangements to concentrate these extracts so
as to see if they would, if used in greater strength,
induce divisions by themselves. At first it was thought
better not to boil down the extracts for fear that the
boiling might spoil the substance which augmented
the action of the dye. The extracts were therefore
placed in test-tubes, which were lightly plugged and
put aside in the laboratory. As already mentioned,
it was necessary to test these extracts from time to
time to see whether they might become more effective
as concentration occurred. When they were originally
made they were sterile, because it may be remembered
that they had been kept at 60° C. for twelve hours after
filtration. Repeated examination of some of the tubes,
however, caused them to become infected, and in
consequence putrefaction set in in those tubes. After
they had all been kept for three weeks it was noticed
that the augmenting action of the contents of one
of the infected tubes of suprarenal extract seemed
to be increased. One cc., or even a few drops, of
this extract, if added to the azur dye and made up
in a jelly, caused advanced mitosis in lymphocytes,
whereas with the other sterile tubes it seemed to require
about the usual quantity of extract to augment the
action of the dye. It was particularly noticed that
this tube which contained so efficient an extract had

DECOMPOSITION OF EXTRACTS 299

been examined on several occasions, and owing to the
infection of its contents the latter was in a foul-smelling
condition. The increased augmentation when this
decomposed extract was used was so remarkable that
we decided to try its action by itself without any azur
or other stain.

The jelly was made up thus: To 5 cc. of coefficient
jelly 3 cc. of the putrid extract, and 0.8 cc. of
5-per-cent solution of sodium bicarbonate (8 units of
alkali) were added. The alkali was present in order
to cause the contents of the jelly to diffuse into the
cells. The jelly wTas made up to a total of 10 cc. with
1 . 2 cc. of water. In order to prevent coagulation of the
extract a film was prepared from the jelly in the
following way: The coefficient jelly was melted and
boiled, and it was only as it cooled that the extract
was added, the film being made immediately before
the jelly had set in the test-tube. Fresh blood from
the finger was spread on the jelly in the usual manner
under a cover-glass. After incubation for ten minutes,
an examination showed that some of the lymphocytes
appeared to be in an early stage of mitosis. Now,
we could not be very certain about this point, because
no stain was present and consequently the chromo-
somes were unstained and almost invisible. If mitotic
divisions are sometimes difficult to see in stained
specimens, they are much more difficult to distinguish
when no stain is employed. Still, the cells looked
rather as if they were attempting to divide (fig. 96).

A fresh jelly was made, but it contained 1 cc. of
alkali solution instead of 0.8 cc. ; and now there was

300 THE CAUSE OF HEALING

no doubt about it — this extract did actually induce
mitotic figures in lymphocytes in the experimental
ten minutes (figs. 97, 98). No azur stain, atropine, or
other "augmenter" was added; the decomposed
suprarenal extract induced mitosis by itself.

Of course we thought at first that this result
was due to the concentration of the extract; but
this thought was soon dispelled by trying some of
the other tubes which had been kept alongside of
the ones wrhich had so often been examined, the
contents of which had decomposed. The sterile
extract contained in these tubes would not induce
divisions by themselves. Moreover, at the temperature
of the laboratory, since they wrere kept in plugged
tubes, the extracts did not evaporate very fast, and
it was appreciated that they could not be so very
concentrated. Some of the effective putrid extract,
therefore, had water added to it, so that it was again
made up to its original strength of 100 per cent. It
wras then made up in a jelly as before, and to our
astonishment again it induced divisions in lympho-
cytes; and what is more important, it induced the
asymmetrical one-sided mitosis in many instances
(fig. 99).

A series of control experiments \vas then made.
.Jellies which contained only the salts sodium citrate,
sodium chloride, and the 1 cc. of alkali were first
tried, and no divisions could be seen. Then yet
another series of experiments with fresh extract of
suprarenal gland wras made, once more without re-
sult, and so it wras ultimately proved that it was

AUXETIC IN EXTRACTS

301

FIG. 96. — Very early stage ot mitosis in a lympnocyte induced by decom-
posed extract of suprarenal gland. No stain.

FIG. 97.— Mitosis of a lympnocyte induced by decomposed suprarenal
extract. No stain.

AUXETIC IN EXTRACTS

303

FIG. 98. — Mitosis induced in a lymphocyte by decomposed extract. No

stain.

FIG. 99. — Asymmetrical division induced by decomposed extract. No
stain or atropine is present.

AUXETIC IN EXTRACTS 305

unquestionably due to the putrefaction that this one
tube of extract induced divisions in the experimental
ten minutes.

Now, this fact required very careful consideration.
A 100-per-cent solution of extract would not in itself
induce divisions in lymphocytes unless it was putrid.
When it is fresh this extract is not effective in the
experimental ten minutes. It appeared probable that
the extract does in itself contain some substance
which causes cell-division, but in the strength of the
extract of 100 per cent this substance is not present
in sufficient quantity for it to induce divisions in the
experimental ten minutes unless the whole extract is
putrid. The first thing to do was to concentrate
the extract and see if this theory was right. It was
appreciated that the concentration process at the
room temperature was a most unsatisfactory pro-
cedure, for if the extracts were tightly plugged
they did not evaporate down, but if they were left
open they became putrid. One of the jellies which
induced divisions by virtue of the putrid extract was
therefore boiled and tried again. Still it induced
divisions in lymphocytes. It was submitted to pro-
longed boiling, and yet it was effective. So it was
proved that the substance which it contained which
caused cell-division wras thermostable. We can
boil these extracts with impunity, and their auxetic
action is not impaired. Hence we made some fresh
extract of suprarenal gland and evaporated it down
to dryness by boiling. It is, when dry, a hygroscopic
brown mass which is readily soluble in water. One

306 THE CAUSE OF HEALING

hundred grammes of sheep's suprarenal glands yields
about 4 grammes of dry extract.

A series of jellies were prepared which contained
0 . 8 cc. of alkali solution (8 units) , variable quantities of
solutions of the extract, and they were always made up
to the total of 10 cc. with water. At first a 5-per-cent
solution of the extract was made ; and it was found that
if the jelly contained 1 cc. of this extract, very early
divisions can be induced in lymphocytes. With 2 cc.
later stages of mitosis will appear (figs. 100, 101) ; and
if instead of the 5-per-cent solution a 10-per-cent one
is made, even more marked divisions can be induced
by this fresh extract alone in the experimental ten
minutes. The best jelly to make in order to cause
suprarenal extract to induce divisions in lymphocytes
in the ten minutes is: 5 cc. of coefficient jelly, 1 cc. of
alkali solution, 2 cc. of a 10-per-cent solution of dried
suprarenal extract, and 2 cc. of water. By means of
this jelly advanced mitotic figures can be induced in
lymphocytes.

So it was proved, therefore, that this extract of dead
suprarenal gland contains a substance which will cause
the divisions of lymphocytes. A fresh jelly was then
prepared the same as the last one, except that it had
added to it four more units of alkali solution. Now, as
we anticipated, the polynuclear leucocytes also divided
on the microscope slide (fig. 102).

But the question was then asked, How was it that
the original extract, although it was not strong enough
to induce divisions by itself in ten minutes, did become
effective when it was decomposed by putrefaction ? It

AUGMENTED BY PUTRIFICATION

307

FIG. 100. — Mitosis induced by fresh extract of suprarenal gland. No
stain or augmentor present.

FIG. 101. — Mitosis induced by fresh suprarenal extract. No stain is present.

AUGMENTED BY PUTREFACTION 309

was evident that the first extracts which we tried were
not strong enough to induce divisions in the ex-
perimental ten minutes. If they became putrid,
however, they apparently were.1 The putrid solution
was again tried, and again the asymmetrical divisions
were seen. Now, these asymmetrical divisions are
frequently induced by azur dye when it is augmented
by atropine, and therefore we thought that it might
be possible that the putrefaction of the extract might
produce in it an augmenting substance which acted
like the atropine.

Fresh suprarenal extract was then made, and after
it had been dried it was redissolved in water. It was
made up in a 10-per-cent solution, and various quanti-
ties of it were added to jellies which contained 1 cc.
of alkali solution (10 units), and it also had added to it
0.7 of a 1-per-cent solution of atropine sulphate. It
was now found that the atropine augmented the action
of the suprarenal extract five-fold, in the same way as it
augmented the action of the azur dye — that is to say,
with suprarenal extract by itself, and no atropine, the
10 cc. of jelly, if it contains alkali to the extent of
10 units, must contain at least 0.05 gramme of dried
suprarenal extract before the earliest sign of cell-
division can be induced in ten minutes. To obtain well-
marked divisions the jelly should contain 0.2 gramme
of the extract.

If atropine is added, however, in the strength of

1 Some 100-per-cent suprarenal extract has been purposely allowed to
become infected, when it induced divisions in lymphocytes (figs. 103, 104).
Control tubes of extract not so infected had not this action.

310 THE CAUSE OF HEALING

0.007 gramme of atropine sulphate to the 10 cc. of
jelly which has 10 units of alkali, divisions in lympho-
cytes can be induced if the jelly also contains no more
than 0.01 gramme of dried suprarenal extract. Once
more we tried to induce divisions with the alkaloid by
itself, but failed; and yet it augmented the action of
the extract five-fold. In addition to this augmentation
it induced asymmetrical mitoses (fig. 105).

To recapitulate: Extract of suprarenal gland of
certain strength will induce by itself mitotic divisions
in lymphocytes; and if more of it is made to diffuse
into cells, it will also cause leucocytes to divide. If a
lower concentration is tried, however, it will not induce
divisions in the experimental ten minutes unless (1) it
has become putrid, (2) its action is augmented by
atropine. In both the latter circumstances asym-
metrical mitosis may be seen.

Other extracts of dead tissues were then tried; but
they would not, by themselves, induce divisions in the
experimental ten minutes. Realising that this might
be due to the detrimental experimental conditions
(corollary 2), we tried them again with atropine to
augment their action. Now, as surmised, all the
extracts of dead tissues which we tried induced
divisions in lymphocytes on the microscope slide. To-
induce divisions in polymorphonuclear leucocytes with
them is much more difficult, as atropine does not appear
to augment their action so much with these cells.

The following table gives the strengths of the
various extracts which, with 1 cc. of alkali (10 units)
and 0.007 gramme of atropine sulphate, will induce

AUXETICS ISOLATED

311

FIG. 102. — A dividing polymorphonuclear leucocyte induced by suprarenal
extract alone. No stain.

FIG. 103.— Mitosis induced in a lymphocyte by suprarenal extract which
had purposely been allowed to become putrid. No stain.

AUXETICS ISOLATED

313

FIG. 104. — Mitosis induced in a lymphocyte by suprarenal extract which
had purposely been allowed to become putrid. No stain.

FIG. 105. — Asymmetrical mitosis induced by suprarenal extract augmented
by atropine. No stain.

AUXETICS ISOLATED 315

divisions in lymphocytes in the experimental ten
minutes.

Amount to be contained in the 10 cc. of jelly..

Dried extract of Testis . . . 0.025 gramme

" Pancreas . . . 0.025

" Muscle . . . 0.025

" Spleen . . . 0.01

" Liver . . . 0.002

Experimentally all these extracts were employed in a.
5-per-cent solution. Divisions in lymphocytes were in-
duced with the first three by adding 0 . 5 cc. of the solution
to the 5 cc. of coefficient jelly, together with 0.7 cc. of
1-per-cent solution of atropine sulphate, 1 cc. of the
5-per-cent solution of sodium bicarbonate, and made up
to a total of 10 cc. with 2.8 cc. of water. The jellies
were boiled and films made from them in the usual way.
It is obvious, therefore, that all the extracts contain
some substance or substances which cause cell-division
in lymphocytes and in leucocytes. To induce these
divisions on the microscope slide in the experimental ten
minutes, it is necessary to augment the action of the ex-
tracts with atropine. Suprarenal extract, however, evi-
dently containing more of the active substance than the
others, will induce divisions without any augmenting sub-
stance. Putrefaction will augment the power of the ex-
tracts like the alkaloid, and it was presumed that, this
putrefaction had this effect through the presence of the
alkaloids of putrefaction. This point, however, was not
investigated till afterwards, as we were immediately
concerned in finding out if possible what the agents
were in these extracts of dead tissues which cause the
division of white blood-corpuscles.

316 THE CAUSE OF HEALING

There were two ways in which we might attempt to
isolate this active principle from the extracts. We
might analyse them and try the different substances one
by one. These analyses had, however, often been done
before, and it was considered better, in the first instance,
to try the well-known constituents of these extracts to
see if they would induce cell-division before we under-
took to analyse the extracts ourselves.

We need not detail the vicissitudes of this research,
which occupied a long time. The constituents of the
extracts of the body are well known. It may be
remembered that the active principle in the extracts is
evidently thermostable, and remains in solution after
most of the proteins have been precipitated by heat.
We tried certain salts, and other substances, and we
have also tried urea, and at last kreatin (C4H9N3O2)
was found to be a substance which will induce divisions
in lymphocytes (fig. 106) and leucocytes (fig. 107).
Kreatinin (C4H7N3O) is not effective in the experi-
mental ten minutes; but xanthin (C5H4N4O2) is if
its action is augmented by atropine.

The following table gives the strengths of kreatin
and xanthin required to be contained in the 10 cc. of
jelly in order to induce divisions in lymphocytes in the
experimental ten minutes, no atropine being employed,
but the jellies contained 1 cc. (10 units) of alkali
solution.

Kreatin.

0.02 gramme . . . No mitosis seen.

0.04 " ... Early mitosis.

0.75 " Well-advanced divisions.

KREATIN AND XANTHIN 317

FIG. 106. — Mitosis induced in a lymphocyte by kreatin. No stain or extract.

FIG. 107. — Division in a leucocyte induced by kreatin. Xo stain or extract

KREATIN AND XANTHIN 319

With atropine :

Kreatin .

0.005 gramme . . . No mitosis.

0.01 . . . Mitosis.

0.02 . . . Well-advanced divisions.

Employing xanthin the presence of atropine is necessary :

Xanthin.

0.002 gramme . . . Early mitosis.

If the jelly contained saturated solution1 of xanthin,
well-marked figures were seen.

We had now succeeded in inducing the reproduction
of leucocytes and lymphocytes — first by the aniline dye
azur, then by a substance contained in the extract
of suprarenal gland, and by our experiments we were
able to infer that this substance is contained in the
extracts of other dead tissues. It has just been shown
that this inference is correct, because cell-division can
be induced by the crystalline extractive kreatin, which
is a constituent of the remains of all dead tissues. So
far, of course, we had only induced divisions with these
substances in vitro; but, as already pointed out, the
cells are under very detrimental conditions while being
experimented with, and it is more than probable that, if
they will divide in response to these substances in vitro,
they will more readily respond to them in vivo.

Healing is caused by the proliferation of leucocytes and
lymphocytes, and, judging from the in-vitro experimenta-
tion, this proliferation is evidently induced by kreatin and

1 Xanthin is sparingly soluble.

320 THE CAUSE OF HEALING

xanthin. Hitherto it has been generally supposed that
the cell-proliferation of healing is due to some inherent
propensity on the part of the cells to divide; but now
it is clear, from in-vitro experimentation, that these
cells divide when they absorb a definite quantity of a
chemical agent, and two of these auxetics are kreatin
and xanthin, wrhich are contained in the remains of
dead tissues. When a tissue is damaged anywhere,
cell-death is occasioned, and the dead cells liquefy.
The products of this death have as constituents the
extractives kreatin and xanthin, and we know that the
neighbouring living cells must absorb the liquefied
remains of their dead neighbours, for it has been shown
that the diffusion of substances into living cells is a
physical process over which the cells themselves can
exercise no control. When a tissue is damaged, there-
fore, the direct result of that damage will be to make
the neighbouring living cells reproduce themselves in
response to kreatin and xanthin, and bring about the
cell-proliferation of healing.

Here, then, is the solution of the first part of our
problem. We now know the nature of the physiological
cause of the cell-proliferation of healing, and we submit
that this knowledge reveals a fresh vista in pathology.

But it must not be supposed that kreatin and
xanthin are the only agents contained in the remains
of dead tissues which cause cell-reproduction. They
are two of the active principles which we have so far
succeeded in isolating. It is probable that there are
others; in fact, we know that there must be. Supra-
renal extract will induce divisions very readily, and the

THE NH2 GROUP 321

amount required to do so is so small that the kreatin
and xanthin which it contains will not account for the
divisions it induces. These bodies are amido-acids, and
we think that the NH2 group of the molecules may be
responsible for the auxetic action. In this respect it
is interesting to note that alkaloids, which augment the
action of auxetics, are compound ammonias; but it
must be remembered that we have never yet been able
to induce a division with an alkaloid by itself, although
we have tried literally hundreds of times.

In the next chapter we shall show that there is
another great and very important source of the " causes
of the cell-proliferation of healing" contained in a
substance we call "globin," a histone derived from
haemoglobin.

CHAPTER XIV

THE AUXETIC ACTION OF GLOBIN

THE fact that in-vitro experimentation has shown that
cell-division is directly caused by certain constituents
of the soluble remains of dead tissues made us consider
the possibility that there might be other sources of
these or similar agents. It was remembered how
frequently old chronic ulcers, when they heal, leave
the tissue pigmented, and it was considered possible
that this pigmentation might in some way be asso-
ciated with the healing process and its cell-prolifera-
tion. The pigment in ulcers is supposed to be derived
from haemoglobin.

Melanotic sarcoma is generally accredited to be
the most prolific of all malignant growths. It is
characterised by the pigmented cells of which it is com-
posed. We have not been able to obtain a case of
melanotic sarcoma, for such cases are rather rare, but it
is generally the case that the pigment is contained in
the cytoplasm of the malignant cells. One of the

322

MALARIA PARASITE 323

commonest sites of melanotic sarcoma is in the choroid
coat of the eye, where the cells are normally pigmented. , \

The pigment of these cells is called melanin, and it
is supposed to be derived from hemoglobin.

Professor Ronald Ross suggested that some experi-
ments might be made with auxetics on the malaria
parasite, and in one case a "crescent" was apparently
made to flagellate prematurely with a jelly containing
azur dye, extract, and atropine, although repetitions
of the same experiment were not successful. Still,
the consideration of the life-history of the malaria
parasite has been — as it turns out — germane to our
researches. The parasite enters the body from the
mosquito as a minute unpigmented amrebula, which
straightway enters a red blood-corpuscle. While in
the red cell it gradually becomes pigmented, and it
proliferates by exporulation. The daughter parasites
have no pigment until they enter fresh red cells,
when in their turn they become pigmented and ulti-
mately proliferate again.

There is the so-called sexual form of the cycle,
however, which probably does not proliferate within
the body. The crescent or gametocyte only pro-
liferates after the blood containing it has been shed.
The crescent is also deeply pigmented; and it is a
most interesting point to remember that when the
crescent stage of the parasite is reached, the red cell
appears to be depleted of haemoglobin, and merely
surrounds the parasite as an empty cell. The parasite,
when it has reached the crescent stage, has apparently

324 THE AUXETIC ACTION OF GLOBIN

devoured all the haemoglobin; the hsematin derived
from the haemoglobin has collected in the parasite as
a pigment known as melanin; and the parasite will
no longer proliferate until the blood is shed. // the
blood is shed, however, whether it is shed on to a
microscope slide or into the stomach of the mosquito,
the parasite again becomes prolific almost immedi-
ately, and flagellation occurs.

Now, when blood is shed, no matter how it is shed,
whether it be on a microscope slide or into the
stomach of the mosquito, haemoglobin must be set
free, for the red corpuscle is a very delicate cell, and
many of them must be ruptured when any injury
occurs in a tissue. The question therefore arises, Does
haemoglobin have any function in inducing the pro-
liferation of the malaria parasite ? From circumstantial
evidence it would appear that it does, for so long as
the parasite is absorbing haemoglobin from the red
cell in which it lives, so long will it continue to
proliferate by exporulation ; but when it has finished
the contents of the cell, proliferation ceases until
more haemoglobin can be absorbed by it when the
blood is shed.

In the malaria parasite, in the cells of melanotic
sarcoma, and in the neighbourhood of old healing
ulcers the haemoglobin is evidently decomposed because
the haematin collects as insoluble pigment.

Haemoglobin is fairly soluble, but when it is de-
composed into haematin and globin the haematin is
insoluble in water except in the presence of dilute
alkalies. Globin is readily soluble. Hence it cannot

EXPERIMENTS WITH GLOBIN 325

be the haematin part of the haemoglobin molecule which
has any function in causing proliferation; it must be
the globin part if it is either of them.

In the first instance we tried the effect of haemo-
globin on blood-cells. Jellies were made which con-
tained 1 cc. (10 units) of alkali solution, and after they
had been boiled various quantities of a saturated solution
of crystalline haemoglobin were added before the jellies
cooled too much for them to set on a slide. But haemo-
globin never induced divisions in lymphocytes or
leucocytes in the experimental ten minutes. Nor did
it excite amoeboid movements in them.

We next made a saturated solution of haemoglobin
and then boiled it, thereby decomposing it and pre-
cipitating the haematin. The filtrate is a straw-coloured
liquid when it is dilute. It was evaporated down by
prolonged boiling, and at the saturation point, which
is about 4 per cent, the solution becomes a deep red
colour. On evaporation to dryness, a sticky residue
remained. Very little is known (about globin. For
years it was thought to be a globulin, but this has
been shown not to be the case. Globin is a histone—
a protein which is not precipitated by boiling. In the
dry state it is a glutinous mass of a deep brick-red
colour, and it has a characteristic sweet smell some-
thing like licorice. If it is very dry, globin can be
ground into a brown powder. It is at all times ex-
tremely hygroscopic, and therefore if it is not kept in
solution it must be placed either in a desiccator or in
sealed tubes. If it is kept in solution and exposed
to the air, it rapidly decomposes owing to putrefaction,

326 THE AUXETIC ACTION OF GLOBIN

and gives off a foul smell, reminding one of that of
the alkaloid neurine.

Jellies were made which contained various strengths
of globin, and, of course, certain quantities of alkali
solution were also added. It was found that globin
by itself would never induce divisions in lymphocytes
in the experimental ten minutes, so we tried it again
with the addition to the jellies of 0.7 per cent of
atropine sulphate, and then globin induced divisions
in lymphocytes (figs. 108, 109). This is the best strength
to employ: In 10 cc. of jelly containing 10 units of
alkali and 0 . 007 gramme of atropine there should also
be 0.0025 gramme of globin. The best divisions are
obtained with 0.025 gramme of globin; but if the
content of it exceeds 0.05 gramme, the cells appear
to be poisoned, because they shrivel up and frequently
burst.

Some globin in solution (1 per cent) was allowed to
putrefy for a fortnight, and, like extracts of dead tissues,
it was then found that its action was so augmented that
it also would (in the strength of 0 . 005 — or better 0 . 01
gramme — in the 10 cc. of jelly) induce divisions by
itself (without atropine) in the experimental ten minutes
(fig. 110).

When putrefaction occurs in a solution of globin
a precipitate falls, and yet it is now more effective in
inducing divisions than it was before. It is clear,
therefore, that it is not actually globin which induces
divisions, but it is some constituent of it which is effec-
tive. Putrefaction decomposes globin, and the active
agent plus some augmenting substances are produced.

EXPERIMENTS WITH GLOBIN

327

FIG. 108.— Mitosis in a lymphocyte induced by glpbin augmented by
atropine. No stain, extract, or kreatin.

FIG. 109. — Asymmetrical mitosis induced by globin augmented by atropine.
No stain, extract, or kreatin.

EXPERIMENTS WITH GLOBIN

329

FIG. 110. — Mitosis induced in a lymphocyte by means of decomposed globin
solution. No stain, extract, kreatin, or atropine.

EXPERIMENTS WITH GLOBIN 331

It must be understood that if the jelly on which the
cells are resting contains 0 . 02 gramme, or more, of
globin, the red cells become distorted and the white
cells are killed without divisions being induced in
them.

We think that it should be mentioned that it is
quite within the realms of possibility that the malaria
parasite proliferated in response to the active agent
contained in globin; but although we have tried a few
experiments to endeavour to prove the point, we have
not succeeded in determining it. Malarial crescents
frequently flagellate in any case within ten minutes of
their being shed; and although we have mixed the shed
blood containing them with citrated solutions of globin,
it has been impossible for us to satisfy ourselves that
the flagellation has been accelerated by its action. In
the cases of malaria at our disposal there have not been
a very large number of parasites in the blood, and time
was lost during the experiments in finding them.
Hence we cannot speak definitely on this point, but
it was the consideration of the life-history of the malaria
parasite which \vas the chief factor which led us to
investigate the auxetic property of globin; and there is
no doubt whatever that globin contains some auxetic,
although it is not so powerful as that contained in
suprarenal extract.

Globin contains no kreatin so far as we can ascer-
tain, and the solution of globin which we have used is
free of haematin, as proved by spectroscope examina-
tion, and there are only traces of lipochrome. What
the exact nature of the auxetic substance contained in

332 THE AUXETIC ACTION OF GLOBIN

globin is we do not know, but possibly it is allied in
some way to the molecules of kreatin and xanthin. It
should also be remembered that we do not know what
the substance is in the azur dye which induces divisions.
We think that they will not be difficult to isolate; but
we ourselves do not feel competent to undertake
chemical analyses of this nature.

CHAPTER XV

THE PROOF THAT THE REMAINS OF DEAD TISSUES AND
GLOBIN CONTAIN THE CAUSES OF THE CELL-
PROLIFERATION OF HEALING AND OTHER CELL-
REPRODUCTION- — EXPERIMENTATION in VIVO CON-
FIRMS in-vitro OBSERVATIONS — THE CAUSE OF

BENIGN TUMOURS

THE foregoing experiments show that some of the
causes of human cell-division are now known. On
the stage of the microscope white corpuscles can be
made to undergo the stages of cell-division in direct
response to certain chemical agents, two of which have
been isolated, and which can be employed in crystalline
form to induce cell-division. What is far more im-
portant, however, is the source whence these chemi-
cal substances are derived. They are contained in the
soluble remains of dead tissues. Another source of the
cause of cell-division is in globin, which is derived
from the decomposition of haemoglobin.

It should be remembered that so far the experimen-
tation has been confined to testing the action of the
active substances and the sources of them on individual

333

334 THE CHEMISTRY OF PROLIFERATION

cells which have been removed for the purpose from
the body; and, as already pointed out, the cells in this
in-vitro experimentation are not by any means in
conditions similar to the natural ones under which they
normally exist. Still there is no question whatever
that the cells do divide in response to these agents ; and
if they will do so under detrimental experimental
conditions, it is obvious that they will be far more
likely to divide and respond to the same agents ; in their
normal conditions. The agents we know do not exist
in the body, and therefore it is practically a certainty
that these substances will cause proliferation there.
For reasons already given, on the microscope slide one
cannot induce more than one generation of cells by
chemical agents, because premature death cannot be
prevented; but in the body the premature death need
not necessarily occur, for its cause is absent, and hence,
provided the causes of cell-division are being constantly
supplied to cells, generation after generation must be
produced.

On the microscope slide cells will not divide, so far
as can be seen, unless they absorb definite quantities
of the agents which cause cell-division. We do not
say that there are no other substances wrhich cause cell-
division besides those which have been mentioned—
in fact, we know that there must be others; but what
we think is now becoming evident is the fact that cells
will not divide at all unless they receive some chemical
agent which makes them do so. That is to say, we
think that there is strong evidence in support of the
view that cell-division in the body is entirely caused by

AUXETICS ARE NECESSARY

335

chemical agents; and if these agents are not present,
there will be no cell-division.

In the case of leucocytes. For nearly a century
and a half these cells have been observed in the blood.
Every doctor and student of medicine must have seen
them alive repeatedly, and yet not a single person had
ever seen them divide. Now, however, if one makes
them absorb certain chemical agents the cells divide im-
mediately; and what is more, we have shown that the
rapidity of onset and the time occupied by each division
varies directly with the quantity of the substances
absorbed by the cells. Cell-division appears to be a
physical phenomenon which can be measured in the case
of each cell in proportions of grammes of the chemical
auxetics absorbed by them. We have shown how it
can be set down as a simple mathematical equation.
It must be admitted that in spite of the fact that
blood-cells have not been seen to divide without an
auxetic, there is no actual proof that a cell cannot
divide without one. It has yet to be proved that
human leucocytes have no inherent power to multiply
"when they feel so inclined," but it is a remarkable
thing that no single leucocyte, out of the many millions
which have been seen by men, should ever have de-
veloped this inclination during nearly a century and a
half. On the other hand, we know that if we cut our
fingers and so produce the remains of dead tissues
containing kreatin and xanthin, proliferation of leuco-
cytes occurs immediately; and the greater the injury,
the greater the cell-proliferation.

We think that if the problem is carefully con-

336 THE CHEMISTRY OF PROLIFERATION

sidered, and, better still, if these mitotic divisions
are actually seen as they occur in response to chemical
agents, it will be appreciated that there is a strong
probability that cells only divide when they are made
to do so by an exciter of reproduction.

The active auxetics are contained in "the remains
of dead tissues." Globin is in reality "the remains of a
dead tissue," for it is obtained by the decomposition
of haemoglobin, and haemoglobin is contained normally
in living red cells. Doubtless the constituents of the
molecules of kreatin, xanthin, and the active principle
of globin are present in living protoplasm; but they
may not be present, presumably, in the same combina-
tion or form as they exist in kreatin and xanthin.
Possibly it is only after death that these substances
are produced, in which case it would follow that a cell
will not reproduce itself by virtue of the constituents
of its own living protoplasm; but it is necessary for it
to absorb fresh active agents from the dead remains of
its' neighbours.

Many points are now explained. When it is re-
quired that an indolent healing surface shall heal well,
we scarify it, as exemplified in the operation of Thiersch
grafting. If a fractured bone will not unite, the ends
are rubbed together or actually "freshened" by opera-
tion, to produce callus; and callus is really a tissue
made by the proliferation of cells. When we scarify
or freshen a surface, we merely cause destruction, and
thereby set free exciters of reproduction. If a part of
the body is bruised, haemorrhage occurs; and, as is
shown by the pigmentation, the haemoglobin set free

CELL-DEATH INDUCES CELL-BIRTH

337

from destruction of red cells which have been shed
into the injured tissues is decomposed, and globin is
thus locally produced. The cell-proliferation of healing
must then occur in response to it, and the remains of
other tissues which have been killed in the injury.

The proliferation of cells, however, is not confined
to the cell-proliferation of healing. It will be shown
that epithelial cells will also respond to auxetics, and
probably some if not all other cells also respond to
the soluble remains of their neighbours by reproducing
themselves. It is true that globin does not exist in
the cornea, for here there is no blood supply, and con-
sequently no haemoglobin until some time after the
injury. Still, if the cornea is injured the corneal cells
must be injured, and the cell-proliferation of healing
occurs in response to the remains of the injured cells.

Irritation is always followed by cell-proliferation.
Irritation means damage, and damage means cell-death.
Cell-death sets free kreatin, xanthin, and other auxetics,
and the cell-proliferation is caused by their absorption
by the neighbouring living cells. The greater the
damage, the greater will the cell-proliferation be.

Cell-division is apparently an automatic phenom-
enon— not in the sense that it is due to some in-
trinsic function or duty of a cell's protoplasm, but
automatic in that the death of one cell will cause the
reproduction of its living neighbours. If we may
speak of the act of cell-division by mitosis as the
"birth" of cells, then we may say that the number
of births of cells in the body depends on the number of
deaths. The greater the number of deaths, the greater

5'''

338 THE CHEMISTRY OF PROLIFERATION

the number of births. If an individual cell dies, its
death causes its neighbours to multiply to supply the
deficiency; but if the cell-death is extensive owing to
damage, the proliferation of those cells which have not
been killed will also be extensive, and this proliferation
will now be extended to that of the white blood-cor-
puscles which have been shed during and after the
injury; and the result will be the cell-proliferation of
healing.

Judging from the experiments which have been
made, it may also be assumed that since the number
of cell-births depends upon the number of cell-deaths,
and since an increase in the number of births must
increase the number of deaths, it follows that the
number of deaths must also depend to some extent
on the number of births. Presumably, if once cell-
division is set going in a tissue or in a part of a tissue,
that cell-division will go on increasing until something
restrains it. Elimination from a tissue of tissue fluids
would restrain it; for if the soluble remains of dead
tissues become quickly eliminated, the diffusion of the
constituents of these fluids into the cells would also be
arrested, for that diffusion varies directly with the
factor time. In a damaged tissue the vessels and
lymphatics are also damaged, and elimination may be
impaired ; hence the remarkable cell-proliferation which
leads to "granulation tissue." In an injury of any part
except the cornea, coagulation of the shed blood occurs ;
the red cells become laked, and ultimately the haemo-
globin is evidently decomposed, as evinced by the
pigmentation which will always be seen even in a

ORIGIN OF BENIGN GROWTHS 339

bruise. The globin so produced will assist in pro-
moting the cell-proliferation of healing.

Such is the explanation of the cause of cell-division
in the human body as demonstrated by in-vitro experi-
mentation. But we think that we may go farther,
and suggest that the initial multiplication of the cells
in the human embryo may also be caused by a chemical
auxetic. Spermatazoa contain extractives. Possibly
it is these extractives, set free from this spermatozoa,
which, after fertilisation, give rise to the subsequent
cell-division in the ovum from which the embryo is built
up. Once the cell-division has started, it will go on in
response to the cell-deaths which sooner or later must
occur.

As we have pointed out, kreatin is not by any
means the only auxetic contained in the remains of
dead tissues, and it is yet to be proved that there is
not some specificity in cell-reproduction due to some
at present unknown substance. We know from the
study of heredity that certain characteristics are car-
ried in the ovum and in the spermatozoon, and if they
are so carried, doubtless other chemical auxetics, far
more complex than kreatin, may be carried too.

In the meantime we think that the knowledge that
dead tissues cause cell-proliferation is sufficient to give
an inkling as to the cause of benign growths. A
sudden cell-death occurring in a tissue will cause pro-
liferation of neighbouring cells. Of course, if the
initial cell-death is extensive, the cell-proliferation of
healing will occur which ultimately leads to the pro-
duction of connective tissue, which in itself may

340 THE CHEMISTRY OF PROLIFERATION

prevent undue extension of the proliferation of the
normal tissue-cells. But supposing for some reason,
such as a slight injury, a local cell-death takes place:
it would cause increased proliferation of local cells,
and so form the basis of a tumour. Once this growth
is started, it will go on until, by causing "irritation"
or, to be more accurate, extensive cell-death, it may
now induce the cell-proliferation of healing round it,
and so, by the formation of connective tissue, cause its
progress to be arrested by a capsule. A benign tumour
is probably due merely to some localised cell-death in
the first place, and it is remarkable how frequently
there is a history of injury in these cases. But there
is also no doubt that the onset of benign growths, and
other cell-proliferation too for that matter, must be con-
trolled to some extent by nervous influence. Possibly
this nervous influence may be actuated by the nervous
control over local elimination. Quite recently a paper
appeared in The Lancet on a case of bilateral benign
tumours;1 and this can only be due to some central
control over the local causes of cell-division.

Fibroids of the uterus occur only during the years
of menstrual activity. During this time the uterus
periodically becomes enlarged, followed by reduction in
size. This reduction and quiescence must be accom-
panied by death of living cells, and presumably it is
this death which, if elimination of the products of
katabolism is impaired, may lead to excessive pro-

1 See a paper on Bilateral Tumours by W. Roger Williams in The Lancet,
Feb. 12, 1910.

ULCERS TREATED WITH AUXETICS 341

iiferation of the remaining living cells, and so cause
the growths known as fibroids.

The foregoing conclusions and deductions have
been arrived at from experimentation in vitro with
individual cells. As pointed out in a former chapter,
conclusions derived from in-vitro experimentation are
not in themselves sufficient to prove a point. Because
we can induce cell-division in individual cells on the
microscope stage with certain chemical agents does
not prove that the same division will necessarily occur
in vivo in the same cells in response to the same
agents. But, fortunately, in-vivo experimentation with
these agents has not been impossible, and the proof
that these agents, or rather some of them, do actually
cause proliferation in the body is now at our dis-
posal. In the wards of the Royal Southern Hospital
at Liverpool cases of chronic callous ulcers of the
legs were admitted, and have been treated in the first
instance with saturated solutions of globin. The
globin was applied to portions of the ulcers by dip-
ping pieces of sterile gauze in the solution and applying
it direct to the ulcerated surfaces. Granulations im-
mediately appeared in response. In the short space
of three or four hours a difference appeared between
the extent of the granulations in the treated as com-
pared with the untreated portions of the sores. In
twenty hours the difference was marked. Granulomata
have been produced in a day or two by means of
globin.

Others suggested that the proliferation was not
necessarily due to the globin, but to the "irritation"

342 THE CHEMISTRY OF PROLIFERATION

of the gauze, in spite of the fact that ulcers have been
treated with gauze all over, but only a part of them
with globin added, and the proliferation occurred to
the marked extent only where the globin was. We
therefore discarded gauze or dressings altogether, and
repeated the experiments. In a case where there were
several ulcers on one leg the surfaces of them all were
scarified, and small pieces of dried globin were "dotted"
all over one ulcer. The cell-proliferation occurred to a
marked extent in that ulcer, but only to a much less
extent in the others which were not so treated.

Globin thus applied to a healing surface causes a
scab to form very rapidly (figs. Ill, 112), and the
cell-proliferation goes on beneath it. This scab forms
in an hour or two, whereas, if no globin is applied, it
takes several days for a scab to form on an ulcer
which has no dressing on it. Globin also causes ex-
tensive proliferation of the epithelium from the sides
of the ulcer.

Unfortunately suppuration occurs under the scab,
no matter how "clean" the ulcer may be when the
globin is applied. The onset of suppuration, how-
ever, has been delayed by preparing the globin with
aseptic precautions throughout, thus: A solution of
haemoglobin is decomposed by boiling, and filtered, and
the globin solution is concentrated until it precipitates
by further boiling. It is evaporated to dryness at a
temperature of 60° C. and immediately sealed into sterile
glass tubes. Even with these precautions, suppuration
usually occurs under the scab in the course of a few
days. The scab is then removed with fomentations,

CLINICAL OBSERVATIONS

343

FIG. 111. — To show the way in which globin is "dotted" over the surface of an ulcer.

FIG. 112. — To show the scab formed by the application of globin to an ulcer.

CLINICAL OBSERVATIONS 345

and when the sore is clean it is once more scarified,
and fresh sterile globin is again "dotted" over its
surface. This procedure can be repeated until the
ulcer heals. During the scarification it is better to
draw blood. Latterly this treatment of ulcers has
been improved by using powdered globin (five parts),
mixed with two parts (by weight) of kreatin, a mix-
ture which produces more marked proliferation than
pure globin.

Many ulcers have now been treated by this method,
and we think that we can say safely that it causes
more rapid healing of them than if they were treated
in the usual way. Callous ulcers will usually heal
by themselves if the limbs are kept at rest, and it
was suggested to us that the cell-proliferation pro-
duced by globin was in reality due to the fact that the
patients were kept in bed. This suggestion was dis-
proved, however, by the production of extensive pro-
liferation in one part of an ulcer by means of globin
in a patient who was made to walk about during the
treatment. Lastly, granulations have been induced by
extracts of suprarenal gland.

It should be mentioned that globin, kreatin, etc.,
when applied to a healing surface will not only cause
proliferation during the application; but once the mul-
tiplication has started, it will continue "automatically,"
even though the application of the auxetic is discon-
tinued. This point has frequently been seen during
the experimentation with ulcerated legs, and it is proof
that the proliferation of cells is "automatic." There
can be no doubt that once proliferation is started

346 THE CHEMISTRY OF PROLIFERATION

in an ulcer, an increased number of deaths is occasioned,
which in its turn still further increases the proliferation,
as seen in the ulcers once treated with globin.

The application of dry globin to a scarified sore has
elicited the interesting fact that it will convert the dark
venous blood drawn by the scarification into the bright
and red arterial variety, and the scars resulting from
the treatment appear to be exceptionally firm and un-
likely to break down again.

This form of treatment, however, must be carried
out with care, and suppuration not allowed to continue
for long in the presence of an auxetic, for, as will be
shown in the next chapter, there is a possibility of
malignant proliferation occurring in place of the normal
one if the products of decomposition become pent up
in the neighbourhood of proliferating epithelial cells.

These experiments afford conclusive proof that the
cell-proliferation of healing can be caused by the chem-
ical auxetics, kreatin and globin, and that the deduc-
tions made from the prolonged experimentations with
the in-vitro method described in this book are correct.
The possibility of the mitotic divisions induced on the
microscope slide being in the nature of "freaks" or
being due to death-struggles is disproved. As a matter
of fact, these possibilities practically fell to the ground
when mitoses were induced by extracts of dead tissues.
One could conceive that a purely artificial substance
like azur dye might cause mitosis by exciting the cells
greatly just before death; but we think that in all
probability the aniline dye contains some constituent

CLINICAL OBSERVATIONS 347

which possibly resembles the molecules of the natural
auxetics.1

The fact that the cell-proliferation of healing is
caused by chemical agents contained in the soluble
remains of dead tissues will, we confidently believe,
be the means of solution of many problems which at
present confront the investigator in pathology and
perhaps in physiology also. It is a fact about which
there can be no doubt whatever.

1 The formula of azur dye (Cent. f. Bakteriologie, Bd. xxix., 1901) is:

S02
(CH3)2NV /\ /\ /\ y>N(CH,)2.Cl

CHAPTER XVI

THE AUGMENTED DIVISIONS INDUCED BY PUTREFACTION
OF THE EXTRACTS ARE DUE TO THE ALKALOIDS OF

PUTREFACTION A THEORY THAT CARCINOMA AND

LYMPHADENOMA MAY BE CAUSED BY THE MIX-
TURE OF THE AUXETICS OF CELL-PROLIFERATION
WITH CHOLINE OR CADAVERINE AN EXPLANA-
TION OF THE AGE-INCIDENCE, METASTASES, AND

OTHER FACTS KNOWN CONCERNING CANCER THE

NECESSITY FOR A CRUCIAL EXPERIMENT TO PROVE
THE THEORY

IN Chapter IX. it was pointed out that there is an
intimate association between "chronic irritation" and
the onset of cancer. As just shown, "irritation" means
cell-death, and cell-death is followed by cell-prolifera-
tion. When a tissue is the seat of chronic irritation,
the cell-proliferation of healing must be going on in the
damaged site owing to the presence of the remains of
the dead cells. The proliferation occasioned by irrita-
tion is in reality due to the auxetics, some of which are
kreatin, xanthin, and that contained in globin, which
are set free by the death of some of the cells. This
will explain why an ill-fitting boot will give rise to

348

THE AUXETICS OF IRRITATION 349

a "corn," and to the "induration" of a tissue which is
under pressure or being chronically irritated. In reality
"irritation" must be followed by chronic cell-prolifera-
tion due to the auxetics produced.

Now, the chief characteristic of cancer is that it
consists of a growth of cells which are proliferating
excessively. Every cancer is a growth which infiltrates
the surrounding tissues; and this growth occurs pro-
bably in every instance in a site in which there is
chronic irritation — or rather where there is chronic
cell-proliferation of healing due to auxetics.

One may suggest, therefore, that since the prolifera-
tion of chronic irritation is due to the auxetics produced
by cell-death, the proliferation of cancer is also associated
with them. The proliferation of chronic irritation,
however, is a normal one, whereas that of cancer is
a malignant one. If the cause of the normal prolifera-
tion is removed, then ultimately proliferation ceases;
but if the irritation which predisposed to cancer is
removed, the malignant cells appear to continue to
multiply until the patient dies. Yet cancer-cells are
cells of the body. They are not foreign parasites, and
hence it may be that in a cancerous growth there is
some other factor in addition to the normal ones.
Therefore it may also be suggested that the onset of
cancer in a normal healing site may be brought about
by the presence of another agent in addition to the
normal auxetics produced by cell-death.

Now let us return to the "augmenting" of the
action of auxetics in promoting cell-division by putre-
faction and by the alkaloid atropine. It is well known

350 THE PROLIFERATION OF CANCER

that certain putrefactive bacteria in decomposing dead
organic structures produce ptomaines and leucomaines.
These substances are in the nature of alkaloids. The
following are common ones:

Choline .... C5H15NO2.

Cadaverine . . . C5H14N2.

Neurine .... C5H13NO.

Putrescine . . . C4H12N.

Choline will, like other alkaloids, excite amceboid
movements in leucocytes and lymphocytes, and so will
cadaverine. In fact choline is just as effective as
atropine in this respect. The best strength of choline
to employ to excite amceboid movements in leucocytes
and lymphocytes is one in which 10 cc. of jelly con-
tains 0.01 gramme of the alkaloid in addition to the
10 units of alkali. Choline, however, is not very poison-
ous to leucocytes, and even 0.04 gramme will not kill
them. Cadaverine also excites leucocytes, and 10 cc.
of a jelly containing 1 cc. of a 1-per-cent solution of
it is suitable for this purpose if 10 units of alkali are
also present, the jelly-film being examined, of course,
at the room temperature.

It may be remembered that it was through the
accidental putrefaction of the extract of suprarenal
gland that we were enabled to induce divisions with
it by itself for the first time, and wre now know that
the reason for this was that the putrefaction produced
the alkaloids choline and cadaverine in the solution
of the extract, and that they, like atropine, greatly
augment the action of auxetics in inducing cell-division.

THE EFFECTS OF ANIMAL ALKALOIDS 351

In order to prove this point we now, in the first instance,
used these pure alkaloids, choline and cadaverine,
added to the extracts, and afterwards we combined
them with kreatin and xanthin to induce augmented
divisions.

If a jelly contains 0.01 gramme of choline and
10 units of alkali solution, divisions in lymphocytes
can be induced if only 0.02 or even 0.01 gramme of
kreatin is present (fig. 113). In fact, this alkaloid of
putrefaction choline, like atropine, augments the action
of auxetics about five-fold.

Using cadaverine in the strength given above,
divisions in lymphocytes w^ere induced if the jelly con-
tained only 2 cc. of a 1-per-cent solution of kreatin.

It has already been mentioned that a mixture of
atropine and an auxetic will give rise to asymmetrical
mitosis in lymphocytes, and we have also found that
these remarkable mitoses also are frequently induced
by the augmenting action of choline and cadaverine
(figs. 114-16). This point is of great importance,
because it is well known that asymmetrical mitoses
are frequently seen in cancerous growths.

So far the augmented divisions had only been
induced in lymphocytes. It is true that there is a form
of cancer which occurs in the lymphocyte class of
cells of the lymphatic glands (lymphadenoma) ; and if
it is a criterion that because a lymphocyte divides by
an augmented asymmetrical division it is necessarily
malignant,1 then the combination in certain proportion
between the causes of the proliferation of healing plus
an alkaloid of putrefaction like choline must be a cause

1 It has not been proved to be a criterion.

352 THE PROLIFERATION OF CANCER

of lymphadenoma. In connection with this it is
interesting to note that many years ago Trousseau1
stated in his book that lymphadenoma often follows
on a suppuration focus, and this view is upheld by
many to this day. At the same time it must be re-
membered that no alkaloid has yet been made to induce
a division by itself; it is essential for an auxetic to be
present also. Alkaloids appear to be augmenters only
of cell-division.

But our object was, if possible, to find the cause
of carcinoma, and we therefore tried to see if our
chemical agents would induce divisions in epithelial
cells. Considerable difficulty was met in investigating
this point. Epithelial cells will not live long in vitro; in
fact, they usually die in a few moments, as far as can be
seen. But at last we did succeed in inducing an early
mitotic figure in two epithelial cells (as shown in the
photographs, figs. 117, 118) from the vaginal secretion.
We did not succeed in inducing the divisions with an
entirely "natural" agent, for epithelial cells evidently
require more auxetics than even leucocytes. The
figure induced wTas seen when the epithelial cells were
placed on a powerful jelly which contained azur stain,
putrid extract of suprarenal gland, and atropine.
In vivo, also, epithelial cells undoubtedly proliferate
in response to globin and kreatin.

The fact was, therefore, proved that epithelial cells
respond to the chemical exciters of reproduction, and
it is possible that they may be subject to the same
conditions as lymphocytes, and only respond to them.

1 Trousseau's Clinical Medicine (Sydenham Society), 1872, vol. 5, p. 207.

AUGMENTORS OF AUXETICS 353

FIG. 113.— Mitosis induced by a mixture of kreatin and choline. No stain,
extract, or atropine.

FIG. 114.— Asymmetrical mitosis induced in a lymphocyte by a mixture
of suprarenal extract and globin, augmented by choline. No stain or

AUGMENTORS OF AUXETICS

355

FIG. 115. — Mitosis in a lymphocyte induced by globin and choline. No
stain or other auxetic.

FIG. 116. — Mitosis induced in a lymphocyte by suprarenal extract and chol-
ine. No stain or other auxetic.

AUGMENTORS OF AUXETICS

357

FIG. 117. — Mitosis induced in an epithelial cell by a mixture of stain and

extract.

FIG. 118. — Early mitosis in an epithelial cell from the vagina induced by

stain and extract.

AUGMENTORS OF AUXETICS 359

The action of these auxetics upon lymphocytes
is greatly augmented by the alkaloids choline and
cadaverine, which are produced in a solution of an
extract of a dead tissue as it decomposes. We now
made some experiments to see if putrefaction also
augmented the auxetic action of globin, and a solution
of it was therefore allowed to decompose at the room
temperature for about three weeks. A solution of
globin of no matter what strength will not (as already
noted) induce divisions by itself in ten minutes; it is
necessary to add atropine. But when a 1-per-cent
solution of globin had decomposed, it was found that
now it would induce divisions in lymphocytes by itself
in the experimental ten minutes. Whether this aug-
mentation of the action of globin by decomposition
is due to the production of choline and cadaverine or
not, we are uncertain; for although we have tested1 the
decomposed solution for the presence of alkaloid, only
a negative result has been obtained. This matter will
require further investigation, for it is possible that other
substances besides the alkaloids of putrefaction may
augment the action of exciters of reproduction. We
have so far obtained the best divisions by making up
the 10 cc. of jelly with 0.5 cc. of a 2-per-cent solution
of globin which had been kept open to the air of the
room for three weeks.

Decomposition of organic solutions which contain
exciters of reproduction will augment the action of the
latter agents up to as much as five-fold; and in this
case the divisions induced in lymphocytes are fre-

JThe iodine and the mercuric chloride tests were employed.

360 THE PROLIFERATION OF CANCER

quently of the asymmetrical variety. Cancer is a
growth of cells which supervenes on an old irritated
site where the cell-proliferation of healing has been
going on for some time. Since cancer is an exu-
berant growth of epithelial cells— which will respond
to auxetics — it is obvious that the quantity of the action
of the normal auxetics present must be augmented in
some way so as to give rise to the exuberant prolifera-
tion of malignancy. And lastly, the mitoses in a
cancerous growth are frequently of the asymmetrical
type. The combination of a normal auxetic plus an
alkaloid of putrefaction and decomposition will cause
not only augmented divisions, but it is important to
note that these divisions tend to be asymmetrical in
character.

Having arrived at this stage of our researches, these
new facts were carefully considered to see how they
harmonised with the well-known features which are
associated with cancer, and we shall now discuss them.

The fact that cell-division is caused by substances
contained in the remains of dead tissue throws light
on the age-incidence of the disease.

In a paper published by us in The British Medical
Journal on October 23, 1909, when we were aware of
the fact that cell-division could be induced by an aniline
dye, and that its action could be augmented (which
was all we knew then) by the remains of a dead tissue,
we appreciated that the remains of dead tissues might
be a predisposing factor in the cause of carcinoma. We
may as well quote the passage from that paper, for it
shows the possible relationship between the remains of

AGE-INCIDENCE 361

dead tissues (the products of katabolism) and the age-
incidence of cancer.

The body is mainly composed of living cells, and
they constitute an elaborate combination of living
factors. We know that in certain tissues these cells
are continually dying and being replaced, so that it
is evident that birth and death must be going on
incessantly in the body. What happens to the dead
cells ? They of course liquefy and become dis-
organised, and their constituents are presumably ex-
creted or converted into other compounds. While
this is happening it seems probable that some of the
products of the remains of dead cells may be absorbed
by their neighbours, for it must be remembered that
the diffusion of substances into living cells appears to
be a physical process over which they exercise no
control. There are doubtless some cells which remain
alive for long periods; for instance, it has been
estimated (and we are informed that it is practically
certain) that some cells of the central nervous system
live throughout the life of a man. Many cells, how-
ever, only live a very short time, the length of their
lives perhaps varying in different parts of the body, so
that the remains of dead cells are probably always
present in the body fluids. In this connection, how-
ever, we have to keep in mind the physiological curve
expressive of the relationship between anabolism and
katabolism. There are only three stages of life if "it is
viewed from this point of view, the first terminating at
about the thirtieth year, when a man reaches his prime,
and up to which period cellular birth must preponderate
over its death-rate. For some years it may be sug-
gested that anabolism and katabolism remain balanced ;

362 THE PROLIFERATION OF CANCER

and that after the age of 40, quite physiologically, so
that nothing occurs to make a man aware of it
physically, these conditions begin to be reversed and
more of the products of katabolism — that is, the
remains of the dead cells — tend to exist in the body-
fluids than was the case before middle age.

Here we have a fact incidental to the cancer period
which suggests the possibility that these products of
katabolism may in some way predispose to the onset of
malignancy. It cannot possibly be suggested that they
are the cause of the disease, for if such were the case
everybody over the age of 40 would die of cancer; but
assuming that some product of katabolism may possibly
favor the onset of the disease, we may enlarge upon the
speculation and say that it is a certain morphological (or
chemical) element in a dead cell which may be the agent.
For the sake of argument it may be derived from either
the cytoplasm, the cell-wall, the nuclear wall, or the
linin, or it may be the chromatin itself.

It is now known, of course, that the products of
katabolism actually contain causes of cell-reproduction ;
and it follows that if it is correct that these products are
in excess after the age of 40, there must be, ceteris
paribus, a greater inclination to cell-proliferation in a
tissue after that age than before it. To produce cell-
division it is necessary for a cell to absorb a certain
quantity of an auxetic, and to produce augmented
asymmetrical divisions by an alkaloid it requires a
certain combination (as already specified) between the
alkaloid and the auxetic. If cancer is due to this
combination, it is possible that before the age of 40
there is not usually sufficient free auxetic to produce

SITE-INCIDENCE 363

the right combination, for we have shown that alkaloids
by themselves are not effectual in inducing cell-division.
On the other hand, after the age of 40 the slight
physiological increase in the quantity of auxetics
present in a tissue, owing to excess of products of
katabolism, may just supply the required quantity of
auxetic to produce the right combination between
them and an alkaloid, should the latter be present.
It has already been suggested that the onset of
cancer may be partly due to the oversetting of a
normal balance.

In connection with this point we would recall the
fact that cancer seems to attack persons who are
prematurely aged, especially those who are subject to
such diseases as the atrophic form of osteoarthritis—
a fact which seems to bear out this explanation of the
age-incidence of cancer.

Conversely, it has been shown that cells with a
lowered vitality require more of an auxetic to produce
cell-division in a given time than normal cells; and
this may explain why cancer does not so commonly
occur in the aged and infirm, for although the right
combination of the cause of the disease may be present,
it is not present in sufficient strength to produce
malignant proliferation in cells which have lost their
vitality to some extent.

The suggestion that cancer may be due to putre-
factive decomposition of the remains of dead tissues
in a chronic healing site will harmonise with the fact
that cancer occurs commonly in certain sites. Carci-
noma occurs most frequently in the breast, uterus,

364 THE PROLIFERATION OF CANCER

mouth, stomach, intestinal tract, and rectum. A
chronic healing focus in the rectum, mouth, or intestine
may readily be associated with decomposition products.
As already pointed out, chronic irritation means chronic
cell-proliferation of healing due to the auxetics con-
tained in dead cells, and in the rectum, intestine, and
mouth further decomposition with the gradual produc-
tion of alkaloid must easily occur in a chronically
injured site in these regions. It is interesting to note
that cancer of the pancreas nearly always attacks the
"head" of that gland, namely, that part of it which has
nearest access to the intestine. In the rectum "irrita-
tion" must be of frequent occurrence by the impaction
of faeces, and this in itself will obviously render this part
of the alimentary canal a common site for malignant
disease — and it is one of the commonest places for it.
In the mouth decomposition readily occurs, and how
commonly one sees carcinoma of the tongue, lips,
fauces, etc.

Syphilis is undoubtedly a predisposing factor. Syphil-
itic lesions of the mouth, which, of course, are ac-
companied by healing, are of very common occur-
rence, and it is possible that choline is produced in
tertiary syphilides. If choline is produced by the action
of Trypanema pallida, then the cause of syphilis may
be a predisposing cause of cancer^that is, if the
argument is correct that the alkaloid choline in cer-
tain combination with auxetics gives rise to malignant
proliferation.

The breast and cervix uteri are localities which are
very prone to cancer, and in these organs destruction

PRODUCTION OF ALKALOID 365

of tissues occurs to some extent every month until
the climacteric, when great involution takes place. It
is during this latter period that the onset of carcinoma
is favoured. It is a remarkable thing that cancer
almost only occurs in these parts in parous women,
whereas in nulliparous women they are comparatively
free. We do not think we are going too far in suggest-
ing that in parous women, when the ducts of the glands of
the breast and uterus have been — so to speak— opened
up, access is now afforded to the organisms of decom-
position and putrefaction. In nulliparous women, when
these organs have remained functionless, the ducts of
their glands are more likely to be closed to invasion
from without.

In any site, however, the products of katabolism
may determine the age-incidence of carcinoma.

One cannot assert that the alkaloid choline, or
cadaverine either, are only produced in a damaged
site by the action of putrefactive organisms.. It was
owing to decomposition by putrefaction of extracts
of dead tissues and globin that we were enabled to
obtain augmented asymmetrical cell-divisions with these
alkaloids, but it is possible that these alkaloids — or
others equally effective — may be produced by other
agencies; and if so, provided the contention is correct
that the alkaloids help to cause carcinoma, "other
agencies besides putrefactive organisms may cause the
disease. The point is an important one in view of
the controversy as to whether cancer is a "parasitic"
disease. In any case these alkaloids can be produced
by more than one class of organisms, and we have

366 THE PROLIFERATION OF CANCER

pointed out that the cause of syphilis may also
produce one of them — -in fact, General Paralysis of
the Insane has been said to be due to choline.
Hence, if our contention is correct, cancer can hardly
be said to be due to a specific parasite.

The mere fact that the alkaloids were being pro-
duced in a chronic healing site would not necessarily
cause in it augmented proliferation. The alkaloids
and the auxetics will have to be present in certain
proportion; and since the production of this pro-
liferation necessitates the diffusion of the combination
into the cells, time is an essential factor. In all prob-
ability it would be necessary for the decomposed
remains of dead tissues to be pent up to some extent
for a considerable period.

The suggestion that cancer may be due to the
combination of auxetics and ptomaines will offer an
explanation of the cause of death from the disease.
If the malignant cells and healing site are completely
removed, the patient may recover; but if this is not
the case, recurrence will, of course, take place, for in
removing the growth a fresh healing site is produced,
and the original decomposition may go on in it.
Putrefaction of the remains of dead tissues may occur
in a healing site without visible suppuration. The
^Bacillus subtilis does not produce pus, yet it will
produce choline. It may be these ptomaines which
ultimately cause the death of the patients by poisoning
them; for if decomposition sets in in a damaged site,
unless steps are taken to remove it, doubtless the
decomposition will usually go on.

METASTASIS 367

The possibility of carcinoma being due to the
combination of alkaloids and auxetics will also
explain the reason for the way in which malignant
growths frequently "break down." As shown by
in-vitro experimentation, cells can only withstand a
certain quantity of the combination. If excess is
forced into them, they will die. Even globin itself
is very poisonous to leucocytes and lymphocytes if it
is in excess. If this excess was present in the body,
it would cause cell-death unless the excess was
removed, and the cell-death would only aggravate
the trouble, especially if the lymphatics were blocked
by malignant cells. Ultimately, of course, the growth
would "point" and break down. A certain amount
of local cell-death will cause increased cell-proliferation ;
but after a certain stage is reached, breaking down must
occur with subsequent ulceration.

Up to a certain point, therefore, the greater the
malignant proliferation, the more cell-death will there
be, and the more will the disease be aggravated.
And the aggravation may be increased by the
chromosome granules which cells appear to discard
when they are excessively prolific. These chromatin
granules may contain kreatin, and they will therefore
merely supply more auxetic for the neighbouring
cells.

The phenomenon of metastasis in cancer is an im-
portant factor to be considered in conjunction with the
other facts. The invasion of lymphatics by cancer may
be due to the combination of auxetics and alkaloid
being passed through them from the original healing

368 THE PROLIFERATION OF CANCER

site. We have seen amoeboid movements in cancer-
cells in response to alkaloids, and possibly this may
assist in the infiltration of vessels and tissues and so
predispose to metastasis. A striking fact known about
secondary growths is that in the arrangement of the
cells they resemble the primary ones. We think that
this can be explained only by embolism. If a second-
ary growth in another organ was a fresh cancer, it
is difficult to imagine how it could possibly resemble
the primary ones in the arrangement of the cells.
Metastases practically only occur in the later stages of
carcinoma when the lymphatics have been extensively
invaded. In benign growths one rarely if ever sees the
vessels invaded by cells, and presumably this is the
reason why secondary tumours do not follow. The
extensive researches which have been done by others
in transplanting tumours in mice have thrown con-
siderable light on the nature of secondary growths.
In transplanting a tumour from one animal to another,
it seems to us that one is in reality producing a secondary
tumour. Now, to effect this, as is well known, it is
necessary that the cells of the tumour should be alive;
the transplanting of dead cells will not cause a second-
ary growth. This knowledge harmonises with our
suggestions as to the cause of cancer. If one inoculates
an animal with dead cells, although the organisms of
putrefaction may be present among them, the remains
of the dead cells are soon removed from the inoculation
site and the production of augmented auxetic must
cease. Normal healing will take place before the
putrefactive organisms have had time to restart and pro-

METASTASIS 369

duce choline, cadaverine, etc., for it is known1 that to
produce these alkaloids it takes at least a fortnight..
If, on the other hand, a portion of a living primary;
growth is transplanted, the living cells will continue
to multiply in response to the auxetics produced by the
cell-death which continues to occur among the malig-
nant cells which have been inoculated. In transplanting
a malignant growth, one must transplant some putrefac-
tive organisms along with the malignant cells, and in
the spaces between the cells the combination of auxetics
and alkaloids must be present from the outset and be
continuously produced without interruption, because a
living growth is transplanted. For a secondary growth
(or a metastatic one) to occur, it is necessary for living
cells to be transplanted; and we believe that it is also
necessary for organisms to be transplanted within it, so
that the causes of the augmented proliferation continue
to be supplied without interruption.

There is another possible explanation of a metastatic
growth which should be mentioned. It has been sug-
gested by others, who, of course, were unaware that
cell-division in the body is caused by chemical agents,
that once a cell becomes a malignant one, its daughter
cells will also be malignant. This would mean that
a cell, in acquiring malignant characteristics, wTould
transmit those characteristics to its progeny. This
would be a "mutation" — an acquired characteristic
suddenly becoming hereditary for all succeeding genera-
tions; an event which we think is most unlikely to

1It must be remembered that these organisms may have nothing to do
with either sepsis or suppuration.

370 THE PROLIFERATION OF CANCER

occur. It is difficult to imagine how a cell, having
started augmented divisions in response to a combina-
tion of alkaloid and auxetics, could in its subsequent
generations continue to divide by augmented divisions
when the cause of the augmentation is absent. We
have shown experimentally that if the supply of auxetic
to a cell ceases, the cell-division also ceases. This
experiment tends to dispose of the expressions "first
(heterotype) divisions and subsequent (homotype)
divisions," which in reality imply a mutation. We
think, therefore, that a metastatic growth consists of
a portion of the primary one transplanted elsewhere
along with some of the original cause of its augmented
proliferation.

It is possible that in the later stages of cancer the
body-fluids may contain considerable amounts of alka-
loid, derived from the primary growths, which might, in
the event of a fresh healing focus occurring anywhere,
be sufficient to act in combination with the new local
auxetics, and so cause another "primary" growth. If
such occurred, it would probably be mistaken for and
called a secondary growth.

Lastly it may be mentioned that if cancer is due to
putrefaction occurring in a chronic healing site, there
may be something in the view upheld by many, that
the disease occurs frequently in certain localities or
even in certain houses. Doubtless putrefaction will
occur more readily in certain places, because the
bacteria of putrefaction may infest the air there. In
connection with this I may recall the remark — al-
ready noted — which was made to me by Sir William

EXPERIMENT REQUIRED 371

MacGregor, that he had never seen a case of cancer
among the Esquimo.

The "error of random sampling," however, must
be considered with the question of the "local inci-
dence" of cancer. Very large figures would have to
be studied before one could say conclusively whether
the incidence of the disease is actually greater in some
localities than in others, and experimentation with
animals in the confines of the laboratory cannot, we
think, determine whether putrefaction is more likely
to occur in one place than in another. Still, the
remark of Sir William MacGregor is striking, because
it is clear that putrefactive bacteria cannot be present
to so great an extent in the Arctic regions as in
temperate and tropical climates.

The above consideration led us to believe that our
researches did harmonise with the facts known about
carcinoma. The fact that cell-proliferation is caused
by auxetics contained in the soluble remains of dead
tissues offers for the first time an explanation of
the remarkable age-incidence of the disease; and the
augmented asymmetrical division induced by these
auxetics combined with alkaloids of putrefaction
seemed to be a reasonable explanation of the cause
of cancer. Proof wras wanting, however. Cancer-cells
have been seen frequently to divide by asymmetrical
divisions, but because one can induce these mitoses in
cells is not proof that one is necessarily inducing
malignant proliferation.1

1 As a matter of fact, the five-fold augmentation by alkaloids is a more
important consideration than the asymmetrical mitoses induced by them.

372 THE PROLIFERATION OF CANCER

Deductions from experimentation in vitro, no
matter how well they may harmonise with know^n
facts, are not sufficient to act as a basis on which to
proceed to find the prevention and cure for the disease.
It is necessary at least to try to prove one's work
definitely. To accomplish this would not be, we knew,
an easy matter. It would be necessary to produce
a cancerous growth in healthy animals with the sub-
stances which were believed to be the cause of the
disease. The chemical auxetics, in correct combination
with an alkaloid of putrefaction such as choline, would
have to be inoculated into or applied to an animal, and
before one could say that the combination is a cause of
cancer a malignant growth would have to appear at the
site of inoculation. The experiment would have to be
frequently repeated, and careful precautions would have
to be taken against possible fallacy.

It was realised that it would be quite useless merely
to inoculate a solution of, say, kreatin and choline sub-
cutaneously into an experimental animal, because it is
obvious that such a solution would rapidly be excreted,
and we know from in-vitro experimentation that before
a cell can divide, either by a normal or an asymmetrical
division, it must be subjected to the chemical agent for
a certain length of time. It would be necessary to
create a sore, because a chronic healing site is essential;
and this would not be readily accomplished in experi-
mental animals, which are not easy to keep quiet, and
in which the local application of substances to sores
offers practical difficulties.

EXPERIMENT REQUIRED 373

Moreover, the question whether the lower animals
suffer from true cancer is still controversial. I
therefore considered whether it would be possible to
try this crucial experiment on a human being. If it
were possible, and if it were successful, the point might
be proved conclusively. At first sight the suggestion
seems to be an outrageous one, but the experiments to
be related in the next and last chapter, which had been
carried out for several months past, revealed a method
by which I considered that an attempt might be made
to put this crucial experiment to the test.

CHAPTER XVII

INHIBITORY ACTION OF BLOOD-SERUM ON AUXETICS—

MEASUREMENT OF THIS ACTION THE TREATMENT

OF SOME CASES OF CANCER BY THE ADMINISTRA-
TION OF DEFIBRINATED BLOOD DESCRIPTION OF

THE CASES THE TREATMENT OF A MALIGNANT

ULCER BY MEANS OF GLOBIN AN ATTEMPT TO

MAKE THE CRUCIAL EXPERIMENT— CONCLUSION

IT is now (August, 1910) more than six months since
it was ascertained that leucocytes and lymphocytes
divide in response to the auxetics contained in the
remains of dead tissues and in globin. When this
fact was appreciated, the question arose as to why
these cells, when they are removed from the peripheral
circulation, had never been seen- in the act of cell-
division. White blood-corpuscles were discovered by
Hewson in 1773; in 1846 Wharton Jones first described
them as granular and nucleated cells (Buchanan).
Since then they must have been seen by every student
of medicine, but no one, until divisions were induced
in them by us, had ever seen one of these cells divide.

374

NO MITOSIS IN BLOOD STREAM 375

Hence it is obvious that these cells do not divide in
the peripheral circulation, for their mitosis occupies
a certain amount of time; and if this mitosis occurred
at all in the peripheral blood, it must have been seen
during the century and a half in which these cells have
been constantly examined by many thousands of workers.
Now, the division of these cells is caused by the
auxetics contained in the remains of dead tissues and
in globin, and it also is certain that the peripheral
blood must contain some free remains of dead tissues
and globin. Hence white blood-corpuscles ought to
be frequently seen in the act of division when they are
removed from it. But they are not so seen. Had it
been seen, the real nature of the Altmann's granules
and the "lobes of the nuclei" would have been apparent
many years ago.

We think that there can be only one explanation
for this, which is that the action of the auxetics in the
peripheral blood is restrained in some way. It appears
to us to be reasonable to suppose that cell-proliferation
in the peripheral circulation must be prevented in some
way. If it were not, the approach of old age or a
chronic suppurative focus with destruction of tissue
might cause indiscriminate cell-proliferation in the
vessels and capillaries, with disastrous results ; for these
vessels might ultimately become blocked. We there-
fore made some experiments to see if blood-serum does
actually restrain cell-division.

. In the first place, 2 cc. of sheep's serum was added
to auxetic jelly composed of azur dye, atropine, and
suprarenal extract. In order to prevent coagulation of

376 PREVENTION OF PROLIFERATION

the serum in this and the subsequent experiments, the
serum was added to the jelly after the latter was boiled
and before it had cooled to such an extent as to prevent
it setting on the slide. It was found that the serum
did not prevent the cell-division induced by the azur
stain.

The experiments were then repeated with a jelly
which contained suprarenal extract, but no stain or
atropine. The jelly was first tested, and mitotic figures
induced in lymphocytes with it. The jelly contained
0.2 gramme of suprarenal extract, and it was found
that if it also contained 0.5 cc. of serum the auxetic
action of the extract was not stopped; but if it con-
tained 2 cc. of serum the auxetic action of the
suprarenal extract was completely inhibited.

The experiments were then repeated with an auxetic
jelly composed of a mixture of 1 cc. of a 1-per-cent
solution of kreatin and 1 cc. of a 1-per-cent solution
of choline; and it also contained 10 units of alkali
solution. With this jelly it required the addition of
2.5 cc. of sheep's serum, to prevent it causing cell-
division.

Using human serum, it required 2 cc. of it to stop
the action of 0 . 2 gramme of suprarenal extract by itself.
1 cc. of serum will stop the action of the combination of
.0.01 gramme of kreatin and 0.01 gramme of choline;
1 cc. of human serum will stop the action of 0 . 5 cc. of a
2-per-cent solution of globin which had been allowed
to become putrid, and which would, by itself, induce
division in lymphocytes.

Hence it is apparent that normal blood-serum

CELL-DIVISION RESTRAINED 377

actually has the power of preventing the "natural"
auxetics from inducing cell-division; but it has no
inhibitory action against atropine or azur dye. The
restraining power of serum can be measured as shown,
and it is possible that this power varies with individuals,
a point which remains to be determined.

It was also ascertained that the restraining body
in serum does not combine permanently with the
auxetic and so prevent its action. Jellies were pre-
pared with suprarenal extract with kreatin and choline,
which induced divisions in lymphocytes. The right
amount of serum was added to them just before the
jellies cooled, and it was noted that they stopped the
auxetic action of the jellies. The same jellies were
then boiled and the serum proteins precipitated. On
making specimens again from these jellies, it was now
found that their auxetic power was re-established.
Hence it is obvious that the restraining body in serum
is not thermostable.

Lastly, it was found out that 1 cc. of serum con-
tained in 10 cc. of jelly which also contained 1 cc. of a
1-per-cent solution of choline stopped the kinetic ac-
tion of the latter in exciting amoeboid movements in
leucocytes. If the jelly was boiled, however, the action
of the choline was restored.

These experiments are very constant in their results.
Careful controls were made throughout. We think
that by means of them the restraining power of
different sera could be measured with a certain amount
of accuracy. What the nature of the restraining body
in serum is we have no opinion to offer. It should be

378 PREVENTION OF PROLIFERATION

noted that some time ago Bashford and Murray showed
that serum had the power of restraining the growth of
secondary transplanted tumours in mice.

In addition to the restraining action of serum
on the cause of cell-division, we also considered
the work of Gaylord and Clowes of the New York
State Cancer Research, Buffalo, and of Bashford and
his assistants at the Imperial Cancer Research in
London, who have shown experimentally that the
transplantation of living growths in mice protect them
to some extent against cancer. It was considered
possible that this might be due to the fresh augmented
auxetic produced by the transplanted growths giving
rise to an increase in the content of the restraining
body. We therefore resolved to try to increase this
body in cancer patients by deliberately injecting them
with augmented auxetic combined with blood-serum.
The way the combination was administered was by
injecting 6 ounces of defibrinated sheep's blood per
rectum every morning. The serum contains the re-
straining body, and it was argued that the red cells
would be destroyed in the rectum, the haemoglobin
decomposed, and in time the globin would become
augmented by the action of the bacteria present. It
was presumed that the restraining body of the serum,
the auxetic in the globin and in the remains of the
white cells, and lastly, the products of the decompo-
sition would be gradually absorbed, and that they might
raise the content of restraining body in the patients ; in
other words they might act as a sort of vaccine.

We must admit that we were not very sanguine of

CLINICAL CASES 379

success when these experiments were first undertaken
six months ago. They were undertaken more with a
view to see what the effect of globin in this way was
than with the object of obtaining a cure of the tumours
from which the patients were suffering. But, as will
be seen from the description of the treatment, the
results have exceeded our anticipations. Unfortunately,
since we did not expect any beneficial results, the cases
were not the most suitable which could have been
chosen, for both of them had "internal" growths which
were inaccessible, and therefore we were at that time
unable to prove conclusively that they were suffering
from carcinoma.

The first patient1 to whom the serum was adminis-
tered was a woman (I. G.) aged 45 (admitted to the
hospital on January 11, 1910), whose left breast had
been removed in November, 1907, for a carcinomatous
growth. She remained well until April, 1909, when she
began to suffer from a severe pain in the region of the
sacrum and left hip. She stated that this pain had
since then become worse and that no treatment had
relieved it. The left lower extremity from the hip to
the ankle had for long been swollen and cedematous,
and there had been swelling also in the abdomen.
Any movement of the limb caused severe pain, and
she had great difficulty in turning herself in
bed. The patient was too ill to be weighed at the
time of her admission, but she was manifestly wasted,

1 These two cases were treated under the supervision of Dr. Macalister,
who has kindly written these descriptions of them.

380

was anaemic, and had a worn expression. On exami-
nation there was manifest swelling on palpation in
the left iliac region, and, examined per rectum under
chloroform, a hard swelling could be felt attached to
the anterior surface of the sacrum and sweeping round
the wall of the pelvis towards the left side. An X-ray
photograph confirmed the involvement of the sacrum.
After treating her with mercurial inunctions and other
remedies for a month without benefit, the defibrinated-
blood injections were commenced on February 21, 1910.
At first they were given in the evening, and were often
followed by sickness and sometimes by actual vomiting ;
it was found that there was less disturbance when the
injections were given in the mornings.1 The sickness
was so troublesome at first that the treatment had to
be abandoned on March 3, and it was not until
March 20 that it was again started and continued
uninterruptedly. Gradually her pains improved, and
the swelling in the leg diminished. (No opiates were
needed after March 29.) On April 20 she could stand
with very little pain, and she was weighed for the first
time (109 lb.). Improvement continued week by
week, she became bright and younger-looking, and
on June 8 she weighed 115 lb. No pain could be
elicited on pressure over the left iliac region, and the
tumour seemed smaller. She maintained her weight,
with some variation, for some weeks, and was able to
walk about the ward without assistance until July 21,

1 In these cases treated with rectal injections of defibrinated blood
there has been sickness following the injection, but this has passed off
as the treatment has been persevered with.

CLINICAL CASES 381

when she sprained her left shoulder and suffered severe
pain in it. On July 19, by the patient's request,
the treatment was discontinued, and an opportunity
thus arose of observing whether the benefit which had
resulted from it was maintained. There had been some
sciatica-like pains in the leg since the beginning of the
month, and during August these increased and the
swelling and pain in the hip returned. Some tender-
ness and a tumour, apparently arising from the medias-
tinum and which grew rather rapidly, appeared in the
mid-sternal region. By the first week in September,
she had relapsed pretty much into the condition in
which we found her at the time of her admission, but
with the added pain due to the thoracic growth. The
treatment has now been resumed.

The second case was that of a woman aged 54,
who had suffered from indigestion for a considerable
period, but severely for three months. There had been
much vomiting, but never any blood. At the time of
her admission (February 9, 1911) ingestion of food
was immediately followed by severe pain, and often
by sickness. She was very wasted, worn-looking,
and anaemic. Weight 94 Ib. On examining the ab-
domen a swelling could be seen and felt above the
umbilicus. It was about the size of a tangerine orange.
It was extremely tender, and moved with respiration.
The stomach was very dilated, and presented peris-
taltic movements. There was pain on pressure over
the pyloric region, but no tumour could be felt there.
The stools contained altered blood (melcena). During
the first fortnight after admission, when she had milk

382 PREVENTION OF PROLIFERATION

and Benger's Food, the vomiting ceased, and she had
less pain, but she lost four pounds in weight (90 lb.).
The defibrinated blood was commenced on February 21,
and until March 23 the weight fluctuated between
88 lb. and 91 lb., there being an occasional gain
and then a corresponding loss; but on March 30
a steady advance commenced, the maximum weight
being attained on May 8, when it reached 101 lb.,
i.e. a gain of eleven pounds since the time of her
admission. From the time of the commencement of
the defibrinated-blood treatment she steadily improved.
She became able to eat fish and a light ordinary diet
without discomfort; but the most striking fact was
the diminution in size of the tumour, which practically
disappeared. As in the former case, after reaching a
climax there was a recrudescence of the symptoms,
and some loss in weight, but the tumour did not
return. The defibrinated blood was omitted on July
20, when she weighed 95 Jb. Subsequently she
mended, and left the hospital on August 9 consider-
ably better and weighing 100 lb. There was un-
doubtedly some real improvement in this case, and
the temporary relapse depended to some extent on
fermentative changes taking place in the dilated
stomach.

Several other cases have been treated with the
defibrinated blood, and in some of them there has
been apparent benefit, although others (a case of very
advanced cancer of the liver, and one of peritoneal
cancer) have not shown improvement.

In addition to the rapid reduction in size of the

CLINICAL CASES 383

growths in these two cases, a striking point was the
improvement in the general symptoms and appearance
of the two patients. Their cachexiae practically dis-
appeared, they became cheerful and seemed to get
younger. In the first case the disappearance of the
oedema of the legs was most remarkable, and never
before had we seen cases of carcinoma, which had been
bed-ridden for months previously and condemned by
surgeons as being inoperable, become able to be up
and about apparently vastly improved in health. It
must be distinctly understood, however, that we do
not assert that this treatment is in any way a cure
for the disease. As mentioned at the outset of the
description of the cases, we have no absolute proof
that they were cases of carcinoma, and it must be
remembered that spontaneous improvement and cure
in some cases of cancer have undoubtedly occurred
without any ^treatment whatever. Gaylord and Clowes
have collected a series of these cases.1 Moreover,
we have been able to deal with only a few cases, and
they have been under observation for only six months;
therefore we cannot say whether the results are going
to be permanent or even maintained for any length of
time.

The reason why these cases are described is that they
suggested to me a possible way in wrhich the crucial
experiment, mentioned at the end of the last chapter,
could be carried out. It appeared reasonable that if
one can cause the reduction in the size of a growth with

1 Seventh Annual Report (Cancer Laboratory, New York State Depart-
ment of Health).

384 PREVENTION OF PROLIFERATION

amelioration of symptoms by general treatment, one
might also be able to improve an accessible growth by
locally inducing the proliferation of healing in it. If
this were possible, and if a local, inoperable, broken-
down scirrhus could be so improved by local treatment as
to replace some of the infiltrating cells by normal ones,
I considered that I should be justified then in carry-
ing out the crucial test on these normal cells, and try to
reinduce the abnormal infiltration amongst them once
more by the direct application of auxetics and choline.
In other words, if a case already has a large "inopera-
ble" tumor and one is able to convert by treatment a
portion of it into normal tissue, it would be useful to try
temporarily to reconvert the normal tissue back into
original condition in order to prove the main point of
our researches. There would already be a neoplasm,
and I proposed thus to test our theory in in vivo on a
portion of it.

A patient suffered from an inoperable, fungating
seirrhus of the breast. The ulcerated surface was about
four inches in diameter. The edges were precipitous
and excavated, and the whole appearance of the ulcer
was typical of carcinoma. The surface was practically
devoid of granulation tissue, and sections of it clearly
showed its nature (figs. 119, 120). A portion of the sur-
face, i.e. about a third of it, was scarified and globin
was applied by being "dotted" over it (fig. 121). The
remaining part of the ulcerated surface \vas untreated.
No dressings were applied. This ulcer had a remarka-
ble propensity for suppurating. No matter what was

LOCAL EFFECTS OF GLOBIN

385

FIG. 119. — Section from the casejsf scirrhus of the breast. Low power.

FIG. 120. — The same as 119. High power.

LOCAL EFFECTS OF GLOBIN

387

FIG. 121. — To show the way in which globin was "dotted" on to a portion
of the malignant ulcer.

LOCAL EFFECTS OF GLOBIN 389

done to try to keep it "clean," pus quickly formed on
its surface — much more quickly than it did on the be-
nign callous ulcers which were also being treated. The
result was that the scab formed by the globin very
quickly broke down. When this occurred, the scab was
removed by fomentations and the ulcer cleaned up as
much as possible. Then the globin treatment was re-
peated, but it was always applied to the same portion
of the ulcer. The other portion never received an ap-
plication. This was repeated many times.

The improvement in the treated portion was gradual,
but it was marked. The precipitous edges appeared
to soften and become flattened. The base no longer
suppurated in a few hours, and the suppuration was
practically confined to the untreated portion. The
glistening malignant surface of the treated portion
gradually gave place to granulation tissue, and after
about a fortnight's treatment there was a contrast be-
tween the treated and untreated portions of the broken-
down surface. A portion of the treated part of the
ulcer was now removed and sections cut from it, which
show that the abnormal cells were now giving place to
normal granulation tissue,

The treatment was continued once more, two parts
of kreatin now being added to five parts of globin,
and soon it was seen that the treated portion became
softer, and the ulcerated edge ceased to extend.
Another section was then cut, which showed that
that part of the ulcer now seemed to be devoid of
abnormal infiltration (figs. 122, 123).

390 PREVENTION OF PROLIFERATION

I considered that the opportunity had now ar-
rived to try the crucial test. A mixture was m.-idt-
of a solution containing five parts of globin and
one part of choline. It was evaporated to dryness
with aseptic precautions, and the dried mixture sealed
up in a glass tube. A minute portion of the edge of
the treated ulcer, from which the last section had been
taken, was nowr scarified and small pieces of the
dried aseptic mixture of globin and choline directly
applied to it. In 48 hours a conical excrescence
appeared at the seat of application. A section has
been cut from it, and the photomicrograph shows
apparently new malignant cells to be infiltrating the
granulation tissue once more in an abnormal manner
similar in appearance to the original infiltration of the
ulcer (figs. 124, 125).

Now, this test is by no means conclusive. I can-
not assert that there were no original carcinomatous
cells at the seat of application, and that I was
not merely producing augmented proliferation of
these cancer-cells. The section may be fallacious
owing to the "error of random sampling," for be-
cause no abnormal cells appear in samples removed
from the treated site it does not prove that none
exist in the neighbourhood. Still, the experiment is
interesting, because the conical excrescence only
appeared at the site of the application of the com-
bined auxetic and the alkaloid of putrefaction, the
rest of the ulcer remaining in statu quo.

This test will have to be repeated many times
before one can speak conclusively on the subject;

ATTEMPTED CRUCIAL TEST

391

•m:-

FIG. 122. — Section of a portion of the ulcer after treatment. Low power.

FIG. 123.— The same as 122. High power.

ATTEMPTED CRUCIAL TEST

393

FIG. 124. — Section of the treated portion of the ulcer after the application
of globin augmented by choline, showing reinfiltration. Low power.

FIG. 125. — The same as 124. High power.

ATTEMPTED CRUCIAL TEST 395

but from the general appearance of the ulcer,
as well as from the microscopical section of it, we
certainly think that it was the mixture of globin
and choline which caused the reinfiltration. The
application of globin alone to the edges of the
sore merely induced the appearance of granulation
tissue, and one would think that if a mixture of
globin and choline does not cause carcinoma, it
would merely have produced augmented proliferation
of the healing cells writh more granulation tissue;
but apparently a new infiltration of epithelial cells
appeared. I hope to be able to repeat this test
under more favourable conditions.

The whole ulcer is now being treated with a
mixture of globin and kreatin, and, although the
edges of it are extending in some places, there can
be no doubt that, on the surface at least, malignant
cells are being replaced by normal granulation tissue.
The whole growth is now comparatively freely mov-
able, and it does not discharge profusely as it did.
The patient no longer complains of pain, and, except
for the extending edges, her general condition has
greatly improved.

As it is possible that carcinoma may be due to
the causes described in this book, and since the
general treatment by defibrinated blood per rectum
and the local treatment by globin and kreatin seem
to have been followed by improvement, we think
that the former might be tried to prevent recurrence
after removal of a growth. Unfortunately we are
not in a position to try this experiment, as early

396 PREVENTION OF PROLIFERATION

cases which have been operated on do not come to
our notice; we therefore take the liberty of suggesting
that the prevention of recurrence might be under-
taken by those who can watch a series of these cases
which have been operated on. We cannot say, of
course, whether recurrence will be prevented by rectal
injections of defibrinated blood, but the treatment
is harmless and it appears to be worthy of a trial.

If cancer is due to the causes which we think it to
be, the reason for recurrence after removal of the
original growth, which occurs in some cases in the
operation scar, is open to two explanations: (1) that it
may be due to portions of the original growth which
have not been removed, and (2) that the healing site,
although healing occurs by first intention, is a fruitful
source of auxetics, and that the operation wound may
easily become infected by the putrefactive organisms
during the removal of the original growths. As already
pointed out, certain putrefactive bacteria do not neces-
sarily also cause suppuration; and therefore recurrence
in the scar may be due to a fresh attack of cancer there.
The proliferation of healing (even in a site healing by
first intention) probably continues for weeks if not for
months after the injury, because the initial proliferation
increases the number of deaths, and possibly it is a long
time before normal elimination is sufficiently restored to
put a stop to the abnormal proliferation.

One frequently hears of cases in which "recurrence"
takes place perhaps ten years after removal of the
original growth. This must be due to a fresh attack of
cancer. One of the commonest sites for it is in the

THE RECURRENCE OF CANCER 397

rectum, the very place where one would expect it to
occur. Now that we know the causes of cell-prolifera-
tion it is difficult if not impossible to believe that a
metastatic growth could remain malignant and quiescent
for ten years without proliferating. We think that the
increase of "restraining body" conferred on a person by
a malignant growth may not last very long, and this,
coupled with advancement of old age and possibly the
existence of physiological excess of general aiixetics
which may occur in some persons, may predispose them
to subsequent attacks of cancer. Our argument is that
cancer is due to a combination of physiological auxetics
and pathological alkaloids of putrefaction. The combi-
nation must be a definite one, or it will not be effectual ;
it must diffuse into the cells to a certain extent for a
Certain length of time, with due regard to the coefficient
of diffusion of the cells; and lastly the vitality of the
cells themselves must not be greatly impaired. We
think that unless all these factors are in correct combina-
tion, malignant disease cannot occur. »

With regard to the cause of sarcoma, we think that
it is probable that the auxetic chiefly concerned in that
disease is that contained in globin. Several 'surgeons
haVe kindly informed us that in almost every case of
sarcoma which they have seen there is .a history of
injury; and it is remarkable that sarcoma occurs .most
frequently in those tissues which are rich in hsemo-'
globin, namely, the choroid coat of the eye (melanotid
sarcoma), the bone marrow, and the neighbourhood of
muscles. The suggestion that globin is the source
of the auxetic in sarcoma will explain the age-incidence

398 PREVENTION OF PROLIFERATION

of the disease; for it probably only follows injury to
large numbers of red cells. The length of life of red
cells in the body is supposed to be only a matter of a
few weeks, so that their anabolism and katabolism is
continuous, and may not depend at all on the age of the
person. Hence sarcoma may occur at any age.

Whether the alkaloids of putrefaction are concerned
in sarcoma or not, we are not in a position to state, but
interesting cases have been reported from time to time
which were associated with suppurative foci. Quite
recently a case was described in The British Medical
Journal,1 of an infant which had been injured in the
neck by forceps at birth. Sarcoma followed on the
injury, which was also complicated by suppurative otitis
media.

The possibility of the alkaloids in both sarcoma and
carcinoma being of the nature of leucomaines which are
supposed to be absorbed from the intestines must not
be forgotten.

The proliferation of leucocytes and lymphocytes
in the leukaemias are also doubtless due to auxetics.
Whether these diseases are caused primarily by injury
to the spleen or not we do not know, but it is possible
that this starts thfe proliferation. The spleen tissue has
direct access, by means of the vessels, with the per-
ipheral circulation, and presumably this is the reason
for the leucocytosis and lymphocytosis in leukaemia.
It is impossible to say whether the proliferation of
leukaemia is of the augmented type, or whether an

1 "A Case of Sarcoma »f the Pectrous Bone," by W. H. Bowen and H. B.
Carlyle (B.M.J., June 25, 1910.)

CONCLUSION 399

alkaloid is present; but we may recall the interesting
fact mentioned by Buchanan in his admirable book on
the clinical pathology of the blood,1 that he had noticed
the discard of granules (flagellation) in the cells of cases
of leukaemia. Possibly the leukaemias may be associ-
ated with the auxetic contained in globin, for the spleen
is a very vascular organ; and if so, it may ultimately
be found that leukaemia is a from of sarcoma of the
spleen.

In concluding these descriptions of the researches
which we have been able to carry out to the end of
the first year and a half of the establishment of the
Research Department of the Royal Southern Hospital
at Liverpool, I wish once more to acknowledge my
indebtedness to all those who have helped me so
materially. I think that the new methods at our dis-
posal have been the means certainly of solving the
problem of the cause of normal human cell-division,
and possibly, if not probably, of the cause of malignant
cell-proliferation also. Much work remains to be done,
however, some of which has already been started.

A series of more than ten "inoperable" cases of
cancer are now being treated by defibrinated blood and
by the local application of normal auxetics. Experi-
mentation is begun to ascertain what organisms produce
substances which "augment" the action of auxetics.
The strength of the body in normal serum which
restrains cell-division is being measured with a view

1 R. J. Buchanan, The Blood in Health and Disease (Oxford Medical
Publication).

400 PREVENTION OF PROLIFERATION

to see if it varies in different persons, both in normal
and in pathological conditions. The lengths of the lives
of leucocytes are being measured in the presence of
various strengths of auxetics and alkaloids of putre-
faction. Many fields of work are now opened by the
knowledge that the reproduction and multiplication of
the cells of our bodies are due to certain known (and
some as yet unknown) chemical agents. The knowledge
that " healing" itself is caused by these agents may ulti-
mately assist the medical man in his work, and I think
that it will be found that trypanosomes amoebae (the
causes of dysentery), and other parasites also multiply
in the body in response to the remains of dead cells.
These paths of research will require many workers, and
I am sure that their investigation will not be wasted.

Whether the benefit derived from the treatment
adopted in two of the cases of cancer will prove to
be of practical value or not remains to be seen. In
any case it is capable of elaboration and further investi-
gation. Even if it confers the smallest amelioration
of symptoms, which it undoubtedly appears to have
done in these two cases, something has been accom-
plished; but whether the benefit is lasting or not time
alone will show.

APPENDIX I

ENUMERATION OF THE NUMBER OP GRANULES CONTAINED IN
EOSINOPHILE LEUCOCYTES

TABLE I1

CONTROLS (healthy and diseases other than cancer). Males

Name.

Age.

Disease (or Health).

Number of
Cells Counted.

Total Number
of Granules
Counted.

Number of
Granules in
Smallest Cell.

Number of
Granules in
Largest Cell.

u
9

is.

d'gS

*go

» C fc,

g» 8,

<U o

Connolly . . .
Parker ....
Edwin

12
14
15
15
15
17
20
21
22
24
24
25
25
25
26
26
27
28
28
32
34
34
34
36
46
47
50
52
60
62
65
66
86

Chorea

5
1
1
1
2
5
2
4
2
1
2
1
3
3
1
1
2
5
21
5
2
4
7
2
5
5
2
6
2
2
2
1
1

811
174
134

180
536
779
291
561
354
227
368
155
738
502
82
175
475
798
3,390
745
325
769
1,086
297
748
840
345
875
318
492
336
165
211

142

175

162
174
134
180
268
156
145
140
177
227
184
155
246
167
82
175
237
160
161
149
162
192
155
148
150
168
172
146
159
246
168
165
211

Healthy

Brewer ....
Hughes ....
Mattison . . .
Duncan . . .
Bradley . . .
Holding . . .
McDonald .
Stevens ....
Ketch

Mitral Disease ... . .

Osteo-arthritis

266
135
141
124
177

169

270
185
150
147
177

199

Malaria

Filariasis

Pneumonia

Healthy

Fracture

Sarcoma . ...

Varicocele . ...

May

Fracture

204
101

287
220

Ball

Pneumonia

Armstrong .
Mahoney . .
Berrv .

Sleeping Sickness . . .
Sarcoma

Floating Kidney ....
Acute Nephritis ....
Healthy

197
125
114
127
146
155
127
105
122
137
151
124
145
230
147

278
200
260
165
179
208
190
192
192
206
194
190
173
262
189

Hankinson .
Cropper . . .
McConnell .
Grue

Fracture ....

Hernia ....

Jones

Healthy ....

Ross

Healthy . . .

Smith

Fracture . . .

Ritchie ....
Morgan ....
Daulby ....
Noble

Hernia

Hydrocele

Empyema

Addison's Disease . . .

Stricture

Braig

Cann . .

Varicose Ulcer

Gould

Healthy

Lowry

Chronic Rheumatism
Healthy

Benn

TOTAL

109

18,282

Averag

e!68

1 In the averages fractions have been neglected throughout.

401

402

APPENDIX I

TABLE II

CONTROLS (healthy and diseases other than cancer) . Females

Name.

Age.

Disease (or Health).

Number of
Colls Counted.

Total Number
of Granules
Counted.

Niiinberof
(iraiuiles in
Smallest Cell.

Number of
Granules in
Largest Cell.

Average Number
of Granules
per Cell.

Shankayne .
Frost

19
21
22
24
27
35
38
56
56
65
90

Chorea

5
6
1
2
1
6
2
2
1
2
2

5
5
2

890
957
117
334
136
994
412
365
201
338
397
801
724

; 390

155
117

155

199
212

179

178
159
117

167
136
166
206
182
201
169
198
160
145
195

Peritonitis

Matthews . .
Farrington .
Simpson . . .
Stone . . . .

Hysteria

Healthy

Osteo-arthritis

Chlorosis

146
168
172

201
244
193

Francis ....
Baker

Healthy

Lymphadenoma ....
Carbuncle

McKey . . : .
Jackson . . .
Harris
Swalwell . . .
Wilson ....
Benn

Myxoma

163
172
120
127
186

175
225
192
165
204

Healthy

Osteo-arthritis . . .

Hernia

Healthy

TOTAL

j 7,056

Averag

e!68

APPENDIX I

403

TABLE III
CANCER CASES

A. Males-

Name.

Age.

Locality of Disease.

Number of
Cells Counted.

Total Number
of Granules
Counted.

Number of
Granules in
Smallest Cell.

Number of
Granules in
Largest Cell.

Average Number
of Granules
per Cell.

Doyle

Stomach

699

113

166

140

Mackie

39!

Lung (Secondary) . .

340

129

211

170

Rhead

Testicle

1,036

104

187

148

Ya Foo ....

Penis

1,201

132

197

171

Nesborough

Lit) . .

468

132

199

156

Donahern . .

Sigmoid

835

106

175

139

Gardiner . . .

Stomach

150

Welsh

Penis

1,013

. . .
132

196

169

Whelan ....

Tongue

928

140

234

186

TOTAL,

6,670

\verag

e159

B. Females —

Duncan . . .
McCann . . .

Evans . .

35
41
42
42
45
45
49
54
58
56
56
56
66

, Stomach

2
2
1
1
6
1
2
5

3
1
6

357
323
144
217
1,201
119
313
767
953
1,046
387
120
804

156
121

201

202

178
161
144
217
200
119
156
153
191
150
129
120
134

Liver

Breast

McQuillian .
Griffiths ....
Jones . .

Uterus

Breast and Pelvis . . .
Breast

163

273

Hiles

Breast

146
147
121
109
96

167
158
247
197
159

Walker ....
Griffiths ....
Griffiths....
Roberts . . .
Hall . .

Stomach

Cervix Uteri

Breast

Stomach

Breast

Cunning . . .

Breast

158

TOTAL

6,751

Averag

e!61

APPENDIX I

TABLE IV

AVERAGE NUMBER OF GRANULES IN (A) LARGEST, AND (B)
SMALLEST CELLS

1. Males and Females separated —

Controls. Controls.
M ales . Females .

Table I. Table II.

Cancer. Cancer.

M ales . Females .

Table III. Table III.

(A) AVERAGE number of

granules in largest cells

204

199

196

(B) AVERAGE number of

granules in smallest cells

150

153

124

127

2. Males and Females combined —

Average Number of
Granules in Largest Cells.

Average Number of
Granules in Smallest Cells.

Controls

202

151

Cancer

196

126

It should be noted that the greatest difference between Cancer
and Control cells is in the smallest leucocytes. ,

APPENDIX I

405

SUMMARY

Total number of persons examined 69

Total number of cells photographed 235

Total number of granules counted 38,759

Table showing differences between the cells of Control (healthy and diseases
other than cancer} persons and Cancer persons

Persons
Examined.

Cells
Photographed.

Granules
Counted.

Average
Granules
per Cell.

Controls

151

25,338

168

Cancer

13,421

160

Males—

Table showing Influence of Sex

Persons
Examined.

Cells
Photographed.

Granules
Counted.

Average
Granules
per Cell.

Controls

109

18,282

168

Cancer

6,670

159

Females —

Controls

7,056

168

Cancer

6,750

161

Number of granules in smallest cell, 82. Number in largest cell, 287.

Variation in the number of granules contained in the cells of
one person

21 cells from Cropper. Smallest cell contained 114 granules;
largest contained 260. The average number of granules in the 21
cells is 161.

APPENDIX II1

SOME COMPARATIVE MEASUREMENTS OF THE LIVES OF LEUCO-
CYTES2 WHEN THE CELLS ARE RESTING IN THE PLASMATA
OF DIFFERENT PERSONS

AND THE POSSIBLE APPLICATION OF SUCH MEASUREMENTS AS AN AID TO
DIAGNOSIS IN INFECTIVE DISEASE

OF recent years I have been endeavouring to ascertain the effect
produced by one person's plasma on the life of another person's
leucocytes. It appeared reasonable to suppose that the plasma of
a person suffering from an infective disease would be poisonous to
the leucocytes of healthy persons. If this is the case it might also
be reasonable to suppose that the same plasma would not be so
poisonous to the leucocytes of another person suffering from the
same disease, because it is probable that the cells would be already
used to, or immune against, the toxin ; and furthermore that if the
toxin of one infective disease differs from the toxin of another
infective disease, it might be inferred that an immunity on the
part of a leucocyte against one disease will not render it immune
against another. Therefore, provided it is possible to tell accurately
when a leucocyte is dead — that is, if one can differentiate a living

1 A method for estimating the number of living and dead leucocytes con-
tained in a given sample of blood; and another convenient formula for the
preparation of "kinetic jelly." Being a paper reprinted from The Lancet of
February 6, 1909, by kind permission of the and Editor of that Journal.

2 The word "leucocyte" refers to the neutrophile polymorphonuclear
leucocyte.

406

APPENDIX II 407

from a dead cell — it also will become possible to measure the
lengths of the lives of leucocytes after they have been removed
from the body. And this will enable us to make comparative
measurements of the lives of leucocytes when they are mixed with
the plasmata of different persons. Supposing, therefore, it is true
that an infected plasma shortens the lives of a healthy person's
leucocytes but does not shorten the lives of the leucocytes of another
person suffering from the same disease, it may be useful to reverse
the process and assist in the diagnosis of infective disease by making
measurements of the lives of such a patient's leucocytes when they
are mixed with dfferent plasmata. For instance, if the leucocytes
of a person suffering from an indefinite infective disease are found
to be easily killed by the plasmata of persons suffering from a
variety of diseases, but are not comparatively easily killed by the
pasma of a person suffering from, say, typhoid fever, it might be
inferred that the patient is suffering from, or has recently suffered
from, typhoid fever, because his leucocytes are used to, or immune
against, that disease.

The above is the enunciation of a problem which I set myself
to solve several years ago, and this paper describes the experiments
which have been conducted to investigate the last part of it — i.e.
with the object of determining the actual measurements of the lives
of leucocytes when they are placed in the plasmata of people who
are suffering from various diseases. The earlier researches made
in order to differentiate living from dead leucocytes have already
been published in the Journal of Physiology (I),1 and the actual
method employed to estimate how many living and how many
dead cells there may be in a given volume of citrated blood has been
described in The Lancet of January 16, 1909 (2). This method
may be again briefly summarised thus:

Method for counting the number of living and dead leucocytes in
a given sample of citrated blood. — The following solutions are pre-
pared and a jelly is made from them. 1. A volume of Unna's
polychrome methylene blue (Grubler) is diluted with two volumes
of water. 2. A solution containing 2 per cent of agar in water,

1 The figures within parentheses refer to the bibliography at the end of the
article.

408 APPENDIX II

filtered and sterilised. 3. An accurately neutralised solution con-
taining 4 . 5 per cent sodium citrate, 1 . 5 per cent sodium chloride,
and 0.225 per cent atropine sulphate. 4. A 5-per-cent solution
of sodium bicarbonate. In a test-tube mix one cubic centimetre
of the diluted stain, two cubic centimetres of the citrate solution,
and three cubic centimetres of the molten agar solution. To
this mixture a quantity of the alkaline sodium bicarbonate
solution must be added in order to cause the excitant for leucocytes
contained in the jelly to diffuse into the cells, and the quantity
added varies with the temperature of the room.1 If measurements
are going to be made in a room with a temperature of between
60° and 70° F., about 0 . 25 cubic centimetre of the alkaline solution
should be added. The mixture is then boiled until it froths up
the tube and a drop poured on to a slide and allowed to set so as to
form a film. Supposing a given capillary tube contains the blood-
corpuscles of one person mixed with the plasma of another, the
average number of living and dead leucocytes in the tube can be
estimated by placing a drop of its contents on a cover-glass which
is inverted and allowed to fall on the agar film. After two or
three minutes the granules but not the nuclei of the living leuco-
cytes will stain and those cells will show exaggerated amoeboid
movements, whereas the dead cells will remain immobile. More-
over, the dead cells may be achromatic (3), in which case they will
not stain. Their nuclei may appear as a single nuclear mass,
or their nuclei may even stain, or the dead cells may have under-
gone other changes which have been described in former papers
(1, 2). Field after field should be rapidly passed in front of
a l-6th inch or equivalent objective and the number of the
living and dead cells counted. Several preparations can be rapidly
examined and an average struck so as to give an estimate of the
number of living and dead cells in the given capillary tube. No
difficulty is met with in making the counts, for living can be
readily differentiated from dead cells by the presence or absence of
exaggerated movements .

• If all the leucocytes appear to be dead, and especially if the
agar jelly has not previously been tested, it is as well to control
1 A scale has been given in the former paper.

APPENDIX II 409

the measurement — that is, to see that the jelly will actually excite
living cells — by placing a drop of fresh citrated blood on to another
part of the same film and noting whether stimulated movements of
all the leucocytes occur.

Procedure for the preparation of capillary tubes containing the
plasma of one person and the leucocytes of another.— It will simplify
description if the details of sterilisation and the precautions for
ensuring asepsis are omitted. Since the presence of bacteria
shortens the lives of leucocytes (2) it is obvious that aseptic pre-
cautions are essential, but the details for sterilisation are so well
known that they need hardly be repeated. A capillary tube of
glass is prepared which has such a diameter that blood will run
into it by capillarity and at the same time its flow can be controlled
by gravity. I use a tube with a lumen of about two millimetres.
15 portions equal to each other are marked off with a pencil. The
marks begin at one end of the tube which is zero, but the tube
is at least two inches longer than mark 15. The portions are
rendered equal by calibration with mercury, and although the
length of each is immaterial, I have found that about half a centi-
metre is a convenient length for practical purposes and I use a tube
about 13 centimeters long. A neutral solution is made which
contains 3 per cent of sodium citrate and 1 per cent of sodium
chloride. Some of this is drawn up into the .tube until its upper
limit or meniscus stands at mark 6. Blood from the finger of the
person whose plasma is going to be tested is added until the
meniscus stands at mark 12, care being taken that no bubble of
air separates the two fluids. Mixture is carried out by allowing
the two fluids to gravitate up and down the tube six times. The
tube is sealed and centrif ugalised ; the blood being driven towards
zero. The end remote from zero is then unsealed and the portion
containing the precipitated corpuscles is separated and discarded
by cutting the tube at 4. Eight portions of the tube now contain
citrated plasma. If, owing to the sealing process, much of the tube
has been occluded at zero the upper meniscus may stand above
mark 12. This can be corrected by tapping out the excess of fluid
on to a sterile slide, controlling the amount removed by the finger on

410 APPENDIX II

the end remote from the mark 4. The lower meniscus standing at
4 where the tube has been cut, and the upper meniscus standing at
12, blood from the finger of the persons whose corpuscles are going
to be tested is added until the upper meniscus stands at 13 (i.e. the
mixture equals 1-9). Mixture is ensured as before and the tube
sealed. It will be seen that although the tube contains the plasma
of both persons the corpuscles are bathed in a solution containing
four times as much plasma of the first person as of the second. A
series of tubes may thus be made.

Appliance to ensure continual mixture and to prevent the cor-
puscles from adlierincj to the glass.- — If a capillary tube prepared
in the way which has been described is laid on a flat surface, the
corpuscles will soon gravitate to the most dependent side and will
ultimately adhere to the glass. The following appliance prevents
this. By means of a simple clockwork movement a split drum is
made to revolve once in about three minutes. The drum is so
adapted thaT the mouth of a long test-tube (having a diameter of
one centimetre and the cavity of which is lined with a roll of blot-
ting paper) fits accurately on to it and revolves with it. The
apparatus is so arranged that the tube is horizontal and is of such
a size that it can be placed in the incubator if necessary. The
capillary tubes inserted into the test-tube are continually tumbling
over each other by gravity as the test-tube revolves, and in so doing
revolve themselves. The blood-cells in their turn are continually
gravitating in different directions through the citrated plasmata.
It has been found that this device prevents them adhering to the
glass and ensures them being evenly distributed through the
citrated plasmata provided the ends of the capillary tubes are not
bent over when sealed. This apparatus also insures all capillary
tubes being subjected to the same conditions of temperature.

Procedure for measuring the lives of tlie leucocytes contained in
the tubes. — Samples of the contents of the capillary tubes are
examined on stimulating agar by the method already described.
If all the cells are alive the tubes are resealed and returned to the
revolving apparatus to be examined later, and so on. By this
means the percentage of living and dead cells in a tube can be

APPENDIX II 411

estimated. It is important to remember that in striking these
averages only an approximate estimate can be obtained, and that
therefore the greater the number of tubes made the better, as the
error decreases with the greater number of leucocytes counted. In
the experiments which I am about to record I have counted about
500 leucocytes in each case by making five films from each of five
tubes, and counting about 20 leucocytes in each film. Since it is
obvious that the greatest error may occur when the number of
living approximates the number of dead cells in a tube, the follow-
ing experiments would appear not to be very erroneous, judging by
the application of Poisson's formula, which shows that supposing
there are half a million leucocytes in the five tubes, which is an
excessive estimate, a count of 500 cells would give a possible error
of not more than about 6 per cent, even when the numbers
approximate.

Before enumerating the actual measurements there is yet
another question to be considered, a point upon which I wish
to lay great emphasis — namely, that all measurements of the
lives of leucocytes should necessarily be comparative. For instance,
it would be fallacious to say that a typhoid plasma killed a person's
leucocytes more rapidly than a septicsemic patient's plasma, when
the typhoid measurement was made to-day and the septicsemic
measurement made three days ago; for even if there was a great
difference in the length of the lives and the same person's leucocytes
were used, one cannot say that that person's leucocytes were in the
same state to-day as they were three days ago, although the person
is apparently in the same healthy condition.

Again, I have shown (4) that the factor Jieat in accelerating
the diffusion of substances into cells also materially affects the
lives of the leucocytes, since the cells are necessarily resting in a
citrate solution which is itself poisonous to some extent, and even
the temperature of incubators is variable. It is thus of the utmost
importance that when the lives of a person's leucocytes, which have
been placed in the plasma of a person suffering from an infective
disease, are measured, a simultaneous measurement of the same
leucocytes shed at the same time must be made in the plasma of
a healthy person. And it is only by the difference between the

412 APPENDIX II

two that the result can be determined. In other words, all
measurements must be simultaneously controlled by other measure-
ments and the contrast is the result. It is also obvious that since
heat and the citrate solution both affect the lives of the cells, all
tubes, whether containing infected or control plasma, must be
subjected to the same conditions as regards temperature. And it
is essential that the same citrate solution must be employed both
for the test and the control. Unless these essential details are
adhered to, any measurements may be considered to be worthless.
Leucocytes are very sensitive to changes in temperature when they
are resting in citrate solution, but if a change occurs and all tubes
are subjected to 'the same change the contrast in the length of life
holds good. The most favourable arrangement of the citrate
solution has already been given. It should be quite neutral,
because if alkaline it shortens the lives of the cells.

Leucocytes appear to live longest at about 20° C. They will
not live very long at 37°, and at 10° will live longer than at 37° but
not so long as at 20° C. I have already suggested (4) that this
may be due to the accelerated absorption of the poisonous salts in
the citrate solution caused by heat, and this will also explain the
early death in the presence of alkali which also accelerates diffusion.
I presume that the reason why they live longer at 20° than at 10°
is because their normal temperature is about 37° C. and that they
die in the cold in spite of the delayed absorption.

In the following experiments a temperature of 30° C. was em-
ployed with the specified citrate solution, and control experiments
were conducted in each case, the results given being the difference
in the measurements between the test and control.

Measurements

Length of the life of healthy person's leucocytes when resting in
their own plasma^ — As has been shown in a former paper (2), an
average shows that all the cells are alive in 24 hours; the majority
are alive in 36 hours; about 50 per cent are dead in 48 hours;
and all are dead in 86 hours.

. Healthy person's leucocytes; other healthy person's plasma. — All
cells were alive in 14 hours; about 50 per cent were dead in 18

APPENDIX II 413

hours ; the majority were dead in 22 hours ; and all were usually
dead in 28 hours. The difference between these avera'ges may be
said to be about 30 hours. I conclude that the plasma of one
person is poisonous to another person's leucocytes.

Healthy person's leucocytes; plasma from cases of .typhoid feber.
— All cells dead in 14 hours. Difference between test and control
about six hours, which is the average out of four cases.

Healthy person's leucocytes; plasma from cases .of malaria\ —
Majority of cells dead in 16 hours, a few alive in 18 ho'urs.
Occasionally 50 per cent were alive in 16 hours. Average differ-
ence between 12 cases and their controls about two hours.

Healthy person's leucocytes; plasma from cases of phthisis. —
Majority dead in 17 hours. Average difference between five
cases and their controls about one hour. Sometimes it was as
much as four hours, but in very chronic cases there Was little
difference.

Healthy person's leucocytes; plasma from a case of osteo-myelitis.
— 50 per cent dead in 14 hours. Repeated with a case of gan-
genous appendicitis the films showed that the majority were
dead in 14 hours. The difference between these cases and their
controls were five hours and three and three-quarter hours
respectively.

Healthy person's leucocytes; plasma from a case of purpura
hcemorrhagica. — Majority dead in 15 hours; all dead in 20 hours.
Difference from controls five hours.

Healthy person's leucocytes; plasma from a case of chorea. — All
cells dead in 14 hours. Difference about six hours.

Leucocytes from cases of typhoid fever; plasma from, other cases
of typhoid fever. — Average from three groups of cases, all of which
reacted to Widal's reaction and were in the third or fourth week
of the disease except one which was convalescent. These groups
include the cases mentioned above. There was never a difference
of more than one and a half hours between the death of the
majority of cells in test and control tubes.

Leucocytes from cases of malaria; plasma from other cases of
malaria. — Five cases. The majority of cells in all cases were alive
in 18 hours. Practically no difference from controls.

414 APPENDIX II

Leucocytes from cases of phthisis; plasma from other cases of
phthisis. — Four experiments. 50 per cent dead was the average
in 18 hours; very little difference from controls, sometimes the
cells lived longer than in the controls.

Leucocytes from cases of malaria; plasma from cases of typhoid
fever. — The majority of the cells in most instances were dead in
14 hours. Differences varied from four to six hours.

Leucocytes from cases of typhoid fever; plasma from cases of
malaria. — About 50 per cent were usually dead in 16 hours and
all were dead in 20 hours in all cases. Five cases tried; average
difference about three hours.

Healthy person's leucocytes; plasma from cases of carcinoma. —
Seven cases; all cells alive in 16 hours; a large number alive in 20
hours. Usually there was little difference between the effect of
cancer plasma and that of a healthy person.

From the foregoing measurements it would appear that in the
cases which have been experimented Avith the plasma of persons
suffering from infective diseases is poisonous to a healthy person's
leucocytes and to the leucocytes of another person suffering from
another disease, but is not so poisonous to the leucocytes of another
person suffering from the same disease. I submit that it may be
reasonable to suppose that such may be the case in other infective
diseases.

Precautions.- —In comparing the lengths of the lives of leucocytes
of persons suffering from chronic infective diseases both in another
infected person's plasma and in healthy plasma, I have frequently
found that such cells will not live so long as the cells of healthy
persons subjected to the same conditions. This was further inves-
tigated by comparing the lives of leucocytes taken from cases of
chronic illnesses in their own plasmata with the length of the lives
of the cells of healthy persons in their own healthy plasmata. In
cases of chronic phthisis, malaria, Hodgkin's disease,1 etc., I have
found that the leucocytes will not live even in their own plasma
nearly so long as if they belonged to a healthy person, as much as

1 It has been noticed that stain will diffuse more readily into the blood-
cells of these patients — that is, that these diseases, and probably other
chronic illnesses, cause a lowered "coefficient of diffusion" in blood-cor-
puscles.

APPENDIX II 415

a day's difference having been observed; and we may infer that
these diseases, and probably others also, cause a loss of vitality in
the patient's leucocytes, so that by this procedure the loss of vitality
can be measured. It is important to remember this point, for if
the making of a measurement is delayed it may be found that all
the cells are dead in both control and test preparations. This
method of measuring the lives of leucocytes may also prove of
value in prognosis as well as in diagnosis.

I do not think that any difficulty will be met with in making
the counts, with the exception of a possible one caused by the
agglutination of the leucocytes. Occasionally large clumps are met
with. If the cells are clumped, however, it does not necessarily
follow that they are dead — far from it, for they may be very active,
though I am of opinion that if clumped death will soon occur.
The cells in a clump can generally be counted. Ruptured cells
are counted as dead. If bacteria are seen in large numbers in a
film the capillary tube is discarded. The revolving apparatus is
not essential, but more constant results have been obtained by its
use. As far as possible I have purposely avoided handling the
blood of the person whose leucocytes are to be tested, for fear of
injuring the cells. The variations of the alkalinity of the plasma
may, I think, be neglected, as it is not sufficient materially to alter
the length of the lives of the cells. This is borne out by the
experiments with cancer plasma, because that plasma is more alka-
line than normal and yet does not shorten life.

Summary

I fear that it is too early to arrive at any definite conclusions
from so small a number of experiments, but I think that there
publication is justified in order to explain the method employed
and because the results are sufficiently promising to warrant further
investigation, though the work must still be regarded as being in
the experimental stage. I hope that this method will be tried by
others, as the problem given in the enunciation may lead to im-
portant developments, and especially as this kind of research in-
volves the striking of averages and a large amount of experiment to

416 APPENDIX II

determine the points. The method may also be useful to others
studying other branches of immunity. As I have already stated,
my aim is to be able to assist in the diagnosis of infective disease
by this method, but a large amount of material will be required
before one can determine its value in this direction, and I have
mentioned its possibilities with reference to prognosis. The stage
in a disease in which measurable immunity appears in a leucocyte
also remains to be determined.

To summarise the method by which I endeavour to assist in
a diagnosis in a case of infective disease, a small quantity of blood
from a patient is mixed with eight times its volume of the citrated
plasma of other persons who are known to be suffering from certain
infective diseases and also with the citrated plasma of a healthy
person. For this last purpose I sometimes use my own plasma.
The method has been described. The capillary tubes are kept
together in the revolving apparatus for about 14 hours. Then
some agar films are prepared from jelly which will excite move-
ments in living leucocytes, and samples of the contents of the tubes
are examined on these films. The number of living and dead cells
are averaged, and the difference between the lengths of the lives of
the cells when resting in healthy and infected plasmata are deter-
mined. When an infected plasma is found which will not com-
paratively shorten the lives of the patient's leucocytes, it seems
probable that the patient is suffering from the same disease as the
person from whom the plasma was taken. I generally confirm
this procedure by reversing the process and trying the patient's
plasma on the leucocytes of other persons suffering from the dis-
ease determined, taking care to make controls in this case as well
as in the first by making measurements with healthy plasma and
with the plasma of persons suffering from other diseases.

The method described in this paper has two disadvantages:
first, in keeping the tubes at 30° C., and, secondly, in counting
500 leucocytes in each case, which is most tedious. The rest of
the method takes very little time; collecting the plasmata and
mixing them with the patient's corpuscles is soon accomplished,
and when the tubes are in the revolving apparatus they require no
further attention until the time has come to estimate the number

APPENDIX II 417

of living and dead cells in them. The agar jelly can be made from
stock solutions as specified and kept in test-tubes for months, as
moulds will not grow on it. Films are rapidly prepared by boiling
the jelly in a tube in a spirit-lamp flame.

With regard to the two disadvantages, an incubator working
at 30° C. is not usually within reach even in laboratories, although
Hearson's apparatus will maintain this temperature if fitted with
a special capsule. Since my aim is to make this possible diagnos-
tic method suitable for practical purposes even away from the vicin-
ity of a laboratory, I dispense with an incubator and employ the
ordinary temperature of a room, say between 60° and 70° F. In
order to do this the citrate solution is modified. If the solution
already specified were used at such a temperature the leucocytes
might live for a long time even in an infected plasma, and a day
or two might elapse before sufficient deaths occurred among the
cells to make a contrast. Consequently I deliberately shorten the
life of the cells by using a solution containing 1 . 2-per-cent sodium
citrate and 1-per-cent sodium chloride. As the same solution is
employed for all tubes the artificial shortening of life does not
appear to vitiate the results. There are several ways by which
this shortening of life can be accomplished, though I consider the
lowering of the sodium citrate content to be the most suitable.
Using this solution it has been found that the majority of healthy
cells in another healthy person's plasma are dead in about 24 hours
if kept at the room temperature, which, of course, may be variable.
So a contrast can usually be obtained within 24 hours of mixing
the bloods. With regard to the second disadvantage, I hope by
experiment to ascertain the minimum number of leucocytes which
it may be necessary to count to obtain a trustworthy average. I
am sure that a smaller number than 500 will be sufficient. I am
also experimenting with a greater concentration of plasma with a
view to obtaining a wider contrast between the length of the
lives of cells in healthy and infected plasma.

In conclusion, I wish to suggest that this method may also be
useful from a medico-legal aspect, for I have found the leucocytes
alive in the blood removed from the hearts of bodies which have
been lying in the mortuary for 24 hours or more, and it may be

418 APPENDIX II

possible to state how long a person has been dead by estimating
the percentage of living cells so many hours after the death of the
subject.

Bibliography. — H. C. Ross: (1) " On the Death of Leucocytes, "Journal of
Physiology, vol. xxxvii., 1908, p. 327; (2) "On a Combination of Substances
which Excites Amoeboid Movements in Leucocytes," The Lancet, Jan. 16,
1909, p. 152; (3) "On the Cause of Achromasia in Leucocytes," The Lancet,
Jan. 23, 1909, p. 226; (4) "On the Modification of the Excitant for Leu-
cocytes composed of Methylene Blue and Atropine," The Lancet, Jan. 30,
1909, p. 313.

APPENDIX III

A METHOD BY WHICH CELLS CAN BE EXAMINED MICROSCOPICALLY
BETWEEN A COVER-GLASS AND A JELLY-FILM WITHOUT
THE FORMER EXERTING ANY PRESSURE ON THEM. (A

"HANGING DROP" PREPARATION WITH THE JELLY METHOD)

Two round cover-glasses are required. One should have a diame-
ter of half an inch, the other of seven-eighths of an inch. The
jelly from which the film is to be prepared is boiled and a drop
of it run on to a slide. Immediately, before the jelly has had
time to set, the small cover-glass is allowed to fall flat on the
centre of the jelly-film on the slide. Since the jelly is not set,
the cover-glass sinks into but not actually through it. The film
with the cover-glass embedded in it is allowed to set for about five
minutes. One needle is then placed vertically against one edge
of the small cover-glass embedded in the jelly, and the point of
another needle is inserted under the opposite edge of the cover-
glass. By a jerk of this needle the embedded cover-glass is lifted
out of the jelly, when it will be found that a shallow circular
depression exactly corresponding to the cover-glass is left in the
jelly-film. The base and sides of the depression will, of course,
be composed of jelly. The cells to be examined are placed in
citrate solution on the large cover-glass, which is inverted and
allowed to fall flat over the depression in the film. By this
means the large cover-glass is resting on the raised sides of the
depression, but the cells are in the depression. They can now
be made to absorb substances from the jelly, but the cover-glass
does not press them into it unless the cells are very large. This
method is useful for the in-vitro staining of motile bacteria, try-
panosomes, etc.

419

APPENDIX IV

A POSSIBLE ASSOCIATION BETWEEN THE AUXETICS OF
HEALING AND IMMUNITY AGAINST INFECTIVE
DISEASE

THE fact that auxetics contained in the remains of dead tissues
and in globin will cause the cell-proliferation of healing has
suggested a new line of research connected with the problem of
immunity. Since the cell-proliferation of healing is caused by
chemical agents, and since the actions of these agents can be
augmented by substances produced by bacteria, and inhibited
by normal serum, it may be useful to ascertain the action of
disease germs on (1) the remains of dead tissues and globin, and the
auxetic it contains, and (2) on serum. It is obvious that if disease-
germs decompose auxetics, there will be less cell-proliferation of
healing; but if they produce substances which augment the action
of auxetics, or if they prevent the inhibitory action of serum, then
they will tend to assist in healing an injury. Before any disease-
germ can obtain a footing in the body it must cause an injury
which is followed by an attempt at healing. If this healing is
prevented, disease will be the result. If, however, healing occurs
successfully, the patient will remain immune. Hence this sugges-
that the action of disease-germs on the sources of the causes of the
cell-proliferation of healing should be investigated. In reality,
the problem is a bacteriological one, but the investigation of it will
not, I think, be very difficult.

420

INDEX.

Achromasia, 4, 14, 21, 33, 43, 55

Acids, 71, 89

Aconitine, 149

Agar, 7, 9, 36, 40, 83

Alkalies, 71, 77, 86, 88

Alkaloids, 149, 157, 352

Amitotic divisions, 297

Amceba Coli, 98, 400

Amoeboid movements, 67

Apparatus, 18, 20, 27, 34, 410

Archoplasm, 112, 123

Artefacts, 17

Asymmetrical mitosis, 10, 12, 44,

132, 235, 240, 300
Augmentation of cell-division, 227,

310

Auxetics, 6, 232, 292
Azur dye, 2, 237, 347

Bacillus subtilis, 366
Bacterial toxins, 8
Basophile leucocyte, 96, 274, 281
Benign tumours, 7, 339, 340
Blood alkalinity in cancer, 176
Blood-platelets, 5, 102, 113, 117, 124
Blood-serum, 9, 374, 376
Brownian movements, 38, 39, 52
Brucine, 149

Cadaverine, 350, 365

Callus, 336

Cancer, aetiology of, 161, 164, 360

age-incidence of, 161, 360

cases of, 379

causes of (theory), 360

climatic incidence of, 164, 371

place-incidence of, 371

plasma, 159

prevention of recurrence of, 395

recurrence of, 397

site-incidence of, 363
Capillary tubes, 18
Centrosomes of leucocytes, 258

of lymphocytes, 186

Chemo taxis, 150

Choline, 352, 365

Chorea, 8

Chromosomes of leucocytes, 258

number of, 239

of lymphocytes, 186
Citrate solution, 41, 82
Citric acid, 85, 88
Cocaine, 149
Codeine, 149

Coefficient jelly, 85, et seq.
Coefficient of diffusion, 61, et seq.
Concentration of substances, 68
Cornea, 337
"Corns," 337
Crenation, 38
Crescent, malarial, 323
Cytogeny, causes of, 158, 166
Cytoplasm, 52, 59, 62, 82

Death, delay of, 107

in cancer, 165, 366
Death-struggles, 147, 246, 346
Decrepitude, 163
Defibrinated blood, 378
Diffusion of substances, 67, et seq.

of two substances, 145

vacuoles, 103, 111
Doses of alkaloids, 148
Drugs, 61, 63

Elimination, 338

Embryo, 7

Eosinophile granules, 274, 281, 401

leucocytes, 94, 293
Epithelial cells, 73, 352
Erythrocytes, 72
Examination of specimen, 102
Excess of diffusion, 103
Excitation, 130

"Experimental ten minutes," 249
Extracts, 7, 298, 300, 315

Fertilisation, 167, 339

421

INDEX.

Fibroids of uterus, 340
"Flagellation," 175, 272

of malaria parasite, 323
Focusing, 33

Gametophytic tissue, 178
Globin, 322, 342
Granulation tissue, 338, 341
Granules, Altmann's, 52, et seq.. 223
counting, 273, 284

Haemal glands, 180, 183
Haematin, 324, 331
Haemoglobin, 128, 322, 397
Haemolysins,' 38, 98
" Hanging drop." 150, 419
Healing, 41, 168, 176, 319
Heat, 70, 78, 88
Hyaline cell, 4, 60
Hydrocyanic acid, 132, 148

Immunity, 9, 420

Incubator, 94, 98

Index of diffusion, 82, 94

Injury, 41

Irritation, 166, 170, 337, 348

Karyokinesis, 6
Katabolism, 361
Kinetic jelly, 133, 377, 407
Kreatin, 316

Lantern slides, 14
Leucocytes, Altmann's granules of,
256

bursting of, 107

divisions of, 252
Leukaemia, 175, 398
Life, climax of, 162
Lives of leucocytes, 132, 406
Lymphadenoma, 351
Lymphocytes in cancer, 393
Lymphocytes, mitosis of, 186

Maiotic divisions, 177, 239
Melanin, 322, 324
Malaria parasite, 323, 331
Malignant proliferation, 12
Methylene blue, 3
Metastasis, 165, 367
Mitosis, 185, 252
Morphine, 40, 118, 149
Moulds, 85
Mutations, 369

Neutral point, 89

red, 3
Nitro-benzol, 132, 148

Nuclei, lobes of, 5, 10, 43
Nuclein, 178

Opsonins, 130, 156
Osmosis, 59, 129
Osteo-arthritis, 163
Ova, 176
Oxygen, 148

Pancreas, 364
Phagocytosis, 155
Photomicrography, 3, 14, 17
Pigmentation, 322
Pilocarpine, 149
Plimmer's bodies, 181
Poisons, 8, 64
Potassium oxalate, 39
Precautions, 99
Protoplasm, 64, 69
Pseudopodia, 51, 56, 134
Putrefaction, 300, 350, 366
Pyridine, 149

Quinine, 149
Quinoline, 150

Red cells, 82, 117, 127, 294
"Red spots," 103
Reduction divisions, 239, 271
Reproduction divisions, 239, 271
Reproductive cells, 240
Resting-stage, 97
Rheumatism, 8

Salts, 37, 59, 86, 97
Sarcoma, 161, 397

cell-granules in, 291

melanotic, 322
Serum, 66, 374
Sexual form of malaria parasite, 323 ,

331

Sodium chloride, 38, 39, 78, 84, 88
Sodium citrate, 38, 44, 66, 76, 78, 88.
Somatic divisions, 239
Spermatozoon, 167, 339
Spindle, 12, 185, 257
SpirocJueta refringens, 97
Staining, extent of, 75
Strychnine, 149
Suprarenal gland, 306, 345
Syphilis, 164, 364

Technique, for inducing mitosis; 245
Time, a factor of diffusion, 69, 78, 88
Tissue cells, 41

Transplantation of tumours, 369
Trepanema pallida, 364

INDEX.

Ulcers, 341, 345 Variables, equation of, 145

malignant, 384 Vitality, 163, 363
Units, 76, 88, 97

Unna's stain, 43, 52, 68, 77 Wandering cells. 240
Urea, 99

Vacuoles, diffusion, 104 Xanthin, 316

ordinary, 103 X-ray cancer, 170

423

University of Toronto
Library

DO NOT

REMOVE

THE

CARD

FROM

THIS

POCKET

Acme Library Card Pocket
LOWE-MARTIN CO. LIMITED

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Books

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "Induced cell-reproduction and cancer; the isolation of the chemical causes of normal and of augmented, asymmetrical human cell-division"

See other formats