demonstrably of high import. 1 And, as we shall see
further on, variations in colour cannot be considered
as mere freaks of nature, however abundant they may
be, for where colour varies, there is also a more or
less pronounced variation in other characters, and
more especially in some interior and less easily
appreciated characters of chemical nature. And no
one can dispute the import of chemical characters,
when one knows the influence of chemical media on
most organisms. So, while recognizing with Linnaeus
that colour is certainly in many cases a very variable
character, I would refrain from repeating after so many
others nimium ne crede colori. For colour is a specific
1 Concerning the uses of colour see especially A. R. Wallace's
Danvinism, which contains an excellent account of the matter.
E
So EXPERIMENTAL EVOLUTION LECT.
character, or at least it -must be one to most of the
present systematic naturalists (though I doubt if this
state of things is eternal), and as evolutionists, we
cannot allow this character to be lightly disposed of,
when it is precisely one of the most variable in some
cases. 1 In some cases, I repeat, not in all. For
while some instances of colour variation may be
observed in the state of nature among animals or
plants, while some special variations are more especially
met with, such as albinism and melanochroism, it must
be noted that colour variations are especially frequent
among cultivated plants and domestic animals, and are
thus due to the results of changing environment. We
do not always perceive how far there may be a change
in environment, but colour variations show that it
exists in many cases where we do not readily detect it.
Among animals in their natural condition, colour
variations are of no very rare occurrence. It is known
that the common fox in the same country offers
marked variations in colour, which are illustrated by
the different names which have been conferred upon
the principal forms ; Vulpes alopex, melanogaster, and
crucigera. The beaver also offers important colour
1 Of course facts concerning colour variation are to be found in a
large amount of works. But I would recommend, concerning colour
variation in insects, two recent works. The one is Mr. S. H. Scudder's
magnificent work on the Butterflies of New England ; the other is the
Entomologist' s Record and Journal of Variation.
COLOUR-VARIATION
variations, and its fur is in some cases of much lighter
colour, in others, of deeper. Colour variations are of
no scarce occurrence among insects and fishes, and in
a recent number of the Entomologists Record and
Journal of Variation - 1 an interesting coloured plate
may be seen, illustrating more vividly than whole
volumes of description, the colour variations which
Mr. J. A. Clark has met with among the British
species of SmerintJius.
Lacordaire records similar facts concerning the
Sphinx elpenor, Audouin has some concerning Pyralis
vitis, and Duges concerning Phasma. Among insects,
again, Hulst has noticed a large amount of variation.
From one and the same Arctia excelsa he has obtained
a number of eggs and larvae which have yielded adult
butterflies belonging to eight or nine different varieties
or, to speak correctly, possessing eight or nine
different specific names (Arctia phalerata, pallida,
phyllina, flammea, decorata, nais, etc.). Of course this
merely shows that the makers of these species were
wrong in establishing species where mere varieties
exist if even varieties may be spoken of in this case
but does this not show also that variation may be
very important ? 2 The common cray-fish is a well-
1 Edited by J. W. Tutt, London. March i6th, 1891.
2 Cf. Hulst: Variation in the Arctias. American Naturalist, 1884,
P- J 93-
E
52 EXPERIMENTAL EVOLUTION LECT.
known instance of colour variation, being generally
brown, but sometimes blue, and even red in its
living state. Leeches offer a large amount of colour-
variation ; most Helices do the same, and in fact, it
may be said that in all groups of animals varia-
tions are met with in the colour of their garment.
I refer here merely to occasional variations, for it is
well known that a large number of mammals, birds,
and other animals offer periodical or seasonal colour
variations, especially in northern climates, being
brown or grey during the summer, and becoming
white during the winter. Such seasonal variations
Wallace, in his recent and excellent book on
Darwinism, ascribes to natural selection and to pro-
tective necessities. Very numerous instances thereof
might be adduced ; and Godron in his DcFRspeceet des
Races dans les Rtres organises (1859, two volumes), gives
a list which might be extended of course of the
mammals and birds and other animals which show
this seasonal variation, and also a list of animals which
offer instances of albinism, melanism, and crythrism.
But I maybe allowed to refer to Wallace's Darwinism
for all seasonal colour variations, and for the investi-
gation of the use and origin of colour generally.
As the last named cases of colour variation, such as
albinism and melanism, cannot be interpreted in a
quite satisfactory manner, we had better leave them
COLOUR-VARIATION 53
out. That which most interests us, so far as colour
variation is concerned, is the evidence showing
that a change of environment causes a change in
coloration. Some instances may be adduced : for
instance, Gerard states in the Dictionnaire dHistoire
naturelle of D'Orbigny 1 that when the small brown
honey bees from High Burgundy are transported into
Bresse although not very distant they soon become
larger and assume a yellow colour ; this happens
even in the second generation. The same author
gives some instances from the vegetable kingdom.
As he rightly remarks, the roots of beet, carrot,
radish, and other plants, are colourless in the wild
and natural state, and as soon as they are subjected to
the process of culture they become red, or yellow, etc.
and Vilmorin in his Notice sur V Amelioration de la
Carotte sauvage, originally published in the Transac-
tions of the Horticultural Society (1840), has noted the
same fact, the red and yellow colours, as well as a
peculiar violet hue which has not been permanent,
appearing only in cultivated carrots after some time
their appearance being at first irregular and transitory.
Moquin Tandon 2 records some instances of change
in colour which are due to the influence of environ-
mental change. For instance, he has seen gentians
which are blue in valleys become white in the
1 Article Espcce. 2 Elements de Ttmtologie vcgetale.
54 EXPERIMENTAL EVOLUTION LECT.
mountains. Similarly Oxytropis montana and Trifo-
lium pratense are white in the Alps and Pyrenees, and
Geranium batrachoides, which is commonly bluish,
becomes variegated, and turns generally white when
it grows in unpropitious soil. There are white
varieties of many plants, such as Latnium purpureum
and Erica vulgaris, while Verbascum lycJinis, and
Campanula Trachelium bear flowers which are blue,
violet, or white, according to circumstances.
Such instances might be given by hundreds, as is
well known. In some cases it would seem that the in-
fluence of environment is very plain, although difficult
to explain, for there are places where some natural
colours of plants or animals disappear soon, and are
replaced by lighter tints, or in many cases by white.
M. d'Apchier de Pruns records the fact as having been
noticed by himself on his own land, and it seems that
at Brassac les Mines, in central France, while oxen
become of lighter hue, and pheasants, pigeons, ducks,
&c., have more or less white feathers, plants with
variegated leaves soon become uniformly green. 1 And
some horticulturists and amateurs have complained,
similarly, of their garden or grounds, saying that they
find it impossible to keep variegated plants for
all return to the ordinary type. The causes of these
facts are difficult to ascertain, as the circumstances
1 Revue Horticole^ 1883, p. 316.
II COLOUR-VARIATION 55
which determine variegations are themselves not
known ; but the facts are numerous, well authenti-
cated, and must be taken into account. Climate
certainly has some influence on the colour of flowers,
and although we do not exactly know yet what we
mean when we speak of differences of climate, as
Naudin aptly remarks, and though climate includes a
large number of very different factors which are
combined in different proportions according to
localities, there are influences which may be ascribed
to it, /;/ toto. G. Bonnier and C. Flahault have per-
formed interesting experiments on this point. G.
Bonnier 1 has compared flowers of the same species
and age, from different altitudes, in the Austrian Alps
and Carpathians, and the result has been that while
some plants, such as Rosa alpina and Erigeron alpinus,
have the same colour at different heights, others are
slightly different : such is the case with TJiymus
serpyllum and Geranium sylvaticum ; others are very
different, such as Myosotis sylvatica, Campanula ro-
tundifolia, Ranunculus sylvaticus, Galium cruciatum.
Of course the colour is not radically changed
a pink flower does not become yellow, but it
grows deeper and richer in plants of higher altitude.
Microscopical investigation shows that the pigment
1 De la Variation avec F Altitude des Matieres colorees dcs Fleurs chez
line menu Espcce vegelale. Bull Soc. Botanique, 1880, . 103.
56 EXPERIMENTAL EVOLUTION LECT.
granules are more numerous in the flowers from high
altitudes. C. Flahault's 1 experiments are more con-
clusive, and the conditions under which they have
been performed are more satisfactory. His experi-
ments have been made on plants grown in Upsala
and in Paris from seeds of same origin. One half of
the Parisian seed has been sown in Paris, and the
other in Upsala ; one half of .the Upsala seed has
been sown in Paris, and the other in Upsala. With
the two experiments the result has been the same, the
flowers have always been more vividly coloured in
Upsala than in Paris, and the same holds good when
flowers of plants spontaneously growing around Paris
and around Upsala are compared. In some cases,
however, there is but a very slight difference. M.
Flahault has had the exact colours represented in his
paper, and the comparison of the Upsala and Paris
flowers is thus shown to the reader as if he had the
flowers themselves.
Concerning colour variation in animals, I must be
content with calling attention to some principal facts.
One is, that while animals in their natural wild
state offer but very slight colour variations in the same
region, these variations become very numerous under
1 Nonvelles Observations sur les Modifications des Vegetanx suivant
les Conditions physiques du Milieii. Annales des Sci. Nat. (Bot.) t. ix.
1880, p. 159.
COLOUR- VARIATION 57
domestication. Of this, horses, oxen, cats, rabbits,
guinea-pigs, &c., are instances. And this may be
explained by natural selection, at least if colour
is always of positive use in some way or other to
animals, in escaping dangers which are of daily
occurrence in the wild state of life, but which dis-
appear under domestication. Under domestication
colour variations, to which a more or less marked
tendency may always exist, are of no inconvenience,
unless positively repelled by artificial selection, and
thus such variations are often present. On this point
I may refer to the works of Darwin and Wallace.
Another fact to be taken into account is that of
the influence of food on colour. Many bird-fanciers
think that by appropriate colour-feeding, as they call
it, they can help the production or intensification of
colours. For instance, they believe that canary birds
can be made to become of a bright yellow when fed
with egg, mustard seed, curcuma powder, saffron
water, and alcohol, in definite proportions ; they even
consider it useful to put yellow flowers around the
bird's cage. But exact experiments, scientifically
conducted, are yet wanting on this subject, as I
have but second-hand and rather untrustworthy in-
formation concerning the investigations conducted by
Dr. Sauermann, which are alluded to in the previous
sentence.
58 EXPERIMENTAL EVOLUTION LECT.
A third fact, which must be noticed here, is the posi-
tive influence of the colour of environment on that of
animals. Mr. E. B. Poulton has recently studied this
matter, showing, in his important memoir published
in 1887, that many lepidopterous larvae are strongly
affected by the surrounding colour. The plates
which accompany his memoir illustrate the fact
very precisely.
If we consider those freaks of colour which are
familiar to horticulturists under the name of variega-
tions, some interesting facts may be noticed. It has
been often questioned whether variegations are not
pathological symptoms, and whether variegated plants
are not more or less diseased. M. E. A. Carriere * has
carefully considered the matter, and his considerable
horticultural experience does not make him feel in-
clined to consider variegated plants as being diseased
at all. It even seems that in many cases, variegated
forms are healthier and stronger than the non-
variegated : for instance, the variegated Euouymus of
the Duke of Anjou variety. There is however one
fact which has been noticed concerning variegations
it is the impossibility of maintaining such plants with
their variegation in some localities. Many horticul-
turists have recorded the fact ; and while some com-
1 Lcs Panachures sont dies des Maladies? Revue Horticole, 1884,
p. 198.
ii COLOUR-VARIATION 59
plain that all variegated plants, when grown on their
grounds, soon revert to the ordinary type, even when
they belong to the most stable varieties, others notice
that their garden seems very propitious to the produc-
tion of variegations. There is some unknown influence
at work in these cases, and experiments might show
what it is.
But it certainly seems that variegated plants cannot
be considered as diseased, and M. Lebas x says posi-
tively that Euonymus sulfurea, Euonyinus radicans
variegata, and Thujopsis dolabrata variegata are cer-
tainly stronger and hardier than the common non-
variegated varieties; and, on the other hand, MM.
Carriere and Andre 2 notice that while Aspidistra
clatior variegata has a strong tendency, in most
places, to revert to the non-variegated type, there
are places where it remains quite constant, and where
even non-variegated forms become spontaneously
variegated. In some cases variegation comes on
slowly, and Vilmorin 3 has studied the process with
care, but in others it comes on all of a sudden.
Carriere 4 has noticed a case of this sort in a garden
where thousands of celery plants were growing, and
1 De qtielqttes Fusains du Japan a Feuilles panachees. Rev.
Horticole, 1872, p. 139.
-' Revue Horticole, 1888, p. 124.
;i Stir les Panachures des Fleiirs. Ibid. 1852, p. 128.
4 Panachurc dn Ccleri. Rev. Horticole, 1882, p. 541.
60 EXPERIMENTAL EVOLUTION LECT.
where all, in a more or less marked degree, at the
same time became variegated. This fact, with others
which might be quoted, goes to show that variegations
depend on some environmental influence.
Colour variations may, however, be noticed in cases
where no environmental influence can, as yet, be
traced. Every one has seen cases where the same
rose-bush yields flowers of dissimilar colours. Carriere
and Andre, 1 to take an instance among many, have
noticed a rose of the Mabel Morison variety carrying-
white flowers and one single pink one. The branch
bearing the pink flower has been grafted on another
bush, and it has maintained its special character,
yielding always pink roses. Such cases are not of
rare occurrence. But how can we understand the
cause of this variation ? Environmental influence
seems out of the question, and we are at a loss to
account for this important variation.
Similar colour variations are often noticed in fruits,
and have often been recorded in connection with
grapes. Carriere has quoted a case of this sort, and
given a good coloured plate showing well how things
stand. In the same bunch of grapes some are black
or red, some colourless, and many variegated in
different manners. It may seem that these cases are
3 C<zs dc Dichroisme dans la Floraison tfun Rosier. Rev. Horticole,
1888, p. 74.
II COLOUR-VARIATION 61
of slight importance, and that I am losing much time
in bringing them before you. Truly, if we were to
view the things in the light of Linnaeus nimiinn
ne crede colori such would be the case ; but recent
facts have gone to show that colour variation is not
merely what it seems to be, a variation in pigment
deposits ; there are more important variations which
underlie or accompany it, and these are of much
interest in showing that colour variation is of greater
import than might be thought at first glance. These
variations are of chemical order, and, so far as I know,
little has yet been done to investigate this delicate
matter beyond a short but excellent paper by Pro-
fessor Armand Gautier. 1 The author begins by re-
calling some instances which have a remote relation
to this matter, and some of which are familiar to all.
It is known, for instance, that when a mare has once
given birth to a mule, it may happen that in after years
her foals, produced after the usual impregnation from
a stallion, present peculiarities which belong to asses.
It seems that the single fecundation by the ass has in
some manner impregnated the whole maternal organ-
ism upon which it has stamped itself. Similar cases
are to be observed in the human race as well as
among horses or dogs, and these show that a foreign
1 Du Mecanisinede la Variation des Etresvivants, dr-'<:., \\\Hommage
a Monsieur Chevreul a V Occasion de son Centenaire. F. A lean. 1886.
62 EXPERIMENTAL EVOLUTION LECT.
influence, which determines no variation at all in the
maternal organism, has, however, effected a deep
change on the germ-cells, and has operated on an
important part of the organism. Instances of the
same sort may be met with among plants. It
happens, for instance, that after a branch of a varie-
gated variety has been grafted on a non-variegated
plant, some variegated branches sprout from the
latter. And this shows, as M. Armand Gautier con-
tends, and as all naturalists must admit, that race-
variation, or, generally speaking, variation of any
sort, or of any importance, is not a mere external
fact a mere external modification but that the
modification makes itself felt in the utmost depths
and intimacy of the cells. Otherwise stated, there
is not only a mere difference of form ; underlying
the formal or external difference, there are modifi-
cations of much greater importance in the chemistry
or physiology of the cell-plasma. These differ-
ences may be localized instead of being general.
For instance, the same orange-tree may bear oranges
and lemons, if some flowers have been impreg-
nated with pollen from lemon-trees ; and it may
happen even that the variation is more localized
still when one half of the fruit is orange and the
other lemon when, as Naudin has seen, one half is
Datura stramonium, the other D. lacvis. Of course,
CHEMICAL VARIATION 63
by reasoning we also come to the conclusion that
there must be physiological or chemical differences
accompanying the anatomical or morphological varia-
tions. ^ But Prof. Gautier has tried to appreciate
these differences, and to measure them in some
way. Notwithstanding the Linnaean axiom con-
cerning colour, and the fact that many consider
colour variation as of very little import because it is
so frequent, Gautier has studied with great care the
intimate changes which accompany colour variation.
The common grape has been selected by him as
his subject, all varieties of grape being varieties
of one common stock, however different they may be
in many respects ; and after a careful investigation of
the colouring matters, he has come to the conclusion
that, although they all belong to the same type, and
have all been considered as one and the same, they
offer important differences. For instance, while some
of the pigments are soluble in water, others are not ;
while some yield green precipitate with lead salts,
others yield a blue one ; while some contain nitrogen,
others do not. For instance, again, grapes of the
Carignan variety contain a colouring matter of the
type C 21 H 20 O 10 , and another which contains C 22 H 24 O 10 .
The Grenache variety contains a different matter,
C, 3 H 22 O 10 ; Aramon contains C 23 H 18 O 10 ; Teinturier
answers to C 22 H 20 O 10 , Petit-Bonschet to C^H^O^,
64 EXPERIMENTAL EVOLUTION LECT.
and Camay to C 20 H 20 O 10 . This shows that the slight
external differences which are peculiar to each sort or
variety, derived from the common stock by selection
and culture, arc accompanied by important chemical
variations. The above-mentioned differences illus-
trate the fact. It must be added, also, that other
differences are to be met with differences in the
amount and nature of sugar, tannin, &c.
These results determined Prof. Gautier to investi-
s gate an important matter which is closely related
to the foregoing topics, and he has examined the
colouring-matter of hybrid forms. For instance,
Petit-BouscJiet is the result of the impregnation of
Teinturier by Aramon pollen ; and the question to be
solved is the following : Is Petit- Bouscket colouring-
matter identical with that of one of its parents, or
intermediate between both, or is it entirely different ?
Direct analysis shows that the second supposition
is true. Petit-Bonschet colouring-matter is nearly
exactly intermediate between those of its parents,
and its composition is the arithmetical mean between
those of the two stocks from which it is derived.
Another important fact which has been pointed out
by Gautier is that all colouring-matter which may
be found in grapes, when subjected to the influence
of potash, splits into two sorts of chemical com-
pounds ; two of them are constant, and invariably
CHEMICAL VARIABILITY 65
met with phloroglucin and protocatechuic acid ; the
other is variable, and consists of potash united to
some fatty acid, so that while in all cases phloro-
glucin and protocatechuic acid are present, the third
element varies, and these variations are the cause
of the peculiarities which characterize each grape
colouring-matter ; or at least, if they are not the
cause, they are concomitant phenomena. Similar
facts have been ascertained by Prof. Gautier con-
cerning a chemical substance which is very abundant
in nature, and which would seem to be constant and
invariable. Chlorophyll, according to his investiga-
tions, is not identical among Acotyledons, Monoco-
tyledons, and Dicotyledons ; there are marked chemical
differences, marked variations in chemical constitu-
tion, and even between two species of two families
(Spinach and Mallow) there is a difference. Of
course these variations are not very intense, and the
different sorts of chlorophyll are always homologous,
but they are not identical. Similar instances are to
be met with among animals. While we commonly
speak of albumen, ossein, syntonin, and other com-
pounds of animal tissues, as being constant and
identical, direct analysis and investigation show that
all these substances, when considered among different
animals, do present common characters, which are
permanent and ubiquitous ; all possess also special
F
66 EXPERIMENTAL EVOLUTION LECT.
features which distinguish them from the similar
compounds. Albumen is not the same in all eggs,
for instance, in those of the hen, turkey, and duck ;
and the flesh of fishes is different from that of
mammals or birds. It even follows from Prof.
Gautier's experiments that there are differences in the
same animal at different times of life, and differences
according to the mode of life and nutrition. For
instance, if the meat of oxen, fed in the ordinary
manner upon hay and grass in the pasture, is com-
pared with that of oxen which have been fed and
rapidly fattened with the refuse of beet-sugar fac-
tories, as is often done in France and Germany, there
is not only a difference in the taste and flavour of the
steaks or roasts, there is also a chemical difference
which is easily detected. While the flesh of the
pasture-oxen rapidly dissolves in water and hydro-
chloric acid, that of the oxen fed on beet-refuse