imprisoned between the cork and surface of the
water hardly amounting to 50 cubic centimetres.
Even if it is argued that some air may pass in and
out, through the cork, the quantity is very small, and
we may consider the renewal of the air as very incon-
siderable so far as penetration from the atmosphere is
concerned. Of course the air does and must remain
quite suitable for the animals, since they thrive, and
the plants are the agents of this continued purifi-
cation. If the animals can and do live under such
conditions, and even live as well as they do when
the communication with the atmosphere is not inter-
rupted, does this not show that aeration must be
considered as quite sufficient even when the surface
is small ?
G 2
84 EXPERIMENTAL EVOLUTION LECT.
In a second series of experiments I varied the
volume of the water, but allowed the surface to
remain the same. For instance, into two vases of iden-
tical form and diameter I poured unequal volumes
of water. The surface was the same, but the volumes
were very different. In such cases, while the animals
were certainly larger in the larger volume, the differ-
ence was not considerable. The influence of volume-
variations is thus seen to be much less important than
that of surface-variations.
These experiments seem to me to call for the
following interpretation. The volume of water is in
itself of comparatively small importance, especially
for some species of pond-snails, and the real influence
is exerted by surface. And surface operates in this
manner only : the larger it is, the more exercise the
animals are able to take. The L. auricularia, to which
the above-mentioned experiments refer more especially,
seems to dislike deep vessels, and moves usually
in the horizontal plane near the surface. If the
surface is small it moves but little, while if it is
large the animal moves a great deal. On the other
hand L. stagnalis seems generally to care less about
surface or depth, and to prefer living in the deep
parts of the vessels where it is always moving about. I
have seen it in one case live almost all the time in the
deepest part of its prison (a glass balloon with a long
SEMPER'S EXPERIMENTS 85
narrow neck), and it acquired dimensions certainly
equal to those of another one living in a large-
surfaced vessel. There is some difference certainly
between L. stagnalis and auricularia in this respect,
and I call attention to this point.
I have tested Semper's interpretation in another
manner. I have caused pond-snails of the same age and
brood to live in unequal volumes of the same water in
the following manner. In some eases I have used
one, or two, or more glass tubes, of same length and
diameter exactly (two pieces of the same tube),
which were closed at one end with some muslin
stretched over the aperture, and made fast by means
of a string or thread wound around the tube. The
tubes were suspended in a large vessel containing
three or four litres, by means of a string, in such a
manner as to allow the other end to rise, say two
centimetres, above the surface (to prevent the animals
from getting out of the tube and going into the
vessel). In each tube I put one pond-snail, with a
sufficient quantity of aquatic plants (submerged al-
ways), and one in the vessel, outside of the tubes.
Every day, and many times a day, the tubes were
lifted so as to empty them of water, and immediately
replunged, so as to ensure the mixture of the water
inside them with the water outside; moreover the
water in the tubes was in constant communication
86 EXPERIMENTAL EVOLUTION LECT.
with the water around them, through the muslin, whose
only function was to prevent the animals from
escaping from the tubes into the vessel, or vice versa.
In all such experiments, the water being the same in
both tubes and vessel, food being superabundant,
temperature identical, and surface and volume only
being different, I have seen the pond-snails in the
tubes remain much smaller than those in the vessel. I
may even add that in some cases I have had one tube
as above described, and another stopped at the lower
end with a good cork and wax around it, so that the
water in the tube never got mixed with that of the
vessel, and have hardly if at all noticed any difference
in the dimensions of the animals of both tubes.
This experiment may be performed in another
manner by fixing up a sort of cage (with muslin and
glass rods), which affords more space than the tubes,
and more surface and volume ; the communication
between the water inside and the water outside is
still better, since it is effected through all the available
sides of the cubic cage ; the results are the same,
and the animal inside the cage remains much smaller
than the one outside it. This series of experiments
answers the objections which might be raised on the
ground that in the smaller volume of water the
proportion of waste products might be larger, and
exert a noxious influence. But these waste products
II SEMPER'S EXPERIMENTS 87
do not require to be taken into account in such cases.
I have also seen that unless a given volume of water
has been inhabited for a long time or by a large
number, it exerts no bad influence on the growth of
other pond-snails. For instance, take two identical
vessels, and use, in one, one litre of pure fresh water,
in the other the same quantity of water in which a
pond-snail has been living two or three months ; in
each put one young Lymncza, of same age and brood ;
kill both after the same time (three or four months) ;
there is no observable difference. Of course, if the stale
water has been much inhabited by pond-snails, the
growth of the fresh ones is impaired. But such
impairment does not occur in my experiments, and I
do not well see how waste products could accumulate
more in a narrow-surfaced vessel where aeration is
very good, as the plants show, than in a large- surfaced
vessel, where it must be also good, the quantity of
water being equal in both, or even, as is the case in
many of my experiments, superior in the former. 1
So it seems that Semper' s interpretation has to be
dismissed as unnecessary, and that a simpler expla-
nation is furnished by the results of my experiments
an explanation which depends upon known principles
1 All these experiments shall be related in greater detail in a forth-
coming memoir, as soon as I have completed the experiments
which are yet being continued (March, 1892).
88 EXPERIMENTAL EVOLUTION LECT.
of known influence. It is quite natural that exercise
should have an influence upon growth and development,
and that in cases where there are physiological or
mechanical impedimenta to movement, dwarfing should
be the result. I think that this is the explanation
which must be accepted ; and if animals living in
confined spaces remain small, this is due to the fact
that they cannot move enough. At all events Prof.
Semper's interpretation seems to me not acceptable.
Further experiments will yield new facts, and time
will tell whether this explanation is sufficient
In this connection I may call attention to the cir-
cumstance that an observer who has given some study
to dwarfing in Lymncea has pointed out a singular
fact connected with this process, but one which
requires to be confirmed by new investigations.
It is the fact that dwarfed forms are generally
exclusively female, and that their liver offers a con-
siderable amount of degeneration.
So much for dimensional variation.
If we now pass on to consider the integument, we
perceive that in this part, and in its appendages,
variations are numerous and also important. Many
animals, when transferred to warm climates, lose
their wool, or their hairy covering is much reduced.
In some parts of the warmer region of our earth
sheep have no wool, but merely hairs like those of
II INTEGUMENTARY VARIATIONS 89
dogs. Similarly, as Roulin notices, poultry have, in
Colombia, lost their feathers, and while the young are
at first covered with a black and delicate down, they
lose it as they grow in great part, and the adult fowls
nearly realize Plato's realistic description of man
a biped without feathers. Conversely, many animals,
when transferred from warm to cold climates, acquire
a thicker covering, dogs and horses, for instance, be-
coming covered with wool, &c. Such cases are easily
observed in Europe when animals from the warm
regions are sent to our zoological gardens. In
Paris, for instance, we have seen sheep from Senegal
acquire, in the course of two years, a long and grizzled
cover of hair, while at first they had but a short one.
Similar modifications have been observed in a large
number of animals, and more precise data could have
been obtained if more attention had been paid to the
subject. As M. Faivre rightly remarks in his La
Variabilite des Especes et ses Limites (1868), while
" no truth is better established in natural history
than the influence of climate on the superficial cha-
racter of animal species, on the dimensions, colour,
form, nature of integuments and hairs, none has been
less investigated and discussed by the naturalists
whose business is to distinguish one might even
say, to multiply species."
Variability and variation of such superficial cha-
90 EXPERIMENTAL EVOLUTION LECT.
racters sometimes go to such an extent that animals
of the same species have been at times considered as
belonging to different species, and even to different
genera. Such has been the case, for instance, with
two fishes Abramis versicolor and Stilbe americana
which C. C. Abbott recognizes as one and the same
species which has a great tendency to variation,
not only as concerns colour, but in respect of fins and
scales, according to its environment. 1 This is doubt-
less an extreme case, but its interest is considerable,
in that it exemplifies, on the one hand, the importance
of variation, while, on the other, it shows once more
how very artificial and unsound our specific and even
generic distinctions in some cases are. While we
ascribe most of the superficial, or integumentary,
variations to that general and complex factor which
we call change of climate although we cannot in all
cases tell which particular factor of the complex
operates there are cases where we can trace the
variation to one determined cause. Such is the case
with the variation in length of wool. There is a direct
relation between the abundance of food and the length
of the wool of sheep, for instance. Krocker, 2 in
Proskau, has shown that the amount of wool yielded
1 C. C. Abbott : Notes on the Cyprinoias of Central New Jersey.
American Nattiralist, vol. viii. p. 326.
- His paper has been published in the Annalen der Landwirthschaft
in den Koeniglich Preussischen Staaten for 1869.
n INTEGUMENTARY VARIATIONS 91
daily per 1,000 kilograms of sheep varies in the
following proportions according to the food :
Kg.
o'69i of wool : scanty winter food.
0*870 ,, plenty of hay.
0*958 ,, good pasture.
ro8o 1*240 ,, fattening process.
If we now turn to plants, we perceive the same
variability in the superficial integumentary organs.
I merely recall here because I shall have to refer to
it at greater length later on the considerable differ-
ences which many observers have recently noticed in
the anatomy and characters of the same parts which
successively lead aquatic and aerial lives. In these
cases the influence of environment is easily to be
traced and appreciated. It is also well known that
where mountain-plants are transferred to the valleys
and plains they lose the hairy covering which they
generally possess, while valley-plants transferred to
the mountains acquire this same covering. Linnaeus
noticed that Persicaria is devoid of this sort of down
when living in humid places, while it becomes very
villous in dry stations. The same is noticed of
Thymus serpyllum. Many plants, in short, exhibit
two varieties which are readily distinguishable the
glabrous and the villous ; such are Prismatocarpus
speculum, Isatis tinctoria, Jasione montana, Onopordon
acanthium.
92 EXPERIMENTAL EVOLUTION LECT.
Similar cases are met with among spiny plants.
While some plants, which possess no spines, become
markedly spiny when they grow in some localities,
others, which are spiny, lose their appendages. Such
is the case with Capparis spinosa for instance, of which
Turrel l has described a variety without spines in the
Balearic Islands. In all other respects this variety
exactly resembles the common form. Whether a
lusus naturae or not, this peculiar character is heredi-
tary, as the seeds of this form always yield non-spiny
plants, in France as well as at Mahon. Another case
is that of Ulex europaeus, of which there exists a non-
spiny form, as Trochu has shown. 2 This form is
seldom met with, as it has less chances of success in
the struggle for life, for while oxen, rabbits, hares,
and other animals are respectful and deferential
towards the common spiny form, they have a great
liking for the other one, and eat all they can of it. It
must be added also that the last-named bears but
few seeds, and thus cannot become very abundant.
De Jussieu considers this Ulex nanus as a variety of
Ulex europaeus, and Vilmorin has made some in-
vestigations concerning this form, 3 which may become
1 L, Turrel : Sur le Caprier sans Epines. Bull. Soc. Zool. Acclima-
tation, vol. viii. p. 448.
2 L. Vilmorin : De PAjonc sans Epines. Revue Horticole, vol. xii.
p. 151.
3 Note sur un projet d? Experience ayant pour but de creer line Variete
li FORM-VARIATIONS 93
very useful, as it can be given as food to animals
which will not eat the common spiny form. His first
experiments have not proved satisfactory, for the seeds
of Ulex nanus have always yielded plants of Ulex
europaeus. But since the tendency to vary is strong
enough in Ulex europaeus to afford some plants vary-
ing in the direction of non-spinosity, we may hope
that by means of careful selection Ulex nanus may
become an abundant and permanent form.
So much for integumentary variations. While con-
sidering variability of external and superficial charac-
ters, we may now say a word of form-variations.
These are very frequent among many groups of
animals, and particularly among molluscs. Locard,
in his interesting and valuable Variations malacolo-
giques (vol. ii.), has collected many instances of form-
variations noticed by himself and by others. In
his opinion, Lymnaa frigida and thermalis are mere
varieties of L. peregra, while Ancylus rupicola and
thermalis are varieties of A. simplex, the only differ-
ence being a matter of mere form. Brot has noticed
that in the cool mountain waters, Lymncea auricularia
has only four whorls to its shell instead of five, and
the Marquis de Folin observes that the pond-snails of
iCAjonc sans epines, se reproduisant de graines. Bulletin de la
Socitte Industrielle d' 'Angers, 1851. Also in Notices sur I' Amelioration
des Plantespar le Semis, 1 886.
94 EXPERIMENTAL EVOLUTION LECT.
the Lake of Constance are less regular in form, more
abrupt, he thinks on account of the movements of the
water of the Lake.
Baudin also considers Pisidium pulchellum and
cinereum as two forms of the same species, and
Locard himself has discovered through experiments
that L. turgida and elophila are mere varieties due
to environment of the common Lymncea stagnalis.
He says : " These are not new species, but merely
different aspects of a common type, which is capable
of modification and of adaptation according to the
nature of the media in which it has to live." Bateson
has recently observed similar facts concerning Cardium
edule ; Locard shows how extensively any one species
Unio rhomboideus for instance varies in forma
and in colore according to its habitat, lake, river, or
torrent, and an indefinite number of such instances
might be quoted here. The same may be said con-
cerning plants. All know that in different stations the
same species exhibits considerable differences in form,
in the comparative height of the stems, in the number,
length, distance of the branches, and so on, and ex-
perienced practical botanists easily recognize through
these differences the origin of an individual plant,
detecting whether it has grown in a valley or on the
Alps, in dry or in moist soil, in an exposed or in a
protected station, As the well-known fungologist,
ii FORM-VARIATIONS 95
M. Boudier, of Montmorency, says, in valuable notes
which he has kindly written down for me in answer
to many queries, " plants growing in dry, unprotected
soil are small and dwarfed, while the same species
living in moist soil are more vigorous, more developed,
and especially much taller. A common species,
Serratula tinctoria, grows indiscriminately in dry and
in moist soil ; in dry and unprotected stations it seldom
is over ten or twenty centimetres' in height, while
in moist soil it easily attains one metre (100 centi-
metres). The common dandelion '(Taraxacum dens
leonis] has in dry soil leaves which are much more
irregular and incised, while they are hardly dentate
in marshy stations, when it is called Taraxacum
palustre" Individuals of the same species grow-
ing near the sea-shore differ markedly from those
growing far inland. Similarly species, such as some
Ramuiculus^ which can live under water as well as in
the air, exhibit marked differences when considered in
their different stations, as is well known to all.
These differences may be important enough to induce
botanists to believe in the existence of two different
species when there is only one. A century and
a half ago, G. Bauhin and Tournefort described two
different species of Coriander. But Fabrejou, a
botanist of that time, who has written a large treatise
on systematic botany, under the title, Description des
96 EXPERIMENTAL EVOLUTION LECT.
Plantes qui naissent ou se renouvellent aux environs de
Paris} was able to show that the two so-called species
are one and the same. " Here is the proof," he says :
" When the same seed, sown in fertile and infertile soil,
yields the two so-called species, one must conclude
that there is but one single species, and that that which
seems to establish a difference between the two so-
called species, can only come from the climate and
culture. It is certain, as I have often witnessed the fact,
that the same seed, sown in fertile and infertile
soils, produces the two alleged species." And Prof.
Bonnier, recently, in his Etudes sur la Vegetation de la
Valle'e de Chamounix et de la Chaine du Mont-Blanc
(1889), says, corroborating others, that "in high alti-
tudes the appearance of the same species is dissimilar :
the stems straggle on the ground, leaves are narrower
and thicker, flowers are comparatively large and of
higher colour, and most of the plants even lose many
morphological characters which they possess in the
plains. . . . The characters of plants of high altitudes
are even different enough to have induced many
writers to describe these alpine forms as particular
species." Prof. Bonnier's statements are of especial
value from the fact that they are based on facts derived
from experiments made in stations situated at different
altitudes ; they are not facts of mere observation.
1 1740, 6 vol. in 1 8, vol, iii. p. 244.
LEAF-VARIATION 97
I have alluded to the considerable morphological
variations which are observed in Ranunculus aquatilis,
Godron, who, in 1839, published an important mono-
graph/ of the Ranunculus group, has studied these
variations with great detail. When the plant develops
wholly under the surface of water, all its leaves are
delicately laciniated. If the plant is able to send
some of its leaves to the surface, they float and assume
a very different form, being kidneyvshaped and lobed.
The same plant when growing entirely out of water
presents a very different appearance : the stem is
short, much divided into branches, which bear a large
number of small leaves, cylindrical, much divided,
and somewhat thick. If it were not for the floral
organs, one would certainly believe in two or three
species. There is but one, however, which varies
greatly according to external circumstances, and this
is shown by the fact that the same individual plant
under different circumstances presents the different
appearances which have been mentioned. Lamarck
believed that Ranunculus aquatilis might be trans-
formed into R. hederaceus through changes in the
environment, but Godron denies the fact. Rubus
fruticosus seems also to vary considerably. Sagittaria
sagittcefolia, when growing in deep water has also
ribbon-shaped leaves, while in shallow water it has
also arrow-shaped leaves, which rise vertically instead
H
98 EXPERIMENTAL EVOLUTION LECT.
of floating horizontally. Similar variations in form
are to be observed in Myriophyllum verticillatum and
Juncus supinus, and many other plants. Polygonum
amphibiiun also exhibits important morphological
variations. When growing out of water it has lanceo-
lated, downy leaves, with short stalks, and covered
with stomata on both faces, while the same leaves, if
the plant is growing under water, are deprived of hair,
have a long stalk, a,nd are obtuse, without stomata on
the lower side. These two forms of leaves are often
met with on the same plant, where it has been,
through accidental circumstances, growing for some
time under water and for some time out of water.
Ch. Martins 1 notices similar facts concerning Jussicea
grandiflora, where the variations are even more im-
portant. Every one may notice in our common ivy
considerable variations in the form of leaves, and these
variations are also to be seen in other plants, in fact
they are more or less present in most plants, and
careful investigation will disclose their number and
importance. 2 These variations, which seem to be of
no account as far as the general life of the plant is
concerned, may however be accompanied by important
1 Observations sur la Jussicea grandiflora, in Bull. Soc. Botanique
de France, vol. xiii. p. 176.
2 G. Fournier : Recherches Anatomiques et Taxonomiques stir la
Famille des Crucijeres, 1868, and also Faivre : La Variabilite des
Especes et ses Li mites, 1868.
FRUIT- VARIATION 99
differences in the physiology of the plant. For in-
stance, Carriere, 1 after having noticed the formal
variations of the leaves of the ivy according to its
mode of life (climbing, or entirely independent and
tree-like), adds the significant fact that the leaves of
the climbing plant when inserted in the soil readily
start an independent life, and emit roots very soon,
while those of the independent form do so only with
great difficulty. Here is certainly an inportant
physiological difference ; not perhaps in itself, but as
indicating differences in the structure and life of the
whole plant.
Fruits vary as well as leaves ; the same branch of a
peach-tree, for instance, bears peaches and nectarines ;
the same branch of an orange-tree bears oranges and
lemons ; the same branch of an apple-tree bears quite
different varieties of apples. Prof. Decaisne, who was
an authority in the matter of fruit trees, especially
apple and pear, observed 2 considerable variations
among the descendants of seeds of the same sort ; in the
course of a few years, from the same seeds, he obtained
six different forms of pear-tree, in which the fruits
were unlike, while differences also existed in the
general morphology of the plant.
1 Polymorphisme des Vegetaux. Revue Horticole, 1886, p. 209.
2 De la Variabilite de F Espece dans le Poirier. C. R. Acad. des
Sciences, 1863.
ioo EXPERIMENTAL EVOLUTION LECT.
Among flowers the same variability obtains. Colour
may be different, but there are also important varia-
tions of a morphological order, and many botanists
have pointed out the more interesting cases in all
parts of the world. Maxwell Masters has collected
a number of them in his Teratology, and Udo
Dammer has added many in the German edition
of this work; more recently, Dr. O. Penzig, of
Genoa, has collected all known cases anew, in his
important Pflanzen-Teratologie (1890). In this book,
of which only the first half has yet been pub-
lished, we find a very complete list of teratological
cases, of cases of variation in all parts of the plants,
and of every sort, so that I may refer to this book
once for all, as concerns all plant variation. Some
idea of its value may be gathered from the fact that
1 66 large octavo pages are filled up with the mere