titles of papers referring to variation, and that the
whole work is devoted strictly to facts, so that it may
really be considered as a list as complete as possible
of all departures from the normal types. Of course
variations of floral structures are numerous, and cases
abound in this work, but I prefer referring, as an
instance, to a case which is not noted by Dr. Penzig,
and which is of great interest, as it concerns import-
ant variations observed in the floral structures of one
and the same individual plant, a Tradescantia virgi-



FLOWER- VARIATION 101



nica. 1 "This plant," says Mr. G. A. Brennan, "presents,
as the result of thirteen years' cultivation, the curious
aspect of a monocotyledonous plant bearing in
bloom at the same time flowers of dimerous, trimerous,
tetramerous, pentamerous, hexamerous, and hepta-
merous types respectively, each flower bearing twice as
many stamens as sepals, petals, or carpels of the ovary.
The plant was set out in 1872, and received very rich
treatment, so that it gave forth blossoms measuring
two inches in diameter. In 1874 it began to depart
from the original trimerous type and to assume the tetra-
merous one, by developing another petal, and instead
of doing this at the expense of the pistil or stamens, it
added another sepal, another carpel with style, and two
stamens, thus making a typical tetramerous flower.
The plant has since then continued to differentiate in
a greater degree each succeeding year." In 1876 it
became pentamerous, in 1879 hexamerous, in 1882
dimerous, in 1886 heptamerous ; thus you perceive that
there has been no regular order in the course of
differentiation. At present, while the pentamerous
type is dominant in this plant, dimerous and hepta-
merous flowers are scarce. It seems that further
variation is forthcoming, for an octamerous ovary has
been detected in one flower. This fact is certainly

1 G. A. Brennan : Variations of Tradescantia virginica. American
Naturalist, vol. xx. 1886, p. 55.



102 EXPERIMENTAL EVOLUTION LECT.

one of the most striking which can be quoted in
respect of floral morphological variation. Such varia-
tion is always present among different varieties of the
same species. For instance, C. E. Bessey l has investi-
gated the floral structures of different varieties of
apples, and while it is generally thought that no
difference obtains, he has detected considerable
differences in the form of the stigmas and styles, and
found that the pistil varies much in length, breadth,
hairiness, &c. And the proof thereof is seen in the
engravings which accompany his paper, and refer to
five well-known varieties of apples red Canada,
Talman Sweet, Rambo, Wagner, &c.

The very smell of flowers is also subject to variation,
as Dalibard 2 showed by direct experiment nearly a
century and a half ago. He planted mignonette in
different soils, using seeds from the same mignonette
plant, possessing its well-known fragrancy. While
the seeds sown in rich garden soil became vigorous,
and were well perfumed, the seeds sown in sandy soil
produced plants which remained weak and small, and
had no perfume. It even seems that the latter
did not acquire any odour when transferred to rich

1 Can Varieties of Apples be distinguished by their Flowers ?
American Nattwalist, vol. xx. 1886, p. 162.

2 Observations stir le Reseda a few odorante, in Me moires de
Mathematiqucs et de Physique de V Academic des Sciences, 1750,
P- 95-



OSTEOLOGICAL VARIATIONS 103



garden soil. Similar facts have since been repeatedly
observed and noticed.

In the more internal functions and organs of animals
and of plants as well the same variability shows
itself. In man himself, as Mantegazza has shown, teeth
vary considerably, and a careful study of the third
molar tooth has shown that there is a strong tendency
towards the disappearance of this part, and while among
inferior races all that concerns it is normal in 50 to 54
per cent., abnormality becomes considerable among
superior races, where the normal state is only met in
37*09 per cent., leaving 62'9i abnormal in one way or
other. No doubt, we could find numerous cases of
variation in the dentition of mammals, although
the number and form of teeth is considered as a
specific character. But teeth may be considered as
external organs in some sense, just as fur or feathers ;
and it is even more interesting to see that more
internal parts vary perhaps as much, if not more.
Such is the case with the bones which go to make
the skeleton of mammals and other animals. Some
instances are referred to by Darwin, and by Wallace
in his recent and valuable Darwinism ; St. George
Mivart has shown that the number of ribs varies among
the apes ; in man himself the number varies from
twelve to thirteen ; and concerning whales, Georges



io| EXPERIMENTAL EVOLUTION LECT.



Pouchet l says : " Balaenidae certainly are among
those higher vertebrates whose skeleton exhibits
least fixity ; this is a peculiarity which cannot be
denied." The same writer also says that in many
species which live in limited regions the same skeletal
variability exists in a marked degree, and although
the individuals are absolutely similar so far as exterior
characters are considered, they may display a varia-
bility which may be said to be unlimited, in the
number and relations of the bones. And Pouchet and
Beauregard 2 say also that it would prove difficult
to meet with two skeletons of the Anteater which
were exactly similar as concerns the number of the
ribs or vertebrae, or the connections of the ilium or
ischium with the vertebral column.

As to differences in weight and length of the skeleton
in different individuals of the same species, Darwin
and Wallace have also collected numerous data of
which all are cognizant. The soft parts of the body
display the same tendency. All have heard of John
Hunter's experiments on the sea-gull (Lams tridacty-
lus}. He fed it during a year on grain, with the result
of hardening to a large extent the inner coat of the

1 A propos de deux Photographies de Baleine Franc he. Compt. Rend,
Soc. Biologie, 1890, p. 705.

- Traite d? Osteologie Comfaree, 1889, p. xii.



II VISCERAL VARIATION 105



stomach of this animal, which, being a flesh-eater, does
not require to have the hard and horny coating of the
gizzard of the pigeon or fowl. This experiment is
repeated each year in nature, and without man's
operation, by another gull (Larus argentatus}, of the
Shetland Islands, which, according to Dr. Edmonstone,
changes the structure of its stomach twice every
year, according to its food, which consists of grain
during part of the year, and of fish during the other
months. So the stomach may vary considerably in
its use and functions, and Holmgren's experiments
show that the gizzard of a grain-eater, such as the
pigeon, may be converted into a carnivorous stomach,
such as that of one of the birds of prey.

I have already said that there is great variability in
the muscular system. Some anatomists have made
a special study of this variability : Wenzel Griiber
in Germany, Testut in France, Cunningham of
Dublin, and many others. Not only are there
variations in the mode of attachment and course of
every muscle of the human body which has been
more especially studied in this connection but super-
numerary muscles are often found which are all exactly
similar to muscles which normally exist in lower
animals, but do not as a rule exist in man. Testut
has dwelt upon this fact, which is of great significance
in the evolution theory, and a very large number of



io6 EXPERIMENTAL EVOLUTION LECT.

instances might be given of man having muscles which
are considered as peculiar to the ape, horse, dog,
bear, &c. 1

Variations also occur frequently in the anatomy of
many internal organs. Wallace refers to the
variability in the length of the digestive system in
the giraffe and other animals, and in the nature and
position of the gall-bladder, which in the same species
is sometimes present, either single or double, some-
times absent. These variations are not confined
to higher animals. Claus observes that ^Equorea
forskalea, a Ccelenterate, varies much in the number
of the radiating canals 2 ; and many botanists have
noticed the important structural variations which
obtain in plants. E. Mer has carried his investigations
into great detail in regard to Isoetes laciistris, and
other plants. 3 It results from these investigations
that the internal anatomy of plants may vary
considerably. This variability displays itself also in
regard to sex ; for it has been shown that external
influences play a large part in the determination

1 Cf. R. Wiedersheim, Der Bail des Menschen ah Zeugniss seiner
Vergangcnheit (Freiburg i. B., 1880).

2 American Naturalist, vol. xvi., 1882, p. 147.

3 De F Influence exercee parle Milieu stir la Forme, la Structure, et le
Mode de Reproduction de F Isoetes lacustris. Ccmptes Rendus 1881, p.
94 (Jan.-July). See also Des Causes qui modifcnt la Strttcture de
ccrfaines Plantcs aquatiques vegetant dan F Eau. Bull. Soc. Botanique^
1880, p. 194.



SEXUAL VARIATION 107



thereof. Whilst among tadpoles left to themselves,
the females are in a slight majority, the proportion
increases from 54 to 78 per cent, when the
tadpoles are fed with beef, to 8 1 per cent, when fed
with fish, and, when fed with frog-flesh, to 92 per cent. 1
Thus food, and the nature of food, has much to do in
the determination of sex. The same is the case with
bees, where the production of queens, workers, and
drones is in great part a matter of nutrition. A
worker-larva may be reared into a queen, if royal food
is provided. Other facts show similarly that external
influence must be at work to operate in the deter-
mination of sex. Fisch has noted the sex of 66,327
plants of hemp, and he finds there are 154 female
against 100 male plants. Among other plants, such
as Spinacia oleracea and Ruinex acetosella, the
proportions vary much, as in some cases the one sex,
in others the other, is predominant. In the human
species males are constantly in slight excess over
females 105 against 100. The same condition
obtains among oxen, sheep, hogs, and domestic
birds. But in the case of the latter, the constancy
is less, and during some particular years there is
a very large number of individuals of one sex
against a small number of the other. There

1 See Yung, Propos Scientijiques, 1890, Reinwald, Paris, quoted in
Evolution of Sex, P. Geddes and J. A. Thomson, Lond. , 1889.



io8 EXPERIMENTAL EVOLUTION LECT.

are some external causes in operation which are
not yet detected. M. C. Cornevin thus summarizes
the proportion of males to 100 females in the
following species :

Horses 101 males against 100 females.

Oxen 104-6

Sheep 115-4

Hogs 104-9 .

Turkeys 120

Guinea-fowl 102

Common fowl 101 ,,

Duck 115

That external influences do play an important part
in the determination of sex is shown by numerous
facts. Spallanzani, Bernardi, and Autenrieth have
shown that female plants of hemp when mutilated
bear male flowers, and M tiller has in some cases seen
male plants bear female flowers. M tiller has also ob-
served female plants of Zea Mays bearing male flowers
when nutrition was deficient. Hoffmann has noticed
that in Lychnis, Spinacia, and Rumex the proportion of
sexes varies according to the greater or less interval
between each individual plant ; and Cornu, Giard, and
Magnin have shown that in Lychnis vespertinci, under
the influence of parasitic " rust " ( Ustilago antherarum\
female flowers bear stamens. Prantl has seen similar
facts among Cryptogams. While the (seeds/ of ferns



SEXUAL VARIATION 109



develop into male plants when the soil is poor in
nitrogen, or when the seeds are very near each other,
they yield female plants when nitrogen is abundant,
and the seeds somewhat distant.

Yung's experiments on tadpoles had already been
performed by Born with similar results, and it seems
that in the human species, a change of climate is
often conducive to a larger production of females.
In Java, for instance, European or white children are
born in the proportion of five females against two ,'
males ; in Yucatan, in the proportion of eight /
females against two males.

All these facts go to show that sexuality is in
great part determined by external factors, whatever
these may be, and that much variability is here
present. This variability may be readily seen in the
same species, under different conditions, and even
in the same individual. For instance, Carriere has
pointed out that variability is common in Ailanthus
glandulosa,%. well-known plant in which the distribution
of sexes is very inconstant. While some individuals
bear a large number of female flowers, many bear but
few, and it is a curious fact that they are all to be
found on the same branch, instead of being on different
branches, interspersed with male flowers. There
seems to be some special condition in one or more
branches which determines the production of female



i io EXPERIMENTAL EVOLUTION LECT.



flowers. 1 This condition may vary from one year to
another, in the course of the lifetime of the plant. It
even seems that in normally male plants, this condition
may put in an appearance. Ch. Martins 2 observed at
Moritpellier a male Chamcerops humilis which yielded
only male flowers from 1851 to 1861 ; in 1861 this
plant produced some female flowers, quite normal,
since the seed from these flowers yielded vigorous
young plants; and in 1862 a large proportion of
female flowers were to be seen. This last fact is of
real importance, as showing that sexual variability
may exist to a high degree.

Another very interesting form of variability is that
which may be observed in individual evolution or
development. Although there are numerous cases of
this sort, and although a large number of instances
might be quoted where the individual evolution is
readily arrested or modified through different circum-
stances, none seem more carefully ascertained than
those which Camerano has published. This writer has
investigated what Kollmann has called Neotenia in
Amphibians. Neotenia is the lengthening (for an
indefinite time) of the period during which Amphibians
are gill-breathers. Every one knows that, at first,

1 Carriere, Sur FAilanthus glandulosa a propos des Sexes. Rev.
Horticole. 1872, p. 234.

Ch. Martins, Transformation d?un Chamcerops humilis male en
polygame. Rev. Horticole, 1862, p. 353.



ii NEOTENIA in

frogs, toads, &c., breathe as tadpoles, by means of
gills, and that after a few weeks the lungs develop
and the gills disappear, while the animal becomes an
adult, and acquires new characters and organs. But as
every one can ascertain, this gill-breathing period may
be considerably lengthened under natural or artificial
and experimental circumstances. I have myself kept
toads in the tadpole state for over two years, merely
by feeding them very scantily. They were born in
the spring of 1889, and remained all the time in an
aquarium in the laboratory, having water enough at
their disposal, being always sufficiently provided with
aquatic plants, and enjoying heat enough : it cart by
no means be said that their evolution was arrested by
the cold of winter, as often happens in mountain ponds,
when the cold of autumn sets in before the tadpoles
have achieved their development, so that they become
frogs or toads only in the course of the following
year. In the case of my tadpoles, it seemed that the
completion of development was due to my imprudent-
ly feeding them in the spring of 1891 on the very sub-
stantial flesh of their congeners ; and in the course
of some three weeks at most, the limbs were evolved,
the long tail disappeared gradually, the very colour
and appearance of the skin underwent considerable
changes, and my superannuated tadpoles became
toads at last. This Neotenia has been observed by



112 EXPERIMENTAL EVOLUTION LECT.



many physiologists in different Amphibians. Mile, de
Chauvin has reared larvae which were the progeny of
Amblystoma, and while some of them became
, Amblystoma, others remained Axolotls in con-
sequence of being kept in very well aerated water,
where the gills had no tendency to atrophy or retro-
gression. Similar experiments have been performed
by a large number of naturalists on different species
of Triton, Salamandra, Pelobates, Alytes, Hyla, Rana,
and Rnfo ; and the result is that while there arc
Amphibians, such as Salamandra atra, in which the
length of the branchial or gill-bearing period is very
short, and others, such asfroteus anguineus, and some
Tritons and Axolotls, where gills exist normally in
adult and even in aged individuals, there exist
also a number of Amphibians among which the
gill-bearing period, normally short, may be much
lengthened. But in Urodela (newts and sala-
manders) this lengthening 1 may and does occur
without seriously modifying the evolution of the
remainder of the body ; and the result is that these
tadpoles are sexually mature, while among Anura
(frogs and toads) this lengthening interferes with the
general development, and sexual maturity does not
seem to occur among the tadpoles. New investigations

are required to ascertain how far this sexual immaturity

>**

exists, and to what extent it may be retarded by the



II NEOTENIA 113

lengthening of the gill-breathing period. 1 If it could
be shown that sexual maturity may occur although the
tadpole state is lengthened, and that sexual reproduc-
tion may take place, although this is on obvious a
priori grounds very improbable, we might perhaps try
to obtain a new species which would exhibit very
marked physiological features.

1 Camerano, Stir le Developfenient dcs Amphibiens, et sur ce que ton
a nomme chez eiix la Neotenie. Arch. c 'ltaliennes de Biologic, vol. v.,
1884, p. 27. Recherches sur la Prolongation de la Periode branchial e
des Amphibiem. Ibid., p. 29. Recherches sur le Developpement et les
Causes du Polymorphisme des Tetards des Amphibiens anoiires. Ibid.,
xv. 1891, p. 165.



LECTURE III

Summary : The Facts of Natural or Spontaneous Variation (concluded)
Physiological or Chemical Variation Not always easily detected.
May be noticed in all parts of the Body, even between very
closely related Forms Exists not only between different Species,
but between Varieties of the same Species, Individuals of the same
' Variety, and even different ages of the same Individual Chemical
Variation explains Racial Immunity to peculiar Diseases This
Chemical or Physiological Variation in some cases of much higher
import than any Morphological Variation Chauveau's Experi-
ments on Bacillus ant hr ads Physiological Differences between
Brown and Green Frog towards Poisons and Heat Tarchanoff's
Experiments Variation generally exists at all Ages, in all Groups
of Beings, at all Geological Epochs Sudden Variation.

THERE is a last form of variability to which I wish to
call attention, and which has not been enough taken
notice of up to the present time. I refer to what I have
mentioned under the name of chemical or physiological
variability. Under this name I include all facts which
indicate a difference in chemical or physiological con-
stitution, expressed through differences in the reaction
of the organisms towards definite and common external
influences. Such chemical variation must certainly
exist at the basis of all specific or even racial characters,



LECT. in VARIATION IN PHYSIOLOGY 115



and if I dwell somewhat upon the topic, it is owing to
the fact that this sort of difference has not been as much
investigated as it ought to have been. Between two
species, however closely allied, between two varieties or
races of the same species, there are not only those slight
external differences upon which so much stress is laid
by morphologists ; there are internal, chemical and
physiological differences which are most likely of
greater importance. For instance, Naudin cultivates in
his gardens at Collioure, in the south of France, a
number of plants of a species of Echium ; part are
indigenous, part come from the Canary Islands ; they
all exactly resemble each other, no external difference
is perceptible ; they differ in origin only. During the
night, the frost comes ; all the Canary plants die,
while the plants of France resist. There is some
difference in their constitution or physiology, some
difference due to habit, to adaptation, however it
may be called, and the result is that life may continue
under circumstances which cause it to cease when this
difference does not exist. This is one fact among
a thousand, and horticulturists and breeders could
provide many similar examples. It shows that,
even in cases where no external differences are
perceptible, variations do exist in given circum-
stances which may be of the highest importance, and
decide life or death. Though they are not always of

I 2



ii6 EXPERIMENTAL EVOLUTION LECT.

equal magnitude, they admit of being detected
with real accuracy by appropriate methods. While
calling this variation physiological, I understand it
to be, in fact, chemical ; the difference is in chemical
constitution most probably but it displays itself
mostly in physiological differences. I have collected
some cases of this variability, endeavouring specially
to obtain widely different instances, in order to show
the extent to which this kind of variation occurs.

Of positive chemical variability among animals, I
meet with a good instance provided by two well-
known savants Ch. Robin, the histologist, and
Sainte-Claire Deville, the chemist who, at Sanson's
request, examined comparatively the structure and
composition of the bony structures of the common
breed, and of a perfected breed, of sheep. While the
microscope detected no difference at all between the
two breeds, chemical analysis showed considerable
variations in the respective percentages of organic
and inorganic matters, as follows :

Organic Substances. Inorganic Substances.

Perfected breed 32*3 per cent. 677 per cent.

Common breed 38*6 61*4

Similarly, considerable variations obtain in the
chemical constitution of the integumentary append-
ages of different varieties of animals. Here follows,



in CHEMICAL VARIATION 117



for instance, the percentage of the principal com-
ponents of the wool of some breeds of sheep
(after Miintz and Girard) :



Dfchley. Merino, So log not.

Water .................. 15 10 14 15-5 16

Nitrogenous matters 63 48 53 64 53

Fatty substances ... 8 30 19 6 10

Ash ..................... ii 10-5 15 13 18

Potash .................. 6 4-5 7 6-5 4-5

While some components vary but slightly, such as
water and potash, others, such as fatty substances,
are found in very different proportions ; the difference
being from six to thirty, or from one to five.

Similar variations are to be observed in the mus-
cular or fleshy parts. Sir R. Christison made
chemical analyses of salmon in good health and con-
dition, and of salmon after spawning. The results
are as follows : 1

Healthy Salmon. Salmon out of Season.
Oil ............ 18*53 per cent. 1*25 per cent.

Nitrogen ... 1970 17-07 ,

Salts ......... 0-88 0-88

Water ...... 60-89 8o'8o

It is not necessary to have studied physiology very
deeply to understand that such differences in the flesh
or skeleton, as are shown above, may be of great im-

1 From a paper read in the Royal Society, quoted in the American
Naturalist, vol. vii., p. 372.



i IS EXPERIMENTAL EVOLUTION LECT.

portance under definite conditions, and may deter-
mine life or death in some circumstances, while at all
events they must, in every-day life, put the animals
which exhibit them in very different positions as
regards the struggle for life and success in it.

Such differences are common in the chemical consti-
tution of the different species of the same genus, and
the following analysis by Forchammer well illustrates
this :



Fitcu
Potash


s digit alus.
20-66
7-65
6-86
10-98
2-36

0-57
1-44
26-18


F. vesiculosns.
13-01

9'54
6-12
836
1-16
24*06
0-28
1-15
21-45


F. serrati

18-67
10-29

14-41

18-59
0*30
0-38
16-56


Soda


Magnesia


Lime


Phosphoric acid ...
Sulphuric acid
Ferric oxide


Silica


Sodium chloride ...



And again, among different individuals of the same
species considerable differences may obtain according
to the mode of life, and particularly, as Hermbstaedt
has seen, according to food. This fact is well dis-