HOME UNIVERSITY LIBRARY 
OF MODERN KNOWLEDGE 
No. 107 
HERBERT FISHER, M.A., F.B.A. 
Prof. GILBERT MURRAY, Litt.D., 
LL.D., F.B.A. 
Prof. J. ARTHUR THOMSON, M.A. 
Prof. WILLIAM T. BREWSTER, M.A. 
Editors: A complete classified list of the volumes of The 
Home University Library already published 
will be found at the back of this book. AN INTRODUCTION 
TO THE STUDY OF 
HEREDITY 
BY 
E. W. MACBRIDE, M.A., 
D.SC., LL.D., F.R.S. 
Professor of Zoology in the Imperial College of 
Science and Technology, London. 
ILLUSTRATED 
NEW YORK 
HENRY HOLT and COMPANY 
LONDON 
Williams and Norgate Copyright, 1924, 
BY 
HENRY HOLT AND COMPANY 
PRINTED IN 
THE UNITED STATES OF AMERICA PREFACE 
In offering this little primer to those 
interested in learning something about the 
present state of our knowledge of the laws 
which govern heredity, it is hardly necessary 
to say that space only allows the briefest 
outline of the subject. The principle which 
has been followed is to give as far as possible 
one sample of each type of research bearing 
on the subject, and in selecting that sample 
to be guided by considerations of clearness 
and simplicity. So in dealing with statistical 
researches we have confined ourselves to the 
exposition of the normal curve of error and 
have omitted all references to those " skew 
curves " to which Professor Pearson has 
devoted so much research, since the mathe- 
matical treatment of these is too complex for 
an elementary text-book. In dealing with 
experiments on Mendelian inheritance those 
were selected which exemplify in the clearest 
way the principles involved, and exceptional 
and irregular phenomena were left out of 
account. The purpose for which this book 
5 6 
PREFACE 
has been written will have failed of achieve- 
ment if its readers are content with the 
account of heredity which it gives and are 
not spurred on to undertake further and 
more advanced reading on the subject. 
E. W. M. CONTENTS 
PAGE 
PREFACE V 
CHAPTER 
I. INTRODUCTION 9 
H. THE NUCLEUS AS THE BEARER OF 
THE HEREDITARY QUALITIES . 37 
IH. THE DEVELOPMENT OF THE GERM 
INTO THE ADULT .... 53 
IV. THE INHERITANCE OF HABITS AND 
OF THEIR EFFECTS ... 77 
V. THE INHERITANCE OF FLUCTUATING 
VARIATIONS 112 
VI. THE INHERITANCE OF " MUTA- 
TIONS " OR " SPORTS " . . 139 
VII. THE INHERITANCE OF SEX . .168 
VIK. THE CAUSES OF MUTATIONS . . 212 
IX. THE BEARING OF THE LAWS OF 
HEREDITY ON HUMAN AFFAIRS . 240 
BIBLIOGRAPHY 251 
INDEX 253 
7 LIST OF ILLUSTRATIONS 
FIG. PAGE 
1. Vegetable and Animal Cells 16 
2. Germ Cells of Animals and Plants .... 20 
3. Transverse Section through the Sorus of a Fern . 28 
4. Section of the Prothallium of a Fern . . 29 
5. Germinating Macrospore and Microspore of a Water- 
fern (Salvinia) 32 
6. Longitudinal Section through the Ovules of a Pine 
and of a Flowering Plant 34 
7. Phases of Mitosis 40 
8. Ripening of the Nucleus of the Male Germ-cell . 43 
9. Ripening and Fertilization of the Egg of Crepidula 49 
10. A Comb-bearer (Ctenophore) 64 
11. Stages in the Development of the Egg of a Sea- 
urchin 71 
12. Growth of a Salamander on a Yellow Background 89 
13. Growth of the Second Generation of Salamanders 
on a Yellow Background 90 
14. Growth of a Salamander on a Black Background 
which was the Offspring of Parents reared on a 
Yellow Background 91 
15. Growth of the Second Generation of Salamanders on 
a Black Background 92 
16. Curve of the Distribution of Stature among Ameri- 
can Soldiers 116 
17. Diagram showing the Correlation of the Heights of 
Children and of their Mid-parents .... 119 
18. Carcinus moenas 126 
19. Daphnia 132 
20. Pure lines of Paramecium 136 
21. Different Types of Combs of Cocks .... 156 
22. Mutants of Drosophila melanogaster . 174 
23. Second Maturation Division of the Male Germ-cell 
of Euschistus variolarius 181 
24. Male and Female Boneilia 187 
25. Inachus scorpio 189 
26. Abdomens of Normal and of Diseased Specimens of 
Inachus scorpio . . . . . . . 191 
27. Males of Euschistus variolarius of Euschistus 
servus . . > 205 
28. (Enothera lam.arckiana and (Enothera gigas . . 215 
29. Carassius auratus, the Wild Progenitor of Gold-fish 220 
30. The Ryukin or Veil-fish 222 
31. The Lion-head or Oranda Shishigashira . . . 223 
32. The Telescope-fish or Deme-Ranchu .... 224 
33. Head of Normal Carp, of Bull-dog Carp and of 
Round-head Carp 225 
34. Development of Normal and of Weakened Axolotl 
Eggs 230 
35. Dissections of Normal and Abnormal Axolotls . 232 
36. Skulls of Wild Boar and of Yorkshire Pig . . 235 
8 AN INTRODUCTION TO 
THE STUDY OF 
HEREDITY 
CHAPTER I 
INTRODUCTION 
The word " heredity " is derived from a 
Latin word " hereditas," which meant origi- 
nally a piece of property belonging to a man 
which passed to his son when he died. Now 
everyone knows that it is from their parents 
that children derive their bodily and men- 
tal qualities; and this handing on of the 
parents' very essence to their children has 
come metaphorically to be called " heredity." 
Since in the human race a child has two 
parents, its nature tends to exhibit the 
features of a compromise between the qualities 
of father and mother. But the child's nature 
is not an exact mean between the natures of 
its two parents. If it were so, all children 
of the same family would be alike save for 
9 10 
AN INTRODUCTION TO 
differences depending on age and sex. The 
science of heredity consists in an investigation 
of the causes which bring it about that one 
child often resembles one parent in a certain 
feature and another parent in another, and 
differs from its brothers and sisters in this 
respect; why, further, a child occasionally 
reproduces the characters of a grandparent 
or of an aunt or uncle even when none of 
these characters is obvious in either parent. 
Our science then starts, like every other 
science, from the basis of common everyday 
knowledge; physics, for instance, began with 
the knowledge gained from handling of heavy 
objects, etc., through thousands of years, 
whilst chemistry commenced with observa- 
tions on the effect of fire on stones, and 
the experience of the value of the various 
" simples " which could be extracted from 
herbs. Science, after all, consists in looking 
closely and consistently at the same things 
which the ordinary man regards carelessly 
and superficially. 
The first and most fundamental problem 
which presents itself in the science of heredity 
is the manner in which the bodies of children 
are derived from those of their parents. In 
the lowest and simplest organisms, the parent THE STUDY OF HEREDITY 
11 
divides in two, thus yielding up its own 
existence in giving rise to two children. This 
phenomenon is known as reproduction by 
fission. When we ascend somewhat higher 
in the scale, we find that the offspring is often 
produced as an excrescence or bud on the 
body of the parent, from which it later be- 
comes separated. In the animal kingdom 
when the bud grows and begins to exhibit the 
characteristic features of the parent, but does 
not separate from it, the compound system is 
spoken of as a colony. Perhaps one of the 
most familiar examples of propagation by 
means of buds is the strawberry plant. As 
all gardeners know, procumbent horizontal 
branches are given off from the stem of this 
plant. These branches are known as runners. 
At the extremity of each runner there is 
formed a bud which proceeds to grow up into 
a new strawberry plant, and the runner 
connecting this with the parent plant dies 
away. 
But in the human race and in the higher 
animals most nearly related to it, in which 
we take most interest, we find that repro- 
duction is effected neither by means of fission 
nor buds, but only by the formation and 
expulsion of small fragments of living matter 12 
AN INTRODUCTION TO 
which are called germ-cells. This method is 
the most widely distributed mode of repro- 
duction in both the animal and vegetable 
kingdoms. In most cases it coexists with 
budding and fission wherever these methods 
of reproduction are found, and it is in every 
way more interesting than they are. It is 
indeed apparently absent in only the very 
lowest plants and animals; we use the word 
" apparently " advisedly, for in many cases 
amongst these the equivalents of germ-cells 
are known to be formed, and he would be a 
rash man who would assert that in the case 
of any species of plant or animal germ-cells 
were never formed. It is therefore of the 
greatest importance to the scientific student 
of heredity to obtain a clear conception of 
the nature and origin of the germ-cell, and for 
this purpose a short digression on the funda- 
mental nature of the structure of the bodies 
of animals and plants will be necessary. 
All who have the slightest acquaintance 
with natural history are aware that it is an 
accepted doctrine amongst naturalists that 
both animals and plants are composed of 
cells, but few have any but the vaguest ideas 
as to what cells are like. It is therefore 
necessary to clarify our ideas on the subject THE STUDY OF HEREDITY 
13 
of cells, and to do this, let us reflect for a few 
moments on the historical origin of the word 
cell. This word was originally used to desig- 
nate the chamber which housed the image 
of the god in a temple, but had long been 
employed as a name for one of the chambers 
in a honeycomb in which the bee stores honey, 
which was supposed to resemble in shape the 
" cella " of a temple. In the seventeenth 
century, Hooke, one of the pioneer English 
naturalists, whilst examining thin slices of 
cork by means of the primitive microscopes 
then available, discovered that these slices, 
viewed by transmitted light, showed a struc- 
ture which he compared to that of a honey- 
comb, and to the cavities which he saw in 
them he gave the name of cells. A short 
time afterwards, it was discovered-chiefly 
by the English naturalist Grew, and the 
Italian naturalist Malpighi-that all living 
plants possessed the same structure, and the 
conception of a plant as a spongework of 
hollow cells filled with watery fluids obtained 
wide recognition. As investigation advanced 
it was pointed out that all young and active 
cells exhibited a layer of clear jelly-like 
material lining the cell-wall, and that inside 
this layer was the watery fluid which consti- 14 
AN INTRODUCTION TO 
tuted a drop called the vacuole. In order 
to distinguish it from this, the layer of jelly 
was termed the primordial Utricle. Further 
observation showed that when this primordial 
utricle disappeared, the cell exhibited no 
further changes and could rightly be described 
as dead, although it might continue to serve 
a useful purpose in the plant owing to the 
mechanical strength of its cell-wall. Atten- 
tion was thus focussed on the primordial 
utricle, and the name protoplasm was invented 
by Von Mohl in 1846 for the material of which 
it was composed. When various staining 
fluids were employed to colour this proto- 
plasm, and thus render it more conspicuous, 
it was discovered that in the protoplasm of 
each cell there was embedded a body, usually 
of oval shape, which absorbed the stain more 
strongly than the rest, and held it more 
tenaciously. On this body the name nucleus 
was bestowed. It was discovered further 
that in young and actively growing parts of 
the plant the cells increased in numbers by a 
process of division, and that this division of 
the cell was in all cases preceded by a division 
of the nucleus. Meanwhile investigators had 
been endeavouring to discover whether the 
bodies of animals showed any analogous THE STUDY OF HEREDITY 
15 
structure to that exhibited by plants; and 
it was soon found that these bodies were made 
up of a repetition of units, each of which 
might be described as a mass of protoplasm, 
with a nucleus embedded in it, and to these 
units the name cell was extended. The cell- 
theory, as an hypothesis of the fundamental 
similarity of the ultimate structure of plants 
and animals, was published by Schwann and 
Schleiden in 1837; but their conception of the 
nature of the cell was entirely erroneous, and 
it was not until 1853 that the real relationship 
between animal and vegetable cells was clearly 
and correctly stated by Cohn. 
It is a familiar truism, that when the ap- 
plication of a concept is widened its content 
is diminished; and to this rule the word 
" cell " is no exception. If we take a broad 
survey of the matter, we find that whereas 
the majority of vegetable cells might fairly 
be described as hollow cases with thick, tough 
walls filled with watery fluid, the majority 
of animal cells, on the contrary, have cell- 
walls of excessive tenuity and consist of com- 
pact masses of protoplasm without any large 
central inclusions of water; further, that 
many of the larger cells of animals, such as 
muscle-fibres, may contain many nuclei. In- 16 
AN INTRODUCTION TO 
deed, if the study of the structure of living 
beings had commenced with animals, the con- 
ception of the unit would not have been that 
of a cell, but something more analogous to a 
brick; and one would have compared the 
Illustrating the differences between vegetable and 
animal cells. A. Cells from the vine. B. Cells from 
the larva of the star-fish. 1. Nucleus. 2. Cell-wall. 
Fig. 1. 
body to a wall composed of bricks united by 
mortar rather than to a honeycomb (Fig. 1, 
A and B). 
Different from one another as are the great 
masses of adult animal and vegetable cells, 
very young animal and vegetable cells 
resemble one another more closely, for in THE STUDY OF HEREDITY 
17 
the young cells of plants the cell-wall is 
thin and the central vacuole has not as yet 
appeared. 
The simplest animals and plants do not 
exhibit cell structure and have been supposed 
to be equivalent to single cells, but this is not 
a just way of looking at the matter. The 
smallest may, perhaps, be regarded as equiv- 
alent to single cells of the higher animals, 
but the larger, which have many nuclei, 
cannot be regarded in this light. They must 
be rather described as destitute of cellular 
structure. Such simple animals and plants 
are known collectively as Protista, because 
in regard to many of them it is very difficult 
to say whether they are plants or animals, 
and the whole class may perhaps be regarded 
as the seed-ground out of which the higher 
animals and plants have developed. 
If we confine our attention for the present 
to the higher animals and plants, we may 
define germ-cells as cells which, when separated 
from the parent, can grow into new organisms, 
and these cells exhibit a considerable uni- 
formity of structure. In the vast majority 
of cases they are of two kinds, termed male 
and female respectively, and in order that 
a young organism should be produced, it is 18 
AN INTRODUCTION TO 
necessary that a male cell should unite with a 
female cell, and so produce a compound cell 
which is termed a zygote. This union is 
termed the Sexual process, and in it the nuclei 
of the two cells unite to form a single nucleus. 
The zygote begins, either at once or after a 
period of rest, to develop into the young 
animal or plant, as the case may be, and the 
first step in this development is the division 
of the zygote into a mass consisting of 
numerous cells. The distinction between a 
member of the class Protista and one of the 
higher organisms is that in the Protista all 
these cells separate from one another, and 
each one grows into a new adult Protist, 
whereas in the case of a higher plant or ani- 
mal, the mass of cells continues to cohere and 
becomes moulded into the tissues of the new 
individual. 
In the group of plants to which the Fungi 
or Moulds belong, great numbers of simple 
germ-cells are produced which can give rise 
to young organisms without the sexual process, 
as well as germ-cells which must unite with 
one another before development can take place. 
Such simple germ-cells are termed conidia 
(Fig. 2, A), whilst to germ-cells which unite 
with one another in pairs the term gametes THE STUDY OF HEREDITY 
19 
is applied, from the Greek word signifying 
marriage. 
" Conidia " or their equivalent are not 
found among the higher animals, but it occurs 
exceptionally amongst animals that gametes 
acquire the power of developing without 
union. This phenomenon is known as 
parthenogenesis. We distinguish such per- 
verted gametes from true conidia by the fact 
that in shape, size and origin they resemble 
the normal gametes in closely allied genera, 
and we know that they are perverted and not 
original from the fact that parthenogenesis 
is the exception and not the rule, and appears 
sporadically in different cycles of animal 
affinity, and is by no means characteristic of 
primitive forms. Lastly, gametes which should 
undergo the sexual process may be induced 
to develop without it by applying artificial 
stimuli to them. 
Returning now to the consideration of 
gametes which undergo the sexual process, 
we find that in a few cases in the lowest plants 
and animals they are all alike in appearance, 
so that male and female varieties cannot be 
distinguished; but, as we have already men- 
tioned, in the overwhelming majority of cases 
they are divided into two categories, male and 20 
AN INTRODUCTION TO 
female. The male germ is exceedingly small, 
and it is motile; in almost all animals, and 
in the lower plants, it is provided with one 
Illustrating the different types of germ-cells. A. The 
conidia of a mould. B. The spermatozoid of a fern. 
C. The oosphere of a fern. D. The spermatozoon of a 
man. E. The egg of a woman. 1. The head (in D), the 
nucleus (in E). 2. The middle-piece (in D), the nucle- 
olus (in E). 3. The tail (in D), the vitelline membrane 
(in E). 
Fig. 2. 
or more hair-like filaments capable of rapid 
vibration, which propel it from place to place. 
In plants it is termed an antherozoid or THE STUDY OF HEREDITY 
21 
spermatozoid (Fig. 2, B) and it is provided 
with several of these hairs, and there is 
sometimes a small cell-body in which a nucleus 
is embedded. In animals the male germ-cell 
is provided with a single vibratile hair which 
is termed the tail, and it consists, besides 
this, of a rod-like nucleus without any dis- 
coverable cell-body at all. The nucleus is 
termed the head, and the whole germ-cell 
is termed the spermatozoon (Fig. 2, D). The 
structure of spermatozoids and spermatozoa 
indicates that they are adapted to swim in 
fluids, and union with female gametes, or 
fertilization, usually does not take place in a 
fluid medium. But in the higher plants this 
is not the case, and here the male cell consists 
of a small naked drop of protoplasm, with a 
nucleus which becomes pressed against the 
female cell. Again, certain animals, such as 
crabs, lobsters, and threadworms, are con- 
stitutionally incapable of producing typical 
spermatozoa owing to the tendency of their 
protoplasm to produce horny matter. In 
these animals the male germ-cells are either 
simple naked cells (threadworms) or cells 
consisting of a nucleus, having attached 
to it a complex horny disk which absorbs 
water and swells and thus pushes the 22 
AN INTRODUCTION TO 
nucleus into the female gamete (crabs and 
lobsters). 
The mass of developing male gametes sur- 
rounded by a wall of protecting cells is 
known as an antheridium in plants, and as a 
testis in animals. 
The female germ-cell in all the higher plants 
and animals is a motionless cell, and it is 
always considerably larger than the male 
germ-cell. In plants it is called the egg-Cell, 
oosperm or oosphere (Fig. 2, C), in animals 
the egg or OVUm (Fig. 2, E). The mass of 
developing female germ-cells surrounded by 
a protective wall is known in plants as an 
archegonium, and in animals as an Ovary. 
The female germ-cell contains embedded in 
its substance a certain amount of unworked-up 
food material, which serves as capital on 
which the zygote lives until it can feed itself. 
The size of the egg-cell depends chiefly on the 
quantity of this reserve material contained in 
it, which in the animal kingdom is known as 
yolk. In birds the true egg is popularly termed 
the " yolk," and this in the case of the ostrich 
reaches the size of a cricket-ball and is the 
largest cell known. On the other hand, the 
egg of a woman when it is discharged into 
the womb is about %5o inch in diameter, and THE STUDY OF HEREDITY 
23 
the egg-cell or oosphere in one of the higher 
plants is not of any greater size. The reason 
for the disparity in size in these cases is the 
difference in the source from which the young 
organism derives the greater part of its 
nourishment. In the egg of the ostrich nearly 
the whole of this nourishment consists of 
the yolk embedded in the protoplasm of the 
egg, and hence the large size of the latter. 
The " white " of the egg constitutes a some- 
what unimportant addition to this store. 
But in the case of the higher plant and of 
the woman the nourishment contained in the 
protoplasm of the egg only suffices for the 
first few steps in the development of the young 
organism. In the woman additional nourish- 
ment is supplied to the egg from the walls 
of the womb, and it is at the expense of this 
nourishment that the baby is enabled to reach 
its full size. In the plant the additional 
nourishment is supplied by a series of cells 
surrounding the egg-cell which are termed 
endosperm, and which are absorbed as the 
seed develops. 
Now common experience, to which we re- 
ferred as the starting-point of the science 
of heredity, teaches unequivocally that the 
father is just as potent in transmitting his 24 
AN INTRODUCTION TO 
qualities as the mother. The essential 
hereditary element in the female germ-cell 
cannot, therefore, consist of the yolk or of 
the protoplasm in which the yolk is embedded, 
but must consist of the nucleus, since in the 
spermatozoa of animals both yolk and proto- 
plasm are absent. This comes out quite 
clearly when the sexual process is studied in 
detail. This can be easily done in the case 
of creatures like star-fish and sea-urchins, 
which produce quantities of minute trans- 
parent eggs and shed them into the sea, where 
they are fertilized by the spermatozoa of the 
male. If a sea-urchin egg be placed in a drop 
of sea-water and some spermatozoa added, 
it is not difficult to watch the process of 
fertilization through a microscope. The sper- 
matozoa propel themselves in a spiral course 
through the water by lashing their tails. One 
at last strikes the surface of the egg and 
proceeds to bore into it by a continuance of 
the same spiral motion. As soon as the 
" head " becomes embedded, the superficial, 
layer of the protoplasm of the egg undergoes 
a change and becomes converted into a skin 
called the vitelline membrane or primary 
egg-shell. This skin cuts off the " tail " of 
the spermatozoon and prevents any other THE STUDY OF HEREDITY 
25 
spermatozoon from entering. Then the head 
pursues its way through the protoplasm of 
the egg, swelling up as it does so, and taking 
on the appearance of a nucleus. It gradually 
approaches the nucleus of the egg and the 
two join to form the nucleus of the zygote 
(Fig. 9, p. 49). 
Since the father is just as potent as the 
mother, we arrive at the remarkable conclusion 
that the nuclei of the germ-cell are the trans- 
mitters of the hereditary qualities and 
constitute the bridge between parent and 
offspring. 
When we look closely at a spermatozoon, 
we can see that between head and tail there 
intervenes a short neck, which is termed the 
middle-piece. The middle-piece follows the 
head into the protoplasm of the egg and there 
acts on this protoplasm so that it becomes 
the centre of a clear sphere with radiating 
streaks. This sphere is known as the '' aster," 
from its resemblance to the conventional 
figure of a star; it comes to lie at the side of 
the compound nucleus of the zygote, and it 
plays an important part in the division of 
that nucleus, as we shall see later. The 
middle-piece of the spermatozoon only gives 
rise to a minute particle called the centrosome, 26 
AN INTRODUCTION TO 
which lies in the centre of the aster, but it is 
the action of the centrosome on the maternal 
protoplasm which produces the aster (Fig. 9, 
B, C). We have termed the protoplasm a 
jelly-like substance; in life it is constantly 
altering from this condition, in which it is 
termed a " gel," to the condition of a thick 
syrup, in which we call it a " Sol," and vice 
versa. The action of the centrosome consists 
in liquefying some of the protoplasm round it, 
which becomes a sphere of thick sol, into which 
radial currents of fluid (" clear streaks ") 
flow from the rest of the egg. These clear 
streaks are separated by semi-rigid bars of 
" gel." The centrosome is normally produced 
from the middle-piece of the spermatozoon; 
but if the egg just after the spermatozoon has 
entered it be subjected to violent centrifugal 
force it is possible to tear away the middle- 
piece; the head of the spermatozoon-that is, 
the male nucleus-then gives rise to another 
centrosome. In eggs which have not been 
fertilized at all it is sometimes possible to 
start the formation of one or of several asters 
by subjecting them for a short time to the 
action of over-concentrated sea-water, and 
in this way an unfertilized egg may be caused 
to develop. This phenomenon is known as THE STUDY OF HEREDITY 
27 
artificial parthenogenesis; and its possibility 
shows clearly that the centrosome must be a 
comparatively simple substance readily pro- 
duced in different ways, and cannot have any 
part in transmitting the peculiar qualities 
of the parents. The division of the nucleus 
is preceded by the division of the centrosome, 
and one half goes to each daughter nucleus. 
The course of events leading up to the 
union of the germ-cells of plants differs in 
many important respects from that of animals, 
and as the higher plants have been the subjects 
of most important researches on the laws 
of heredity, a short account of the origin of 
the germ-cells in these plants must be given. 
In order to make this as clear as possible, 
we must commence by the consideration of 
the case of the common fern. The ordinary 
gardener is well aware that on the back of 
the graceful fronds of this plant there are 
to be found at certain seasons multitudes of 
very small dark kidney-shaped patches. Each 
patch is termed a SOruS, and is covered by a 
brown scale called the indusium. Underneath 
this scale are to be found multitudes of minute 
bodies of microscopic dimensions which are 
termed Sporangia. Each sporangium is a 
tiny box having the shape of two watch- 28 
AN INTRODUCTION TO 
glasses pressed together, mounted on a slen- 
der stalk. Inside the box are multitudes of 
excessively minute rounded bodies termed 
spores (Fig. 3). Each spore is in fact a non- 
sexual germ-cell or conidium, and consists 
of a small mass of protoplasm, with a nucleus 
Fig. 3. 
Transverse section through the sorus of a fern leaf. 
1. Sporangium. 2. Spores inside the sporangium. 3. 
Indusium. 4. Bundle of sap-conveying cells in the leaf 
(fibro-vascular bundle). 5. Green cells of the leaf. 6. 
Skin cells of the leaf. 
and a thick brown cell-wall. If these spores 
be sown on a damp brick and exposed to 
moderate light they will germinate after 
a somewhat prolonged period of rest, and 
give rise to, not new ferns, but minute 
plants termed prothallia, which resemble THE STUDY OF HEREDITY 
29 
little pieces of duckweed (Fig. 4). If we 
examine the structure of these prothallia, 
we shall find that they show on the under- 
surface numerous minute cavities, and that 
these cavities are of two kinds: one sort is 
hemispherical and the other sort is flask- 
shaped. The outer walls of both kinds pro- 
Fig. 4. 
Section through the prothallium or " oophore " genera- 
tion of the fern. 3. Archegonium. 4. Oosphere, or 
egg-cell in the archegonium. 5. Antheridium. 6. Sper- 
matozoids, or male cells, in the antheridium. 7. Root- 
hairs of the prothallium. 
ject very slightly above the ground level of 
the surface of the plant. The hemispherical 
cavities are antheridia, and contain the male 
germ-cells, which in the case of the fern are 
spermatozoids, differing from spermatozoa in 
possessing several whip-like tails instead of 
one. The flask-shaped cavities are archegonia, 
each of which contains one female germ-cell 30 
AN INTRODUCTION TO 
or oosphere surmounted by several cells situ- 
ated in the neck of the flask, which break 
up into a kind of slime. When the male cells 
are ripe, the lid of their case breaks, they 
swarm forth and, swimming in the film of 
moisture underneath the prothallium, are at- 
tracted by the substances contained in the 
slime above the egg-cells. One spermatozoid 
reaches and fertilizes each egg-cell (the spiral 
portion of the spermatozoid which enters 
the egg-cell is the metamorphosed nucleus). 
As the egg-cells ripen later than the sper- 
matozoids, and usually many prothallia are 
found near each other, it will generally happen 
that a spermatozoid belonging to one pro- 
thallium will fertilize the egg-cell belonging 
to another prothallium. The fertilized egg- 
cell then begins to divide and develops into a 
young fern plant, and as this grows the 
prothallium withers away. It follows that 
one prothallium can give rise to several fern 
plants. 
In the fern plant we are thus confronted 
with an alternation of generations. A genera- 
tion which produces spores, or a spore-bearer 
(sporophore), is succeeded by a generation 
which produces egg-cells (oophore). A similar 
alternation of generations can be traced in all THE STUDY OF HEREDITY 
31 
the higher plants, although modified and ob- 
scured, and we can trace a series of stages in 
this modification. Thus in certain water-ferns 
(Salviniacea) there are two kinds of spores 
produced, viz. small spores, termed micro- 
spores, and large spores, termed macrospores. 
The sporangia containing these two varieties 
of spores are termed microsporangia and 
macrosporangia respectively. The micro- 
spores on germination give rise to prothallia 
which bear only spermatozoids, whereas the 
macrospores give rise to prothallia which pro- 
duce only egg-cells. These may be termed the 
male and female prothallia respectively. The 
male prothallium sometimes consists of a single 
filament at the end of which two cells with 
granular contents are cut off. These cells 
constitute the antheridium, and in them the 
spermatozoids are developed (Fig. 5, A). 
The female prothallium (Fig. 5, B) consists 
of a mass of tissue projecting only a little way 
beyond the burst wall of the spore. In this 
mass are formed one or more archegonia, in 
each of which an egg-cell is developed. The 
macrospores, which produce female prothallia, 
are larger than the microspores, which pro- 
duce male prothallia. Both macrospores and 
miscrospores are produced in fours by two 32 
AN INTRODUCTION TO 
successive divisions of a mother cell, but 
whereas in the formation of microspores all 
Fig. 5. 
Illustrating the germination of the two kinds of spores 
of the water-fern (Salvinia). A. The germination of the 
microspore. 1. The spermatozoids in the vestigial 
antheridium. 2. The burst microspore. B. The germina- 
tion of the macrospore. 1. The burst macrospore. 2 
The vestigial prothallium. 3. The archegonium. 4. The 
egg-cell in the archegonium. 7. The wall of the burst 
macrosporangium. 
four sisters become spores, in the case of the 
macrospores only one of the four becomes a THE STUDY OF HEREDITY 
33 
spore, whilst the other three cells remain small 
and remind us of the polar bodies in animal 
eggs. 
In that strange tropical plant the cycad 
there is essentially the same arrangement, 
but the female prothallium (termed the 
endosperm) remains entirely within the 
macrospore, only one of which is produced 
in each macrosporangium. The macrospore 
is now termed the embryo-sac. We find the 
further modification that the filamentous male 
prothallium is produced only when the 
microspore, or, as it is now termed, the 
pollen-grain, is wafted by the wind so as to 
alight on the macrosporangium, or, as it 
is now termed, the OVllle. This male pro- 
thallium consists of a filament termed the 
pollen-tube, at the apex of which the male 
germ-cells are formed, which, as in ferns, are 
spermatozoids, each provided with several 
flagella. In the pines and their allies the male 
cells are no longer typical spermatozoids, 
but small rounded naked nuclei, which when 
they escape, fall directly on the slime cells in 
the necks of the archegonia in the female 
prothallium. They easily traverse the slime 
and fertilize the egg-cells. Several young 
pine plants are thus formed in each embryo- 34 
AN INTRODUCTION TO 
sac, but only one survives; the rest perish in 
consequence of the withdrawal of nourishment 
from them by the successful one (Fig. 6, A). 
Fig. 6. 
Longitudinal section through the " ovules " or macro- 
sporangia of the seed-plants. A. Section through the 
ovule of a pine. 1. Macrospore or embryo-sac. 2. 
Endosperm or prothallium inside embryo-sac. 4. Arche- 
gonium. 5. Pollen-tube in male prothallium. 6. Male 
nucleus in pollen tube. 7. Macrosporangial wall. 8. 
Envelope surrounding macrosporangial wall. B. Section 
through the ovule of a flowering plant. 1. Embryo-sac. 
4. Egg-cell and its two sister cells. 5. Antipodal cells. 
6. Endosperm-forming cells. 7. Macrosporangial wall. 
8. Inner envelope of macrosporangium. 9. Outer enve- 
lope of macrosporangium. 
Finally, in ordinary flowering plants mat- 
ters are very much as they are in the pines, 
but the macrosporangia or ovules are enclosed 
in a box or case formed by conjoined leaves THE STUDY OF HEREDITY 
35 
called the ovary (not to be confounded with 
the ovary of an animal). The tips of these 
leaves form the sticky surface or Stigma on 
which the microspores or pollen-grains fall, 
and where they germinate and produce the 
male prothallia or pollen-tubes. Each pollen- 
tube produces only one male cell, which is to 
all intents and purposes a naked nucleus. 
The female prothallium is represented only 
by eight cells, devoid of cell-walls, embedded 
in the protoplasm of the macrospore or 
embryo-sac; these are produced by three suc- 
cessive divisions of the macrospore nucleus. 
Of these, three are situated in the upper part 
of the macrospore where the contact with the 
pollen-tube takes place, three, termed " antip- 
odal " cells, at the lower end, and two in the 
middle. These last two coalesce to form a 
single cell, termed the mother-cell of the endo- 
sperm. The endosperm is a nutritive tissue 
formed by the division of this mother-cell 
after fertilization, which fills up the cavity 
of the macrospore, and very possibly does not 
correspond to the endosperm of cycads and 
pines. Of the six remaining cells only one is 
a ripe egg-cell, and is fertilized by the male 
cell; the remainder are to be regarded as 
vestigial egg-cells and disappear (Fig. 6, B). 
In the case, therefore, of the flowering plant, 36 
THE STUDY OF HEREDITY 
only one young plant is formed in each 
prothallium, that is to say, in each macro- 
spore; and as one macrospore is formed in 
each macrosporangium, we can further assert 
that only one baby plant is formed inside each 
macrosporangium. The macrosporangium 
grows rapidly and forms the seed, and the 
seed when fully formed drops off the plant 
and falls to the ground. By the time that 
this has occurred the baby plant has also 
grown enormously, and in many plants, 
such as peas and beans, it has completely 
absorbed the endosperm; but in other plants, 
such as wheat, maize and the cereals generally, 
it lies embedded in an unconsumed mass of 
endosperm. The wall of the macrosporangium 
and that of the macrospore crushed together 
form the seed-coat or testa. It follows, 
therefore, that what we call a seed consists of 
parts belonging to two different generations; 
the seed-coat or testa is part of the tissues of 
the mother plant, whilst the baby plant or 
embryo within represents the next generation. 
When endosperm is present we may assert 
that three generations are represented, for 
the endosperm represents the prothallium or 
oophore, which is produced by an asexual 
spore from the mother plant or sporophore. CHAPTER II 
THE NUCLEUS AS THE BEARER OF THE 
HEREDITARY QUALITIES 
In Chapter I we saw ourselves driven to 
the conclusion that the nucleus of the male 
germ-cell was the sole channel through which 
the qualities of the father were transmitted 
to the child. Since father and mother are 
equally potent in transmitting their qualities 
to their offspring, it is practically certain 
that it is through the nucleus of the egg that 
the mother also perpetuates her individual 
peculiarities. As soon as these conclusions 
were seen to be inevitable, the minute struc- 
ture of the nucleus assumed an enormous 
importance in the eyes of all students of 
Biology, and a corresponding amount of 
study was devoted to its elucidation. 
In some Protista, and in some cells in the 
higher animals, the nucleus can be seen during 
life; it then appears as a vesicle surrounded 
by a thin wall and containing a clear fluid, 
37 38 
AN INTRODUCTION TO 
which is known as the nuclear Sap. In small 
transparent eggs the nucleus can be easily 
seen and it is relatively very large; it was 
called by the older authors the germinal 
vesicle, and it contains a conspicuous solid 
inclusion which they called the germinal Spot, 
but which is now known as the nucleolus. 
It is only, however, when the cell is pre- 
served and stained that the full particulars 
of the structure of the nucleus can be 
observed. It is then seen to stand out 
from the rest of the cell by reason of its 
greater power of absorbing colouring matter, 
and when this colouring matter is removed 
from the rest of the cell by dilute acid, the 
nucleus still holds on tenaciously to it. The 
nuclear wall and the nuclear sap are not 
coloured, but the coloured portion consists 
of the nucleolus (when this is present) and of 
a series of grains of peculiar substance called 
chromatin. This substance is sometimes 
arranged in streaks along the nuclear wall 
and sometimes in lines on what appear to 
be cords traversing the nuclear sap. These 
cords have been termed linin. In some large 
nuclei they can be seen in the living condition, 
and they can be investigated by means of a 
long, excessively fine needle of glass. It THE STUDY OF HEREDITY 
39 
then transpires that they are not cords but 
sheets of denser material traversing the fluid 
sap. When a nucleus divides into two the 
lines of chromatin contract, and become 
arranged in a number of bands called chromo- 
somes. It is a most wonderful thing that 
whenever they can be counted these chromo- 
somes appear in the same number in all the 
cells of the body when the nucleus divides. 
This number is typically the same for all 
members of the same species of animal, but 
varies from species to species. For reasons 
which will soon transpire it is termed the diploid 
number, and is denoted by the symbol 2/r. 
The chromosomes continue to contract, and 
as they grow shorter and thicker the nucleo- 
lus, if there is any, becomes dissolved in the 
cell sap and totally disappears, and a little 
later the nuclear wall is also dissolved and the 
nuclear sap mingles with the protoplasm. It 
has been proved in several cases that the 
material of the nucleolus is ejected from the 
nucleus into the protoplasm. Whilst these 
changes have been going on the centrosome, 
which lies outside the nuclear wall, has divided 
into two and the daughter centrosomes have 
moved away from each other, each of them 
producing round itself an " aster " with out- 40 
AN INTRODUCTION TO 
streaming rays. Then each of the chromo- 
somes is seen to be split longitudinally, and 
soon to each half of each chromosome a ray 
from one of the two asters is seen to be 
Illustrating the process of karyokinesis (mitosis). 
A, B, C. Prophase. D. Metaphase. E, F. Anaphase. 
G. Telophase. 1. Chromosome. 2. Centrosome. 
Fig. 7. 
attached. As the daughter centrosomes con- 
tinue to move apart, these rays seem to 
pull on the halves of each chromosome. 
The first effect of this pull is to cause THE STUDY OF HEREDITY 
41 
all the chromosomes to be arranged in a 
plane at right angles to the line joining the 
two centrosomes. In this way a disk of 
chromosomes is formed which is known as 
the equatorial plate, and it is in this phase 
of division that the chromosomes are most 
easily counted. The fibres converging from 
each side of the plate to an aster consti- 
tute a biconical figure known as the " spin- 
dle." 
As the process of nuclear division proceeds 
the halves of each chromosome move apart 
as if each was being pulled towards one of 
the two asters. As they separate longi- 
tudinal fibres of linin are seen to connect 
them. Eventually a group of chromosome- 
halves becomes aggregated in the neighbour- 
hood of each aster. Then a nuclear wall is 
formed round each group, and so two daughter 
nuclei are constituted. It only remains for 
the chromosomes in each daughter nucleus 
to lengthen and gradually to lose their dis- 
tinctness in order that the division should be 
completed and each nucleus should pass back 
into the resting condition. 
This method of nuclear division is known 
as mitosis or karyokinesis. It is much the 
most frequent method, but occasionally when 42 
AN INTRODUCTION TO 
multiplication takes place in a group of 
specialized cells, which have no power of 
developing into anything unlike themselves, 
the nucleus divides directly by simple con- 
striction into two parts without any rear- 
rangement of its chromatin. In contrast to 
karyokinesis this method is sometimes known 
as " direct division." 
Mitosis is divided into phases. The gradual 
differentiation of the chromosomes out of 
the resting nucleus from the time of the 
separation of the centrosomcs up till the 
dissolution of the nuclear wall is known as 
the prophase; the formation of the equatorial 
plate and of the spindle is known as the meta- 
phase; the separation of the halves of the 
chromosomes is known as the anaphase; and 
the passing of the daughter chromosomes 
back into the resting daughter nucleus is 
known as the telophase. 
We may now turn to consider the features 
of special interest which are to be observed 
in the growth and development of the germ- 
cells. The germ-cells, both male and female, 
when they first make their appearance are 
small and rounded, and consist of somewhat 
denser and more deeply staining protoplasm 
than their neighbours. In this condition 
they are known as primitive germ-Cells, or, THE STUDY OF HEREDITY 
43 
when sex can be distinguished, as Spermato- 
gonia and OOgOHia respectively. They divide 
Fig. 8. 
Illustrating the stages in the ripening of the nucleus 
of the male germ-cell. A. Prophase of the first matura- 
tion division. B. Metaphase (i. e., the equatorial plate) 
of the first maturation division seen from above. C. The 
same a little later, showing the beginning of " pairing " 
or " meiosis." Eight " paired " and six as yet unpaired 
chromosomes are shown. D. The first maturation divi- 
sion seen from the side. E. The equatorial plate of the 
second maturation division seen from above. F. The 
second maturation seen from the side. 
repeatedly and multiply rapidly in number 
by ordinary mitosis. After a time these di- 44 
AN INTRODUCTION TO 
visions cease and the cells enter on a period 
of growth and rest, during which the distinc- 
tions between male and female become 
obvious. The male cells remain small, and 
are known as primary spermatocytes; but 
certain of the female cells enlarge by devouring 
and absorbing many of their less fortunate 
sisters, and these cells are known as unripe 
eggs or primary oocytes. In both sperma- 
tocyte and oocyte the nucleus contains much 
sap, and there is a well-marked nucleolus 
(compare Fig. 9). 
When growth is finished both spermatocyte 
and oocyte ripen by undergoing two divisions 
termed the maturation divisions. In the 
first of these the nucleolus disappears and the 
usual (diploid) number of chromosomes char- 
acteristic of the species is differentiated out 
of the chromatin grains. This number is 
in the vast majority of cases an even one. 
Then the chromosomes join together in pairs, 
so that the number is halved (Fig. 8, C). 
The half number is known as the reduced 
or haploid number, and is designated by the 
symbol x. In the best attested cases the 
junction takes place in such a way that the 
two chromosomes lie alongside one another 
and fuse into one. Sometimes it is true that THE STUDY OF HEREDITY 
45 
the members of a pair join end to end, but 
this is only the beginning of a side-by-side 
union, since the two attached chromosomes 
subsequently swing round so as to be parallel 
with one another. Later a split appears in 
the centre of the region where the junction 
took place, and as the two halves of the com- 
pound chromosome still adhere at the ends, it 
takes the form of a ring. When the equatorial 
plate and the spindle are formed, the two 
halves of the ring are dragged apart, and one 
half-ring, which corresponds to one of the 
original diploid chromosomes, is incorporated 
in each daughter nucleus. This nuclear di- 
vision, therefore, differs from every other 
nuclear division in this, that instead of one 
half of each diploid chromosome passing into 
each daughter nucleus, some whole chromo- 
somes go to one daughter nucleus, and some 
to the other. This division constitutes, there- 
fore an exception to the rule that there is 
an equal distribution of material to each 
daughter nucleus; it is called the reducing 
or meiotic division, and the process of the 
union and subsequent separation of two 
chromosomes is known as meiosis. Meiosis 
of perfectly typical character is found in the 
Protista, in the higher animals and the higher 46 
AN INTRODUCTION TO 
plants; it must, therefore, be of enormous 
antiquity, and must have originated at a time 
when the ancestors of existing animals, plants 
and Protista formed one uniform stock of 
lowly organisms. 
In the male cell the meiotic division results 
in the production of two precisely equal cells 
which are known as the secondary spermato- 
CyteS, but in the case of the female cell the 
division is unequal, and results in the formation 
of one large daughter cell, which is known 
as the secondary oocyte, and a very small 
daughter cell, which is termed the first polar 
body. 
The second maturation division now takes 
place. So far as the nuclei are concerned 
this is a quite ordinary case of mitosis in 
which each chromosome is halved longi- 
tudinally in the usual manner. In order to 
mark its contrast with the reducing or meiotic 
division it is sometimes known as the equa- 
tional division, but this name is unnecessary. 
In the case of the male cell each secondary 
spermatocyte divides into two precisely equal 
cells called spermatids; in the case of the 
female cell the secondary oocyte divides 
unequally, giving rise to one large daughter, 
which is the ripe OVUm, and one small daugh- THE STUDY OF HEREDITY 
47 
ter, which is the second polar body. The 
first polar body usually divides at the same 
time into two equal daughters, but sometimes 
this division fails to take place (Fig. 9). It 
is obvious from this description that the three 
polar bodies are reduced and starved sister- 
cells of the egg, and correspond to the three 
vestigial spores which are formed along with 
the macrospore in the macrosporangium of 
the water-fern (Chap. I). This was pointed 
out by Hertwig in 1890; previous to this 
time, mysterious powers had been attributed 
to the polar bodies by various authors, espe- 
cially Weismann. 
When an egg has given off the second 
polar body, it is ready to receive the sperma- 
tozoon; the nucleus, which is termed the 
female pronucleus, does not return to the 
resting condition until it has united with that 
of the spermatozoon, and indeed in many 
eggs the second maturation division does 
not take place until after the spermatozoon 
has entered the protoplasm of the egg. In 
the case of the male cell, however, the sperma- 
tid has to undergo considerable modification 
before it becomes the spermatozoon. The 
nucleus shrinks and expels fluid and becomes 
converted into an apparently solid dense 48 
AN INTRODUCTION TO 
mass of chromatin; the centrosome which 
had occupied one side of the spindle in the 
last maturation division gives rise to the tail 
filament, which appears as a tiny thread 
growing out from it and piercing the proto- 
plasm. Then the nucleus, dragging the tail 
with it, emerges from the protoplasm, which 
forms for a time a mantle round the tail; but 
this mantle eventually disintegrates and is cast 
off, and so the perfect spermatozoon is formed. 
As already mentioned, when the sperma- 
tozoon head penetrates the egg it swells up, 
owing to the absorption of fluid from the 
protoplasm of the egg, and forms the male 
pronucleus, which possesses a definite nuclear 
wall, nuclear sap and has its chromatin ar- 
ranged in chromosomes. Now Boveri showed 
that in the case of some eggs after ferti- 
lization it was perfectly possible to demon- 
strate in the zygote nucleus two groups of 
chromosomes, one derived from the female 
pronucleus and one from the male pronucleus, 
each group consisting, of course, of the 
haploid number. These two groups taken 
together therefore represent the diploid num- 
ber characteristic of the species. 
When the fertilized egg divides for the 
first time so as to give rise to the first two THE STUDY OF HEREDITY 
49 
Fig. 9. 
Illustrating the ripening and fertilization of the egg 
of a sea-snail (Crepidula). A. First maturation divi- 
sion, which gives rise to the first polar body. The 
spermatozoon is just entering the egg. B. Second ma- 
turation division, which gives rise to the second polar 
body. The first polar body has divided into two cells. 
The head of the spermatozoon has become the male pro- 
nucleus. C. Male and female pronuclei approaching one 
another, each accompanied by a large aster. D. The 
first division of the zygote nucleus. The male and female 
chromosomes form two separate groups. 1. The first 
polar body. 2, 3. The two cells into which it divides 
4. The second polar body. 5. (in A) the head of the 
spermatozoon; (in B) the male pronucleus; (in C) the 
aster at the side of the male pronucleus. 6. The aster at 
the side of the female pronucleus. 50 
AN INTRODUCTION TO 
cells of the embryo, each chromosome of 
the zygote nucleus splits as usual longitudi- 
nally, and one half of each chromosome is 
distributed to each cell, so that each receives 
a half of each maternal and each paternal 
chromosome. 
From the fact that the spermatozoon head, 
which is the bearer of all the paternal heredi- 
tary qualities, consists only of chromatin, it 
is obvious that chromatin is the essential ele- 
ment in the nucleus and that nuclear sap and 
nuclear wall are formed by its reaction 
with the protoplasm. This conclusion is 
confirmed by the observation that when, as 
happens in some irregular cell-divisions, one 
chromosome gets separated from the rest and 
does not become included in either of the two 
daughter nuclei, this errant chromosome is 
able to gather round itself nuclear sap and a 
nuclear wall of its own, and so constitute itself 
into a miniature nucleus. 
The diploid number of chromosomes in the 
fertilized egg is, as we have seen, composed 
of two sets of chromosomes equal in number, 
one set being paternal and one maternal. 
As this diploid number reappears in all 
subsequent cell-divisions of the embryo, the 
question arises whether the chromosomes per- 
sist as individuals through the resting stages THE STUDY OF HEREDITY 
51 
of the nucleus when all that we can detect 
are lines of chromatin granules apparently 
irregularly arranged. There is strong reason 
for believing that this is indeed so. When 
the chromosomes disappear after one cell- 
division they present a peculiar bunched 
arrangement (the so-called synaptic knot), 
and when they reappear in the prophase of 
the next division they present the same appear- 
ance. The growth in size of the primitive 
oogonium which is converted into the oocyte 
or unripe egg begins after the prophase of 
the first maturation division. During this 
period the chromosomes become invisible 
owing to the loss of their staining powers, but 
they reappear when growth is completed. 
The chromosomes which can expand and con- 
tract and change their composition without 
losing their individuality, and which repro- 
duce themselves, are really elementary living 
organisms. 
In a later chapter we shall discuss the 
evidence for and against the idea that indivi- 
dual chromosomes contain distinct sub- 
stances destined to serve different ends in the 
building up of the embryo. 
The development of the germ-cells in those 
of the higher plants in which it has been 
thoroughly investigated follows a different 52 
THE STUDY OF HEREDITY 
course from that which has been observed in 
animals. We have already learned that in all 
the higher plants there is a constant alterna- 
tion of two kinds of individual, viz., a small 
oophore which produces egg-cells and sperma- 
tozoids, and a large leafy sporophore which 
produces spores. Now Farmer has shown 
clearly that in all these higher plants the all- 
important meiotic or reducing division does 
not occur during the formation of the sexual 
germ-cells, but during the formation of the 
spores. The nuclei of all the cells of the 
prothallia (oophores) produced by the germi- 
nation of the spores have therefore the re- 
duced number of chromosomes, and this is 
equally true of the comparatively large pro- 
thallia of ferns, and of the vestigial prothallia 
of the pines. The germ-cells produced by 
these prothallia have likewise the reduced num- 
ber, and, as in animals, the diploid number is 
restored by the act of fertilization. 
The endosperm of the flowering plants, 
although it represents the prothallium, has, 
however, the diploid number of chromosomes 
in its nuclei. This is due to the fact that 
the mother cell from which its formation 
starts is formed by the fusion of two of the 
nuclei formed in the embryo-sac or oophore. CHAPTER III 
THE DEVELOPMENT OF THE GERM INTO 
THE ADULT 
When the germ-cells have united with one 
another, and the zygote has been formed, it 
has, of course, a long period of development 
to undergo before it attains the adult form. 
From this development some important facts 
can be gleaned about the nature and mecha- 
nism of heredity. This is especially true of 
animal development, and we may now indicate 
a few general points in the nature of this. 
The egg of one of the higher animals begins 
its development by dividing into a number 
of cells, w hich are termed blastomeres J this 
process is known as Segmentation. Then 
these cells are moulded, by folding, into the 
primary organs or germinal layers. Of 
these there are three, viz., the ectoderm or 
primary skin, the endoderm or primary gut- 
wall, and the mesoderm or primary lining of 
the body-cavity. This process is known as the 
formation of the layers. Finally, these pri- 
mary organs, by further processes of growth 
53 54 
AN INTRODUCTION TO 
and folding, give rise to the secondary organs; 
thus, for example, the ectoderm gives rise to 
the outer layer of the skin, the nervous sys- 
tem and the sense organs; the endoderm to 
the liver, the pancreas and the lungs, as well 
as to the lining of the alimentary canal, whilst 
the mesoderm forms the bones and muscles. 
This process is known as organogeny. 
The growing organism during the first part 
of its existence is sheltered from outside in- 
fluences either by an egg-shell, or by being re- 
tained within the mother's body, or in rare 
cases by being taken into cavities in the 
father's body. During this period it is called 
an embryo. When it finally escapes into 
the outer world it is still in many cases very 
unlike the parent, and has to undergo a series 
of changes or metamorphoses before it at- 
tains the adult form; during this period it is 
known as a larva. 
The best known example of a larva is the 
tadpole of the frog. Now every animal begins 
its existence by being an embryo, and it may 
be fairly said that every animal later becomes 
a larva, for none leaves the egg-shell or the 
body of the parent with the proportions and 
powers of the adult. The small boy can 
be justly characterized as a human larva, THE STUDY OF HEREDITY 
55 
for in body, but still more in mind, he is 
different from the adult. The relative impor- 
tance, however, of the embryonic and larval 
stages varies greatly. Sometimes the develop- 
ment is mainly of the embryonic type and 
sometimes of the larval type. 
Now the startling thing about some larvae 
is that they are unquestionably reproductions 
of a former state of the race to which the 
parents belong. This will become clear when 
some examples are given. Thus, as everyone 
knows, the vast majority of fish are bilaterally 
symmetrical animals, that is to say, that in 
them the right and left sides resemble one 
another as an object resembles its image 
in a mirror. But in a small but very im- 
portant group of fish (the flat-fish) this is not 
so. These fish swim lying on one side. On 
the side directed upwards are both eyes, 
and the mouth is often unevenly placed, the 
greater part being directed downwards. The 
larvae of these fish are, however, symmetrical, 
for in them each side has its eye and the 
mouth is straight in front. Again, the over- 
whelming majority of bivalve molluscs (i. e., 
cockles, mussels, scallops, clams, etc.), bur- 
row in the mud, and possess for this pur- 
pose a muscular tongue-like appendage termed 56 
AN INTRODUCTION TO 
the foot, which they protrude from between 
the valves of the shell. The oyster, however, 
is an exception to the rule, for it lies through- 
out life on one side and possesses no foot. 
But the American oyster, when young, creeps 
about and possesses a tongue-like foot. Still 
one more example of this phenomenon may 
be given. The majority of those exquisite 
marine animals, the feather-stars or " lily- 
encrinites " (Crinoidea), are extinct, but we 
learn from their fossilized remains that nearly 
all of them during life were rooted to the 
bottom by long stalks. Some of the living spe- 
cies show the same structure. The commonest 
form, however (Antedon rosacea), has no 
stalk when adult and possesses the power of 
swimming from place to place, but when it 
is young it is rooted to the bottom by a stalk 
exactly like its fossil relatives. When it 
approaches maturity it breaks loose from its 
stalk and swims free. In the case of these 
larvae we may say that the life history 
" recapitulates " the history of the race. 
The discovery of cases like these fired the 
imagination of zoologists thirty years ago, 
and for some considerable time one main pre- 
occupation of zoological research was the at- 
tempt to reconstruct ancestral history from THE STUDY OF HEREDITY 
57 
a study of the life history. Unfortunately, 
the recapitulatory element, as we may call 
it, is not the only influence which moulds 
the life history; other influences which 
modify and obscure the ancestral element are 
at work, and the disentanglement of the 
parts played by the various influences is 
no easy matter. The attempt to do so led 
to much wild speculation and mutually con- 
tradictory hypotheses, and the apparent im- 
possibility of arriving at certainty in the 
matter resulted in the disparagement of the 
whole idea. 
But if there be any truth in the idea that 
there is a recapitulatory element in develop- 
ment, we simply cannot afford to ignore it 
if we are attempting to elucidate the laws of 
heredity. For the recognition of this element 
implies this: that when an animal changed its 
mode of life and sought a new environment 
and new ways of acquiring food, this usually 
happened at adolescence, and the new struc- 
tures which were at first acquired as a reaction 
to the environment became, in course of time, 
so fixed in the animal's constitution that they 
appeared at progressively earlier periods in 
the growth of the animal, and eventually in- 
dependently of the environment altogether, 58 
AN INTRODUCTION TO 
as in the case where the young animal com- 
pletes the greater part of its development 
within the mother's womb. 
But we must remember that when the young 
organism passes its early life in the open as a 
larva its circumstances may change, so as to 
produce modifications in its structure which 
were utterly unknown to the adult animal when 
the race was in the stage now represented by 
the larva, and such modifications can be com- 
pared to falsifications of an old record. Only 
when we base our conclusion on a comparison 
of the life histories of many allied forms can 
the true record be discriminated from the 
secondary falsifications, just as an historian 
bases his views as to what actually happened 
in the past more on the points on which old 
documents agree than on those on which they 
differ. 
Of course a critic might maintain that we 
have no proof that the ancestors of the flat- 
fish ever were like ordinary fish, or that the 
ancestors of the oyster ever burrowed, or, 
finally, that those of the feather-star were 
ever permanently fixed. The only conclusive 
proof would be furnished if some angelic re- 
corder had watched the transformation slowly 
proceeding, and had left a duly attested THE STUDY OF HEREDITY 
59 
account of the whole thing. As Huxley has 
said: "If a man chooses to assert that a 
fossil shell is a mineral concretion it is im- 
possible to drive him from his position by 
logic." But the fact that in the three cases 
mentioned no naturalist, however much he 
might inveigh against recapitulatory inter- 
pretations in general, would question the an- 
cestral significance of the larva is proof 
enough for most sensible people. Exactly 
the same kind of proof convinced people that 
Frauenhofer's lines in the solar spectrum and 
similar lines seen in the spectra of heated 
metals were due to the same cause. But if 
we once admit that in the life histories of 
some animals there are reminiscences of a for- 
mer condition of the race, it becomes almost 
certain that there is a recapitulatory element 
in all life histories, however obscured and 
falsified. It is in the highest degree unlikely 
that the development from egg to adult in 
different species of animals is a phenomenon 
of a totally different nature in different cases. 
Once it is understood that the larval phase 
is often a partial reproduction of some bygone 
condition of the race, the relation of the 
embryonic phase to it becomes plain. The 
embryonic phase is a modification of the 60 
AN INTRODUCTION TO 
larval phase. Taking the tadpole as a typical 
larva, we find that all its organs are adapted 
for life in water. Now there is in the island of 
Martinique a toad (Hylodes) which is hatched 
from the egg as a toad. If we examine what 
goes on within the egg-shell we find that 
the egg develops into a tadpole, some of 
whose organs are useless under embryonic 
conditions and only comprehensible if we 
allow that in a former condition of the race 
the tadpole escaped from the egg and lived 
in the water. A still more striking instance 
of the same thing is shown by the salamanders 
of Europe. There are two of these, a yellow 
and black Salamandra maculosa, which lives 
in the valleys, and a black one, Salamandra 
atra, which lives high up on the mountains. 
In both cases the eggs are fertilized before 
they are laid and development begins whilst 
they are still inside the mother. The young 
of Salamandra maculosa when born possess 
feathery gills, a flattened tail, and they im- 
mediately take to the water; those of Sahi'- 
mandra atra, on the other hand, resemble 
their parents from the first. But if a preg- 
nant Salamandra atra be laid open there will 
be found inside her embryos provided with 
beautiful feathery gills, and these embryos if THE STUDY OF HEREDITY 
61 
taken from the mother and placed in water 
will occasionally survive and take up their 
life there as creatures breathing by means 
of gills. Who can entertain any doubt that 
the ancestors of Salairiandra atra once pro- 
duced tadpole-like larvse like those which its 
gaily-coloured cousin bears, but has been 
forced by the unfavourable circumstances 
under which it lives to nourish them in its 
body instead of casting them forth to seek 
their fortunes in the water? 
It is often asked, if the development of 
animals be a partial recapitulation of the 
history of the race, why the ancestral history 
of plants should not be revealed by their 
life history. The answer to this is twofold: 
first, sharply marked larval phases are found 
where an animal changes its environment and 
lives during the earlier part of its life under 
one set of conditions, and the later part 
under another set, and the earlier set of con- 
ditions then represents what was the environ- 
ment of the adult in the ancestral state of 
the race. A plant cannot move from one 
environment to another during its growth, 
and so when the surrounding conditions 
change they generally leave their impress on 
the plant from its first appearance above 62 
AN INTRODUCTION TO 
ground, and leave little or no room for a larval 
stage. Secondly, in some few cases plants 
have a larval stage, or, as botanists term it, 
a " youth-form." The common gorse, in which 
green thorns replace leaves, has, when young, 
leaves like other plants, but these drop off as 
it grows older, and the Australian acacia, in 
which flat green leaf-stems take the place of 
leaves, has likewise leaves like those of other 
acacias when young. 
If we then accept the view that in its 
broad outlines the life history of an animal 
is an epitome of the past history of the race, 
it becomes a matter of interest to investigate 
how this development is brought about in 
the individual. In a few cases we might 
plausibly accept the view that the develop- 
ment was caused by the reaction to new con- 
ditions. If the tadpole of a frog be kept 
cool and in the dark and prevented from 
rising to the surface of the water, it will, if 
well fed, grow very much in size, but it 
will never change into a frog. In order that 
it may metamorphose, it must have sunlight 
and be allowed to come up to the top of the 
water and swallow air; but when an embryo 
undergoes a parallel development to this 
inside the egg-shell or its mother's womb no THE STUDY OF HEREDITY 
63 
such explanation is available. The endeavour 
to analyze the causes of development in the 
individual, whether these be external stimuli 
or internal causes, forms the object of the 
science of experimental embryology. The 
principal objects which have been experi- 
mented with are the eggs of frogs, sea-squirts 
(Tunicata or Ascidians), molluscs, thread- 
worms (Nematoda), star-fish, and sea-urchins 
(Echinodermata), and comb-bearers (Cteno- 
phora). 
The eggs of the sea-urchin have formed 
the corpus vile for such experiments more often 
than those of any other form. The reason 
for this will readily be understood when the 
fact is grasped that our large British sea- 
urchin will produce in one season about 
20,000,000 eggs, which can be easily got to 
develop into a characteristic larval form 
termed the Echinopluteus, although to rear 
them into the adult condition is a very 
difficult matter. 
Amongst those who have investigated the 
causes of the development of the sea-urchin's 
egg no one has done more brilliant work than 
the German zoologist Driesch. Driesch has 
shown that when a sea-urchin's egg divides 
into two cells, if these cells be separated by 64 
AN INTRODUCTION TO 
violent shaking, each will develop into a perfect 
larva of reduced size. The same result is ob- 
tained if one of the first four cells be separated 
Fig. 10. 
Illustrating the structure and development of a comb- 
bearer (Ctenophore). A. Adult form. 1. "Rib." 2. 
Tentacle. 3. Mouth. 4. Apical sense organ. B. The 
eight-cell stage in the development of the embryo. 
from the rest, and even occasionally if one of 
the first eight cells be separated from the rest. 
But, on the other hand, if we take the eggs of 
the comb-bearers (Ctenophora) (Fig. 10) we THE STUDY OF HEREDITY 
65 
get a different result. These animals are 
shaped like elongated grapes. They pass their 
lives swimming in the sea; they are perfectly 
transparent, and they move by the aid of 
eight meridians of delicate comb-like struc- 
tures which converge to meet at the upper 
pole, where there is situated a delicate brain 
and sense organ. Each " comb " consists 
of a short transverse row of cilia or vibratile 
hairs united at the base but free at the tips. 
Each row of combs is known as a rib. As 
the Ctenophora come to the surface in a calm 
sunlit sea, these combs glisten with the most 
lovely iridescent colours, and the whole crea- 
ture presents a spectacle of great beauty. 
The first stage in the development of these 
animals is that the egg divides into eight cells 
ranged in a circle (Fig. 10, B). From each 
cell one-eighth of the animal, including one 
rib, is developed. If we remove one of the 
eight cells the remaining seven give rise to 
a creature with seven ribs; if we remove two 
a creature with six ribs is produced, and so 
on. 
How are these differences in the constitu- 
tion of the two eggs to be accounted for? The 
most plausible explanation which has been 
put forward is that there exist organ-forming 66 
AN INTRODUCTION TO 
substances, and that in each cell of the 
Ctenophore there is a substance which con- 
tains the materials necessary for the formation 
of a particular rib. In each of the first eight 
cells of the segmenting sea-urchin egg, on 
the contrary, there is the same kind of sub- 
stance, so that any one of them contains the 
necessary materials for constructing the whole 
body. 
The question then arises, in what part of 
the cell are these materials to be found, in 
the protoplasm or the nucleus? Now Driesch 
has shown that if a fertilized egg of a sea- 
urchin be allowed to develop under such 
pressure as is produced by being squeezed 
between two glass slides, the first eight 
cells, instead of forming two tiers of four 
each, as they do in the normal egg, are all 
spread out in one plane. If now the pressure 
be removed the eight cells divide into two tiers 
and thus the sixteen-cell stage is reached, and 
this mass of cells is capable of developing 
into a perfectly normal animal, although the 
nuclei which should have been situated at 
its lower pole are now placed at its sides. 
The nuclei can therefore be juggled about 
like a handful of marbles without altering 
the course of development, and therefore the THE STUDY OF HEREDITY 
67 
peculiar organ-forming substances must be 
in the protoplasm and not in them. This 
result is confirmed by some experiments 
which Driesch made on the Ctenophore egg. 
He cut pieces from these eggs before they had 
divided into cells at all, and he showed that 
when these mutilated eggs survived they gave 
rise to imperfect animals, in which the full 
number of ribs was not produced. The specific 
organ-forming materials for the ribs must be 
in the protoplasm, for there was as yet only 
one nucleus and it was not touched by the 
operation. 
These quite unexpected results were seized 
on by those who opposed the view that the 
nucleus was the sole bearer of the hereditary 
qualities as confirmatory of their objections. 
Of course, even if we were forced to admit 
that the bases for the qualities which the young 
animal inherits from its mother were situated 
in the protoplasm and not in the nucleus, we 
should still have to face the difficulty that 
the evidence compels us to believe that all 
qualities inherited from the father are carried 
in the head of the spermatozoon, which is a 
nucleus. 
Fortunately, further experiments enable us 
to get over this apparent contradiction in 68 
AN INTRODUCTION TO 
the constitution of the two kinds of germ- 
cells. The egg of the threadworm is a very 
specialized object. At its first division it 
gives rise to two cells, one of which is clear 
and the other granular. The clear cell gives 
rise to a part of the skin and to the nervous 
system; whilst the granular cell produces 
the internal organs. If the fertilized egg, 
however, be fixed to the rim of a horizontal 
wheel so that its axis of symmetry points 
towards the centre, and be allowed to develop 
whilst the wheel is kept in rapid rotation, 
it will divide into two granular cells which will 
develop into a double monster with two sets 
of internal organs, which cannot survive. It 
is obvious that the centrifugal force has caused 
the equal distribution of the organ-forming 
materials between the two cells, so that each 
in developing must give rise to the same 
result. 
But if the same treatment be applied to 
the unfertilized egg we find that it can lose 
large portions of its protoplasm and become 
much reduced in size, but that nevertheless 
if it be afterwards fertilized it will develop 
into a normal embryo. Its protoplasm at 
this period of its development is, therefore, 
homogeneous, and we can only attribute its THE STUDY OF HEREDITY 
69 
subsequent heterogeneity to its infiltration by 
materials given off from the nucleus, and we 
have already learnt that the material of the 
nucleolus is ejected into the protoplasm; 
these materials obviously give rise to the 
different organ-forming substances. 
This hypothesis is completely borne out 
when we examine the egg of the sea-squirt 
Cynthia. Here also we find strongly localized 
organ-forming substances. Indeed, in one spe- 
cies they are distinguishable in the living egg 
by their different colours. If we examine the 
ripening egg we find that streams of minute 
granules of chromatin are pouring into the 
protoplasm from the nucleus. Further, in 
the development of the species (Cynthia 
partita) in which the different organ-forming 
substances are distinguished by different 
colours it has been shown by an American 
zoologist (Conklin) that the ectoderm of the 
embryo is formed from the fluid contents or 
sap of the nucleus of the egg. When the egg 
ripens and the nuclear membrane is dissolved 
the escaping sap forms a cap of clear material 
at one side of the egg, and when its develop- 
ment is followed it is found to be changed 
into the ectoderm or primitive skin of the 
young animal. 70 
AN INTRODUCTION TO 
We therefore arrive at two conclusions: 
first, that the nucleus of the female germ-cell 
is just as much the bearer of the hereditary 
qualities as is the nucleus of the male cell; 
and, secondly, that the protoplasm of the 
growing egg becomes, in many cases, organized 
into substances which are the rudiments of 
the organs of the future animal by the 
emission of materials into it from the nucleus. 
Once the protoplasm has been organized 
in this way, the nucleus ceases for a time to 
have any further influence, but after the 
first-formed organ-forming substances are used 
up the nuclei at a later period of develop- 
ment resume their active role. 
If we now turn to examine the develop- 
ment of the sea-urchin we find that when the 
egg has divided into about 1000 cells these 
form a little hollow balloon or vesicle known 
as the blastula (Fig. 11). Then one side 
of the blastula becomes flattened and cells 
termed mesenchyme cells are budded from 
the flattened area and pass into the interior. 
Immediately afterwards the rudiment of the 
future gut is produced by the pushing in of 
one side of the blastula as if by an invisible 
finger. The in-turned portion forms the lining 
of the gut, and the embryo is now termed a THE STUDY OF HEREDITY 
71 
gastrula. Now Driesch has shown that the 
blastula may be cut in pieces with a fine pair 
of scissors, and each piece if above a certain 
minimal size will develop into a perfect larva. 
But if the operation be repeated after the 
mesenchyme has been produced, this is no 
Fig. II. 
Three stages in the development of the egg of a sea- 
urchin. A. The blastula stage. B. The formation of 
mesenchyme cells. C. The gastrula. 1. The mesenchyme 
cells. 2. The gut. 
longer the case; then fragments of the 
blastula will, indeed, heal up and produce 
smaller blastulas, but these are incapable of 
further development-only the piece taken 
from the region where the mesenchyme has 
been produced can form a gut. The wall of 
the blastula, which was at first homogeneous, 
has now become heterogeneous, and one 
possible explanation of this is the assumption 72 
AN INTRODUCTION TO 
that both skin-forming and gut-forming sub- 
stances are at first found in every cell of the 
blastula, but gradually the gut-forming sub- 
stances flow down, as it were, towards one 
end and become concentrated there, and the 
skin-forming substances remain in a pure 
state at the other end. The cells forming 
the blastula are, it may be remarked, connected 
with one another by bridges of protoplasm. 
That this is the true explanation is ren- 
dered almost certain by the result of the 
experiment of allowing sea-urchin eggs to de- 
velop in sea-water to which a small quantity 
of the salts of lithium have been added. 
Under these circumstances the gut grows 
outwards instead of inwards, and a very 
large portion of the blastular wall is used 
up in forming the gut, so that the skin- 
forming portion is reduced to a mere button. 
This result is only obtained if the eggs are 
placed in the lithium solution soon after 
fertilization and kept there till the blastula 
stage has been reached. It is obvious, then, 
that it is during this stage that these sub- 
stances are formed and that the absorption of 
lithium compounds causes a great deal more 
gut-forming substance to be made than is usu- 
ally the case. THE STUDY OF HEREDITY 
73 
These organ-forming substances are, indeed, 
amongst the most striking discoveries of bio- 
logical science, and we are as yet only on the 
threshold of our knowledge of them. Whether 
a certain mass of such a substance will give 
rise to one organ or two of the same kind 
seems to depend on its shape. If, for instance, 
we take the egg of a frog and allow it to di- 
vide into two cells, and then clamp it between 
two slides and reverse it, we may get a 
double-headed or a double-tailed tadpole. If 
we do the same with the egg of the newt we 
may get a complete double embryo. What is 
the explanation of this phenomenon? We have 
added nothing and removed nothing, we have, 
in fact, only turned the embryo upside down. 
The answer to this conundrum is that in the 
frog's egg there are two organ-forming sub- 
stances which stand in a fixed spatial relation 
to one another, because of their different 
specific gravities, so that if the egg be turned 
upside down they sort themselves out under 
the influence of gravity. But if the egg be 
turned upside down when it has divided into 
two cells, then, since the contents of these 
cells are to a certain extent cut off from one 
another, the materials in each tend to sort 
themselves in the same way as if each cell 74 
AN INTRODUCTION TO 
were a separate egg, and hence the tendency 
to produce two embryos. We can attain the 
same end by constricting not the egg, but the 
blastula of the newt, by a fine hair. If this be 
done in such a way as to constrict the front 
end longitudinally a two-headed monster is 
produced. In this case there can be no 
possible doubt that shape is the determining 
factor in deciding whether a normal or a two- 
headed embryo is to be produced. 
Another instance of the same thing can 
be seen in the tail of the lizard. When this 
is broken off a little bud is formed on the 
surface of the stump by which a new tail is 
produced. If the surface of this bud be 
slightly indented by the prick of a knife it will 
produce two tails. 
When once the primary organs have been 
formed, they seem to emit substances which 
act on each other and produce further modifi- 
cations by which the secondary organs are 
formed. 
One instance of this must suffice. All 
students of animal biology are aware that the 
type of eye found in man and the higher 
vertebrates is a complex organ consisting 
of two main portions, one formed as an out- 
growth from the brain and constituting the THE STUDY OF HEREDITY 
75 
retina, and one formed as an ingrowth from 
the skin and constituting the lens. The 
retinal portion appears first in development, 
and if this portion be cut off in a young 
unhatched tadpole and planted under the 
skin in another part of the body, it wTill cause 
the skin above it to form a lens. We are, 
therefore, driven to the conclusion that the 
formation of a lens in the normal develop- 
ment of the tadpole is not due to a special 
quality of the skin in the region of the head, 
but is the result of the modification of the 
skin in this place under the influence of the 
growing retina beneath it. 
These influences are known as formative 
Stimuli. Even less is known about them 
than about organ-forming substances, but 
their further investigation would be a most 
fascinating study. Their existence, however, 
shows quite clearly that Weismann's con- 
ception of the growth of the body as the 
development of a mosaic pavement-like ar- 
rangement of rudiments, each predestined to 
form a distinct organ, is fundamentally 
wrong. Just as the sea-urchin blastula in 
its earlier stages is a mass of relatively homo- 
geneous material, any part of which can form 
the whole animal, so the primary tissues are 76 
THE STUDY OF HEREDITY 
relatively homogeneous, and what organs they 
will give rise to depends to a large extent 
on the influences to which they are subject. 
To sum up, we conclude that the bearers 
of hereditary influence are the chromosomes 
of the nuclei of the male and female germ- 
cells. These chromosomes are brought to- 
gether in fertilization, and cause the develop- 
ment of the offspring by the emission of sub- 
stances into the protoplasm which change 
its character and confer on it the power to 
develop into the primary organs. These 
primary organs in turn emit materials into 
the fluids of the body which affect parts of 
other primary organs so as to give them the 
power of developing into the secondary organs, 
and so the adult form is attained. Therefore 
we must always remember that in looking 
for the reappearance of parental characters 
in the offspring, we must not only compare 
with one another corresponding stages of 
development, but also the offspring must be 
exposed to the same conditions during its 
development as the parent. Adverse con- 
ditions by interfering with the complex 
mechanism of development may altogether 
prevent the appearance of parental characters. CHAPTER IV 
THE INHERITANCE OF HABITS AND OF THEIR 
EFFECTS 
In the three preceding chapters we have 
studied the material bridge between parents 
and offspring. We have seen ourselves driven 
to the conclusion that the bearers of the 
hereditary qualities are to be found in the 
nuclei of the germ-cells, and that the essential 
parts of the nucleus are constituted by the 
chromosomes, which undergo great changes 
during the growth and -division of the cell, 
but which retain their individuality through- 
out them all. If, as we have already said, 
like always produced like, the problem of 
heredity would be a simple one, but children 
do not completely resemble their parents, and 
the task of the 'science of heredity is to find 
out the causes of this divergence. 
Now one of the oldest ideas on this subject 
was that the experience which the parent 
underwent, especially during the time shortly 
before the birth of the child, affected the 
nature of the child. Thus a pregnant mother 
77 78 
AN INTRODUCTION TO 
who had constantly seen a cripple passing 
and repassing her house gave birth to a 
deformed child, and this pre-natal impression 
was given as the reason. In its extreme form 
this doctrine finds few defenders to-day, but 
the question whether the habits and activities 
of the parents have any effect on those of the 
child is at present one of the most hotly 
debated questions in zoology and it is one of 
fundamental importance. The doctrine that 
habits and their effects are inherited was first 
clearly stated by the French naturalist 
Lamarck in 1810, and this theory is usually 
called Lamarckism. Lamarck stated his views 
as follows: 
(1) Every fairly considerable and perma- 
nent alteration in the environment of a race 
of animals works a real alteration in the needs 
of that race. 
(2) Every change in the needs of animals 
necessitates new activities on their part for 
the satisfaction of those needs and so new 
habits are formed. 
(3) Every new need which necessitates new 
activities for its satisfaction requires the 
animal either to make more frequent use of 
some of its parts which it previously used 
less, and thus greatly to develop or enlarge THE STUDY OF HEREDITY 
79 
them, or else to make use of entirely new 
parts to which the new needs have imper- 
ceptibly given rise by inner feeling. 
(4) All the acquisitions or losses wrought 
by nature on individuals through the influence 
of the environment in which their race has 
long been placed-all these are preserved by 
reproduction to the new individuals which 
arise. 
If Lamarck's views were to be accepted a 
great part, if not all, of the evolution of 
animals and plants from simple ancestors 
could be explained, for the principal differ- 
ences between allied organisms consist in their 
different habits and the structures associated 
with those habits, and the more closely animals 
are examined in their natural surroundings, 
the more clearly does this fact come out. 
But Lamarck's theory has been vehemently 
denied; and other alternative explanations of 
evolution have been put forward. Of these 
there are two which may be termed the 
Darwinian and the Mendelian, although Dar- 
win accepted the truth of Lamarck's theory 
and Mendel did not deal with evolution at all, 
but only with the laws governing the charac- 
ters of hybrids produced when two races of 
the same species were crossed. 80 
AN INTRODUCTION TO 
Darwin pointed out that if members of the 
same litter, all of which are produced at the 
same time, be closely examined it will be 
seen that no two are exactly alike, but that 
each one differs slightly from its fellows in 
its shape or colour. He assumed that these 
differences were inheritable and that some 
characters would better suit the environment 
to which the young animal was to be exposed 
than others; those fortunate -members of the 
litter which chanced to possess these char- 
acters would survive and propagate the race, 
whereas the others would perish prematurely. 
Thus 'the " fittest " would survive, and this 
process was likened by Darwin to the selection 
which a breeder practises when he chooses the 
individuals of his stock which he wishes to 
mate together in order to obtain progeny, and 
so he termed it "natural selection.'' To 
this selection he attributed the major part of 
evolution, the inheritance of the effects of use 
and disuse (Lamarck's principle) occupying 
only a subordinate place in his idea of the 
causes of evolution. 
Mendel, as we have seen, dealt only with 
the qualities of hybrids, but when that con- 
siderable school of modern biologists who are 
usually termed Mendelian began to repeat THE STUDY OF HEREDITY 
81 
Mendel's experiments, they soon discovered 
that amongst any large collection of domesti- 
cated animals or of cultivated plants indi- 
viduals showing marked divergences from the 
type " crop up " occasionally, and that when 
these individuals are mated with typical ones, 
the offspring obey the Mendelian rules govern- 
ing hybrids. No reason has been assigned for 
the appearance of these abnormal individuals 
by Mendelians other than " chance " or " acci- 
dent." Nevertheless to the occurrence of such 
" accidents " they do not hesitate to attribute 
the whole process of evolution. 
We must postpone to a later chapter the 
critical examination of the efficiency of na- 
tural selection and of Mendelian " accidents " 
or " mutations," as they are called, as agents 
in evolution, and return to the question of the 
inheritance of habits-or, as they are usually 
termed, " acquired, characters." 
The early exponents of evolution, such as 
Darwin and Haeckel, accepted Lamarck's 
principle. It was first explicitly denied by 
Weismann, and that on two grounds: viz., 
(1) Weismann could imagine no mechanism 
by means of which changes in the structure 
of the body resulting from changed habits 
could be transformed into representative 82 
AN INTRODUCTION TO 
changes in the germ-cells; (2) the experi- 
ments which Weismann instituted seemed to 
prove that acquired characters were not in- 
heritable. 
We shall deal with (2) first. Weismann 
cut the tails off young white mice, reared 
them to maturity and bred several genera- 
tions of offspring from them. None of these 
offspring showed the slightest sign of any 
deficiency in the tail. But it may be at once 
objected that to lose a tail by mutilation is 
not to acquire a habit; and the experiment 
is therefore entirely irrelevant. As to (1), 
we may say that it is likewise irrelevant, for 
no one has any clear idea as to the mechanism 
by means of 'which the chromosomes of the 
paternal and maternal nuclei in the fertilized 
egg give rise to the bodily organs of the adult; 
this undoubtedly occurs, and yet it is as 
incomprehensible as the transference of the 
qualities of the parent to the egg. 
Inconclusive as was Weismann's reasoning 
it was widely accepted, and it became for a 
time almost a biological dogma that under no 
circumstances could the effects of habits be 
handed on to posterity. It is therefore not 
to be wondered at that when new evidence 
was offered to the effect that, after all, THE STUDY OF HEREDITY 
83 
acquired characters are passed on to pos- 
terity, this evidence aroused the most violent 
opposition; for scientific men dislike as much 
as religious men to be awakened from a 
dogmatic sleep. The researcher who first 
questioned the accepted belief was openly 
charged with fraud; and this facile way of 
dealing with the subject, adopted by a 
leading English biologist, was at once copied 
by a large number of his less original fol- 
lowers. During the recent war the question 
could not be probed further, as Kammerer, 
the author of the new researches alluded 
to, is a native of Vienna and all his work 
has been done in the Institute of Experi- 
mental Biology in that city. But quite re- 
cently Kammerer has visited England, bring- 
ing his critical specimens with him, and 
has demonstrated these to English scientific 
men, and all who have discussed the matter 
with him are convinced that he is an upright 
and skilled investigator. It is still more 
important to observe that precisely similar 
results to those obtained by Kammerer have 
been reached by two other investigators who 
dealt with widely different animals, one of 
them working in Moscow and one in Breslau. 
In neither of these two latter cases would it 84 
AN INTRODUCTION TO 
occur to anyone to bring forward the hypoth- 
esis of fraud to account for the results, the 
more so as the Russian scientist before taking 
up the question of inheritance at all had ac- 
quired world-wide fame as one of the leading 
physiologists of the day. The reader will thus 
observe that a great mass of careful scientific 
experiment has been accumulated which dem- 
onstrates the inheritability of the effects of 
habit. When this mass of evidence has had 
time to penetrate the scientific world and be 
appreciated at its proper value it will revolu- 
tionize many of the current ideas on heredity. 
We shall now give an account of some of 
Kammerer's experiments which demonstrate 
Lamarck's principle in the clearest way. We 
have already mentioned (see Chap. II) that 
there are two species of salamander commonly 
found in Europe, viz., Salamandra maculosa 
and Salamandra atra. The former is of a 
black colour spotted with patches of yellow, 
but the relative quantities of these two 
pigments vary very much; some specimens 
are nearly black with a few minute dots 
of yellow, whereas in others the yellow pre- 
dominates so much that the black is nearly 
suppressed. A fairly common variety has 
received the name toeniata, because in it the THE STUDY OF HEREDITY 
85 
yellow forms two conspicuous longitudinal 
stripes on the back separated by a median 
black stripe, but this variety is not found wild 
in the neighbourhood of Vienna, for it only 
occurs in the Jura mountains and westward 
of them, and when Kammerer began his 
researches he had never seen it. The adult 
animal lives on land in damp nooks, under 
moss or among stones; it is often difficult to 
find, but sometimes after a heavy thunder 
shower hundreds will be found where not one 
could be seen before. It is an inhabitant of 
the warmer regions near the sea-level in 
Austria, Italy and Southern Europe generally. 
The animal is born as a larva which lives in 
the water. This is of the same general shape 
as the adult, but is distinguished by the 
possession of three feathery gills on each side 
of the neck between which are slits leading 
into the throat, and by having a flattened tail 
suitable for swimming. The larva is of a 
yellowish colour owing to a plentiful develop- 
ment of yellowish-white pigment, but this 
pigment is quite distinct from the adult pig- 
ment, which only becomes conspicuous after 
the creature leaves the water and the meta- 
morphosis has been accomplished, although 
traces of it can be seen in the larva. Traces 86 
AN INTRODUCTION TO 
also of the adult black can be detected in the 
larva, but it, like the yellow, increases in 
quantity only after the metamorphosis. 
Salamandra atra is an inhabitant of the 
colder upland regions among the Alpine 
meadows. It is of a uniformly black colour, 
and, as already mentioned, brings forth its 
young in a shape completely resembling the 
adult, without gills and with a rounded tail. 
Preliminary experiments showed that 
young, just metamorphosed specimens of Sala- 
mandra maculosa slowly altered their colora- 
tion as they grew older, if they were kept living 
on black or yellow backgrounds strongly illu- 
minated from above. In those kept on black 
backgrounds the black increased in quantity 
and overshadowed the yellow, whilst in those 
kept on yellow backgrounds the reverse change 
took place. But if the background were of 
any other colour, or even of black and yellow 
patches in even proportion, or, lastly, if it 
were not brightly illuminated, no change 
whatever took place; so that when an experi- 
ment was being carried out it was only neces- 
sary to enclose a similarly coloured salamander 
in a neighbouring cage from which bright light 
was excluded in order to establish a perfectly 
satisfactory " control " experiment. THE STUDY OF HEREDITY 
87 
Long experience and many unfortunate 
failures were required before the best way of 
keeping the beasts healthy was discovered. 
The larvae were always kept under neutral 
conditions, i. e., they were kept in diffused 
light and were not exposed to the influence of 
black or yellow backgrounds. They were fed 
with blood-worms (Tubifex), which could be 
procured in large quantities. The adults, on 
the contrary, were kept in boxes of consider- 
able size, lit from above. These were kept 
damp by basins of water, one of which was 
placed in each of them, and in this the sala- 
mander could bathe itself; in each there 
was also a " den " of moss and stones in 
which the animal could hide. The food given 
consisted of meal-worms (the grub of a beetle 
which infests flour mills) ; these were placed 
in a little dish at some considerable distance 
from the " den," so that the salamander had 
to make frequent journeys across the open to 
get them. 
The whole series of experimental cages were 
placed in a sunny gallery in the laboratory of 
the Institute of Experimental Biology at 
Vienna. It is therefore clear that for the 
successful performance of an experiment of 
this kind a very considerable " plant " is 88 
AN INTRODUCTION TO 
necessary, and it will demand practically the 
whole of the experimenter's time. Indeed, 
Kammerer says expressly that anyone who 
desires to repeat his experiments and test 
their validity must be prepared to devote a 
large part of his life-time to the work. 
The nature of the experiments performed 
was as follows: 
The floors of some of the cages were covered 
with yellow loam and the sides were painted 
yellow. The animals confined in them were 
thus exposed from the beginning to adult life 
to the influence of a yellow background. The 
floors of other cages were strewn with black 
garden earth and their sides were painted 
black, so that the salamanders in them lived 
on a dark background. 
When individuals chosen for their predomi- 
nantly black colour, which had been reared 
on a background of yellow loam, gave rise to 
young-which only happened when the parents 
were four years old-these young were reared 
as larvae on neutral backgrounds. When they 
had undergone metamorphosis they were 
divided into two lots, one lot being reared 
further on a background of yellow loam, and 
the other on a background of black garden 
earth. Those which were reared on yellow THE STUDY OF HEREDITY 
89 
earth became increasingly yellow. The yellow 
pigment, which in the parents had increased 
till it formed two longitudinal rows of large 
Fig. 12. 
Illustrating the effect of a background of yellow loam 
on a comparatively black salamander. A. Young form 
just after metamorphosis. B. Fully-grown form. 
spots (Fig. 12, B), grew in the children to 
such an extent that the rows of spots became 
amalgamated so as to form longitudinal 90 
AN INTRODUCTION TO 
stjeaks, and finally the streaks became con- 
nected by transverse ridges until the whole dor- 
pal surface was predominantly yellow (Fig. 
Fig. 13. 
Illustrating the effect of a background of yellow loam 
on the offspring of salamanders born from those repre- 
sented in Fig. 12. A. Young form just after meta- 
morphosis. B. Adult form. 
13, B). In those young which were reared on 
black garden earth, the yellow pigment made 
its appearance as two longitudinal lines of THE STUDY OF HEREDITY 
91 
spots, and these spots for some considerable 
time increased in size owing to the persistence 
Illustrating the effect of a background of black garden 
mould on the offspring of salamanders which had been 
reared on a background of yellow loam. A. Young 
salamander five months after metamorphosis, showing 
influence of the yellow background to which its parents 
were exposed. B. Adult salamander, showing influence 
of the black background. 
Fig. 14. 
of the influence of the yellow background on 
which their parents had lived (Fig. 14, A); 92 
AN INTRODUCTION TO 
but as the young generation grew older the 
spots began to diminish in size, and finally, 
when maturity was attained, the yellow pig- 
Illustrating the effect of exposing the offspring of 
salamanders reared on black earth to the same influences 
as those to which their parents were exposed. A. Just 
metamorphosed form. B. Adult form. 
Fig. 15. 
ment was reduced to two longitudinal rows of 
small spots (Fig. 14, B). 
If, on the contrary, salamanders were THE STUDY OF HEREDITY 
93 
chosen which had comparatively large yellow 
spots, and these were reared on a background 
of black garden earth, the yellow spots 
diminished in size till by the time that ma- 
turity was attained many of them had en- 
tirely disappeared, and those which remained 
had become very small. The offspring of 
such salamanders, just before and at the time 
of metamorphosis, were distinguished by the 
possession of a single median row of small 
yellow spots (Fig. 15, A). If these young 
were allowed to grow up on a background of 
garden earth the yellow spots became smaller 
and finally almost disappeared-indeed, on the 
ventral surface they totally disappeared, and 
an almost perfectly black salamander was 
produced, in appearance closely resembling 
Salamandra atra (Fig.15, B). If, on the con- 
trary, the offspring of salamanders reared on 
black earth were transferred to yellow loam, 
the median row of spots slowly enlarged and 
finally fused to form a single longitudinal 
stripe of yellow, but it will be noticed that 
this distribution of yellow and black was 
totally different from that observable in the 
parent with which the experiment started. 
As in the case of the experiment which began 
by placing salamanders on yellow earth, the 94 
AN INTRODUCTION TO 
influence of the environment made itself felt 
in the constitution of the young, although 
these might be born and pass the first part of 
their life under neutral conditions. 
It occurred to Kammerer that some part 
of the influence of the yellow and black back- 
grounds might be due to the amounts of 
water they contained-in a word, to their 
comparative dampness. He determined ex- 
perimentally that the yellow loam was much 
more hygroscopic, and therefore damper than 
the black earth. In order to get rid of this 
factor he replaced in some of his experiments 
the yellow and black earth with yellow and 
black paper. For the experiment in which 
yellow paper was to be the background sala- 
manders were chosen in which black predomi- 
nated; the result of the experiment was that 
the yellow increased in amount-but solely 
by the increase in size of spots which were 
already in existence-practically no new yellow 
spots made their appearance. The offspring 
of these parents reared also on yellow paper 
showed the same phenomenon; the large spots 
amalgamated into broad yellow stripes, but 
the yellow was of a dusky appearance, show- 
ing that it was mixed with the black pigment; 
in a word, as no new yellow pigment spots had THE STUDY OF HEREDITY 
95 
been generated, the existing pigment had been 
spread out in a very thin layer. When the 
young of parents reared on yellow paper were 
reared on a background of black paper, very 
much the same effect was produced as when 
they were reared on a background of black 
earth, but there was this difference, that the 
yellow spots diminished in size, not by a 
gathering together of the yellow pigment and 
a drawing in of the root-like outgrowths of 
the yellow spots, which is what happens over 
a background of black earth, but by a darken- 
ing of the borders of the spots due to an 
invasion of them by the black pigment 
(melanin). It was thus obvious that some 
other factor in the background besides that of 
mere colour played a part in bringing about 
the result; and it appeared probable that 
this was the moisture of the environment. In 
order to test this a background was chosen 
which produced no colour reaction in sala- 
manders which were raised on it. A very 
suitable one was found in grey river sand, 
which could, of course, vary very much in 
the quantity of moisture which it contained 
without sensibly altering in colour. When 
young salamanders, chosen for the small num- 
ber of yellow spots which they possessed, 96 
AN INTRODUCTION TO 
were placed on sand so wet that the air in the 
cage had a humidity of from 80 to 90 per 
cent, of saturation, then the yellow spots did 
not increase in size, but a great number of 
small new spots made their appearance, espe- 
cially on the under surface, where ordinarily 
few yellow spots are to be found. The off' 
spring of these salamanders, just after meta- 
morphosis, were distinguished from all other 
salamanders by the great number of minute 
yellow spots peppered over them. When these 
salamanders were transferred to a background 
either of yellow earth or yellow paper all the 
existing spots began to increase rapidly in 
size. On the other hand, when salamanders 
characterized by having numerous yellow spots 
were reared on a background of relatively dry 
river sand in an atmosphere the humidity of 
which did not reach more than 50 per cent, 
of that required for saturation, these spots all 
diminished in size, not, as on black earth, by 
contraction, but by the invasion of their bor- 
ders by black pigment, so that the yellow 
colour became fainter and eventually dis- 
appeared. 
It appears, therefore, that the presence of a 
yellow background causes the yellow spots to 
increase in area, to throw out roots and THE STUDY OF HEREDITY 
97 
eventually to join with one another, whilst 
the presence of a damp atmosphere causes 
the production of new yellow spots, and that 
both these changes are to a considerable 
extent inherited, insomuch that the offspring 
start their career at a phase nearly, but not 
quite, as advanced as that at which their 
parents left off. 
The question now arises as to the channel 
through which the environment acts on the 
animal. Does it take place through a direct 
action on the skin, or is it indirect through 
the influence of the environment on the eyes 
of the animal? That such influence can take 
place through the eyes is known from the 
behaviour of the plaice and other flat-fish. 
These fish are able to alter the coloration of 
the upper surface according to the colour of 
the ground on which they lie. Now it is well 
known that when a flat-fish is blinded it 
remains pale and does not take on the colour 
of its surroundings. The answer to the ques- 
tion so far as the salamander was concerned 
was found by Kammerer by a series of the 
most ingenious experiments. 
He first found out what occurred when a 
piece of skin was removed and the wound 
became covered with new skin. If a piece of 98 
AN INTRODUCTION TO 
yellow skin which had been originally black 
and had been turned yellow by the environ- 
ment was removed and the animal continued to 
live on a yellow background, then the new 
skin when first formed was of an olive-green 
colour, due to the presence of both yellow 
and black pigments in equal quantities, the 
black pigment being situated in a more super- 
ficial layer and the yellow pigment showing 
through it, but soon the new skin became 
bright yellow under the continuous influence 
of the yellow background. If a piece of black 
skin which was formerly yellow were removed, 
the new skin began by being yellow and re- 
mained so if the animal was placed on a yellow 
background. If a piece of yellow skin which 
was formerly black were removed and the 
animal were placed on a black background the 
new skin began by being black and remained 
black. If, however, the skin removed were 
black, but had been yellow, then, if the animal 
continued to live on a black background, the 
new skin when first formed was flecked with 
dusky yellow spots, but soon became totally 
black. Kammerer concludes that the regener- 
ated skin tends to repeat the history of that 
which it replaces, but that this history is 
shortened and may be even entirely sup- 
pressed. THE STUDY OF HEREDITY 
99 
Then the experiment was made of trans- 
planting a black piece of skin from a black 
salamander living on a black background to a 
bare place on a yellow salamander living on a 
yellow background. The foreign skin throve 
in its new situation and the wound healed. 
If the animals were transferred to neutral 
backgrounds no change at all took place in 
the colours of the transplanted skin; but if 
the animal carrying the black skin were 
removed to a background of yellow loam, 
then quickly numerous small yellow spots 
began to appear, but they did not grow any 
bigger. After a period of fourteen or fifteen 
months of living on this background the spots 
suddenly began to enlarge and eventually the 
whole patch of skin became yellow. Kam- 
merer interprets this remarkable result as 
follows: The appearance of yellow spots on 
black skin was due to the effect of the humidity 
of the loam acting directly on the skin, but in 
order that the yellowness of the loam should 
produce its effect on the black skin, new 
nervous connections must be formed, and this 
was a process which required time. He con- 
cludes, therefore, that the background, so far 
as it is yellow, produces its effect through the 
eye of the animal. The correctness of this 
conclusion was confirmed by keeping yellow 100 
AN INTRODUCTION TO 
animals, and to each of which a piece of black 
skin had been transplanted, on backgrounds 
of wet sand and yellow paper respectively. 
On the first background the black skin began 
at once to develop small yellow spots, due to 
the direct influence of the humidity of the 
background on the skin; on the second no 
change took place for a year; but then, if 
there had been small yellow spots on the 
black skin when it was removed from its 
former host, these began suddenly to increase 
in size, owing to the influence of the yellow 
colour of the background indirectly on the 
skin through the eye. 
But Kammerer's experiments were by no 
means limited to the investigation of the 
inheritance of colour. He succeeded in alter- 
ing the whole of the breeding habits of both 
Salamandra maculosa and Salamandra atra. 
When Salamandra atra was kept in warm and 
moist surroundings, instead of producing only 
two youngsters at a birth, one from each side 
of the body, it began to produce more, first 
three, and then at subsequent seasons four or 
five or even six young. This change is ren- 
dered possible by the mode of embryonic de- 
velopment. At each breeding period about 
six eggs pass from each ovary into the egg- THE STUDY OF HEREDITY 
101 
tube beneath it, but of these six only the 
lowest on each side develops into an embryo, 
the remainder degenerate into a kind of yolky 
soup, which serves as food for this embryo. 
When, however, the parent was kept warm and 
moist, more of the eggs developed and the 
young which developed from them were born 
whilst they were still provided with gills; these 
young took to the water and lived in it like 
the young of Salamandra maculosa. When 
these young were reared to adults under the 
same conditions and allowed in turn to pro- 
duce young, they gave birth to still larger 
broods than did their parents, and these young 
of the second filial generation were born at as 
early a stage as those of Salamandra macw- 
losa. Conversely, if Salamandra maculosa was 
kept in comparatively cool dry surroundings, 
it produced at each breeding season fewer 
young, and these were born at a later stage of 
development. These young when reared under 
similar conditions to those to which their 
parents were exposed produced still smaller 
broods, and eventually gave rise at each birth 
to only two or three fully-developed young in 
which the gills were almost completely absorbed 
-only small stumps of them remained, just as 
is the case with Salamandra atra. It is still 102 
AN INTRODUCTION TO 
more interesting to find that if Salamandra 
maculosa be kept in semi-tropical conditions in 
which the warmth and moisture are both 
greater than those to which it is accustomed, 
its young become more numerous and are born 
at still earlier periods of development, and 
these young, reared to adults and allowed to 
pair, give rise to eggs from which issue larvae 
with only the forearms developed. Such is 
normally the case in the allied genus Triton 
(Molge), to which our English newts belong. 
The so-called midwife toad Alytes was also 
the subject of another of Kammerer's experi- 
ments. Ordinary frogs and toads pair in the 
water, the male clasping the female round the 
waist, and remaining in this position sometimes 
for weeks, till the female emits the eggs, when 
he fertilizes them by a simultaneous emission 
of sperm. The young are hatched from the 
eggs as very small larvae, which at first carry 
on each side of the neck three feathery gills 
like those of the salamander larva, but these 
gills soon become covered up by a fold of 
skin growing back from the side of the head, 
and the larvae then assume the rounded plump 
form so familiar to all as the tadpole. Alytes, 
however, pairs on the land, and when the 
eggs are emitted, as they are in the case of all THE STUDY OF HEREDITY 
103 
kinds of toads, in long strings, the male winds 
these strings round his legs after he has 
fertilized them, and he carries these strings 
about with him till the young hatch. These 
hatch as tadpoles which have passed through 
the stage with feathery gills whilst still 
embryos within the egg membrane, but whilst 
in this stage the embryo only develops one 
large feathery gill on one side. If, now, Alytes 
be kept in a warm place, however dry it be, 
it will thrive provided it is given access 
to a pool in which it can occasionally bathe 
itself. Under these circumstances it breeds 
in the water and the egg-strings slip off the 
legs of the male and lie in the water. Most 
of the eggs perish, but by transferring them 
to sterilized water some may be reared. Alytes 
in Vienna breeds twice a year, and it is soon 
noticed that at each successive breeding period 
more eggs are produced than usual and these 
are of a smaller size than those normal to the 
species. From these smaller eggs tadpoles 
emerge in the stage in which each carries one 
large feathery gill on each side. If these 
young be reared to adults they at once resort 
to the water to pair and produce eggs from 
which young are hatched with three feathery 
gills on each side of the neck, as occurs in 104 
AN INTRODUCTION TO 
other forms of toad. In this generation also 
the males develop a horny pad on the inner 
side of the palm of the hand and of the 
fingers. This pad enables the male to retain 
his hold on the female during the nuptial 
embrace. A similar pad is developed by the 
males of all other species of frogs and toads 
which pair in water; but the normal Alytes 
male does not develop it because, since he 
pairs on land, the skin of his partner remains 
dry and horny, and he does not need it for the 
purpose of maintaining his embrace. An 
indication of this pad is seen in males of the 
second generation reared from eggs lying in 
water, but only becomes marked in the third 
generation of such males. Kammerer ex- 
hibited a male with such a pad in London in 
1923. 
It has been suggested that Kammerer's 
results on Salamandra maculosa may have 
been due to the influence of yellow light 
reflected from the yellow background pene- 
trating the body of the animal and acting 
directly on the germ-cells. In order to test 
this a pupil of Kammerer, Secerov, dissected 
out the ovaries of female salamanders and 
inserted in their place sealed glass tubes con- 
taining each a roll of photographic paper. THE STUDY OF HEREDITY 
105 
After remaining exposed to yellow light for a 
period of months the salamanders were killed 
and the tubes extracted and examined. The 
photographic paper showed a very faint indi- 
cation of darkening. It is thus seen that the 
light which reaches the ovaries through the 
tissues is of minimal amount and utterly 
unlikely to have any effect on the egg-cells. 
Kammerer himself brought further proof that 
the egg-cells are affected through the medium 
of the body in a series of most painstaking 
and difficult experiments. He transplanted 
ovaries from females which were striped to 
females which were spotted, and vice versa, 
and he crossed these females with males 
similar to them in colour. The transplanta- 
tion of ovaries rather than testes was under- 
taken, because the testis of the salamander is 
an elongated, lobed organ, and it was difficult 
to make certain that it W'as completely 
removed. When the young produced by 
these crossings were examined it was found 
that they were sometimes influenced by what 
we may term the foster-mother and sometimes 
not. Upon examining the matter more closely 
it was found that if the foster-mother were 
the natural striped variety, and if she were 
supplied with the ovary of the spotted variety, 106 
AN INTRODUCTION TO 
then she had no influence on the offspring 
whatever, which turned out exactly as if the 
eggs had still been carried by the female which 
produced them. If, however, the female were 
the artificially produced striped variety which 
was supplied with the eggs of the spotted 
variety, then the young turned out striped 
like the foster-mother. This result was pro- 
duced even though the foster-mother during 
the time of the experiment was kept in dim 
light, so that all direct action of the environ- 
ment was excluded. In the results of these 
experiments then we have a proof of the direct 
action of the body on the germ-cells, which 
is, after all, the crucial point in the question 
of the inheritance of acquired qualities. It 
will be observed that in Kammerer's experi- 
ments the newly-acquired qualities are those 
which are most potent in influencing the germ- 
cells. 
Kammerer's experiments are unique in their 
extent and in the length of time (eleven to 
twelve years) wrhich has been devoted to their 
execution, but other workers have obtained 
confirmatory results in the case of other 
animals. 
Thus Durkhem showed that pupse of the 
common white butterfly are usually greyish THE STUDY OF HEREDITY 
107 
in colour owing to the deposition of a quantity 
of black and opaque white pigments in the 
pupal integument. But when these pupae are 
exposed during the latest caterpillar stage to 
the influence of orange light, then the deposi- 
tion of these pigments is prevented and the 
pupae appear green owing to the green blood 
of the insect shining through. Not all the 
pupae show this reaction; ordinarily only about 
65-70 per cent, of them turn green. When 
however, these pupae give rise to butterflies 
and these lay eggs, and the pupae reared from 
these eggs are exposed to orange light, about 
95 per cent, turn green. If, on the other 
hand, these pupae of the second generation are 
allowed to grow in ordinary white light, then 
about 40 per cent, turn green, whereas in 
control pupae, the parents of which were never 
exposed to orange light, the number of green 
caterpillars did not exceed 4 per cent. 
Pavlov experimented with white mice, to 
which he endeavoured to teach an association 
of ideas. They were fed once a day, and 
each time food was put into their cages an 
electric bell was rung. It required 300 lessons 
to impress on these mice the meaning of the 
ringing of the bell. When, however, a first 
" filial " generation of mice was bred from 108 
AN INTRODUCTION TO 
them, 100 lessons sufficed to teach these 
younger mice to seek their food when the 
bell was sounded. In the second filial gener- 
ation this result was accomplished in 30 
lessons and in the third filial generation in 
5 lessons. 
Kammerer's, Durkhem's and Pavlov's ex- 
periments thus all give the same result, viz., 
that the exercise of a habit on the part of an 
animal renders the formation of the same 
habit easier in the offspring. The pigmentary 
reaction to light and the association of food 
with sound are both habits. Of course the 
newly-inherited habit does not persist in its 
full strength if the young are not exposed to 
the same conditions as their parents, because 
the transference of the young to new con- 
ditions sets up an opposing reaction and begins 
the formation of a new habit. But never- 
theless, Durkhem's and Kammerer's experi- 
ments show that traces of the effect of parental 
habit still persist, especially in young animals, 
and this persistence is the basis of that " re- 
capitulation of ancestral history " which is such 
a prominent feature in animal development. 
The question as to how the tendency to 
form a habit is transferred from the body of 
an animal to its germ-cells is one on which THE STUDY OF HEREDITY 
109 
we can only speculate. Rignano has sug- 
gested that every reaction of an animal tends 
to deposit in the tissues, and incidentally in 
the germ-cells, a substance which acts as a 
stimulus towards the performance of the same 
reaction in the young which develop from the 
germ-cells. We have already seen that in the 
formation of the bodies of animals a stage is 
arrived at where the primary organs act on 
each other and determine each other's future 
development. We can only picture such an 
action as the emission of substances from these 
organs into the body fluids. Some such sub- 
stances have been detected and they are 
known as hormones. Thus if the thyroid 
gland which lies beneath the throat be re- 
moved from a tadpole, it will not metamor- 
phose into a frog, and if this gland is diseased 
in a human being the mental powers do 
not develop and he or she becomes an idiot. 
If we assume that all the tissues are con- 
tinually throwing off hormones into the blood, 
and that these hormones are stored in the 
germ-cells, then when a new habit is learned 
we may imagine that an altered product 
is given off into the blood and stored in 
the germ-cells; and so when the germ-cell de- 
velops into a young animal this is rendered 110 
AN INTRODUCTION TO 
more sensitive to the change that calls forth 
the original formation of the habit. The 
tendency of the young to repeat the parental 
habits may be likened to memory. In Kam- 
merer's experiments with Alytes the formation 
of a pad on the hand of the male Alytes in 
the third generation after the animals had 
repaired to the water to mate could be re- 
garded as the revival of a half-forgotten mem- 
ory, for undoubtedly the ancestors of Alytes 
resembled other toads in possessing this pad. 
It has been suggested, and it is perhaps true, 
that the revival of an old habit is an easier 
matter than the formation of a new one, but 
the principle that the experience of the parents 
affects the children is the same in both cases. 
Another good example of the revival of an 
old memory is afforded by another experi- 
ment of Kammerer. In the caves of Carniola 
in Austria there lives a blind, pale, flesh- 
coloured newt called Proteus, which retains 
its external gills throughout life. The eye in 
this species is represented by a minute sac of 
pigmented cells buried beneath the skin. If 
young Proteus are brought out into the light 
the skin acquires pigment but the eye does 
not develop; but by exposing the young 
animal only to red light the formation of skin THE STUDY OF HEREDITY 
111 
pigment is prevented and the eye develops 
into a large normal eye, capable of sight. 
This experiment may be compared to the 
revival of a forgotten childish memory by 
some special stimulus. CHAPTER V 
THE INHERITANCE OF FLUCTUATING 
VARIATION 
We have already mentioned the fact that 
Darwin in constructing the famous theory of 
evolution, although he accepted Lamarck'si 
principle, yet laid far greater stress on the 
small differences of size, shape and colour 
which distinguish members of the same brood 
of nestlings and of the same litter of mam- 
mals from one another. These differences 
are usually known as fluctuating variations, 
and Darwin assumed that they were inherit- 
able, and in this assumption he was followed 
by Alfred Russel Wallace, who shares with 
Darwin the glory of having first clearly 
enunciated the principle of natural selection. 
But Darwin and Wallace, without perhaps 
clearly realizing it, assumed more than this. 
It is not enough to suppose that, for instance, 
the differences in length of beak amongst the 
same brood of nestlings are inheritable. If 
natural selection favours the survival of the 
112 THE STUDY OF HEREDITY 
113 
bird with the longest beak, unless amongst 
the offspring of this bird there appear some 
nestlings with still longer beaks, evolution will 
not proceed. Therefore we must suppose 
that the characters of the offspring of the 
survivor will fluctuate around the characters 
of their immediate parent, just as did the 
characters of that parent and his brothers 
and sisters around those of the grandparent. 
Wallace tried to show that if this assump- 
tion were granted the action of natural selec- 
tion would account for every kind of change 
of which we have evidence in the animal and 
vegetable kingdoms, including those which 
Darwin was inclined to attribute to the effects 
of use and disuse. 
If these individual differences had the 
tremendous significance which Darwin and 
Wallace attributed to them, it was clearly of 
the first importance that they should be 
measured; and this step was taken by Sir 
Francis Galton, Darwin's cousin, who devised 
a scheme for the measurement of the in- 
heritance of individual peculiarities among 
the human population. A short account of 
how this scheme was applied to the question 
of the inheritance of stature will exemplify its 
character. Galton offered a considerable sum 114 
AN INTRODUCTION TO 
of money as prizes for good family records of 
stature, and in this way he secured a large 
supply of data. When he came to analyze 
these he found that on the whole adult women 
were shorter than adult men, and that the 
ratio of the average male stature to the 
average female stature was as 13 to 12. For 
this reason he felt justified in multiplying all 
his figures for female stature by 1.08, thereby 
implying that each woman if she had been a 
man would have been taller in that proportion. 
In this way he got rid of the disturbing factor 
of sex. He next sorted his records into 
groups, in which no individual differed from 
another in height by more than half an inch. 
Then a distance (abscissa) was measured from 
one fixed point along a horizontal line, pro- 
portional to the mean stature of each of these 
groups, these ranging from 62 to 76 inches. 
At every point on this line corresponding to 
the stature of each group, a perpendicular 
(ordinate) was erected, proportional in length 
to the number of individuals comprised in 
that group, and through the summits of these 
upright lines a curve was then drawn. The 
more numerous the measurements upon which 
the curve was based the smoother it became, 
and as the number of measurements was THE STUDY OF HEREDITY 
115 
increased the curve continually approached a 
geometrical form well known to mathema- 
ticians, called the curve of error. If x repre- 
sent the length of the abscissa and y the 
length of the corresponding ordinate, then the 
relation between these two for all points lying 
on the curve of error is defined by the equation 
y=ke-ft2aT2, where h and k are constants which 
define the slope of the curve, and e is the 
logarithmic base. 
To those unversed in mathematics it may 
be explained that the shape of this curve is 
that of a summit having a slope of the same 
steepness on both sides, the inclination of 
which diminishes as it approaches the hori- 
zontal axis with which it tends to coincide, 
but never quite succeeds in reaching. Trans- 
lated into ordinary language this curve in- 
dicates that in a sufficiently large sample of 
the population taken at random, a consider- 
able proportion may be described as being of 
medium stature (this is generally indicated 
by the height of the ordinate at the summit of 
the curve), that the number of those who are 
taller and of those who are shorter than the 
mean stature by a given amount is approxi- 
mately equal, and that the number of giants 
and of dwarfs is extremely small. The extent 116 
AN INTRODUCTION TO 
to which the mean stature of any of the 
classes into which the population has been 
divided differs from the mean stature of the 
whole population is known as the deviation 
Fig. 16. 
Diagram to show the distribution of stature amongst 
20,000 recruits of the American army. The figures on 
the base line indicate height in inches. The dotted line 
indicates the continuous curve to which the records 
approximate. 
of that class. If we now select two individuals 
from the population, one of whom is taller 
than one-quarter of the population and shorter 
than the remaining three-quarters, and the 
other of whom is taller than three-quarters THE STUDY OF HEREDITY 
117 
of the population and shorter than one-quar- 
ter, then the deviations of the statures of these 
individuals from the mean are known as the 
quartiles, and are designated by the symbols 
Q and Q1. If the curve were perfectly 
symmetrical Q would be equal to Q1, but in 
practice the smoothness and perfect propor- 
tions of the ideal curve are never attained, 
and so Galton used the mean of the quartiles 
Q Q-as a measure of the variability. The 
deviation of this mean from the mean of the 
whole population is known by the extremely 
unfortunate name of probable error-a name 
against which Galton emphatically protests. 
For mathematical reasons, however, workers 
on heredity who use Galton's method now 
employ another measure of the variability, 
which is termed standard variation. 
It is arrived at in this way. We take the 
deviation of each group, square it and multiply 
it by the number of individuals in the group. 
We add together all the values so obtained, 
divide the sum by the total number of in- 
dividuals measured, and take the square root 
of the whole. Suppose the letters 
^3, etc. represent the values of the deviations 
corresponding to the different groups, nx. 118 
THE STUDY OF HEREDITY 
n2, ^3, etc., the numbers of individuals in those 
groups, and n the total number of individuals 
measured, then the formula for the " standard 
deviation " is 
+ n2e2 + n2e2 + 
or more briefly 
■ /^ne2. 
y n 
Having discovered a method of expressing 
at the same time the value and frequency of a 
variation, the next step was to investigate 
how far it was inherited. By adding together 
the statures of the father and of the mother 
(reduced to the male standard) and dividing 
the sum by two, a value is arrived at which 
Galton denominates the stature of the mid- 
parent. First we take the statures of the 
children of the group of mid-parents having 
the same stature and determine their mean. 
Then the statures of each parental group are 
plotted out along a vertical line, and the means 
of the statures of the children belonging to 
each mid-parental group are indicated by dis- 
tances measured along a horizontal line. If 
lines perpendicular to the vertical and hori- 
zontal lines respectively are drawn at the Fig. 17. 
Diagram illustrating the correlation of the height of chil- 
dren with their mid-parents. The figures along the vertical 
line indicate the heights of various mid-parental groups in 
inches, those along the upper horizontal line the mean 
heights of the corresponding groups of children. A B is a 
vertical line drawn through the point indicating the height 
of the group of children of medium stature. C D Is a hori- 
zontal line drawn through the point indicating the height 
of the mid-parental group of medium stature. These two 
lines meet at the point M in the mathematical centre of the 
square, since the heights of the medium groups of parents 
and of children are identical. U V is the line along which 
the points indicating the mean heights of the children of 
each parental group lie. X Y, the diagonal of the square, is 
the line on which the points indicating the heights of the 
children would He if they corresponded exactly to those of 
the parents. 
119 120 
AN INTRODUCTION TO 
points corresponding to the stature of each 
mid-parental group, and the mean stature of 
its children, each pair of such lines will meet 
in a point indicated in Fig. 17 by a dot. If 
the average stature of the child corresponded 
exactly to the stature of the mid-parental 
group from which it is descended, a series of 
points would be defined which would lie on 
the diagonal of a square, the side of which 
was equal to the differences between the two 
extreme statures of parental groups measured. 
This, however, is not so. The points indicat- 
ing the mean heights of the children of each 
parent group lie on a line UV which cuts that 
diagonal at its centre, since parents of medium 
stature have children of medium stature, and 
cuts the upper side of the square at a point 
some distance from its end. The relation of 
the distance between the mid-point of the 
upper side of the square (which corresponds 
to the mean stature of the children of the 
parental group of medium height) and the 
point where this line cuts it (which corresponds 
to the mean height of the children of the tallest 
parents), and of the distance from the mid- 
point to the end of the side of the square, is 
what Galton called the coefficient of regres- 
sion of filial stature on mid-parental stature, THE STUDY OF HEREDITY 
121 
and its value in this instance is about 
two-thirds; this ratio may be expressed by 
AU 
the symbol - Translated into ordinary 
language it means this: that people of average 
stature have children also of average stature, 
that tall people on the whole have tall children, 
but not so tall as the stature of the parents 
would lead one to expect, and similarly that 
the children of short people are not pro- 
portionately as short as their parents. 
The relation between the deviations of two 
related variables is now usually computed by 
a different method, which may be termed 
algebraical as contrasted with the geometrical 
method of arriving at the coefficient of 
regression which has just been outlined. 
Suppose we desire to know how far the 
variability of any one organ in an animal is 
dependent on that of another organ, let us 
term the deviations of the sizes of these organs 
in the various members of the population 
from their respective means xlf #2, ^3, etc., 
and ?/i, 2/2, 2/3, etc., respectively. Then let 
us note that the sum of each of these sets of 
terms, i. e., both ^x and is equal to zero, 
because a deviation in size above the mean is 
reckoned as positive, and one below the mean 122 
AN INTRODUCTION TO 
as negative, and the sum of the positive 
deviations cancels the sum of the negative 
deviations. But of course in each case the 
sum of the squares of the deviations will be 
positive owing to the well-known algebraical 
rule that the square of a minus quantity is 
positive. (Incidentally this is one reason for 
adopting the standard deviation as a measure 
of the variability.) If we now multiply 
together the deviations in the two organs of 
each individual of the population we get a 
series of products the sum of which is repre- 
sented by the symbol ^xy). The mean of 
S/rw 
these products will be - if n be the number 
of the population. Let mr and m2 represent 
the standard deviations of the two organs, 
then the ratio of the mean of the products 
of the two deviations determined for each 
individual to the product of the two standard 
deviations is termed the coefficient of corre- 
lation, and is used as a measure of the varia- 
bility. A couple of examples will serve to 
show how this ratio measures the correlation 
of the two variabilities. It is represented by 
the formula --Let us first suppose that 
the variabilities of the two organs are per- THE STUDY OF HEREDITY 
123 
■fectly independent of one another; then a 
large size in one of the organs is just as likely 
to be associated with a small size in the other 
as with a large size. But a large size is a 
positive deviation from the mean and a small 
size a negative deviation, and the product of 
two such deviations gives a minus quantity. 
Therefore, of the terms included in the formula 
^(xy), half will be positive, and half negative 
quantities, and the sum of the whole will be 
practically zero, for the negative terms will 
cancel the positive ones. Therefore the value 
of the coefficient of correlation will be zero, as 
it should be. 
Now let us make the opposite assump- 
tion, viz., that the deviations of the one 
organ are strictly proportional to those 
of the other. In this case every value of 
x will bear the same ratio to the corre- 
sponding value of y. Let us represent this 
ratio by the symbol 8; then y = ^x and 
our formula becomes ■ -, that is --- - • 
nmxm2 nmtm2. 
Now the standard deviation mt is, of course, 
and the standard deviation m2 is 
/|/ n 
that is j/ 2^1, or S and our 124 
AN INTRODUCTION TO 
formula for the coefficient of correlation 
, &2 
becomes ■ - - b - 4/ -_, «• e. -- - ., 
n \ n V n » 
V- V XL lx 
, or unity which expresses the complete 
n 
correlation of the two organs. When the 
deviation of one organ is only partly de- 
pendent on that of the other, the coefficient 
of correlation has a value situated some- 
where between unity and zero. When it ap- 
proaches unity it is said to be high and the 
two organs are said to exhibit a high degree of 
correlation; when the coefficient is represented 
by a small fraction, the two organs are said 
to exhibit a low degree of correlation. 
The methods of Galton were applied by 
Weldon to the measurement of the variations 
of animals, but few cases present such simple 
conditions as those of the human stature. 
We may take, for instance, the common 
shore-crab, Carcinus moenas. This little green 
crustacean, familiar to all visitors to our sea- 
coast, occurs between tidemarks, hiding in 
pools and under seaweed left bare by the 
receding tide. It varies considerably in shape, 
some specimens being comparatively broad 
and others rather slim-waisted. But if we 
wish to determine the curve of error for this THE STUDY OF HEREDITY 
125 
variation, it is of no use plotting out absolute 
measurements of the breath of the carapace 
along our abscissa line, for the crabs vary 
so much in size that a given value might 
correspond to the breadth of the carapace of 
a small broad crab or that of a larger slim 
one. If, however, we divide each absolute 
value of the breadth of the carapace by its 
length, then we obtain a fraction which indi- 
cates the broadness or slimness of the crab, 
and which is independent of its size. This is 
the procedure which was adopted by Professor 
Weldon in dealing with this case, and the frac- 
tional values which he obtained by measuring 
the carapaces of 500 crabs when plotted out 
constituted a curve of error as perfect as that 
yielded by the records of human stature. It 
is not a practicable thing to rear shore-crabs 
in the laboratory under experimental con- 
ditions, so that Weldon could not directly 
determine how far variations in the breadth of 
carapace were inherited, but he was able to 
show that variations in the size of other 
organs, such as that of the great claw, were 
correlated with the breadth of the carapace, 
just as is the stature of children with that 
of their parents; there is a correspondence 
between changes in sizes of the two organs, 126 
AN INTRODUCTION TO 
but it is not a complete correspondence. But 
Professor Weldon was able to discover many 
other interesting things by measuring the 
carapaces of crabs of different ages. If the 
" curve of error " for the carapace of young 
crabs be compared with the similar curve for 
The common shore-crab, Carcinus mccnas. The dotted 
lines indicate the dimensions which were measured. 
Fig 18. 
older crabs, then it will be seen that the former 
curve has a more gentle slope than the latter. 
It is thus seen that there are much more 
marked deviations from the mean among 
younger than among older crabs; and the 
most probable explanation of this is, that the 
more extreme types become weeded out by 
natural selection. In confirmation of this 
idea Professor Weldon tried the experiment THE STUDY OF HEREDITY 
127 
of mixing the water in which the crabs lived 
with fine china clay. A great many died, but 
when the carapaces of all were measured it was 
found that those which survived had narrower 
carapaces than those which died; that the 
cause of death was the covering of the gills 
with silt, and that a broader carapace was 
associated with a broader ingress into the 
gill-chamber, and consequently a greater 
clogging of the gills with silt. It may be 
explained that in a crab the gills are contained 
in two chambers roofed over by the lateral 
parts of the carapace, and the entrances into 
these chambers are slits between the in-turned 
edges of the carapace on either side and the 
bases of the legs. 
Professor Davenport, who carried on work 
on the lines begun by Weldon, was able to 
show how the statistical methods outlined 
above enable us to find out when a given 
population consists of two races. For when 
he measured various dimensions in certain 
small North American mammals, such as 
the mole, he found that the resulting curve of 
error had two summits, showing that there 
were two values about which the other values 
tended to group themselves-in a word, that 
the samples examined consisted of two races 128 
AN INTRODUCTION TO 
having different median values for the char- 
acter measured. 
The general conclusion which was drawn 
from Galton's researches was, that the differ- 
ences between members of the same family 
were inherited, but that there is regression 
towards mediocrity, that is, that the deviation 
of the child from the mean was on the average 
less than that of the parent. This regression 
would check the gradual modification of a 
character by natural selection, but would not 
entirely prevent it, because the number of 
progeny which survive is only such a small 
proportion of those which are born, that 
some few children of an exceptional parent- 
if only enough children were produced-would 
be sure to be more exceptional than the parent 
himself, even though the average of the 
children was less exceptional than the parent; 
if these few alone were preserved by natural 
selection and became the progenitors of the 
next generation the race would be modified. 
In a word, the mean value of the character 
in the survivors would exhibit a greater 
deviation from the average than the value of 
the character in the parent, although the 
mean value of the character in all the children 
would deviate less. THE STUDY OF HEREDITY 
129 
A great deal of importance was formerly 
attributed to the " regression to mediocrity '* 
by workers on heredity, but nowadays it is 
only of historical interest. The reason for 
this change of attitude is the recognition of 
the fact that a human population selected at 
random consists of a mixture of different 
races with different hereditary potentialities 
and that before statistical inquiry can give us 
any useful help in determining the laws of 
heredity, the material with which it deals 
must be selected and purified; in a word, the 
conditions of the experiment must be standard- 
ized. 
A much better means of investigating the 
laws of heredity is afforded by what are 
termed pure line investigations. By a pure 
line is meant a group of children which are 
the offspring of a single parent. The three 
cases in which a pure line can be produced 
are as follows: (1) The vast majority of the 
higher plants and a certain number of animals 
belonging to different classes are hermaphro- 
dite, i. e., the same individual produces both 
kinds of germ-cells, and in most cases self- 
fertilization is possible, so that from one 
parent a large population can be produced. 
(2) A certain number of animals exhibit 130 
AN INTRODUCTION TO 
parthenogenesis-i. e., produce ova which will 
develop without fertilization, and in these 
cases also a single parent can become the sole 
progenitor of a large family. (3) The simple 
organisms termed Protista reproduce by 
fission, and one specimen by repeated divisions 
can give rise to an enormous progeny. 
The first to study the inheritability of 
variations within a pure line was the Swedish 
investigator Johannsen. 
Johannsen worked with the common 
kidney-bean plant (Phaseolus). He found 
that ordinary beans differ from one another 
in weight, beans of medium weight being the 
most numerous, whilst unduly light and 
heavy specimens were comparatively rare. 
If the heavier beans were sown and plants 
raised from them, when these plants produced 
beans the average weight of these was greater 
than the average weight of the beans pro- 
duced by the plants raised from the beans of 
lighter weight, whilst beans of medium weight 
gave rise to plants bearing beans of medium 
weight. Nevertheless the same " regression 
to mediocrity " which was noted in the case of 
human stature can also be observed here. 
The average weight of beans produced by 
sowing the heavier beans, though greater than THE STUDY OF HEREDITY 
131 
the average of the beans produced from sowing 
the beans of medium weight, was not pro- 
portionately greater, and the same was true, 
mutatis mutandis, of the beans produced by 
sowing the beans of lighter weight. But 
when the beans produced by self-fertilizing 
the flowers of a single plant were separated, 
and heavier, lighter and medium beans were 
selected from the crop thus produced, and 
separate " pure lines " produced from single 
samples of such beans and the resulting 
progenies weighed, then a totally different 
result was obtained. It then appeared that 
curves of error calculated from the weights 
of the offspring of a heavy bean and of a light 
bean respectively had the same median values, 
provided that heavy and light beans were sisters 
descended from the same plant. In a word, 
it appeared that differences in weight 
between sister beans were not inherited; 
that such differences to which Johannsen gave 
the name Fluctuations were due to differences 
in nutrition, and did not affect the hereditary 
mechanism, which was the same for all. 
Agar took up for examination the partheno- 
genetic water-flea, Simocephalus. This little 
crustacean is about %3 inch long, and is an 
inhabitant of our ponds and ditches. The 132 
AN INTRODUCTION TO 
hinder part of its body is enveloped by two 
flaps of skin stiffened by horny matter, which 
constitute a bivalve shell or carapace. Its 
head projects in front, and on it is situated a 
single compound eye, whilst from it spring 
Side view of Daphnia, a crustacean closely allied to 
Simocephalus. B. The brood-pouch. 
Fig. 19. 
two great forked feelers or antennae, by 
strokes of which it swims. The two flaps of 
the carapace are united in the middle line 
above the back of the animal. The legs 
behind the mouth are thin, leaf-like and 
flexible, for which reason Simocephalus is 
classed with the Phyllopoda or Leaf-footed THE STUDY OF HEREDITY 
133 
Crustacea. The animal usually swims lying 
on its back, and the space between the back 
and the united valves of the carapace forms a 
convenient pocket into which the eggs pass as 
soon as they are laid. In this pocket, or 
' ' brood-pouch " as it is termed, they develop 
until they are ready to hatch and take up a 
free life. The young feed from the beginning, 
as did their parents, on minute plants found 
swimming in the water; they moult four 
times, growing considerably at each moult, 
and their attainment of the adult condition is 
marked by their giving birth to living young. 
In the autumn some of the eggs develop 
into males and these males then fertilize the 
eggs which the females have produced. These 
fertilized eggs surround themselves with thick 
egg-shells, and lie dormant during the winter. 
The skin lining the brood-pouch of the mother, 
becomes thickened by a stiff deposit of horny 
matter. This deposit is eventually moulted 
and falls to the bottom, enclosing the fertilized 
eggs in the brood-pouch. Such a structure is 
called an ephippium. In the spring these 
eggs develop and hatch out as young Simo- 
cephalus, but these young Simocephalus, so 
long as the temperature is mild and food 
abundant, produce eggs which develop into 134 
AN INTRODUCTION TO 
females directly without any fertilization at 
all, this being the parthenogenesis. 
Agar began by collecting ephippia from a 
stream near Glasgow University. These were 
allowed to hatch out in the laboratory, and 
every young Simocephalus was isolated by 
itself in a corked tube, three-quarters filled 
with water, containing an abundant supply of 
microscopic plants. Here it lived, grew and 
produced young, two females of which were 
selected to continue the line and removed to 
separate tubes. Males were not produced 
under the conditions of the experiment. The 
character, whose variation Agar measured, was 
the length from the anterior apex of the head 
to the hinder end of the united line of the 
carapace valves. By selecting those speci- 
mens to continue the race which were the 
largest and smallest respectively of the brood 
produced by each parent without fertilization, 
Agar was able to determine whether such 
differences were or were not inherited. He 
arrived at the conclusion that they were not 
inherited. The offspring of the shorter speci- 
mens were at birth actually rather longer on 
the average than the offspring of the longer 
specimens, whilst when they had attained the 
adult condition they were about the same size. THE STUDY OF HEREDITY 
135 
It must be mentioned that these crustaceans 
are extremely susceptible to changes in temper- 
ature, and in the quality of the water, which 
may favour or arrest their growth, and Agar 
had not only to strive to keep the conditions 
constant, but to make corrections for altera- 
tions due to the altering environment. 
Finally, we must mention the experiments 
of Jennings. This observer chose as the 
subject of his experiments the Protist Para- 
mecium, well known to all students of elemen- 
tary biology as the Slipper Animalcule. This 
little creature is shaped somewhat like a 
bedroom slipper without a heel. The part 
corresponding to the toe is thick, the rest of 
the animal is like a thin scoop. It is covered 
all over with short vibratile hairs termed cilia, 
by means of which it propels itself about. It 
reproduces by dividing itself into two by a 
line of fission at right angles to its long 
diameter. At considerable intervals it under- 
goes a process known as conjugation. In this 
process two specimens approach one another 
and adhere together. Then an interchange of 
nuclear matter takes place between the two 
and after this they separate, being now known 
as eX-COnjugateS. This process has been com- 
pared with the sexual process amongst higher 136 
AN INTRODUCTION TO 
animals. Jennings selected from a culture 
six conjugating pairs and two ex-conjugates- 
Fig. 20. 
Showing three " pure lines " of Paramecium. The 
sign (-}-) marks the median value in each strain; the 
vertical line passes specimens in each of the three strains 
which are of the same absolute size. 
fourteen specimens in all. Each of these was 
isolated in a separate culture-dish and allowed 
to reproduce itself by fission. The quality THE STUDY OF HEREDITY 
137 
the variation of which Jennings measured 
was the length. He found that the offspring 
of his fourteen selected specimens included six 
different strains of Paramecium differing from 
one another in length, for if the offspring of 
each of them was measured so as to give rise 
to a separate curve of error, it was found that 
no less than six different median values were 
to be found. It is true that a giant might be 
selected from a strain of low median value, 
which was longer than a dwarf belonging to a 
strain of high median value, but the offspring 
of such a giant were no longer on the average 
than those of a dwarf belonging to the same 
strain, and were decidedly shorter than those 
of a dwarf belonging to a strain which had a 
higher median value. It follows that in the 
case of Paramecium, as in those of Simo- 
cephalus and the bean, variations appearing 
within the pure line are not inheritable. It 
is to be noted that Jennings, like Agar, had 
to introduce numerous corrections; for the 
length attained by a given specimen of Para- 
mecium is not only determined by the strain 
to which it belongs, but also depends on tem- 
perature, crowding, and the chemical nature 
of the solution in which it is situated. 
The concordant results obtained in three 138 
THE STUDY OF HEREDITY 
sets of experiments in different types of pure 
line may well give us food for serious reflection. 
The obvious inference seems to be that in each 
natural population there exists a large number 
of strains, each with its fixed unalterable 
quality of heredity, and that the apparent uni- 
formity in character of the species is due to 
the constant intercrossing of these strains. 
The appearance of the possibility of modifying 
the species by selection would, on this hypoth- 
esis, be due to the fact that certain strains 
would be gradually eliminated, and the popu- 
lation " purified " by being gradually reduced 
until it consisted solely of individuals belong- 
ing to one desired strain. 
It is therefore quite clear that the results 
of carefully conducted experiments carried 
out under standardized conditions lend no 
support whatever to the cardinal postulate of 
Darwin and Wallace that natural selection 
alone could effect evolution. The question 
as to how the various strains originated is 
quite unsolved, but we may surmise, in 
accordance with what we have learnt in the 
previous chapters, that they have derived 
their peculiarities from differences in the en- 
vironment to which their ancestors have been 
exposed. CHAPTER VI 
THE INHERITANCE OF " MUTATIONS " OR 
"SPORTS " 
If we survey the animal and plant kingdoms 
we discover that each is divided into a number 
of well-marked groups and that the members 
of each group cross freely with one another 
and produce fertile offspring. These natural 
groups are called species and in each kingdom 
the species show all degrees of difference from 
one another. In order to express these 
differences the species most alike are placed 
together in groups called genera, the genera 
are arranged in families, the families in or- 
ders, the orders in classes, and so on. Darwin 
clearly saw that the first and essential step in 
any theory of evolution was to show how 
several species have been developed out of a 
single species, and for this reason he called his 
great book The Origin of Species. 
When members of different species are 
crossed with one another they are usually but 
not always sterile, but sometimes they produce 
139 140 
AN INTRODUCTION TO 
offspring which are termed hybrids. The most 
familiar example of a hybrid is the mule, which 
is produced by mating a horse and an ass. 
The mule, however, like the vast majority of 
hybrids, is sterile; and in the few cases where 
hybrids between different species are fertile the 
laws governing the hereditary qualities of their 
offspring have not been investigated. But 
when plants are cultivated and when animals 
are domesticated the original species become 
divided into a great number of different 
types which are mutually fertile. These 
types among plants are known as " Varieties," 
amongst animals more usually as " breeds " 
or " strains." Darwin thought that these 
" breeds " or " varieties " had been produced 
by the conscious or unconscious selection by 
man of those fluctuating variations which 
most suited his purpose, and that this action 
on the part of man showed us how Nature 
proceeded in producing new species. But 
the " pure-line " investigations recorded in 
Chapter V have shown us that this view is 
no longer tenable, and there is, in fact, no 
universal agreement on the question as to how 
domestic breeds were produced. Many think 
that they owed their origin to the sudden 
appearance of strongly-marked deviations THE STUDY OF HEREDITY 
141 
from the normal type commonly called 
' ' mutations " or " Sports, which attracted 
the attention of primitive man and led him to 
preserve them, and it is noteworthy that when 
distinct strains are crossed with one another 
the characters of their offspring obey the same 
laws as are shown when a recently produced 
monstrosity is crossed with the type from 
which it has sprung. These laws were 
worked out by Gregor Mendel, an Augustinian 
monk belonging to a monastery in Bronn in 
Moravia; his experiments have been repeated 
and confirmed by other workers all over the 
world, and thus the laws which he discovered 
are usually termed Mendelian. For the reason 
that we have given, we prefer to call them the 
laws governing the inheritance of sports. 
The cross between two breeds-like the cross 
between species-is usually termed a hybrid. 
It is important to remember that Mendel him- 
self in giving an account of his experiments 
expressly stated that he resolved not to deal 
with variations of the " more " or " less " de- 
scription, but to choose for the subjects of 
his researches varieties distinguished from one 
another by some sharply-marked character. 
For this reason he selected the cultivated pea 
(Pisum sativum) because it is easily cultivated 142 
AN INTRODUCTION TO 
and is obtainable in many clearly-marked 
varieties. It is usually self-fertilized. We 
begin our survey of these varieties by taking 
the case of two which are distinguished from 
one another by a single differentiating feature. 
Peas when ripe and dried may be either of a 
yellowish colour or of a pale green tint. If a 
plant raised from a yellow pea be crossed with 
a plant raised from a green pea, a set of hy- 
brid peas will be produced. The crossing is 
effected by opening the flower buds of the 
plant which is to provide the ovules (see Chap. 
I) and removing the stamens before they are 
ripe so that self-fertilization is prevented, this 
plant is termed the Seed-parent. A little 
later, when the stigma is ripe, it is dusted with 
pollen procured from the other plant, which 
is known as the pollen-parent. When the 
hybrid peas ripen it is found that they are 
all yellow, and this is true no matter whether 
the ovules have been derived from the plant 
sprung from the green pea or from the plant 
sprung from the yellow pea-or, to put it an- 
other way, they are yellow whichever kind of 
plant is employed as the seed-parent. This 
generation of peas is known as the first filial 
generation and is designated by the symbol 
F1. When these peas are sown and the plants THE STUDY OF HEREDITY 
143 
to which they give rise are self-fertilized, or, 
as it is termed for the sake of brevity, selfed, 
a generation of peas is produced which is 
known as the second filial generation, and 
is indicated by the symbol F2. 
In this generation-if enough peas are 
produced-it is found that, as nearly as 
possible, one-quarter of the peas are green and 
three-quarters are yellow. Thus the green 
character, which had been entirely suppressed 
in the first filial generation, reappears in the 
second filial generation. Mendel termed the 
character which is thus temporarily sup- 
pressed the recessive character, and to the 
character which alone appears in the first 
filial generation he gave the name of domi- 
nant character. When the recessive green 
peas are sown they give rise to plants which, 
when " selfed," bear only green peas, and this 
is true for however many generations they 
may be cultivated. When the yellow peas, 
however, are sown, they give rise to plants 
which, when " selfed," yield different results 
in different cases. One-third of them-that 
is to say, one-fourth of the total number of 
plants which were raised from the first filial 
generation-give rise to yellow peas only, and 
these yellow peas yield only yellow peas for 144 
AN INTRODUCTION TO 
however many generations the cultivation 
may be continued. The remaining two-thirds 
of the plants raised from the hybrid yellow 
peas yield both yellow and green peas in the 
proportion of three yellow to one green; in a 
word, they behave exactly as did the peas of 
the first filial generation of hybrids, and the 
peas of this third filial generation, produced 
from the hybrids of the second filial genera- 
tion, behave exactly as did the peas of the 
second filial generation, that is to say, the 
green reproduce their like however long they 
may be cultivated, and one-third of the yellow 
do the same, but the remaining two-thirds 
yield yellow and green peas. 
In order to account for this strange phe- 
nomenon Mendel put forward a theory which 
has been widely adopted, and which is termed 
the theory of the purity of the germ-cells. 
According to this theory when two strains are 
crossed their qualities are, of course, mingled 
in the body of the hybrid, but when this hybrid 
in turn forms germ-cells, these germ-cells awe 
of two kinds, each kind carrying the poten- 
tiality of developing the qualities of one 
strain, but not of both. Applying the theory 
to the case in point, we assume that when the 
hybrids of the first filial generation come to THE STUDY OF HEREDITY 
145 
form germ-cells, two kinds of pollen-grains and 
two kinds of ovules are produced, one of these 
kinds of pollen-grain and one of these kinds of 
ovule having the potentiality of producing a 
yellow pea, whilst the other kind, both of 
pollen-grain and of ovule, has the power of 
producing a green pea. It is a further assump- 
tion of the theory of germ-cell purity that the 
two kinds of germ-cells are produced in equal 
numbers. If, for the sake of brevity of expres- 
sion, we bestow the names of green pollen-grain 
and green ovule on the type of pollen-grain and 
ovule respectively which carry the potentiali- 
ties of becoming a green pea, and correspond- 
ing names on those which carry the potentiali- 
ties of becoming a yellow pea, then the problem 
we have to solve is this: given equal quanti- 
ties of green pollen-grains and yellow pollen- 
grains, mingled together and used to fertilize 
equal quantities of green ovules and yellow 
ovules mingled together, what will be the 
result? On the doctrine of chances it appears 
probable that once in four times a green pollen- 
grain will meet a green ovule, for if all the 
pollen-grains were the same, and there were 
equal numbers of yellow and green ovules, half 
the unions would be with each kind of ovule, 
but as only half the pollen grains are green, it 146 
AN INTRODUCTION TQ 
follows that only a half of a half of the total 
unions, that is, a fourth, will be between green 
pollen-grains and green ovules. By a parity 
of reasoning we deduce that once in four times 
a yellow pollen-grain will meet a yellow ovule, 
and that once in four times a yellow pollen- 
grain will meet a green ovule, and that, finally, 
once in four times a green pollen-grain will 
meet a yellow ovule. These last two types of 
union yield hybrid strains of exactly the same 
nature, and these hybrid strains appear to the 
eye like pure yellow peas because of the domi- 
nance of yellow; but the union of the yellow 
pollen-grains and yellow ovules yields pure 
yellow peas, and these added to the yellow 
hybrids account for the three-quarters of the 
second filial generation constituted by yellow 
peas. Since a green pea can only be produced 
by the union of a green pollen-grain and a 
green ovule, it follows that all the green peas 
produced are " pure " and yield only their like 
when progeny is raised from them. 
It will be observed that in the reasoning 
which has just been outlined it is assumed that 
by random mingling of two sorts of pollen- 
grains and two sorts of ovules a definite nu- 
merical result will be obtained. This is only 
true if very large numbers are considered, for if THE STUDY OF HEREDITY 
147 
smaller numbers were examined we might have 
accidental " runs " of one kind of pea. To give 
an idea, however, of the numbers actually re- 
quired, and of the close approximation between 
the results obtained by Mendel and those which 
might be theoretically expected, we may here 
reproduce one of the sets of figures which he 
gives. Two hundred and fifty-eight plants 
raised from the hybrid peas of the first filial 
generation yielded 8023 seeds, of which 
6022 were yellow and 2001 green. The theo- 
retical numbers would be 6018 yellow and 
2005 green. The student will do well to regard 
with the utmost suspicion figures adduced to 
prove the proportions of strains appearing in 
the progeny of hybrids if these figures indi- 
cate comparatively small numbers. 
Besides differences in the colour of the seed, 
or, to speak more correctly, differences in the 
colour of the cotyledons or seed-leaves which 
constitute nearly all the bulk of the ripe pea, 
Mendel found other strains of peas which 
differed from one another in a single character. 
Thus some peas when ripe and dry are rounded, 
whilst others are shrunken and angular, and 
some produce plants about six feet high, whilst 
others give rise to dwarf procumbent plants 
about one foot high. 148 
AN INTRODUCTION TO 
Besides selfing the hybrids of the first filial 
generation or crossing them with one another, 
the experiment is frequently made of crossing 
the hybrid with one of the parental strains. 
If a yellow hybrid pea, for instance, be crossed 
with the parental recessive strain, i. e., with 
a green pea, equal numbers of yellow and green 
peas are produced. A moment's reflection will 
show us why this must be so, for the hybrid 
produces equal numbers of " yellow " and 
" green " germ-cells, but the recessive bears 
only green germ-cells, so that all the germ- 
cells of the hybrid unite with green germ-cells, 
and we have already seen that green germ-cell 
X green germ-cells = green pea, and yellow 
germ-cell X green germ-cell = yellow pea. 
If D denotes the dominant character and R 
the recessive, then the results of crossing two 
strains may be expressed by the formula 
D X R - D(R) where the brackets indicate 
that the character enclosed in them is present 
but not visible; and D(R) X R = D(R) + 
RR indicates the result of crossing the hybrid 
with the recessive parent. Such matings are 
regularly employed in experiments on heredity 
in order to explore the constitution of a sus- 
pected hybrid. They are termed " back- 
crosses." THE STUDY OF HEREDITY 
149 
The question now arises as to what is going 
to happen when we cross two strains of plants 
with one another which differ from each other 
in two differentiating characters. Suppose, to 
take one instance, we cross plants produced 
from round yellow peas with plants produced 
from green angular ones, of what description 
will the first and second filial generations be? 
When the experiment is tried we find that 
the first filial generation consists solely of 
round yellow peas, since roundness is dominant 
over angularity and yellowness over greenness. 
When this generation is selfed we find that in 
the second filial generation three-quarters of 
the peas are round and one-quarter angular, 
and three-quarters are yellow and one-quarter 
green, and each differentiating quality is in- 
herited as if the others were absent. That is 
to say, though three-quarters are round and 
three-quarters are yellow, it is not the same 
three-quarters which are round which are also 
yellow, nor is it the same quarter that is 
angular that is green. It is exactly as if one 
had a bag of white billiard balls and one were to 
select three-quarters of them at random and 
paint them yellow, and then to put the yellow 
balls back into the bag and jumble the whole 
lot up together, and then to select again three- 150 
AN INTRODUCTION TO 
quarters of the whole and paint each of them 
with a black cross. It is obvious that in a 
great many cases a ball would receive both the 
yellow colour and the black cross, but some 
white balls would receive the black cross and 
some yellow balls would escape it. Since 
three-quarters of the whole are yellow and 
since the black cross is distributed evenly 
amongst the whole number of these, being at- 
tached to three out of every four of them, 
it is evident that three-quarters of three- 
quarters will be yellow and have the black 
cross, that is, nine-sixteenths of the whole. 
Since one-quarter are white and one-quarter 
are without the cross, one-quarter of one-quar- 
ter, or one-sixteenth of the whole, will be white 
and unmarked. By a parity of reasoning we 
deduce that one-quarter of three-quarters, or 
three-sixteenths, will be yellow and unmarked, 
and three-quarters of one-quarter, or three- 
sixteenths, will be white and marked with the 
cross. So in the second filial generation of 
the cross between yellow round and green 
angular peas we find that nine-sixteenths of 
the peas are yellow and round, three-sixteenths 
yellow and wrinkled, three-sixteenths green 
and round, and one-sixteenth green and 
wrinkled. We have thus obtained two new THE STUDY OF HEREDITY 
151 
varieties of peas, viz., green round and yellow 
wrinkled, and therein lies one of the most 
important applications of experimental breed- 
ing. We are enabled in this way to combine 
two valuable qualities belonging to different 
strains. When we test the qualities of the 
peas of the second filial generation, by rearing 
from them by selfing a third filial generation, 
we find that all the peas which combine the 
two recessive qualities, viz., greenness and 
wrinkledness, breed true, and are, therefore, 
" pure." But of the yellow round peas only 
one-ninth breed true, i. e., one-sixteenth of the 
whole number of the second filial generation; 
the remaining eight-sixteenths, or one-half of 
the whole, exhibit the quality of hybrids with 
respect to one or both differentiating charac- 
ters, that is to say, they either yield yellow and 
green peas, or round and wrinkled, or, finally, 
peas which may be yellow or green, round or 
wrinkled. Of the three-sixteenths which are 
yellow and wrinkled, one-third, i. e., one- 
sixteenth of the whole, breed true, the remain- 
ing two-sixteenths, or one-eighth of the whole, 
behave as hybrids with respect to yellowness; 
and of the three-sixteenths which are green 
and round, one-third, i. e., one-sixteenth of the 
whole, breed true, while the remaining two- 152 
AN INTRODUCTION TO 
sixteenths, or one-eighth of the whole, act like 
hybrids so far as roundness is concerned. 
These facts are explained by an extension 
of the hypothesis of the purity of germ- 
cells. 
We assume that each hybrid strain produces 
twice as many types of germ-cells as there 
are pairs of differentiating characters. Thus 
yellow round peas of the first filial generation 
produce (using the abbreviated terminology 
explained on p. 145) yellow round pollen- 
grains, yellow wrinkled pollen-grains, green 
round pollen-grains, and green wrinkled pollen- 
grains in equal numbers, and ovules of the 
same four varieties also. When a pollen-grain 
and ovule of the same kind meet-and this 
can only happen once in sixteen times-a 
stable variety is produced. Such a union is 
called a homozygous union, and the organism 
so produced is termed a homozygote. But 
where a pollen-grain and ovule of different 
kinds meet, a hybrid variety, or heterozygote, 
is produced, and such a union is termed 
heterozygous. It is clear, however, that a 
pollen-grain and ovule may be alike with re- 
spect to one character and dissimilar with 
respect to another. When such a pair meet 
the union is said to be homozygous with respect THE STUDY OF HEREDITY 
153 
to one character and heterozygous with respect 
to others. 
Mendel performed some experiments in 
crossing strains which were separated by three 
pairs of differentiating characters. On the 
principles already explained it is a simple 
matter to calculate the distributions of these 
characters in the second filial generation. Let 
us call the dominant members of these pairs A, 
B and C, and the recessive members a, b and 
c. The two pairs of these characters, A, B, 
a, b, will, as we have seen, be distributed 
throughout the offspring in the proportions 
of 9 AB, 3 Ab, 3oB, 1 ab. The third pair of 
characters is distributed throughout the 
generation independently of the others in the 
proportion of the three dominant (C) to one 
recessive (c). We shall, therefore, have 27 
ABC, 9 AbC, 9 aBC, 9 ABc, 3 abC, 3 Abc, 
3 aBc and 1 abc; or, in other words, twenty- 
seven sixty-fourths exhibiting all three domin- 
ant characters, three lots of nine sixty-fourths 
exhibiting two dominant characters accom- 
panied by the one recessive character, and 
three lots of three-sixty-fourths each exhibit- 
ing one dominant character accompanied by 
two recessive characters, and finally, one 
sixty-fourth exhibiting all three recessive 154 
AN INTRODUCTION TO 
characters, and all progeny belonging to this 
last-named category will breed true. But of 
the categories some will be heterozygote, with 
respect to one, two, or all three characters, and 
further breeding will be necessary to separate 
out the pure strains. 
Since Mendel's time the continuous work of 
many zoologists and botanists has shown that 
similar strains, separated by differentiating 
characters, are to be found in a wide range of 
animals and plants, and that these strains when 
crossed behave in the same manner as those 
investigated by Mendel. Comparatively few 
new principles have been discovered, but there 
are one or two extensions of Mendel's prin- 
ciples which have come to light, and we shall 
now give examples to elucidate these. 
The suppression of one character in the first 
filial generation is known as the principle of 
dominance; the appearance of two types of 
descendants in the second filial generation is 
termed the principle of Segregation. It occa- 
sionally happens that in the first filial genera- 
tion the hybrid can be discriminated from both 
parental types by its appearance, and even 
when this is not possible, a microscopic examin- 
ation will sometimes succeed in distinguishing 
the hybrid from the pure dominant. Thus in THE STUDY OF HEREDITY 
155 
the cross between round and angular peas the 
two varieties are distinguished by the size of 
their starch grains, and the starch grains of 
the hybrid, which is also a round pea, are 
intermediate in size between the grains of the 
two parental varieties. In the second filial 
generation, in which both parental types re- 
appear, there will in such cases be one-fourth 
of the offspring reproducing each parental 
type, and one-half showing the hybrid charac- 
ter. One such case is given by Bateson. The 
so-called Andalusian fowl is a fancy breed, of 
a colour described by specialists as blue, but 
which might be more aptly described as slaty. 
It is produced as a hybrid by crossing two 
breeds, one of them of a black colour and the 
other of a white dotted with black here and 
there. When one tries to raise a breed of 
Andalusians one can consequently only obtain 
half the offspring of the Andalusian type-• 
the other half being made up of black and 
white birds in equal numbers. For reasons 
which will be made clear later on, this condi- 
tion of affairs is termed imperfect dominance. 
Characters like the yellow or green colour 
of the cotyledons of the pea, which replace 
one another, are termed allelomorphs of 
one another. Roundness and angularity are 156 
AN INTRODUCTION TO 
another such pair of characters, and it is usual 
to speak of each of them as being a member 
of an allelomorphic pair. One of the most 
Fig. 21. 
Illustrating the types of comb found in different breeds 
of fowl. A. Single comb. B. Rose comb. C. Pea comb. 
D. Walnut comb. (After Bateson.) 
interesting of the newer results is the dis- 
covery by Bateson that the presence in the 
same embryo of two characters belonging to 
two different allelomorphic pairs can produce THE STUDY OF HEREDITY 
157 
a new result different from that which would 
be produced by the presence of either of them 
alone. Bateson discovered this principle also 
by crossing fowls. Breeds of fowl differ much 
in the comb. The original wild type is seen 
in the Jungle Fowl of India, and also in many 
domesticated breeds, such as the Leghorn, and 
is known as a single comb, which is blade-like 
with a serrated edge (Fig. 21, A). Many 
breeds, such as the Hamburg, have what is 
called a rose COHlb, which has a triangular 
horizontal surface covered by small knobs 
(Fig. 21, B). Then there is a type called 
a pea COHlb, which is blunter than the single 
comb and not serrated, but which exhibits 
laterally a few knobs (Fig. 21, C). Finally we 
have the Walnut comb, which is a bulbous, 
misshapen structure, with here and there a 
few tufts of vestigial features (Fig. 21, D). 
Bateson began by crossing breeds with rose 
and pea combs, and found, to his great sur- 
prise, that the first filial generation all possessed 
walnut combs. When, however, he bred from 
these walnut-combed fowls he obtained off- 
spring in the proportion of nine walnut-combed 
fowls to three rose-combed, three pea-combed 
and one single-combed fowl, whereas he ex- 
pected to obtain a generation consisting of two 158 
AN INTRODUCTION TO 
walnut-combed, one pea-combed and one rose- 
combed, out of every four, which is what would 
have occurred had rose comb and pea comb 
been members of the same allelomorphic pair 
and walnut comb the hybrid between them. 
Here, then, the original wild type turned up in 
the second filial generation in such proportions 
as one would expect a type bearing two re- 
cessive qualities to appear when two strains 
differing from one another in two differenti- 
ating features are crossed (see p. 150). This 
circumstance at once suggested to Bateson the 
following explanation of the anomaly. He 
assumed that a rose comb and a pea comb each 
owed its peculiarity to the presence of some 
peculiar factor in the germ-cell which altered 
the typical single comb in a peculiar manner. 
The allelomorphic fellow of the presence of 
this peculiar factor was simply its absence, and 
in its absence the single comb was produced. 
The factors giving rise to rose and pea combs, 
respectively, belonged to different allelomor- 
phic pairs, and when both were present at 
once in the same zygote a walnut comb was 
produced. When, therefore, we cross rose- 
comb- and pea-comb-bearing strains, we have 
unwittingly crossed rose, with a strain (pea) 
characterized by the absence of the rose THE STUDY OF HEREDITY 
159 
factor, and at the same time crossed pea, 
with a strain (rose) characterized by the ab- 
sence of the pea factor, and in this way we 
have produced hybrids possessing two pairs 
of differentiating factors which segregate in 
the next generation. 
The theory that members of the same 
allelomorphic pajr are distinguished by the 
presence in one of something which is absent 
in the other can be used to explain dominance. 
If we consider the case of yellow and green 
peas, we see that all unripe peas are green, but 
that yellow peas possess a factor absent in 
green peas, which changes the green into yellow 
as ripening comes on. When two germ-cells 
meet, if even only one of them possesses this 
factor, the hybrid pea will turn yellow and 
will be distinguishable to the naked eye from 
the pure yellow pea. But, occasionally, the 
presence of the factor in only one gamete- 
or, to use the phrase commonly employed, a 
single dose of the factor-is not sufficient to 
confer on the zygote the same outer appear- 
ance as two doses, and in these cases the hybrid 
is distinguishable by the naked eye, and we 
have a case of imperfect dominance. The 
blue colour of the Andalusian fowl alluded to 
above turns out on closer investigation to be 160 
AN INTRODUCTION TO 
a dilute form of black, due to the zygote re- 
ceiving only one " dose " of the black colour. 
Another principle was discovered by the 
French zoologist Cuenot; this is, that the 
presence of a factor belonging to one pair is 
sometimes necessary in order to allow the 
influence of a factor belonging to another pair 
to be visible. Cuenot crossed ordinary grey 
wild mice with tame white mice, and in the 
first filial generation, as was to be expected, 
all the hybrids were grey, since whiteness or 
albinism is obviously due to the absence of 
colour. When a second filial generation, how- 
ever, was raised from these hybrids, a most 
curious result was obtained. On an average, 
out of sixteen mice, nine were grey, three were 
black, and four were white. Here we have to 
account for the appearance of black, present 
in neither of the parental strains, and for the 
ratio 9: 3: 4. Cuenot's hypothesis was as 
follows: colour in the skin of mice is due to 
the meeting of two factors, one a factor deter- 
mining the kind of colour (grey, black, choco- 
late, etc.), and the other a factor which he 
named chromogen, which must be present if 
the colour factor is to show itself. The 
albinos in this case were black mice devoid of 
the chromogen factor, and in the second filial THE STUDY OF HEREDITY 
161 
generation there were produced, according to 
the rule for the result of crossing two strains, 
distinguished from one another by two 
differentiating factors, nine grey with chromo- 
gen, three black with chromogen, three grey 
without chromogen, and one black without 
chromogen. The grey and black without the 
chromogen component are, of course, white, 
and so we see why four (3 + 1) mice out of 
sixteen are white. Greyness is dominant over 
blackness, the presence of chromogen over its 
absence. From the reasoning which we have 
detailed on the nature of dominance we suspect 
that grey must be a modification of black due to 
the introduction of another factor. On in- 
vestigation this proves to be the case, for if 
the grey hairs be examined under the micro- 
scope they are seen to be coloured black at 
the base, then yellow, and then black, so 
that they may be described as " barred with 
yellow." The technical name for grey is 
agouti. The " barring " factor is obviously 
absent in the purely black mice, but chocolate- 
coloured mice are also known in which the 
whole of the hair is suffused with chocolate 
pigment. When these mice are crossed with 
black, the first filial generation are black, but 
when these hybrid blacks are crossed with 162 
AN INTRODUCTION TO 
another, chocolates reappear in the second 
filial generation. Black is therefore dominant 
over chocolate and on investigation it is found 
that black mice contain chocolate pigment, 
but that this is overshadowed by the black 
pigment so as not to be visible when looked 
at in the mass. When one factor, by its 
presence, conceals another factor, as in this 
case, it is said to be epistatic to the factor 
which it conceals, and the concealed factor is 
said to be hypostatic to the one which conceals 
it. Chocolate indeed appears to be the colour 
produced when " chromogen " is present 
alone, without the addition of another factor. 
When grey is crossed with chocolate it is 
dominant over it; it is therefore dominant 
over both black and chocolate; such a 
relationship is denominated " multiple allelo- 
morphism." 
Bateson was able to get one step further 
than Cuenot, for he discovered a case where 
two factors were present as members of two 
different allelomorphic pairs, neither of which 
could manifest itself without the presence of 
the other. There exist two races of white 
sweet-pea (Lathyrus odorata) which, when 
selfed, yield only white flowers, but which when 
crossed yield purple flowers. The production THE STUDY OF HEREDITY 
163 
of colour in the flower is therefore due to the 
meeting of two factors, one of which was carried 
by each white strain, but neither of which taken 
by itself was able to give rise to colour. When 
a second filial generation is raised from these 
purple flowers, then out of sixteen, on an av- 
erage, nine are coloured and seven are white. 
If A and B denote these two factors, and 
a, b their absence, then the second filial genera- 
tion ought to consist of 9 AB, 3 Ab, 3 «B, 
1 ab. But Ab, aB, and ab will all equally 
yield only white, and hence we have seven 
(3 + 3 -|- 1) white out of sixteen. 
We have so far spoken of the different 
allelomorphic pairs of characters as if each pair 
in its distribution in the second filial genera- 
tion was completely independent of every 
other pair. But this is not always the case. 
Sometimes two characters belonging to two 
different pairs show a marked tendency to 
occur together, and are found together in the 
fertilized germ-cell or zygote far oftener than 
is in accordance with the laws of chance. 
Such an association is known as " Coupling." 
If it were complete, then the two characters 
would always occur together and would be 
regarded as two different manifestations of 
the same " factor." It is, however, of all 164 
AN INTRODUCTION TO 
degrees of completeness, but the most com- 
monly accepted theory of the cause of its 
occurrence will be explained in the next 
chapter, when we deal with the inheritance of 
sex. 
Our survey of what we may term the 
" Mendelian " laws governing the inheritance 
of parental qualities by hybrids enables us to 
come to an important conclusion. It is clearly 
proved that by the crossing of distinct strains 
which differ from one another in two or more 
characters, new strains can be produced, 
which combine certain qualities derived from 
both parental strains and which breed per- 
fectly true. If we could conceive of some 
sudden change in the environment which 
should sweep away both parental strains, but 
which a new hybrid strain could successfully 
resist, then a new species would in this way 
be produced, and one daring speculator 
(Lotsy) has asserted that this is the only way 
in which new species can originate, but in this 
view he has found few followers. Were we to 
accept it we should be committed to the posi- 
tion that from the beginnings of life on the 
earth a large number of independent types of 
organism had existed, each with definite 
hereditary characters, and that all progress THE STUDY OF HEREDITY 
165 
had been due to the combination of these 
factors-such as those for rose comb and pea 
comb in fowls-as a result of the constant 
intercrossing of these forms. We must reserve 
our criticism of such views until a later chap- 
ter, in which we shall sum up the general im- 
pression produced on us by all the evidence. 
In the meantime we may be permitted to 
point out the great advantage which a knowl- 
edge of the laws of Mendelian segregation 
confers on the practical breeder. He can 
combine the advantageous qualities of two 
different strains, and from the numerical pro- 
portions in which different varieties appear 
in the second and third filial generations he 
can select those which will prove stable and 
suitable for further propagation. An instance 
will make the high importance of the matter 
clear. It is well known to most people that the 
red, hard wheat grown in Manitoba makes a 
most desirable bread and commands a high 
price. The grains of this wheat are rich in 
gluten, and when dry and broken across 
exhibit a smooth, glassy surface of fracture. 
On the other hand, the wheat grown in the 
east of England makes a heavy, doughy sort 
of bread, and its grains are surcharged with 
starch, and when broken across exhibit a 166 
AN INTRODUCTION TO 
white, granular surface. Now, Professor 
Biffen, of Cambridge, has shown that when 
suitable precautions are taken Manitoba wheat 
can be grown in England. The farmer, how- 
ever, would find it unprofitable to do so, be- 
cause the gain from the higher price which 
this wheat commands is more than offset by 
the loss due to its much smaller yield per acre 
as compared with the English wheat. This 
lessened yield per acre, according to Professor 
Biffen, is due to the lax system of branching 
of this plant, which is a racial characteristic; 
the English plant, on the other hand, produces 
a thick tuft of close-set branches, which ex- 
plains its much higher yield per acre. 
Professor Biffen, by crossing these two kinds 
of wheat, secured in the second filial genera- 
tion a stable hybrid which combined the hard 
grain of the Manitoba wheat with the thick 
tufted branching and consequent high yield of 
the English wheat, and this new strain has 
proved most successful during a series of years, 
and has gained high prizes at agricultural 
shows. It is obvious that whilst the benefits 
which it confers on English farmers are not 
to be under-estimated, if it were to be intro- 
duced into Canada and were to produce any- 
thing like the high yield there which is obtained THE STUDY OF HEREDITY 
167 
from it in England, the profit which it would 
bring to the Canadian farmer would be 
colossal. For whereas the English farmer 
would benefit to the extent of getting three 
or four shillings more per quarter for his grain, 
the Canadian farmer would have his harvest 
increased from 50 to 100 per cent. We con- 
fidently expect in the near future great results 
will be obtained from experiments like this 
which we have just described. CHAPTER VII 
THE INHERITANCE OF SEX 
In the vast majority of the species of the 
higher animals the individuals are of two 
different kinds, male and female, which differ 
from one another even more than do two 
Mendelian strains of the same species. Indeed, 
in the case of some insects, male and female 
might easily be regarded as types of two 
distinct species if their relationship to 
one another were not known. The decisive 
difference between males and females is, of 
course, that the first carry spermatozoa and 
the latter eggs, but with this primary difference 
are associated, as we all know, a number of 
other differences, as, for instance, the beard 
and moustache in men or the comb in cocks, 
which are known as secondary sexual char- 
acters. The question then arises, whether 
males and females might not have the same 
relation to one another as two allelomorphic 
strains of the same species. Now, of course, 
when male and female unite and produce 
168 THE STUDY OF HEREDITY 
169 
children these are not, like the hybrids of 
yellow and green peas, all alike, but are of 
both sexes, since otherwise the race would 
come to an end, and hence the relation of the 
two sexes to one another cannot be simply 
that of dominant and recessive. Since, on the 
whole, in the human race and in the higher 
animals, the sexes are produced in approxi- 
mately equal numbers, the only assumption 
that is tenable, if the inheritance of sex 
resembles that of Mendelian characters, is that 
one of the sexes is hybrid and that the other 
is a pure recessive; for we have already learnt 
that when a hybrid and a recessive are mated, 
equal numbers of hybrids and recessives are 
produced by the cross. If F denotes female- 
ness and M maleness, then in a case of sexual 
union we must either have the cross expressed 
by the formula F(M) X M = F(M) + MM, 
or by the formula M(F) X F = M(F) + FF, 
the character in brackets being the recessive 
one. In the first case we assume that the 
apparent female is a hybrid in which the female 
quality is dominant over the male quality, but 
that the male is a pure recessive. In the 
second case we assume that the apparent male 
is a hybrid, and that the quality of maleness is 
dominant over femaleness, but that the female 170 
AN INTRODUCTION TO 
is a pure recessive. In either case we must 
assume that the sex which we regard as 
hybrid produces two kinds of germ-cells in 
equal numbers, one sort carrying the male and 
the other sort carrying the female characters. 
It follows that there must be either two kinds 
of eggs or two kinds of spermatozoa. Cu- 
riously enough, evidence has been produced to 
show that both these cases can occur. It is 
the merit of the late Professor Doncaster to 
have been the first to have almost demon- 
strated a case where there are two kinds of 
eggs, and where femaleness is dominant over 
maleness, and we shall give an account of 
his work in order to serve as a type of this 
kind of investigation. 
One of the common moths in England is the 
currant-moth, Abraxas grossulariata. It is an 
undistinguished-looking insect whose wings 
are mottled with blotches of a dull reddish- 
brown colour. There exists, however, a rather 
rare variety, the milky currant-moth, which 
has the same general type of coloration on its 
wings, but the splashes of colour are so pale 
as to be nearly invisible, and have somewhat 
different outlines. This variety is known as 
Abraxas lacticolor, and all the specimens of it 
which have been caught wild have turned out 
to be females. THE STUDY OF HEREDITY 
171 
Male specimens of Abraxas lacticolor have, 
however, been bred in captivity. Now when 
one of these wild female specimens of Abraxas 
lacticolor is mated with a male Abraxas 
grossulariata, all the progeny, males as well 
as females, are of the grossulariata variety. 
We see from this that the type of coloration 
denominated grossulariata is dominant over 
the type denominated lacticolor, and this 
latter type, therefore, owes its peculiarities 
to the loss of some " factor " which confers 
on the offspring the quality of being grossu- 
lariata when it is present. When, however, 
a grossulariata female is mated with a lacti- 
color male, a quite different result is produced. 
In this case all the male offspring bear the 
grossulariata character, and all the females 
the lacticolor character. From this circum- 
stance we are driven to conclude, firstly, 
that the grossulariata female is a hybrid and 
produces two kinds of eggs, one kind carrying 
the grossulariata factor, and one kind devoid 
of the factor because all the germ-cells of the 
recessive lacticolor male parent must be devoid 
of the grossulariata factor; and, secondly, 
that the eggs bearing the grossulariata factor 
always develop into males, and those devoid of 
this factor into females. 
Therefore, in this case at least, there are 172 
AN INTRODUCTION TO 
male-producing and female-producing eggs, 
whilst the spermatozoa of the male grossu- 
lariata, which can carry the paternal character 
to both male and female offspring, are all 
alike. But all the eggs, both male-producing 
and female-producing, of a female lacticolor 
are, of course, of the lacticolor variety, and 
when a male-producing lacticolor egg is 
fertilized by a spermatozoon from a grossu- 
lariata male, a hybrid male is produced which 
gives rise to equal numbers of spermatozoa 
with the grossulariata factor and devoid of it. 
If now we mate this hybrid male with a 
lacticolor female, both male and female eggs 
will be fertilized with both kinds of sper- 
matozoa, and there will result offspring of the 
following types, viz., (1) hybrid male grossu- 
lariata (grossulariata spermatozoon X lacti- 
color male egg); (2) pure male lacticolor 
(lacticolor spermatozoon X lacticolor male 
egg) ♦ (3) female grossulariata (grossulariata 
spermatozoon X lacticolor female egg) ; and 
(4) female lacticolor (lacticolor spermato- 
zoon X lacticolor female egg). By this cross, 
therefore, we are enabled to produce lacticolor 
males. It is a remarkable fact that there is 
apparently no such thing as a pure grossu- 
lariata female; every one, whether caught THE STUDY OF HEREDITY 
173 
wild or raised in captivity when tested by 
being crossed with a lacticolor male, reveals 
its hybrid nature by producing two kinds of 
offspring. 
An instance of the same kind in which the 
eggs are all alike whilst there are apparently 
two kinds of spermatozoa is afforded by the 
little fruit fly, Drosophila melanogaster. From 
the study of the heredity of this insect star- 
tling results have been obtained by Professor 
Morgan of New York and his pupils, who have 
chosen it as their special subject for investiga- 
tion. These results have, in fact, formed the 
basis for a complicated theory of the nature of 
the " factors " and their relation to the chromo- 
somes. We shall have to examine this theory 
critically and in detail. Drosophila has a very 
rapid life cycle, since the period which 
intervenes between the fertilization of the egg 
and the production of ripe eggs by the female 
which grows from the egg is only about twelve 
days. It can be fed on a diet of decaying 
bananas, and hence it is comparatively easy 
to raise enormous numbers of the insect and 
to obtain a considerable number of generations 
within the period of a year. 
The normal colour of the eye in Drosophila 
is red, but Morgan found that when he raised Drosophila melanogaster, viewed from the dorsal sur- 
face. A. Typical specimen. B. White-eyed specimen with 
abbreviated wings. C. Red-eyed specimen with vestigial 
wings (after Morgan). 
Fig. 22. 
174 THE STUDY OF HEREDITY 
175 
thousands of individuals a few white-eyed 
males turned up. These " sports " when 
crossed with the red-eyed type acted like 
Mendelian recessives. All of the first filial 
generation had red eyes, and this was the first 
case observed where an isolated abnormal 
individual which turned up in a culture was 
proved, when mated with the type, to yield off- 
spring following the Mendelian rules. When 
these hybrid red-eyed insects were mated with 
one another, the proportions of red-eyed to 
white-eyed amongst the offspring in the 
second filial generation were, as was to be ex- 
pected, about three to one. All the white- 
eyed specimens in the second filial (F2) 
generation were males, as were also one-third 
of the red-eyed. Since on the average equal 
numbers of males and females were produced, 
half the entire population would be males; 
of these a half-i. e., a quarter of the whole 
population-are provided by the white-eyed 
members, and therefore another half-i. e., 
a quarter of the whole population-must be 
red-eyed. As the red-eyed members altogether 
form three-quarters of the population, it 
is clear that a third of them must be males. 
In order to obtain white-eyed females it was 
necessary to mate the hybrid red-eyed females 176 
AN INTRODUCTION TO 
with white-eyed males, and in this way both 
red-eyed and white-eyed males and females 
were produced. When red-eyed males, whether 
obtained in this way or by selection from 
amongst members of the original strain, were 
mated with white-eyed females, they produced 
red-eyed and white-eyed offspring, amongst 
which all the females were red-eyed and the 
males white-eyed. White-eyed males and females 
mated together gave nothing but white-eyed 
offspring. Adopting the same kind of reason- 
ing which we made use of in the case of 
Abraxas grossulariata, we conclude: (1) that 
there are two kinds of spermatozoa, male- 
producing and female-producing; for we 
know that a hybrid red-eyed male must 
produce what we may term " red-eyed " and 
" white-eyed " germ-cells in equal numbers, 
and when these are used to fertilize the eggs 
of a white-eyed female, all of whose eggs must 
be recessive " white-eyed " germ-cells, all 
the red-eyed offspring, half of the whole 
number, are females; (2) that the eggs of the 
female are all alike with regard to the trans- 
mission of sexual characters, for when hybrid 
red-eyed females, which, of course, produce 
two kinds of eggs, one bearing the " red-eyed " 
factor and another devoid of it, were crossed THE STUDY OF HEREDITY 
177 
with white-eyed males, all of whose spermato- 
zoa were devoid of the factor for red eyes, 
red-eyed males and females were produced 
as well as white-eyed males and females, 
showing that both " red-eyed " and " white- 
eyed " eggs, when fertilized with " white-eyed " 
spermatozoa, gave rise to males and females in 
equal numbers. (3) Since in Drosophila the 
male is a hybrid and produces two kinds of 
germ-cells, yet exhibits male qualities, in this 
genus maleness must be dominant and female- 
ness recessive; in Abraxas, on the other hand, 
where the female produces two kinds of germ- 
cells yet is female in appearance, femaleness 
must be dominant and maleness recessive. 
It is a most interesting thing that there are 
certain types of human disease, or rather 
deformity, which seem to be inherited in 
exactly the same manner as the red eye and 
the white eye in Drosophila Melanogaster. 
One of these is the affliction known as night- 
blindness, the sufferer from which becomes 
totally blind in the dusk. This disease is due 
to the absence from the retina of the eye 
of certain sense-cells which are adapted to 
respond to light of low intensity. If a night- 
blind man marries a normal woman all his 
children are normal, and as the affliction is a 178 
AN INTRODUCTION TO 
recessive, due to the absence of something, 
this is what we would expect. When these 
children in turn marry, the sons have normal 
children, but some of the daughters are likely 
to have night-blind sons. Night-blindness 
does not seem to affect women, except in very 
rare cases. We therefore have the paradox 
that the disease appears in men, but is trans- 
mitted by women. Now let us compare night- 
blindness to the white eye in Drosophila, as 
we may very fairly do, since both are defects 
of the visual organs. We assume, then, that 
there are two kinds of spermatozoa, male- 
determining and female-determining, and that 
consequently the sex of a human being is deter- 
mined by his or her father. We further assume 
that the male-producing spermatozoon is 
devoid of the factor for seeing in the dusk, 
so that every male is a hybrid who derives 
his capacity to see in the dusk from his mother, 
and all his male-producing spermatozoa are 
devoid of the factor for seeing in the dusk. 
When a man is afflicted with night-blindness, 
however, he is a pure recessive, and not only 
his male-producing, but also his female- 
producing spermatozoa are devoid of the 
" dusk-seeing " factor. But still he cannot 
transmit his affliction to his children because THE STUDY OF HEREDITY 
179 
the defectiveness of his spermatozoa is over- 
come by the normality of his wife's ova. But 
their children, both boys and girls, are hybrids, 
and each produces normal and abnormal 
germ-cells in equal numbers. When one of 
these girls marries a normal man who is also 
ex hypothesi a hybrid, then it may happen 
that one of her imperfect eggs is fertilized 
by one of his defective male-producing sper- 
matozoa and so a night-blind male baby is pro- 
duced. 
The reader will see, however, that theoreti- 
cally a night-blind girl baby could be produced 
if the daughter of a night-blind man were to 
select another night-blind man as her husband. 
For all the spermatozoa of her husband, female- 
producing as well as male-producing, would be 
defective, and as she can produce defective 
eggs as well as perfect ones, if such an egg were 
fertilized by one of her husband's female-pro- 
ducing spermatozoa, she should bear a night- 
blind daughter; and such cases, though very 
rare, are not entirely unknown. 
So far as investigation has gone, birds agree 
with moths in producing two kinds of eggs, 
whilst mammals and bugs and most other in- 
sects agree with flies in producing two kinds of 
spermatozoa. 180 
AN INTRODUCTION TO 
This type of inheritance, where certain 
characters appear normally only in one sex 
but are inherited through the opposite sex, is 
termed by Morgan " sex-linked inheritance," 
and it may be defined as a weakness or defect 
normally associated with a certain proportion 
of the eggs or of the spermatozoa which under 
ordinary circumstances remains entirely con- 
cealed, and which can only be rendered manifest 
when the same defect turns up accidentally 
in a germ-cell of the opposite sex. Morgan 
has brought forward a plausible and ingenious 
hypothesis to account for this kind of inherit- 
ance based on his observations on Droso- 
phila. The diploid number of chromosomes in 
this insect is eight, but in the nuclei of the cells 
of the female these eight consist of four pairs 
of similar chromosomes. In the nuclei of the 
cells of the male, on the other hand, we find 
three pairs of similar chromosomes and two 
unlike ones. These aberrant chromosomes are 
termed the X and the Y chromosomes respec- 
tively. The X chromosome is represented by 
a pair in the female nucleus, but the Y 
chromosome, distinguished by its hook-like 
form, is not found in the female. When the 
reducing division takes place, the X and Y 
chromosomes undergo longitudinal splitting THE STUDY OF HEREDITY 
181 
as in an ordinary nuclear division, but in the 
second maturation division of the male nucleus 
they pair with one another and then separate, 
an X chromosome being distributed to half 
the male germ-cells and a Y chromosome to 
the rest (Fig. 23). The same division in the 
ripening egg-cells gives rise to ripe eggs and 
Fig. 23. 
The second maturation division of a male germ-cell of 
Euschistus variolarius showing the separation of the X 
and Y chromosomes. 
second polar bodies each with an X chromo- 
some. Since the nuclei of the male body have 
each an X and Y chromosome, and since 
every egg has an X chromosome, the males 
must have developed from eggs which have 
been fertilized by spermatozoa each of which 
had a Y chromosome; and since the nuclei 
of the female body have two X chromosomes,  AN INTRODUCTION TO 
182 
the female must arise from an egg fertilized 
by a spermatozoon with an X chromosome. 
Now if we assume that the " sex-linked " 
factor for white eyes is based on something 
in the X chromosome, then a white-eyed 
female will possess what we may call white X 
chromosomes in all her nuclei. We further 
assume that the Y chromosome " carries no 
factors," as Morgan says, and is therefore also 
incapable of giving rise to eye-pigment and may 
be called white, so that all spermatozoa which 
possess a Y chromosome will also be white- 
eyed. When, therefore, an ordinary male 
fertilizes a red-eyed female all the young are 
red, because all contain a red X chromosome, 
but the females contain two red X chromo- 
soms, the males a red X and a white Y. 
When, however, a normal male fertilizes 
a white-eyed female all the X-bearing sper- 
matozoa will produce females which have 
red eyes owing to the red X chromosome, 
and all the Y-bearing spermatozoa will produce 
males which are white-eyed owing to their 
possession, in addition to the maternal white X 
chromosome, of only a white Y chromosome. 
The reason, therefore, that the normal male 
behaves like a hybrid is that it possesses a Y 
chromosome. It may seem to many a forced 
and improbable assumption that Y " carries THE STUDY OF HEREDITY 
183 
no factors"; the only justification for it is 
that in some insects we find in the male nuclei 
only one X chromosome and no Y at all, 
whereas the female nuclei possess two X 
chromosomes; we therefore suppose that the 
Y chromosome is something that can be 
present or absent, and hence Morgan assumes 
that it has no function at all. 
On this explanation of these aberrant 
chromosomes Morgan founds his whole system. 
If we go back to the simple case of the crossing 
of two strains of peas, yellow and green, differ- 
ing from each other only in one character, 
Morgan assumes that this difference is due to a 
peculiar factor in one pair of chromosomes of 
the recessive strain and of a different factor in 
the corresponding chromosomes of the dominant 
strain.1 We shall call the former the weak 
chromosome pair, the latter the strong 
chromosome pair. In producing the hybrid 
a germ-cell with one weak chromosome unites 
with a germ-cell with a strong chromosome 
to form a zygote. When the hybrid forms 
germ-cells in* the reducing division, corre- 
sponding weak and strong chromosomes come 
together and then separate, the weak chromo- 
some going into one germ-cell and the strong 
i Morgan does not accept the "presence and absence" 
theory of Bateson, v. page 158. 184 
AN INTRODUCTION TO 
into another, and so we see why the hybrid 
produces equal numbers of germ-cells bearing 
the paternal and maternal characters. The 
formation of four types of germ-cell by hybrids 
produced by crossing two strains differing 
from one another in two characters is explained 
by supposing that each differentiating char- 
acter is due to an alteration of a different 
pair of chromosomes. When the hybrid 
forms germ-cells, corresponding paternal and 
maternal chromosomes approach one another 
as before to form diploid pairs, but the paternal 
ends of different pairs are not all turned in the 
same direction, i. e., towards the same pole 
of the karyokinetic spindle, whence it follows 
that when the members of these pairs separate 
from one another, one germ-cell may receive 
the paternal member of one pair of chromo- 
somes and the maternal one of another; and 
so an independent distribution of the two 
characters amongst the germ-cells of the 
hybrid results. In theory, therefore, there 
should be just as many pairs of independent 
differentiating characters as there are pairs 
of chromosomes and no more. In practice 
it is found that there are more, but so far as 
Drosophila is concerned, Morgan has striven 
to show that all these characters are " coupled " THE STUDY OF HEREDITY 
185 
(see p. 163) into four groups corresponding 
to the haploid number of chromosomes. (It 
is always to be remembered, however, that 
in this computation the Y chromosome is 
ignored, yet this chromosome in Drosophila 
is at least as large as the X chromosome.) 
That is to say, that the characters which are 
" coupled " together are usually found together 
in one germ-cell when the hybrid produces 
germ-cells. That this is not always so is ex- 
plained by means of another daring sup- 
position put forward by Morgan. He assumes 
that when paternal and maternal chromo- 
somes come together they twist round each 
other in a spiral line, and that when they 
separate it often happens that this separation 
takes place in a different plane from that of 
the union, so that the paternal chromosome 
carries away with it a bit of the maternal 
chromosome and the maternal chromosome a 
part of the paternal chromosome. This 
hypothetical occurrence is termed by Morgan 
" crossing over." From the frequency with 
which coupled characters accompany one 
another when the coupled group is broken, 
Morgan draws deductions as to the nearness to 
or distance from one another of the factors 
which produce them on the chromosome; he 186 
AN INTRODUCTION TO 
assumes indeed that the chromosome consists 
of a linear group of factors or, as he prefers 
to call them, " genes." All the characters 
that normally go with the X chromosome 
in the female-producing spermatozoon are 
supposed to have their basis in the " genes " 
in this chromosome. The smallest group of 
coupled characters is supposed to be situated 
in the smallest chromosome, which is a small 
round dot, and proceeding in this way Morgan 
has actually constructed chromosome maps. 
This theory, although it has been readily 
and uncritically accepted by many " geneti- 
cists " both here and on the continent of 
Europe as well as in England, encounters 
enormous difficulties when it is followed out 
in detail. For a full discussion of these diffi- 
culties we must wait till the next chapter. 
In Drosophila, therefore, and in most other 
insects as well as in mammals, femaleness 
is supposed to be due to an excess of one 
particular kind of chromatin (the X chromo- 
some), and maleness to a defect of the same 
material, whereas in birds, butterflies and 
moths, maleness is supposed to be the conse- 
quence of an excess of the X chromosome 
material and femaleness to its defect. But 
femaleness and maleness are essentially the THE STUDY OF HEREDITY 
187 
same kind of quality throughout the animal 
kingdom, and it is inconceivable that an excess 
Male and female specimens of the worm Boneilia. 
A. The female. B. The male. 
Fig. 24. 
of a particular kind of substance should cause 
maleness in one type of animal and femaleness 188 
AN INTRODUCTION TO 
in another. Further, we have strong evidence 
that the powers of developing female and male 
characteristics are present in all animals, and 
that when a particular individual becomes a 
female, all that happens is that the male 
characteristics are suppressed and vice versa. 
A few examples will make this plain. Fig. 24* 
represents an extraordinary marine worm 
called Boneilia, which is characterized by 
having a large forked proboscis in front of the 
mouth. In this species of animal the male 
is a minute parasite which lives in the genital 
duct of the enormous female, which may be 
eighteen inches long. All the larva when 
hatched from the egg are alike. They are 
little spherical larva with a girdle of cilia which 
are termed " trochophores," and for a short 
time they swim freely in the sea. Then they 
settle down and begin their lives as worms on 
the bottom. If, however, a larva settles down 
on the proboscis of a female it becomes a 
male. If a larva be allowed to attach itself 
to the proboscis of a female and remain there 
some little time, and be then detached, it 
develops into a hermaphrodite. Here sex is 
obviously determined by the environment. 
Then, we have the extremely interesting ob- 
servations of Geoffrey Smith on the effects THE STUDY OF HEREDITY 
189 
of certain parasites on their hosts, which, in 
the case we shall select for special description, 
consists of a certain species of Mediterranean 
crab called Inachus scorpio (Fig. 25). The 
parasites belong to a race of degenerate 
barnacles termed Rhizocephala (Gr. = 
Male specimen of the crab Inachus Scorpio, viewed 
from above. 
Fig. 25. 
root), because the stalk by which they are 
attached to the unfortunate crab grows out 
into a series of roots which permeate all 
portions of the crab's body and absorb nourish- 
ment from the blood. The alimentary canal 
and legs of the barnacle, being no longer used, 
have disappeared, and its body has become 
reduced to a shapeless sac containing the 190 
AN INTRODUCTION TO 
generative organs. Like other barnacles, it 
develops from a small, free-swimming larva 
provided with antennas, eyes and swimming 
feet, all of which it loses when it finds its host. 
The particular species which infests Inachus 
scorpio is termed Sacculina neglecta, and it 
appears to attack chiefly young specimens. 
The diseased crabs cease to grow and are rather 
more sluggish in their movements than their 
neighbours, but otherwise they do not show 
any signs of inconvenience. 
Now it may not be superfluous to remind 
our readers that crabs possess strongly-marked 
secondary sexual characters, which allow the 
sexes to be easily distinguished from one 
another by the naked eye. The abdomen of 
the female is broad and hollowed on its under 
surface, which during life is kept bent upwards 
against the thorax. It bears four pairs of 
forked legs fringed with long hairs, and the 
eggs, after being shed, are attached to these 
hairs, and so the mother carries the eggs about 
with her until the young ones hatch out (Fig. 
26, B). 
The abdomen of the male is much narrower 
and is not grooved. It bears only two pairs 
of appendages which are not forked, nor much 
pointed, but are comparatively rigid (Fig. 26, THE STUDY OF HEREDITY 
191 
D). These are termed the copulatory styles, 
because they are used to introduce the male 
germ-cells, which in crabs are motionless (see 
Fig. 26. 
Abdomens of normal and diseased specimens of Inachus 
Scorpio, viewed from below. A. Abdomen of immature 
female. B. Abdomen of mature female. C. Abdomen 
of female which has been infected by Sacculina neglecta; 
the circular disc on the second segment shows where 
the parasite was attached. D. Abdomen of male. E. 
Abdomen of male which has been infected by Sacculina 
neglecta. 
Chap. I), into the egg-tubes of the female. 
The male has also much larger claws than the 
female, which he apparently uses in fights with 
his male rivals. 192 
AN INTRODUCTION TO 
In immature females the abdomen is not 
grooved, and its appendages are not so much 
jointed nor as richly provided with hairs as 
in the adult female (Fig. 26, A). Now when 
a young half-developed female is attacked by a 
Sacculina it assumes at once, without further 
growth, the characters of the adult female, 
that is, the abdomen becomes grooved and the 
appendages long and hairy, although some of 
them may be lost (Fig. 26, C). If we examine 
by dissection the effect of the parasite on its 
host we find that the only organ which suffers 
is the ovary; the eggs mostly degenerate, 
break up and disappear. The parasite, on the 
other hand, after it has produced several 
batches of eggs-a process which occupies 
from eighteen months to two years-dies and 
drops off and its roots become absorbed. 
The place where it was attached heals up, 
leaving a noticeable scar (Fig. 26, C), and in 
some few cases the host still lives on and may 
resume growth, and in this case a remnant of 
the ovary may be left and may produce eggs. 
On the other hand, when a half-developed 
male is attacked, a widely different and 
extremely interesting result is produced. 
Female secondary sexual characters make 
their appearance. The abdomen becomes THE STUDY OF HEREDITY 
193 
broad, although not quite as broad as that of 
a normal female, and additional abdominal 
appendages appear, which, though smaller 
than those of the normal female, have the same 
characteristic form (Fig. 26, E). The claws 
also remain about the size of those of a femaje. 
In the case of the male also dissection reveals 
the fact that the principal, if not the exclusive, 
damage done by the parasite is experienced 
by the genital organ, the lobules of which are 
seen to be broken up and destroyed. Never- 
theless, as in the case of the female, so here, 
the host sometimes outlives the parasite and 
then resumes growth and begins to regenerate 
its generative organs, but these now consist 
of both ovary and testis, so that the crab has 
become a hermaphrodite. Here, again, as in 
the case of Boneilia, an accident of the environ- 
ment has determined sex. How are we to 
conceive of the action of the parasite? Do 
its roots emit a corrosive substance which 
digests and destroys the generative cells of 
the host, or do these cells dwindle and die 
simply because the parasite withdraws from 
the blood of the host the special substance 
necessary for their maintenance? If the 
action of the roots of the parasite were a 
directly digestive one, we could not understand 194 
AN INTRODUCTION TO 
the introduction of female characters into the 
male. Geoffrey Smith's brilliant and sugges- 
tive hypothesis is as follows: he assumes that 
every male is a potential hermaphrodite and 
contains what we may term female substance 
in his composition, which under normal 
circumstances is kept under by the superior 
potency of the male substance; but the female 
is supposed to have only female substance. 
These substances are conceived of by Geoffrey 
Smith as circulating in the body fluids, and 
thus producing the secondary sexual char- 
acters ; and also, by becoming stored up in 
certain of the interior cells, producing the 
primary sexual organs. The roots of the Sac- 
culina are supposed to have a special affinity 
for the male substance, which they absorb in 
far greater proportion than the female sub- 
stance ; and, as a result, in a male attacked 
by a parasite the proportion of female sub- 
stance increases and is sufficient to produce 
the female secondary sexual characters, and 
when the parasite is removed to generate eggs 
also. 
There is, no doubt, a large measure of truth 
in Geoffrey Smith's hypothesis, but one or two 
criticisms may be made. He points out that 
his hypothesis agrees with those cases of sex- THE STUDY OF HEREDITY 
195 
limited inheritance where there are two kinds 
of spermatozoa, and where the male is repre- 
sented by the formula M(F) and is a hetero- 
zygote. He asserts that in all cases of 
hermaphroditism the male elements ripen 
first, and that hermaphrodites only differ from 
ordinary males in having more female 
substance than they. But his hypothesis is 
in complete contradiction to the case of 
Abraxas grossulariata, where there are two 
kinds of eggs and the female is represented by 
the formula F(M), whilst the male is pure 
M. Nor is it universally true that in her- 
maphrodites the male elements ripen first. 
In compound ascidians like Salpa, the eggs 
ripen before the spermatozoa, and there are 
numerous indications from other divisions of 
the animal kingdom that females must contain 
some male substance. 
Thus, in the primitive fish Amphioxus, in 
which the generative organs of both sexes 
consist of two longitudinal series of sacs, 
Goodrich has pointed out that in a female, one 
sac in the series may produce spermatozoa. 
Again, Crew has recorded the case of a hen 
which had laid eggs for several years from 
which chickens had been raised, which turned 
into a cock in its old age. This animal not 196 
AN INTRODUCTION TO 
only acquired the comb and the characteristic 
feathers (hackles) of a cock, but its crow and 
its attitude towards the hens, and it finally 
became the father of several chickens by 
successfully " treading " several hens. When 
it was killed it was discovered that the ovary 
had been destroyed by tubercular disease, but 
that a male-gland or testis had sprouted from 
the membrane lining the body-cavity. If we 
assume that in the female there is always some 
male substance which is inhibited by the 
preponderance of female substance, we shall be 
able to understand the effect of the parasite 
in producing precocious female maturity. 
For we may reasonably regard the male 
substance present in a young female as the 
principal cause of her immaturity, and when 
this male substance, which normally would be 
inhibited by the gradually increasing female 
substance, is suddenly removed by the 
parasite, maturity at once supervenes. 
This modification of Geoffrey Smith's 
hypothesis which we have adopted is strongly 
confirmed by the results obtained quite 
recently by Richard Goldschmidt. Gold- 
schmidt experimented with the moth Lyman- 
tria dispar, commonly known as the gipsy 
moth. This species is divided into a good THE STUDY OF HEREDITY 
197 
many local races, and characteristic races 
inhabit Japan, Europe and North America. 
When a Japanese female is fertilized by a 
European male the progeny consists of normal 
males and females, but when a European female 
was fertilized by a Japanese male a very dif- 
ferent result was obtained. Normal males 
made up half the filial generation, but the 
females presented various mixtures of male 
and female characters. Such abnormal 
females were termed intersexes. Where the 
male admixture of characters was slight it 
concerned only external features, such as the 
colour of the wings or the shape of the antenna?, 
but where the male influence was more 
pronounced, the ovary was partly or com- 
pletely transformed into a testis, and in 
extreme cases the female was almost entirely 
changed into a male. Goldschmidt interprets 
his results in the following manner. Each 
individual has in it the potentiality of develop- 
ing the organs of both sexes, but the develop- 
ment of these organs is dependent on the 
stimulus of certain ferments set free in the 
blood during development, which he calls the 
sexual hormones. The intensity of these 
hormones varies in different races; and when 
the Japanese spermatozoon penetrates the 198 
AN INTRODUCTION TO 
European egg it carries into it an exceedingly 
powerful male hormone. It will be remem- 
bered that moths have two kinds of eggs- 
male-producing and female-producing. Now 
Goldschmidt assumes that the male hormone 
is contained in the X chromosome, but that 
the female hormone is in the protoplasm of 
the egg. In the male-producing egg the X 
chromosome which it contains reinforces the 
action of the X chromosome introduced by 
the spermatozoon, and a male insect is formed. 
In the female-producing egg, which is devoid 
of an X chromosome, the female hormone in the 
protoplasm is stronger than the male hormone 
introduced by the X chromosome of its own 
spermatozoon, but it can be wholly or partly 
overpowered by the stronger influence of the 
male hormone introduced by the X chromo- 
some of the spermatozoon of the foreign race. 
But if this kind of reasoning is valid it must 
apply also to Drosophila, which has one kind 
of egg and male-producing and female-pro- 
ducing spermatozoa. Every egg has an X 
chromosome, and if the action of this 
chromosome is reinforced by the entry of an 
X-bearing spermatozoon a female is formed, 
whereas a spermatozoon without an X leaves 
the X chromosome in the egg unable to contend THE STUDY OF HEREDITY 
199 
with the male hormone, and so a male is formed. 
This male hormone must be either in the 
protoplasm of the egg or in the other chromo- 
somes. Intersexes in Drosophila have been 
actually observed in cases where, owing to 
irregularities of cell-division, the number of 
chromosomes has become multiplied. Thus 
individuals have been noted in whose cells 
there are twice the number of ordinary 
chromosomes and three X chromosomes, and 
these are females with a certain infusion of 
male characters-in a word, intersexes. In 
Drosophila, then, it would seem that the male 
hormone is contained in the ordinary 
chromosomes. 
The sexes in frogs show a series of extremely 
interesting phenomena. The frog mates in 
spring. Both sexes after waking from their 
winter sleep go into the water, and there the 
male mounts on the back of the female, 
clasping his arms around her waist; he remains 
in this position for a considerable period, 
croaking wildly with excitement. At length 
when the eggs are ripe the female discharges 
them into the water in which she is sitting, 
being helped to get rid of them by the pressure 
which the male exercises on her abdomen. 
As they pass out the male fertilizes them 200 
AN INTRODUCTION TO 
by discharging his spermatozoa on to them 
(see Chapter IV). Now Hertwig showed that 
whereas both sexes develop in approximately 
equal numbers from spawn laid under normal 
circumstances, if a female, after having 
laid a few eggs, be forcibly separated from 
her male partner and kept separate from 
him for two and a half days, and then 
allowed to rejoin him, she will lay the rest 
of her eggs, but the proportion of males de- 
veloping from these eggs will be seven times 
as great as the females. One of Hertwig's 
pupils, by keeping male and female apart for 
four days, succeeded in obtaining a culture 
from which only males developed. As the 
rate of mortality in these cultures had been 
kept quite moderate, it is impossible to resist 
the conclusion that eggs which would have 
developed into females under ordinary cir- 
cumstances had been forced to develop into 
males by being retained for two or three days 
too long within the womb of the mother. 
That the effect was due to a change in the 
eggs and not in the spermatozoa was proved 
by allowing males at different periods in the 
breeding season to fertilize the same female, 
and in all cases the same proportions of males 
and females developed from the spawn. The THE STUDY OF HEREDITY 
201 
possibility of performing this experiment was 
afforded by the wide range of the edible frog 
{Rana esculent a} on the continent of Europe. 
It is found everywhere from Northern Ger- 
many to Central Italy. The breeding season 
begins very much earlier in the south than in 
the north, and so when males collected in 
Central Germany were beginning their breeding 
season, males collected on the same day in 
Central Italy were nearly at the conclusion 
of it. Hertwig procured males from all 
quarters, which were sent by express train 
to Munich, where the experiment was per- 
formed. 
Witschi, a pupil of Hertwig, has made some 
still more wonderful discoveries. He showed 
that in certain localities, such as Holland, if 
one examines young frogs which have just left 
the water, one finds that nine out of every 
ten are females; nevertheless amongst the fully- 
grown frogs from the same locality there are 
just as many males as females. It follows that 
four out of every five males must have been pro- 
duced from females which changed as they grew 
up. Further observations proved that this is 
indeed the case. Witschi discovered many 
half-grown females in which the ovaries were 
being invaded and destroyed by testis tubes 202 
AN INTRODUCTION TO 
growing into them from the surface, just as 
happened in the case of the abnormal hen 
mentioned above. Further, a review and 
repetition of Hertwig's experiments showed 
that in the experiments in which all the eggs 
were induced to develop into males a large 
proportion of these when they had reached 
the stage of young frogs were females, and that 
their male character only developed as growth 
proceeded. 
We see, therefore, that if the X chromosome 
exercises an influence in favouring in some 
cases the production of male characters, and 
in other cases the appearance of female 
characters, it is by no means the only cause of 
the production of these characters, and that 
the influences of the environment during early 
growth have also powerful effects in determin- 
ing some individuals to be males and others 
to be females. 
We must now devote a little attention to 
the secondary sexual characters. By this 
term is meant not only the outward marks 
which distinguish the male from the female, 
but also the ducts or tubes which convey the 
eggs and spermatozoa to the exterior. Thus 
a woman is distinguished from a man not only 
by the absence of a beard and by her large THE STUDY OF HEREDITY 
203 
and well-developed breasts, but by the posses- 
sion of egg-tubes and a womb in which the egg 
develops. Amongst vertebrates (fishes, am- 
phibia, birds and mammals) the young embryo 
passes through a stage in which male and 
female sexes cannot be distinguished from one 
another, and the development of the external 
marks of the male or female depends on 
hormones emitted by the genital cells. If in 
a young male rat the testis be extracted and 
replaced by an ovary, the breasts enlarge and 
secrete milk and the animal takes on the 
habits and appearance of a female. Therefore 
in this group the secondary sexual characters 
are dependent on the primary for their 
development. But this is not so amongst 
insects. If the testis be removed from a moth 
caterpillar and replaced by an ovary, the grub 
will still metamorphose into an insect with all 
the external male characters, although its 
body is full of eggs. The question as to how 
these secondary characters are inherited in 
insects has been investigated by Miss Foote and 
Miss Strobeli. 
These ladies chose as the subjects of 
their experiments two species of the genus 
Euschistus. Euschistus is a genus of bugs 
which feed on plants; these insects show 204 
AN INTRODUCTION TO 
well-marked X and Y chromosomes in their 
germ-cells (Fig. 23). Like Drosophila, the 
male possesses in its body-cells an X and a Y 
chromosome, whereas the female has two X 
chromosomes. One of the species which was 
experimented with, viz., Euschistus variolarius, 
had in the male sex a well-defined black 
circular dot on the genital segment, which 
is the sixth segment of the abdomen (Fig. 27). 
The other species {Euschistus serous) was de- 
void of this spot, and in neither species did it 
occur in the female sex. On one occasion, and 
on one occasion only, a male E. servus fertilized 
a female E. variolarius. Fortunately, this union 
proved so prolific that no less than 120 eggs 
were produced, but many of these were abnor- 
mal and failed to develop. Only twenty-seven 
of them succeeded in developing to maturity, 
and of these eleven were males and sixteen 
females, and these constituted consequently the 
first filial generation. Seven distinct pairs of 
this generation were mated with each other and 
produced an abundant second filial generation. 
An attempt was made to induce a male E. 
variolarius to mate with a female E. servus, 
but this was unsuccessful. It was possible, 
however, to induce a male E. variolarius to 
mate with a female hybrid of the first filial THE STUDY OF HEREDITY 
205 
generation, and from this union twenty-six per- 
fect insects were reared. 
Now the results of these experiments were 
somewhat unexpected. It was found that 
many of the F1 hybrids exhibited the spot, 
and, therefore, this essentially male character 
must have been inherited from the mother, 
E.VARIOLARIUS. 
Fig. 27. 
E. SERVUS. 
The males of the bugs Euschistus variolarius and 
Euschistus servus, seen from beneath. 
since the male parental strain belonging to the 
species E. servus is devoid of the spot. Con- 
sequently the " factor " which produces the 
spot cannot be situated in the Y chromosome, 
because there is no Y chromosome in the 
female. If we evade this difficulty by assum- 
ing that the factor for the spot, since it is 
inherited from the mother, is carried by the X 206 
AN INTRODUCTION TO 
chromosome, but can only appear in the male 
sex, then we have to explain the circumstance 
that in the offspring of the cross between the 
pure E. variolarius male and the F1 female 
hybrid the spot appears, and is much more 
strongly marked than in any of the F1 male 
hybrids. This greater strength can only be 
explained as due to the influence of the E. 
wariolarius father, and, therefore, the spot can 
be inherited from the male E. variolarius. It 
follows that the factor for the spot cannot be 
situated in the X chromosome, for the male- 
determining spermatozoon does not possess an 
X chromosome; and, of course, a male mem- 
ber of the hybrid strain is produced by a male- 
determining spermatozoon. 
In the second filial generation some of the 
specimens showed the spot as strongly devel- 
oped as in the original strain, and others were 
completely devoid of it; but in some cases 
the offspring of F1 males, which were devoid 
of the spot, showed it strongly marked. In 
other cases the spot was faintly marked. If 
we are precluded from placing the factor for 
the spot in either the X or the Y chromosome, 
our next recourse would be to one of the 
ordinary chromosomes. But the factors for 
the spot must be situated in both members of THE STUDY OF HEREDITY 
207 
a pair of ordinary chromosomes, since, if 
they were in one chromosome only, there 
would be in the species E. variolarius two 
separate strains, one with and one without 
the spot, and half the males would be devoid 
of it, but this cannot be the case, since the spot 
in the pure species is a constant character. 
If, however, we make the assumption that the 
factors for the spot are situated in an ordinary 
pair of chromosomes, then the sequence of the 
maturation divisions baffles us entirely. For 
we must remember that the first maturation 
division is the reducing division in the case 
of the ordinary chromosomes, and ordinary 
longitudinal division occurs in the secondary 
maturation division, whereas in the case of 
the X and Y chromosomes ordinary longi- 
tudinal division takes place in the first 
maturation division, but union and final 
dej unction of the supposedly male and female 
determining elements occur in the second 
maturation division. Any character situated 
in any ordinary chromosome whatever must 
therefore inevitably be distributed to both 
male-determining and female-determining sper- 
matozoa. A diagram will make this clear. 
Let us call the pairs of ordinary chromosomes 
A and A1, B and B1, etc., and the results of 208 
AN INTRODUCTION TO 
their longitudinal division a, a1, b, b1, etc. 
Let us suppose that the factors for the spot 
are contained in the pair of chromosomes AA1. 
Then the maturation divisions will be repre- 
sented thus: 
female-determining 
abcdx spermatozoon. 
ABCDxt/ 
dbcdy male-determining 
spermatozoon. 
AA'BB'CC'DD'XY 
female-determining 
a'b'c'd'x spermatozoon. 
A'B'C'D'a't/ 
a'b'^d'y male-determining 
spermatozoon. 
We may endeavour to reconcile this con- 
clusion with the facts by supposing that 
though the factors for the spot are situated 
in a pair of ordinary chromosomes, and it is 
therefore inherited by females, yet its appear- 
ance in a female is prevented by the presence 
of an inhibitor. This is a quite reasonable 
assumption, since in the female the shape of 
the genital segment is totally different from 
its shape in the male; it is, in fact, represented 
by a pair of triangular lobes and there is no 
room for the spot to appear on it. But we 
have also to account for the fact that the spot 
is very feebly developed in most of the F1 
hybrid males and completely absent from some THE STUDY OF HEREDITY 
209 
of them, for all these hybrids have received 
the factors for the spot with the chromosome 
which has come to them from their E. vario- 
larius mother. Therefore it would seem neces- 
sary to assume that E. servus males carry an 
inhibitor for a spot which they do not possess 
and transmit this inhibitor to the male hybrid 
offspring, which, to say the least of it, is not a 
very likely assumption. 
We may now sum up our general conclusions 
as to the inheritance of sex. It seems clear 
that the development of maleness and female- 
ness represents two different types of growth 
which affect in greater or less degree all the 
organs of the body. The power of starting 
these two types of growth is inherent in every 
individual of the species; which type will 
gain the upper hand depends on a variety of 
circumstances. Sometimes it is determined by 
external influences (Boneilia), sometimes by 
the constitution of the blood as affected by 
disease (Inachus and Crew's transformed hen), 
and sometimes-it appears, at any rate, a 
plausible supposition-by the proportion be- 
tween two different sets of chromosomes in 
the same nucleus. But in all cases we are 
dealing with a wavering balance between two 
possibilities-it is a question of those varia- 210 
AN INTRODUCTION TO 
tions of " more " and " less " which Mendel ex- 
pressly excluded from his consideration, and 
to talk of sex being determined by a Mendelian 
" factor " is absurd 
In the human race the knowledge that male 
potentialities are inherent in the constitution 
of a woman and that a man has a certain 
feminine element in his nature helps to explain 
a number of puzzling facts. We see why in 
many women when the child-bearing epoch 
is past and the ovaries diminish in size, the 
repressed male factors make themselves felt 
and the grandmother develops a beard and a 
moustache. Further, the slender and almost 
girlish build of many growing boys is due to the 
fact that the feminine element is yet not over- 
come by the growing male substance. But 
human beings-whether men or women-are 
not uniform mixtures of male and female sub- 
stances. The proportion varies from individual 
to individual, and so we find that there are 
womanly women and mannish women, and 
effeminate men and manly men, or, as we 
Americans tersely denominate them, " he- 
men." It has been surmised, and is probably 
true, that in sexual attraction, which is the 
great motive power in life, each of the sub- 
stances in the constitution of an individual THE STUDY OF HEREDITY 
211 
seeks to unite with its opposite, so that the 
forceful man is attracted by the tender, 
shrinking woman, and the effeminate man is 
usually captured by the strong-minded member 
of the opposite sex. CHAPTER VIII 
THE CAUSES OF MUTATIONS 
In Chapter VII we learnt that when 
Morgan raised large numbers of the fly Droso- 
phila melanogaster, " sports " of various kinds 
turned up, and that these " sports " when 
crossed with the type behaved in a Mendelian 
manner-that is to say, they were recessive 
to the type in the first filial generation, but 
segregated out in the correct proportions in 
the second. Morgan's experience in this re- 
spect is similar to that of other breeders. 
In all cases where large stocks of animals or 
plants are kept, sooner or later " sports " turn 
up, which breed true when crossed with 
their like, but which "mendelize " when 
crossed with the type. It is the fashion to 
call these sports " mutations," in order to 
emphasize the theory that their appearance 
is the consequence of a definite change (muta- 
tion) in the heredity of the species; they are 
thus contrasted with the fluctuating variations 
which we discussed in an earlier chapter, 
which were supposed to represent the resultant 
212 THE STUDY OF HEREDITY 
213 
of the effects on growth and development 
of varying external circumstance and of a 
fixed and unchangeable hereditary constitu- 
tion of the germ-cells. We have already seen 
that the environment does produce an inherit- 
able effect on the constitution of living beings, 
but the demonstration of this fact belongs 
to the latest achievements of experimental 
biology, and many still refuse to accept the 
evidence offered. These still cling to the be- 
lief that the evolution of animals and plants 
has been brought about by the sudden appear- 
ance of mutations which have been preserved 
by natural selection. Therefore the research 
into the causes of the mutation assumes the 
rank of a biological problem of fundamental 
importance. 
Strange to say, most supporters of the 
evolutionary importance of mutations have 
only the phrases " chance," " accidents " 
or " internal causes " to put forward to account 
for their appearance. De Vries, a Dutch 
botanist who published a book entitled The 
Mutation Theory, raised this illogical pro- 
cedure to the dignity of a scientific hypothesis. 
He experimented with the garden plant (Enoth- 
era lamarchiana, the so-called evening primrose. 
Selecting three roots which he discovered in 214 
AN INTRODUCTION TO 
a potato-field at Hilversum in Holland, he 
removed them to his botanical gardens at 
Amsterdam, and from them he raised about 
10,000 seedlings a year. Every year a few 
" mutants " turned up, and these usually- - 
though not always-when fertilized with their 
own pollen reproduced their like. Accord- 
ing to De Vries each of these mutants dif- 
fered from the type in several characters 
affecting stem, leaves, flowers and fruit. He 
at once jumped to the conclusion that he had 
surprised a species in the act of throwing off 
fully-developed daughter species, as Minerva 
was fabled to have been born from the head 
of Jove complete in all her armour. It was 
indeed he who invented the term " mutation," 
and (Enothera lamarckiana was regarded by 
him as a species which was indulging in a 
" fit of mutation." Gates, who examined the 
nuclear structure of some of these mutants, 
was able to show that some of them had a 
chromosome fewer than the number char- 
acteristic of the nuclei of the type, whilst in 
one of them, (Enothera gig as (Fig. 28, B), 
distinguished by its greater height, stouter 
stem and larger leaves, the number of chromo- 
somes had been doubled. Gates considered 
that his discoveries supported De Vries's THE STUDY OF HEREDITY 
215 
theory and showed that the chromosomes 
had been altered owing to unknown causes- 
Fig. 28. 
A. (Enothera lamarckiana. B. (Enothera gigas-one 
of its mutants. 
possibly " accidents of cell-division "-and 
that the mutations were the outward and visible 
result of the chromosomal changes. 216 
AN INTRODUCTION TO 
Further research has thrown considerable 
doubt on this easy interpretation of the 
phenomena exhibited by (Enothera lamarcki- 
ana. CEnothera belongs to the Fuchsia order 
(Onagracea), and every genus of this order ex- 
cept the willow herb (Epilobium) is confined to 
North America. Yet when we go to North 
America we fail to find any wild representative 
of (E. lamarckiana. Therefore we are forced 
to the conclusion that this species is nothing 
more than a garden hybrid between two 
unknown wild species which was produced by 
some gardener in the eighteenth century, and 
that the production of " mutants " is a re- 
sult of the hereditary instability character- 
istic of such hybrids. Many of them may be 
mere recombinations of features of the two 
parental species; for others, such as (E. gigas, 
this explanation is inadequate, yet other 
examples of such " tetraploidy " (i. e., doubling 
the ordinary " diploid " number of chromo- 
somes) are known to occur amongst the prog- 
eny of hybrids between different species. 
These doubts are reinforced by the results of 
the experiments carried out by Davies and 
McClelland, two American botanists. These 
workers cultivated a wild American species of 
CEnothera for seven or eight years, and no THE STUDY OF HEREDITY 
217 
mutants appeared amongst the offspring of 
the wild plants. Against this it is urged 
that " mutants " similar to those of (E. 
lamarckiana have been found growing near 
masses of other wild species of CEnothera 
which bear small flowers and are self-fertilizing. 
But amongst self-fertilizing plants occasional 
crosses are nevertheless effected both by wind 
and insects, and these alleged " mutants " 
have therefore no evidential value. When, 
however, we come to consider Morgan's cul- 
tures of Drosophila, there is no evading the 
conclusion that very numerous mutants have 
" spontaneously " appeared. Allied species of 
Drosophila are known, but when these are 
crossed with D. melanogaster, either they pro- 
duce no offspring at all or the offspring that 
are produced are sterile. 
If now we examine the mutants which 
Morgan has obtained, such, for example, as 
those shown in Fig. 23, B and C, we are at 
once struck by their abnormal appearance. 
They are practically all of them cripples, and 
the vast majority are recessive to the type. 
Now we have seen that Bateson and Punnet 
explain the production of recessive strains 
by the " dropping of factors," and so if 
mutations have produced evolution we should 218 
AN INTRODUCTION TO 
expect that when a " factor " drops out a 
previous stage in evolution would be repro- 
duced. We know something about the evolu- 
tion of wings by means of evidence drawn 
from fossil insects, from the embryonic de- 
velopment of insects, and from the compara- 
tive anatomy of living insects, and we are 
confident that wings at no period of their 
history ever resembled the shrivelled vestiges 
represented in Fig. 23, C. 
If dropping of " factors " should reveal a 
previous stage in evolution, the appearance 
of new " factors " should give us a sample 
of how evolution progresses. Now the combs 
of the various breeds of domestic fowl are 
dominant to the single comb characteristic 
of the wild Jungle Fowl. We might therefore 
assume that in the production of these, new 
factors have appeared. But, again, these 
factors are abnormal in character. They 
seem to represent various stages in the loosen- 
ing of the skin of the head, and they culminate 
in the Polish fowl, which here has a huge 
vacuity in the skull. Nothing like them is 
known in any wild species-only the genus 
Gallus possesses a comb at all, and every 
species in the genus has a single comb. 
Morgan too has discovered a few dominant THE STUDY OF HEREDITY 
219 
" mutations." These, however are obviously 
abnormal; so abnormal indeed that when 
crossed with their like they are sterile. Only 
by mating them with the type has it been 
found possible to propagate them at all. 
Now the most eminent palaeontologists 
(students of fossils), embryologists (students 
of development), and systematic zoologists 
(naturalists), are all convinced that mutations 
as found by the breeder and gardener have 
played no part in evolution; they all seem 
to be due to pathological-that is, diseased- 
deviations from the type. The theory of 
Bateson and Punnet, which accounts for the 
formation of recessive strains by the " drop- 
ping " or " absence " of something, seems an 
imperfect representation of the truth. A 
better statement of the matter would seem 
to be this: a mutation is due to some patho- 
logical interference with growth; if it is com- 
paratively slight it does not affect the ap- 
pearance of the child when inherited from 
one parent only, and it is then said to be 
recessive, but if it is severe it makes its effect 
felt even when it comes in only from one side 
of the house, and it is then said to be dominant. 
Is not there a common type of cause, 
physiological in its nature, for all these varied 220 
AN INTRODUCTION TO 
mutations? The only attempt to answer 
this question has been made by a German zo- 
ologist called Tornier, whose work has escaped 
notice almost as completely as did that of 
Mendel when he first published his results. It 
is because we think that Tornier is right in 
Fig. 29. 
Carassius auratus, the wild progenitor of the goldfish, 
viewed from the side. A., Anal fin. C., caudal fin (i. e., 
tail fin.) D., dorsal fin. P.c., pectoral fin. P.v., pelvic 
fin. 
principle, although doubtless his theory will 
require much modification in detail, that we 
devote the remainder of this chapter to a short 
account of his views. Tornier began by 
studying the various races of gold-fish which 
are cultivated by purveyors of ornamental fish THE STUDY OF HEREDITY 
221 
both here and in America. All these races 
have been produced by the domestication of 
a species of carp, Car as sins auratus, which is 
found wild in the rivers of China (Fig. 29). 
This is a very ordinary fish with well-developed 
anal, dorsal and caudal (tail) fins, in which, 
as in many other primitive types of fish, the 
pelvic fins, which correspond to the legs of 
land animals, are placed behind the pectoral 
fins, which correspond to the arms. 
All the domesticated races of this fish come 
from China; for although some have been 
imported from Japan, the Japanese them- 
selves admit that they have derived all their 
gold-fish from China. According to Chi- 
nese records the gold-fish was " discovered " 
in 1200, by which is obviously meant that 
it was then that the characteristic " gold " 
colour turned up amongst the wild grey 
Carassius that were being kept in tanks. 
Now the Chinese even to-day keep their gold- 
fish under the most insanitary conditions; 
in winter these are kept in earthenware pots 
ranged in hundreds on shelves in dark, ill- 
ventilated huts; in summer they are trans- 
ferred to small, dirty, out-of-door tanks which 
are overgrown with weeds. Under these cir- 
cumstances much of the spawn that is laid 222 
AN INTRODUCTION TO 
dies, and amongst the eggs that survive many 
develop into abnormal forms, and by selection 
amongst these the various " fancy " races 
have been produced. These races exhibit the 
most extraordinary deviations from the nor- 
mal type of structure; three of them are 
Fig. 30. 
The Ryukin, or Veil-fish, seen from the side. A., anal 
fin. C., caudal fin (i. e., tail fin.) D., dorsal fin. P.O., 
pectoral fin. P.c., pelvic fin. 
represented in Figs. 30, 31 and 32. In Fig. 
30 a breed is represented which is called by 
the Japanese the Ryukin, and by the Ameri- 
cans the Veil-fish; the body as compared 
with the wild Carassius is very much shorter 
and stouter; the anal fin is very long, and 
the caudal fin is enormously lengthened and 
flexible, so as to be compared to a lady's THE STUDY OF HEREDITY 
223 
veil. The Lion-head, called by the Japanese 
the Oranda shishigashira, is depicted in Fig. 
31. In this fish the head is covered with loose 
folds of skin which are puffed out like bubbles. 
These folds impede the action of the gill-cover 
Fig. 31. 
The Lion-head, or Oranda shishigashira, seen from 
the side. A., anal fin. C., caudal fin (i. e., tail fin.) 
D., dorsal fin. P.c., pectoral fin. P.v., pelvic fin. 
or operculum and the fish has difficulty in 
breathing. The body is short and swollen. 
The dorsal fin is completely absent, but the 
caudal fin is doubled, there being two tail 
fins side by side. Finally, in Fig. 32 the ex- 
traordinary Telescope-fish, or Deme-Ranchu 224 
AN INTRODUCTION TO 
is shown. In this fish the eyes protrude as 
hollow tubes from the sides of the head; the 
caudal fin is doubled and the dorsal fin is 
absent. 
These three breeds are only a selection 
out of a large number of bizarre abnormal 
forms. Now according to Tornier all these 
Fig. 32. 
The Telescope-fish, or Deme-Ranchu, seen from the 
side. A., anal fin. C., caudal fin (i. e., tail fin.) D., 
dorsal fin. P.c., pectoral fin. P.v., pelvic fin. 
mutations are due to the weakening effect on 
the vigour of the embryo of the abstraction of 
oxygen from the water during the first few days 
of its life. He regards each egg as endowed 
W'ith a certain amount of vital energy (the elan 
vital of Bergson), which shows itself both in 
vigorous growth and in vigorous movement. 
One of the first changes which initiates 
growth is an absorption of water by and THE STUDY OF HEREDITY 
225 
through the protoplasm and a consequent 
swelling. Davenport and Parker have shown 
that nearly all the increase in weight which 
a tadpole acquires during the first period of 
Fig. 33. 
A. Head of normal carp. B. Head of bull-dog carp. 
C. Round-head carp. D. Round-head carp, eth., mes- 
ethmoid bone; it is this bone which in B, C, and D is 
buckled up under the longitudinal pressure exercised by 
the swollen yolk in front of the head. 
its growth is accounted for by the imbibed 
water. But this intake of water is regulated 
by the activity of the growing cells. When 
the germ is weakened an excess of water 
is absorbed. Not only are the more inactive 
parts of the egg (loosely called the yolk) 226 
AN INTRODUCTION TO 
unduly swollen, but all the natural cavitieg 
of the growing embryo, the mouth-cavity, 
the gill-cavity and the body-cavity, become 
distended with water. These swelling cavities 
and the swelling yolk press on and impede 
the growth of the embryo. In Fig. 33 four 
diagrams (reproduced from Tornier's paper) 
illustrate his idea, how the swollen yolk in 
front of the embryonic head interferes with 
the growth of the head. In A a diagram of 
the normal head of a carp is given. In B is 
the outline of a " bull-dog carp "; in this 
case the yolk pressed upon the upper part 
of the head, leaving the lower jaw unaffected. 
C and D represent less and more extreme 
pressure respectively, in which the whole head 
is involved and " round-heads " have been 
produced. In the last diagram the enormous 
distension of the mouth may be noticed. If 
we apply this principle to the explanation of 
the three breeds of gold-fish which we have 
described, we may infer that the longitudinal 
pressure exerted by the swelling yolk on the 
backbone has prevented it growing in length 
and so the whole animal has been shortened. 
The swollen body-cavity causes a protuberant 
abdomen. The fins are folds of skin, and when, 
as in the Veil-fish, the skin area from which THE STUDY OF HEREDITY 
227 
the tail fin arises is stretched, this fin is unduly 
elongated, for stretching promotes continued 
growth. At the same time the sluggishness of 
the weakened embryo causes the tail move- 
ments to be feeble. These tail movements in 
the normal fish tend to stretch the membrane 
enclosing the embryo and procure more room. 
In the Veil-fish the elongated tail fin is bent 
against the unyielding membrane and the 
drooping veil form produced. On the other 
hand, where the area of skin from which a fin 
grows is compressed, the fin remains rudi- 
mentary or vanishes altogether, as the dorsal 
fin does in the Lion-head and in the Telescope- 
fish. Sometimes the yolk intervenes between 
paired rudiments which normally coalesce to 
form a single organ. Thus in the Lion-head 
and the Telescope-fish swollen yolk separates 
right and left halves of the developing tail fin, 
and so two tail fins are produced. In the Lion- 
head, again, the bones of the head undergo 
great pressure and are very much shortened, 
whilst the skin retains its normal length and 
so great bulbous folds are produced, just like 
the folds of skin on the face of a bull-pup. 
Finally, in the Telescope-fish the growing 
eyeballs absorb a large amount of water and 
become swollen to preposterous dimensions, 228 
AN INTRODUCTION TO 
whilst the retina or sensitive layer becomes 
enormously stretched and thin and is wanting 
over parts of the internal surface of the eyeball. 
Not the least interesting part of Tornier's 
theory is the way in which he accounts for 
the change of colour which the gold-fish has 
undergone from the grey of the wild ancestor 
to the familiar orange of the cultivated fish. 
When the " yolk " or food-materials embedded 
in the protoplasm of the egg imbibe water, 
they are partly coagulated and rendered 
indigestible by the growing embryo. This 
useless material is sooner or later expelled. 
As a consequence the embryo in its later stages 
of growth suffers from a deficiency of nourish- 
ment and therefore draws on its pigment cells, 
which act physiologically as reserves of food, 
and so all grades of diminished colour from 
a normal orange gold-fish to an almost com- 
plete albino (silver-fish) are produced. 
These abnormalities produced by the weaken- 
ing of the germ are hereditary. These curious 
monstrosities selected from the tanks and 
pots of China have been brought to Europe 
and to America and used to propagate their 
like. Moreover, when crossed with other 
strains they exhibit Mendelian inheritance. 
Tornier insists that what is transmitted to THE STUDY OF HEREDITY 
229 
posterity is not a " gene " or " factor " for 
an enormous fin or a doubled tail, a bulbous 
head or a telescopic eye, but a greater or less 
degree of germ-Weakening, and the structural 
consequences which result from it. He makes 
the additional hypothesis that this weakening 
thus hereditarily transmitted may affect 
different parts of the embryo and that this 
localization of the weakness is also inherited. 
The races of gold-fish when bred yield, as 
we have seen, progeny like themselves, but 
this is accompanied by a greater or less 
amount of reversion towards a more normal 
type. Thus Veil-fish yield a certain propor- 
tion of fish with fairly moderate-sized tail 
fins. This reversion is interpreted by Tornier 
as a gradual healing or recovery of the germ, 
so that as it gains in vigour the abnormal 
structural consequences of its weakening tend 
to disappear. 
In order to establish his theory Tornier 
experimented on the eggs of those large 
newts called axolotls, which ordinarily retain 
their gills throughout life. When the eggs 
were fertilized he put them into strong 
solutions of sugar, which tended to remove 
the dissolved oxygen from the water. Too 
long immersion in these solutions would Fig. 34. 
Diagrams illustrating the development of a normal and 
of a weakened axolotl egg. A, B, C, D represent stages 
in the growth of the normal egg. A', B', C', D' corre- 
sponding stages in the growth of the weakened egg. a., 
anus; bri, br2, br3 the rudiments of the three gills; emb., 
the embryonic rudiment; f.l., rudiment of the fore-limb; 
hy., the hyoid; m., the mouth; mn., rudiment of the 
lower jaw; y., the yolk. 
230 THE STUDY OF HEREDITY 
231 
have killed the eggs, but when removed in 
time and brought into clean fresh water they 
survived and produced weakened embryos. 
In the axolotl, as in the gold-fish, the embryo 
grows on the upper surface of a spherical 
egg, the rest of which is packed with yolk. 
Fig. 34 consists of a series of diagrams show- 
ing side by side the growth of a normal and of 
a weakened embryo. In the latter we see 
all the features familiar to us from the study 
of the races of the gold-fish. The longitudinal 
axis is shortened and eventually bent, the 
growth of the tail and of the tail fin is im- 
peded and the mouth cavity is swollen and the 
mouth gaping Fig. 34, D'). 
Some of the less damaged eggs survived so 
as to hatch and develop into young axolotls. 
These had enormously protuberant bellies and 
were sluggish in their movements. Fig. 35 
represents a dissection of one of these ab- 
normal individuals, contrasted with a corre- 
sponding dissection of a normal one. 
We see at once that the swollen abdomen of 
the weakened form is produced as a result 
of the enormous distention of the body-cavity 
with water. As a consequence nearly all the 
internal organs, which are contained in this 
cavity (bladder, lungs, liver, etc.), have been 232 
AN INTRODUCTION TO 
compressed and impeded in their growth 
and are mere shadows of their normal selves. 
Fig. 35. 
A. Side-dissection of a normal axolotl. B. Corre- 
sponding dissection of an abnormally weakened axolotl, 
al., alimentary canal; bl., bladder; h., heart; liv., liver; 
pul., lung. 
Tornier was able to produce varying shades 
of red and yellow in toad embryos by two THE STUDY OF HEREDITY 
233 
methods, viz., (a) by feeding them on innu- 
tritions food, (6) by stabbing the yolky part 
of the egg with a fine needle so as to allow 
part of the yolk to escape. The reduction in 
pigment was in some cases so great as almost 
to amount to albinism. 
Tornier's theory would have received a com- 
plete proof if he had been able to rear these 
weakened axolotls to maturity, then to have 
induced them to breed and to have shown 
that their abnormalities were transmitted 
to their posterity. But the great extent of 
time required for this experiment prevented 
its being carried out; an axolotl requires 
four or five years to grow to breeding size. 
Nevertheless, as we have seen, there is a great 
body of circumstantial evidence in favour of 
Tornier. The conditions under which the 
abnormal axolotls were produced were similar 
to those under which abnormal gold-fish 
appear, and these last certainly can transmit 
their abnormalities to their young. A pupil 
of Tornier, Milewski, has, however, produced 
further evidence in favour of Tornier's views. 
He mated together two Veil-fish in which 
the veil character was but moderately de- 
veloped. These produced a brood of young, 
10 per cent, of which had well-developed 234 
AN INTRODUCTION TO 
veils, whilst 90 per cent, had moderately 
developed fins. The same pair were induced 
to mate again in water which had been ren- 
dered poor in oxygen by being filled with 
water plants and kept in the dark. A second 
brood was produced; the eggs after two days' 
sojourn in darkness were transferred to clean, 
fresh, aerated water. Of this brood 90 per 
cent, had well-developed veils and only 10 
per cent, showed signs of reversion to type! 
It would, we imagine, puzzle a follower of the 
" factorial hypothesis " of Morgan to explain 
why the immersion of an egg in badly aerated 
water should increase the proportion of the 
progeny inheriting a " factor " ninefold. 
If then the hypothesis of Tornier be accepted 
-and we think that the cumulative evidence 
in its favour is irresistible-it throws a brilliant 
light on the nature and origin of mutations. 
As they are all pathological in character, 
the result of weakened germ energy, they can 
have played no part in the process of evolu- 
tion, for in competition with the type they 
would be inevitably weeded out by natural se- 
lection, and to explain evolution we are thus 
driven back on the inheritance of acquired 
habits and of the structures that result from 
these. When we consider how Morgan reared THE STUDY OF HEREDITY 
235 
his flies, confined in glass tubes with bits of 
rotten banana, we can form a pretty confident 
Fig. 36. 
A. The skull of the wild boar. B. The skull of a 
Yorkshire pig. lac., lachrymal bone-note its com- 
pression in B. 236 
AN INTRODUCTION TO 
guess as to the causes of the production of the 
wretched pathological mutants with which he 
has worked. As Tornier himself has painted 
out, if we contrast the vigorous, healthy life 
of a wild boar with the sluggish, insanitary 
existence of a farmyard pig, we see ample 
reason for the appearance of a germ-weaken- 
ing in the latter; and if we compare the skulls 
of the two animals as shown in Fig. 36, we 
see in the skull of the domestic pig the same 
fore-and-aft shortening which we noticed in 
the skull of the gold-fish contrasted with 
that of the carp. It is true that the mam- 
malian egg contains no yolk, because the 
embryo derives all its nourishment from the 
womb, but the head of the embryo is immersed 
in a sac (the amniotic sac), tensely filled 
with fluid, and abnormally high amniotic 
pressure has already been indicated by our lead- 
ing orthopaedist, Sir Robert Jones, as the cause 
of many congenital deformities. 
As to the difference in vigour between the 
normal and weakened types, anyone who has 
ever handled wild mice and tame white mice 
will require no further demonstration of the 
truth of Tornier's views in this respect. It 
was the great mistake of Darwin in that 
classic book, The Origin of Species, that he THE STUDY OF HEREDITY 
237 
imagined that he was discovering in the breeds 
of domestic animals analogies to the produc- 
tion of the natural races of animals in the field. 
It may be asked how Tornier's principle 
would apply to plants. It is common knowl- 
edge that when gardeners want to cultivate 
a plant they introduce it into rich garden soil, 
and that after having grown in soil of this 
kind for some years it " breaks," that is, it 
begins to " throw " varieties. The life of 
plants is much slower in its reactions than 
that of animals, and we may plausibly compare 
the slow, suffocating effects of over-manured 
soil on the roots of plants, to the results of 
scanty oxygen on the developing eggs of 
animals. 
The really curious thing is that when a 
weakened germ-cell is crossed with a healthy 
normal one, the hybrid produces germ-cells 
some of which are normal and some of which 
are weakened. Further research may pos- 
sibly show that individual chromosomes be- 
have almost like individual germ-cells, and 
may undergo weakening in different degrees; 
that is a question for the future, but 
the " coupling " which induced Morgan to 
postulate the existence of separate " factors " 
for each type of deformity and their arrange- 238 
AN INTRODUCTION TO 
ment along the chromosome is much more 
plausibly explained by different degrees of 
damage exhibited by various organs according 
to the intensity of the weakening. When 
this is slight, few organs show the results of 
it; when it is intense, more organs are affected 
and then there is marked coupling, but the 
whole of the phenomena of hybridization will 
have to be reviewed from the standpoint of 
Tornier's views when these become widely 
known and gain general acceptance. 
The weakness of the " factorial hypothesis " 
is its extreme plasticity. When every ob- 
served change in the appearance of an ani- 
mal which can be transmitted to its off- 
spring is ascribed to a new hypothetical fac- 
tor, and this ascription is regarded as an ex- 
planation, then there is no limit to the number 
of factors which may be postulated, and the 
explanation, as Goldschmidt has wisely re- 
marked, becomes merely a formal one. Again, 
the test of Mendelian inheritance has been not 
dominance in the first filial generation- 
this under the name of prepotency was known 
long before the days of Mendel-but segrega- 
tion of the recessive in the second filial genera- 
tion. If this segregation is complete it is 
regarded as proof of the Mendelian character THE STUDY OF HEREDITY 
239 
of the mutation; if it is not complete, then 
a " modifying factor " is assumed which pre- 
vents the complete purity of the recessive. In 
a word, however the facts may turn out, fresh 
factors are invented to squeeze them into 
conformity with the factorial hypothesis, and 
no crucial experiment can be devised, and 
when a theory reaches this state it has no 
further interest for the seeker after truth. CHAPTER IX 
THE BEARING OF THE LAWS OF HEREDITY 
ON HUMAN AFFAIRS 
We have now come to the conclusion that 
the assumption of new habits, that is, the 
adaptation of the organism to new situations, 
has been the guiding force which has brought 
about evolution, and that when animals take 
on new habits their descendants tend to assume 
these habits more readily and on slighter 
stimulation than their parents. To quote 
Macdougall in the Outline of Psychology, 
" The seal did not take to the water because 
his legs were flipper-like and his body fish- 
shaped ; he acquired these peculiarities of 
structure in consequence of his food-seeking 
instinct having become specialized for the 
pursuit of fish. The evolution of the animal 
world may properly be conceived as primarily 
and essentially the differentiation of instinc- 
tive tendencies (that is, habits) from some 
primordial capacity to strive." In a word, 
240 THE STUDY OF HEREDITY 
241 
change of habit preceded change of structure 
and was the cause of the latter change. The 
various monstrosities which fill our cattle-pens 
and adorn our gardens are the result, as we 
have seen, of germ-weakening with its conse- 
quent loosening of the ' ' regulation ' ' whereby 
in a normal animal one organ is kept in 
functional relation to another. 
Now humanity is divided into a number of 
natural races, each distinguished from the 
other by physical and mental characters; they 
differ indeed as much from one another as 
do many natural species, and they are only 
regarded as races because it is assumed that 
they are all perfectly fertile when crossed with 
each other, though this assumption-so far at 
least as the most widely divergent races are 
concerned-has not been adequately tested. 
These races have acquired their characters 
as reactions to their different environments, 
and their development out of a common stock 
of primitive men has occupied a period so 
long that the entire historical period appears 
as a brief episode in comparison with it. In 
the Northern hemisphere at least two glacial 
ages have passed since the evolution of these 
races began. It follows that the marks which 
distinguish one race from another are the 242 
AN INTRODUCTION TO 
outward and visible signs of differences in 
habits long established and deeply engrained. 
Let us, for instance, contrast the Negroes and 
the Caribs or South American Indians. Both 
these races inhabit the tropics--the first live 
in the Equatorial parts of Africa and Malaya, 
the second in the Amazonian region of South 
America. The Negro is a thoroughly tropical 
animal; his dark skin shields him from the 
deleterious effects of the sun's rays and his 
wide nostrils permit of a large surge of air 
into and out of his lungs, and this surge plays 
an important part in ventilation and the 
getting rid of superfluous heat. He thrives 
in heat and in fact luxuriates in it, and com- 
petes eagerly for positions in the engine- 
room of the steamers on the equatorial lakes 
of Africa. The Carib, although he inhabits 
the tropical forests of the Amazon, dislikes 
and shrinks from the heat; according to Bate 
he perspires profusely in the hottest time of 
the year and seeks to shelter himself in the 
deepest recesses of the forest. This is because 
the Carib is an offshoot of the great yellow 
Mongolian race which is primarily adapted to 
life on the cold Northern Steppes of Asia, 
and has only in the last 3000 years invaded 
South America, and these people are not yet THE STUDY OF HEREDITY 
243 
thoroughly adapted to their new environ- 
ment. 
The population of the British islands is 
composed of a mixture of three races: (a) the 
Nordic, tall, long-headed, fair-skinned and 
blue-eyed, who were evolved round the shores 
of the North Sea; (b) the Alpine or Central 
Europeans, short, thick-necked, stocky people 
with broad heads, brown eyes, straight hair 
and full beards who originated in the steppes 
of Turkestan and made their way westward 
over the European plain; and (c) the Iberian 
or Mediterranean race, short, long-headed, 
rather swarthy people with dark wavy hair 
and dark eyes; this race, as its name implies, 
was the aboriginal population inhabiting the 
shores of the Mediterranean Sea. The blood 
of the three races is hopelessly intermixed, but 
the Nordic is found nearly pure in the north- 
east of Scotland, whilst the Mediterranean 
exists almost unmixed in the south of Wales 
and the south-west of Ireland. 
Each of these races has its inborn psychic 
qualities. The Nordic race, who were origi- 
nally fishermen and mariners, learnt in their 
struggle with the bleak climate of their old 
home the virtues of indomitable courage, bold 
adventure and justice between man and man. 244 
AN INTRODUCTION TO 
Is not the double meaning of the word " fair " 
a tribute to their virtue? This race, as 
Normans and Danes, established and settled 
the kingdom of England and by their oversea 
adventures build up the British Empire. They 
form the bulk of its aristocracy, and are the 
backbone of the middle-class, the sons of 
whom still show the racial enterprise in their 
willingness to seek their fortunes overseas. 
The Alpine race were the first workers in 
bronze and they still form a large portion of 
the artisan class. There is little doubt that 
the traditional superiority of British work- 
manship is due to the inherited skill of this 
type of people. The progress of the Alpine 
race across England can be traced by their 
graves, which contain broad skulls and bronze 
tools. The Mediterranean race form the 
aboriginal stratum in the population. They 
are characterized by a mercurial temperament, 
prone to quarrel and quick to take revenge, 
with musical and poetical gifts, but without 
the courage and organizing power of the 
Nordic race or the plodding industry of the 
Alpine. Their natural organization, as Irish 
legends show, is a division into warring and 
treacherous clans. When they drift into the 
towns they tend to form the " submerged THE STUDY OF HEREDITY 
245 
tenth," i. e., the inhabitants of the slum. 
From this sketch it will be seen that the 
maintenance of the kingdom and of the 
Empire is largely dependent on the uphold- 
ing of a due proportion of the Nordic race 
amongst our population, yet recent legislation 
has been entirely detrimental to this race. 
All attempts to favour the slum population 
by encouraging their habit of reckless repro- 
duction in throwing the support of their 
children on the State places a heavier burden 
on the shoulders of the Nordic race, who form 
the bulk of the taxpayers. The prospect is 
such as to make a patriotic Englishman 
shudder. The history of civilization since its 
beginnings in Egypt is now pretty well known, 
and one civilization after another has perished 
through the dying out of the ruling organizing 
race. 
The tacit assumption of philanthropists 
all through the eighteenth and nineteenth cen- 
turies was that the differences between men 
were due to circumstances and could be 
abolished by education. Whilst in the last 
resort racial differences are due to circum- 
stances operating through thousands of years, 
yet the idea that education and environ- 
ment acting through one or two generations 246 
AN INTRODUCTION TO 
can cancel the work of thousands of genera- 
tions is singularly futile. The inferior races 
can be trained in civilized habits and kept in 
them so long as the superior race is in control, 
but left to themselves they revert to the stage 
in development appropriate to their inborn 
psychic equipment. Of this fact the republics 
of Hayti, Liberia and Mexico furnish ample 
proof. 
The American Republic was founded 150 
years ago, mainly by emigrants of Nordic 
stock. All the Nordic characteristics of in- 
trepid enterprise, high ideals of justice, etc. 
were exhibited in marked degree by the 
founders and by their immediate successors. 
But under the influence of the fallacious 
a priori eighteenth-century doctrine of the 
quality of men and the essential similarity 
of the human races, the doors of immigration 
were opened wide and North America became 
filled with a vast crowd of Mediterranean 
peoples; the inborn tendencies of these people 
are making themselves felt more and more as 
time goes on. Whilst the Nordic element, with 
characteristic prudence, limit their families to 
numbers which they can support, the Mediter- 
ranean races breed recklessly and thus tend 
to form a larger and larger section of the THE STUDY OF HEREDITY 
247 
population; many thoughtful American citi- 
zens view the continuance of this process with 
grave apprehension. 
But in addition to racial differences wre have 
to consider the effects of germ-weakening and 
the consequent production of " mutations." 
This germ-weakening seems to arise most 
readily in large towns under conditions which 
favour overcrowding and unhealthy conditions 
during conception. Several so-called " sex- 
linked " mutations have already been detected 
in the human race; amongst them are colour- 
blindness, night-blindness and haemophilia 
(inability of the blood to form a clot when it 
issues from a wound, and hence the tendency 
of the wounded person to bleed to death) ; 
these, however, are not widespread and it 
cannot be maintained that they do any 
great amount of social damage; but there 
is a mutation known as " mental defect " 
which is of colossal social importance. This 
mutation consists in the inability of the mind 
to expand and develop its powers as the body 
grows. It exists in a great variety of grades 
of intensity, and on being crossed with the 
type, behaves as a recessive. The various 
grades are distinguished from one another by 
the age at which a normal child acquires an 248 
AN INTRODUCTION TO 
intelligence equal to that of the defective; 
thus we have defectives of " mental ages " of 
one, two, three or more years. At fifteen 
years the normal child develops its full mental 
powers, though naturally not its complete 
mental equipment. Now the point to note 
is that defectives of the mental ages of one, 
two and three years are unable to dress and 
feed themselves and must be cared for as 
babies-these are called idiots; those of four, 
five, six and seven years can be taught to 
dress and feed themselves and to perform 
simple tasks under direction-these must be 
treated as school children, and are termed 
imbeciles. Both these classes are utterly 
unable to maintain themselves by their own 
endeavours in competition with their fellows, 
and so they must be segregated in institutions 
and maintained at the public expense-but 
they are prevented from breeding. 
Defectives of the mental ages eight, nine, 
ten, are termed morons by American authors; 
these are just able to support themselves 
in the lowest paid and least skilled occupa- 
tions, and so they form a large element 
in our slum population. For some reason 
which is still obscure, they are exceedingly 
prolific, and as they breed recklessly their THE STUDY OF HEREDITY 
249 
numbers tend continually to increase. In the 
days of our grandfathers, and indeed of our 
fathers, their offspring, owing to inherent 
weakness and want of care, died like flies, and 
hence they did not form a serious social 
menace. But nowadays, with the growth of a 
maudlin, unthinking sentimentality, strenu- 
ous efforts are made not only to keep all their 
offspring alive but to allow them to breed at 
the expense of the more competent members 
of the community. The consequences of such 
a policy can easily be foreseen; in America 
it has already been shown that a very large 
proportion of the criminals, prostitutes and 
habitual drunkards are mental defectives, and 
similar researches in England, if authorities 
were wise enough to permit them, would doubt- 
less yield the same result. It is not that there 
is a criminal type of mind, but moral develop- 
ment, which is essentially self-control, goes 
hand-in-hand with mental development, and 
defective self-control coupled with inability to 
perceive the consequences of impulsive actions 
makes the criminal. 
The emotion of sympathy, developed, as 
Macdougall has shown, out of the parental 
love of the young, is doubtless the noblest 
human sentiment, and it is the principal 250 
THE STUDY OF HEREDITY 
cement which holds society together, but like 
all other human sentiments it must be gov- 
erned by reason. The study of biology 
teaches unequivocally that all progress in the 
animal world has been brought about by 
natural selection, that is, by the weeding out 
of those individuals which refused to adapt 
themselves to altered circumstances. 
If social reformers seek to interfere with 
natural selection by keeping the weak and 
incompetent alive, they must provide some 
device to secure breeding by the fit only. In 
the last resort this must involve sterilization 
of the reckless reproducers, but much may be 
done by the spread of the knowledge of the 
means of birth control. 
The task of the man of science is different 
from that of the politician, but the biologist 
with his wide survey over all the various 
forms of life, extending over periods of 
millions of years, can see the trend of develop- 
ments much more clearly than the politician, 
whose range of vision is confined to human 
affairs for a period of a few decades. It is 
in the hope that the consequences foreseen by 
the biologist may be averted by suitable action 
before they are indelibly engraved in history 
that this book has been written. BIBLIOGRAPHY 
(This list includes only the best standard works on each 
division of the subject.) 
Agar, W. E.: Cytology with Special Reference to the 
Metazoan Nucleus. London: Macmillan & Co., 1920. 
Bateson, W.: Mendel's Principles of Heredity. Cam- 
bridge Univ. Press, 1909. 
Darwin, Charles: The Origin of Species by Means of 
Natural Selection (6th Edition, 1872). London: 
John Murray. 
De Vries (translated by J. B. Farmer and A. D. Din- 
bishue) : The Mutation Theory. Chicago: Open 
Court Publishing Co., 1909. 
Doncaster, L.: The Determination of Sex. Cambridge 
Univ. Press, 1914. 
Galton, Francis: Natural Inheritance. London: Mac- 
millan & Co., 1889. 
Goddard, H. H.: Feeble-mindedness: its Causes and 
Consequences. New York: The Macmillan Co., 1914. 
Goldschmidt, R. (translated by W. Dakin) : The 
Mechanism and Physiology of Sex Determination. 
London: Methuen & Co., 1923. 
Kidd, W.: Initiative in Evolution. London: Witherby, 
1920. 
Morgan, T. H., Sturtevant, A. H., Muller, H. J., and 
Bridges, C. B.: The Mechanism of Mendelian 
Heredity. New York: Henry Holt and Co., 1915. 
Rignano, Eugenio: Sur la transmissibility des caractbres 
acquises. Paris: Alcan Press, 1906. 
Tredgold, A. F.: Mental Deficiency. London: Bailli&re, 
Tindall & Cox, 1920. 
Wallace, Alfred Russel: Darwinism: an Exposition of 
the Theory of Natural Selection. London: Mac- 
millan & Co., 1890. 
Weismann, A. (translated by W. K. Parker and H. 
Romfeldt) : The Germ-plasm: A Theory of Heredity. 
Scott, 1891. 
251 INDEX 
Ahraxis grossulariata, 170- 
172 195 
lacticolor, 170-172 
Acquired characters, inherit- 
ance of, 81 ff. 
Agar, experiments of, 131 ff. 
Agouti, definition of, 161 
Allelomorph, definition of, 
155 
Allelomorphism, multiple, 
definition of, 162 
Alpine (race), 243 ff. 
Alytes, development of, 
102 ff. 
Amphioxus, 195 
Anaphase, 40, 42 
Antedon rosacea, 56 
Antheridium, 22, 29, 31, 32 
A.ntherozoid, 20 
Antipodal cell, 34, 35 
Archegonium, 22, 29, 31-34 
Ascidians, 63 
Aster, 25, 39-41, 49 
Axolotl, experiments with, 
229 ff. 
Back-cross, definition of, 
148 
Bateson, experiments with 
fowls, 135 ff. 
• with sweet-peas, 
162 ff. 
Bergson, theory of, 224 
Blffen, experiments with 
wheat, 166 ff. 
Blastomeres, definition of, 53 
Blastula, definition of, 70 
Boar, skull of, 235 
Boneilia, 187, 188, 193, 209 
Boverl, 48 
Breed, definition of, 140 
Brood-pouch, 133 
Bud, definition of, 11 
Carassius auratus, 220-222 
Carcinus manas, 126 
Carib, 242 
Carp, abnormalities of, 225, 
226 
Bulldog, 226 
Roundhead, 226 
Cell, definition of, 13, 15 
Cell-wall, definition of, 16, 
17 
Centrosome, definition of, 25, 
26, 39-42, 48 
Chromatin, definition of, 38, 
42, 44, 48, 50, 51 
Chromogen, definition of, 160 
Chromosomes, 39-43, 45, 46, 
48-52, 76, 77 
Class, definition of, 139 
Coefficient of correlation, 122 
of regression, 120 
Cohn, 15 
Colony, 11 
Comb (of fowls), "pea," 
156, 157 
" rose," 156, 157 
"single," 156, 157 
" walnut," 156, 157 
Comb-bearers, 63, 64 
Conidia, 18-20, 28 
Conjugation, 135 
Copulatory styles, 195 
Coupling, definition of, 163 
Crabs, 21 
Crepidula, 49 
Crewe, observations on sex 
of fowls, 195 ff. 
Crinoidea, 56 
Crossing over, definition of, 
185 
Ctenophora, 63-67 
Curve of error, 115 
Cycad, 33 
Cynthia pastita, 69 
Daphnia, 132 
Darwin, theory of, 79, 139 
Davenport, observations of, 
127 
252 INDEX 
253 
Davenport, observations on 
growth of tadpoles, 225 
Davies, experiments of, 216 
Deviation, definition of, 116 
De Vries, experiments with 
CEnothera, 213 ff. 
theory of, 213 
Diploid, definition of, 39, 45, 
48, 50, 52 
Direct division, 52 
Dominance, definition of, 154 
Dominant character, 143 
Doncaster, experiments with 
currant moth, 170 
Drlesch, researches of, 63 ff. 
Drosophila melanogaster, 173, 
174, 177, 212, 217 
Intersexes of, 199 
Duckweed, 29 
Durkhem, experiments of, 107 
Echinodermata, 63 
Echlnopluteus, 63 
Ectoderm, 53 
Egg, 20, 22 
of frog, 73 
of newt, 73 
Egg-cell, 22, 30, 34, 52 
Embryo, 50 
definition of, 54 
Embryo-sac, 32-35 
Endoderm, 53 
Endosperm, 33-36, 52 
of egg, 23 
Ephippium, 133 
Eplloblum, 216 
Epistatic, 162 
Equational division. 46 
Equatorial plate, 41, 43, 45 
Euschistus servus, 204, 205, 
209 
varlolarius, 204-207, 209 
Evening primrose, see CEno- 
thera 
Ex-conjugates, 135 
Experimental embryology, 63 
Factor, definition of, 158 
Family, definition of. 139 
Farmer, Professor, 52 
Fasclolus, 130 
Feather-star, 56 
Female, definition of, 17 
Fern, 27 
Fertilization, 21, 52 
Fibro-vascular bundle, 28 
Filial generation, first, 142 
second, 143 
Fission, 11 
Flatfish, asymmetry of, 55 
Flowering plant, reproduc- 
tion of, 34, 35 
Fluctuating variations, 112 
Fluctuations, 131 
Foot, of bivalves, 56 
Of oyster, 56 
Foote (Miss), experiments on 
Euschistus, 203 fit. 
Formation of layers, defini- 
tion of, 53 
Formative stimuli, 75 
Fowls, Andalusian, 155 
Hamburg, 157 
Jungle, 157, 218 
Leghorn, 157 
Polish, 218 
combs of, various, 156, 
157 
Frog, mating of, 199 ff. 
Fuchsia, 216 
Fungi, see Moulds 
Gallus (see Fowl), 218 
Galton, Sir Francis, experi- 
ments of, 113 ff. 
Gametes, definition of, 18, 
19, 22 
Gastula, definition of, 71 
Gates, Ruggles, observations 
on (Enothera Lamarckiana, 
214 ff. 
Gel, definition of, 26 
Genes, definition of, 186 
Genus, definition of, 139 
Germ-cell, 12, 18, 37, 42 
Germinal layers, 53 
spot, 38 
vesicle, 38 
Germ-weakening, definition 
of, 229 
Gipsy moth, see Lymantria 
dispar 
Goldschmidt,experiments with 
Gipsy moth, 196 ff. 
Goodrich, observations on 
Amphioxus, 195 
Grew, 13 
Haeckel, 81 254 
INDEX 
Haploid, definition of, 44, 48 
Head (of spermatozoon), 20 
Hertwig, 47 
experiments with Rana 
esculenta, 200 ff. 
Heterozygote, definition of, 
152 
Homozygote, definition of, 
152 
Hooke, 13 
Hormones, 109 
Hybrid, definition of, 140 
Hylodes, tadpole of, 60 
Hypostatic, 162 
Iberian (race), 243 ff. 
Idiot, definition of, 248 
Imbecile, definition of, 248 
Imperfect dominance, defini- 
tion of, 155 
Inaschus Scorpio, 189, 190, 
209 
Indusium, 27, 28 
Intersexes, definition of, 197 
Jennings, experiments of, 
135 
Johannsen, experiments with 
kidney beans, 130 ff. 
Kammerer, experiments of, 
83 ff. 
Karyokinesis. See Mitosis. 
Kidney bean, inheritance of 
weight, 130 
Lamarck, theory of, 78 ff. 
Larva, definition of, 54 
of flatfish, 55 
Lathyrus odorata, 162 
Lens (of eye), 75 
Lily-encoinite, 56 
Lion-head (goldfish variety 
" Oranda shishlgashira ") 
223, 227 
Lithium, action of, 72 
Lobster, male germ-cells of, 
21 
Lotsy, theory of, 164 
Lymantria dispar (Gipsy 
moth) experiment on 
crossing races of, 196 
Macrospore, 30, 32-35, 47 
Macrosporangium, 30, 32, 33 
Maize, seed of, 36 
Male, definition of, 17 
Malpighi, 13 
Maturation division, 43, 44 
first, 44, 49 
second, 46, 47, 49 
McClelland, experiments with 
CEnothera, 217 
McDougall, theory of, 240 
Meal-worm, 87 
Mediterranean (race) 243 ff. 
Meiosls, 43, 45 
Meiotic division, 45, 46, 52 
Mendel, experiments with 
peas, 141 ff. 
theory of, 79 [247 
Mental defect, inheritance of, 
Mesenehyme, 70, 71 
Mesoderm, 53 
Metaphase, 40, 42, 43 
Microspore, 30, 32, 35. 36 
Microsporangium, 30, 34, 36, 
47 
Middle-piece (of spermato- 
zoon), 25 
Mid-parent, 118 
Milewski, experiments of, 233 
Mitosis, 40-43, 46 
Molge, 102 
Molluscs, 63 
Morgan, experiments with 
Drosophila, 173 ff. 
Moron, definition of, 248 
Moulds, reproduction of, 18 
Mutations, definition of, 141, 
212 
Natural selection, 80 
Negro, racial characters of, 
242 
Nematoda, see Threadworm 
Nordic (race), 243 ff. 
Nuclear sap, 38, 39 
wall, 38, 39, 41, 42 
Nucleolus, 20, 38, 39, 44 
Nucleus, 14, 20, 24, 25, 37, 
42, 44, 45, 47, 48, 51, 52 
CEnothera gigas, 214 
lamarckiana (evening- 
primrose), 213, 214, 216, 
217 
Onagracea, 216 
Oocyte, 51 
primary, 44 INDEX 
255 
Oocyte, secondary, 46 
Oogonium, 43, 51 
Oophore, 30, 36, 52 
Oosperm, 22 
Oranda shlshigashira (" lion- 
head " goldfish), 223 
Order, definition of, 139 
Organ-forming substances, 65 
69, 73 
Organogeny, 54 
Ovary, 22, 35 
Ovule, 33, 34 
Ovum, 22 
Oyster, 56 
Paramecium, 135-137 
Parker, observations on 
growth of tadpoles, 225 
Parthenogenesis, 19, 27 
Pavlov, experiments of, 107 
Pea, 36 
Phyllopoda, 132 
Pig, skull of, 233 
Pine, 33 
Pisum sativum, 141 
Polar body, 33 
first, 46, 47, 49 
second, 47, 49 
Pollen-grain, 33, 35 
parent. 142 
tube, 33-35 
Prepotency, 238 
Primary organs, 53 
Primitive germ-cell, 42 
Primordial utricle, 14 
Probable error, 117 
Pronucleus, female, 47-49 
male, 48, 49 
Prophase, 40, 42, 43 
Proteus, eyes of, 110 
Prothallium, 28, 29, 32, 33, 
35, 36, 52 
Protista, 17, 18, 45, 46 
Protoplasm, 14, 15, 24, 25 
Quartiles, 117 
Qu6no, experiments with 
mice, 160 ff. 
Races (of men), 243 ff. 
Rana esculenta, 210 
Recapitulation, 56, 108 
explanation of, 56 ff. 
Recessive character, 143 
Reducing division, 45, 46, 52 
Regression towards medioc- 
rity, 127, 130 
Regulation, definition of, 241 
Retina, 75 
Rhlzocephala, 189 
Rlgnano, 109 
Root-hair, 29 
Ryukln ("veil-fish"-goldfish 
variety), 222 
SaccuUna neglecta, 190, 191 
Salamandra atra, 84, 86 
development of, 60, 
100, 101 
■ maculosa, development 
of, 60, 84 ff., 101 
toeniata, 84 
Salpa, 193 
Salvlnla, 32 
Salvlnlacea, 31 
Schleiden, 15 
Schwann, 15 
Sea-squirt, 63, 69 
urchin, 24, 63 
Secondary sexual characters, 
definition of, 168 
Secondary sexual characters, 
inheritance of, 202 ff. 
Seed-coat, 36 
parent, 142 
plant, 34 
Segmentation, 53 
Segregation, definition of, 
154 
Sex-linked Inheritance, defini- 
tion of, 180 
Sexual Hormones, definition 
of, 197 
Process, 18, 19 
Silver-fish, 228 
Simocephalus, 131-3, 137 
Slipper-animalcule, see Pa- 
ramecium 
Smith, Geoffrey, observations 
on sex of crabs, 188 ff. 
Sol, 26 
Sorus, 27 
Species, definition of, 139 
Spindle, 41, 45, 48 
Spermatid, 46 
Spermatocyte, primary, 44 
secondary, 46 
Spermatogonium. 43 
Spermatozoid, 20, 21, 29, 
30, 32, 33, 52 256 
INDEX 
Spermatozoon, 20, 21, 25, 29, 
30, 47-50 
Sporangium, 27-28 
Spore, 28, 30, 52 
Sporophore, 30, 36, 52 
Sport, definition of, 141, 212 
Standard variation, 117 
Star-fish, 24, 63 
Stigma, 35 
Strain, definition of, 140 
Strawberry, runners of, 11 
Synaptic knot, 51 
Tadpole of frog, 54, 62 
of Hylodes. 60 
Tail of lizard, 74 
Tail of spermatozoon, 20, 24 
Telescope-fish (" Deme-Ran- 
chu"), 223, 227 
Telophase, 40, 42 
Testa, 36 
Testis, 22 
Tetraploldy, 216 
Thread-worm, 21, 63, 68 
Tornier, experiments of, 
220 ff. 
Triton (see Molge), 102 
Trochophore, 188 
Tubifex, 87 
Tunicata, see Ascidians 
Vacuole, 14 
Variety, definition of, 140 
Veil-fish (see " Ryukin ") 
999 997 9QQ 
Vitelline membrane, 20, 24 
Von Mohl, 14 
Wallace, theory of, 112 
Waterfern, 30, 32, 47 
Weismann, 47, 75, 81 
experiments of, 82 
Weldon, experiments of, 124 
White (of egg), 23 
Witschi, observations on 
frogs, 201 
X-CHROMOSOMES, 180 ff. 
of Drosophila, 198, 199 
of Euschistus, 204-208 
of Lymantria, 198 
Y-chromosomes of Droso- 
phila, 180 ff. 
of Euschistus, 204-8 
Yolk, definition of, 22 
Youth form, of acacea, 62 
of gorse, 62 
Zygote, 18, 48, 49 
development of, 53