EXPERIMENTS ON SELECTION 239



of the variation. On this point I shall merely refer
those who desire more information to Darwin's book
on variation : instances are there most numerous and
convincing, as concerns animals and plants, and are
enough to show the power of selection. I may also
refer to some recently-noticed cases. One con-
cerns the production of a new variety of hornless
oxen. In 1874 a Sicilian farmer noticed among his
herd a young bull which had no horns at all. This
young bull was allowed to mate, and the result has
been the production of other hornless animals, so that,
by selecting at each generation the progenitors of the
following this farmer has obtained a hornless variety.
A similar fact occurred in 1 86 1 in a village of the
Mouse department. A cow gave birth to twins, male
and female, deprived of horns : they were mated
together, and thus, by constant selection of hornless
animals as progenitors, a hornless variety has been
also created. The breed of Mauchamp sheep has
similarly been evolved out of a ram which was born in
1828, in the Mauchamp farm, with the peculiarity of
bearing an even wool, instead of having it frizzled,
merino-like. And M. Cornevin, from whose work I
abstract these facts remarks that any day any
breeder in the south of France can, if he chooses, pro-
duce by simple selection a variety of sheep with four
udders, for these animals often bear four of these



240 EXPERIMENTAL EVOLUTION LECT.



organs. From what we have seen, and that which
has been done, purposely or unconsciously by man,
we may infer that every possible variation may be
permanently consolidated as a normal and fixed
character, through selection, through the choice of
progenitors based on the similarity of the character
which it is required to render permanent.

Some points concerning selection require a passing
notice.

In the first place, scientific investigation being the
only aim, the only point in view, it seems advisable to
undertake the study of the influence of selection be
it on animal or on plant without any particular fore-
thought at all. I mean by this that such investi-
gations should be begun without any view of obtain-
ing a variation and variety in any particular direction.
For instance suppose Lysimachia nummularia
I quote the first name which occurs to my memory,
without the slightest choice among those which
come with it is made the subject of an investigation
in selection. Well, it would not do to decide before-
hand to seek for a new variety having such or such a
peculiarity in the roots, or stems, or leaves : one
should merely cultivate the plants, and if among them
some offered any interesting or curious variation in
any part whatever, one ought to begin the process
of selection, and try to consolidate in the progeny



V EXPERIMENTS ON SELECTION 241

this particular variation. This method offers the
advantage of opening a wider field to investigation,
and on the other hand, animals or plants which,
through culture have been brought to vary in any
direction, are more apt to vary in the one particular
direction which may be desired. Even in cases where
one particular variation is desired in order to create
a special variety, de Vilmorin rightly advises as
follows : " In order to obtain from a yet unmodified
plant, varieties of a determined character, I would
strive at first to make it vary in any manner, and
would choose, as progenitors, not the accidental
variation which would most nearly approach that I
am seeking for, but merely the one which offers the
largest departure from the type. In the second
generation, I would again select as progenitors the
plants which are most different from the normal type,
and, at the same time, from the progenitors selected
in the previous generation." 1 In fact, as Vilmorin
says, we must try to craze the plant, to make it vary
as much as possible, in all possible directions ; and it
is only when this result has been obtained that it
becomes advisable to select the variations in the
desired direction.

We should always remember that plants and

1 Note snr itn Projet d^ Experience ayant pour but de creer ntie
Variete d* Ajonc sans Epines. Loc. cit. p. 35.



242 EXPERIMENTAL EVOLUTION LECT.

animals may vary in a large number of directions, that
variability increases with variation, and that any
desired variation is more apt to occur among plants
whose tendency to variability generally has been
considerably increased through the process which has
been mentioned. And then also, it is well to
remember that while proceeding in this manner,
though we may not meet what we desire, we may
meet with very unexpected variations which may
prove of even greater interest than that which we are
seeking.

In the second place one must not forget that in
experiments of this kind, especially with wild or
uncultivated plants, a long time is sometimes required
before any important variations occur ; the species
seems for a long period to resist all inducements to
variation, and then, all of a sudden, it begins to vary
considerably and in many different directions.

So much for experiments on selection. While
speaking of selection, a word may also be said of the
method which is in some sense exactly the reverse of
selection ; I refer to crossing. While in the course of
selection progenitors are chosen among the animals
which are the most similar to each other, in that of cross-
ing, on the contrary, the progenitors are of different
type, and crossing is performed in order to obtain
animals or plants which combine the character of



v EXPERIMENTS ON CROSSING 243

both parental forms. In natural conditions, crossing
occurs both among animals and plants, although,
generally speaking, among animals at least, there is
a marked tendency against the mating of unlike, and
towards that of similar animals. It has even been
noticed that in the same town I have heard of the
fact in Florence the pigeons congregate in flocks
according to their colour, and keep together, and mate
together, while they seldom do either with the pigeons
of a different flock. In many cases, crossing gives
origin to hybrids which are unable to reproduce their
own form, so that the process is not of much use if
new varieties are required. In other cases, however,
the varieties produced by crossing are fertile inter se,
and it must be noticed, that when crossing does not
seriously impair or totally destroy fertility, it increases
it in a marked manner. It is on account of this
increase of fertility that crossing is often resorted to by
breeders or agriculturists, and this increase has been
demonstrated by Darwin for plants, and observed
by many persons in animals. Cornevin quotes an
instance when a flock of sheep yielding 6/ twin
births had its percentage increased to 1 3 through the
introduction of a ram of different breed. The same
is true of hogs, of pigeons, and other animals.

If crossing is to be used in experiments on Evolu-
tion some points must be particularly attended to,

R 2



244 EXPERIMENTAL EVOLUTION LECT.

Observation has shown that crossing is beneficial in
some cases only, and many persons have spoken
unfavourably of it. This is doubtless due to the fact
that we are yet ignorant of some conditions which are
of high importance in the matter, and it would seem
that facts of apparently little significance have a great
deal to do with the ultimate result. Fertility and
sterility may be induced by very feeble external
agencies, and at the beginning results may be obtained
in crossing, which are quite unfavourable, and quite
unlike those which are obtained later on. On the other
hand some conformity must exist between the plants
or animals between whom the crossing occurs. When
it does exist and, in fact we hardly know when and
where it does or does not the results may be un-
favourable in that the characters of the parents, though
apparent in the progeny, are not blended together, as
was expected ; they are separate, and may be seen side
by side instead of being combined. Such is the case,
for instance, when white-seeded peas are crossed with
green-seeded varieties ; in the progeny green peas arc
met amidst white seeds, some seeds are green striped
with white ; none are between green and white.
Similar instances are met with among animals when
among the progeny of the same parents, some exhibit
the character of the one, and others those of the other
parent. Another point which requires consideration



v EXPERIMENTS ON CROSSING 245

is the fact that of the two progenitors in an intended
crossing the one is generally as a variety possessed
with stronger hereditary tendencies than the other,
and the result is that the progeny takes more after
one of the parents than after the other. For instance,
in the crossing of the Angus bull with the Dutch cow,
the progeny has more characters of the former than of
the latter, and among plants, according to Cornevin,
Vitis rupestris is said to transmit its characters very
markedly when crossed with other varieties.

Such varieties or individuals display what may be
called predominant heredity ; by this is meant that in
crossing the hereditary tendencies of one parent lord
it over those of their mate. This is true of females
as well as of males ; it does not seem to belong to
the one sex more than to the other, and among horses
and oxen Eclipse and Duchess have been renowned
for the predominance of their own qualities over that,
of their mates. Among plants, instances also occur,
and are well known, although in many cases horticul-
turists vainly strive to find individuals thus endowed.
Many have sought to discover Ulex europaeus devoid
of spines, able to transmit this peculiarity to their
progeny this would be a very valuable conquest, as
Ulex europaeus is good fodder but requires to be
crushed so that the cattle do not injure themselves
but none have obtained it yet. And this is but one



246 EXPERIMENTAL EVOLUTION LECT.

instance among a thousand. Why, and how it is that
some individuals are thus endowed, we know not. It
must be noticed that the pre-eminence of one of the
progenitors, as a race or variety is not constant ; while
variety a for instance, possesses stronger hereditary
tendencies when crossed with b and r, its tendencies
are overpowered by those of d in a crossing with this
variety ; and in order to ascertain exactly or more
accurately the relative energy of these tendencies, many
systematic crossings would have to be made. This
pre-eminence of one of the progenitors over the other,
in regard to the character of the progeny may in some
cases be so considerable that no difference is apparent ;
it would seem that no crossing has been made and
that both progenitors belong to the same variety. 1
As M. C. Dareste 2 rightly observes, " in the present

1 It may happen, however, that the difference, as concerns external
characters, is well marked, but then one marked physiological feature
may be transmitted from one of the progenitors to the hybrid ; for
instance, M. Millardet, in his Essai sur F Hybridation de la Vigne
(Paris, Masson, 1891), shows that in some hybrids between two varieties
oi grape-vine, the marked immunity of one of the progenitors the
male especially towards phylloxera, is transferred to the progeny,
notwithstanding the reverse tendency in the other progenitor. The
same writer notices that when the American variety is used as male,
immunity towards phylloxera is considerable, while the amount of
fruit produced is smaller ; the reverse obtains when the American
variety is used as female. Similar facts have been observed by Th.
Niebnei, Die Rose, Berlin, 1880.

2 Noiivdle Exposition d'un Plan d' Experiences sur la Variabilite
des Aniiiiaux, p. 16 (Bull. Soc. Zool. Acclimatation, 1888).



v FORMS OF HEREDITY 247

condition of science it is nearly impossible to foresee

what will be the results of a crossing But we

may assert that this method will yield very interesting
results, and all the more so that the crossed varieties
will be more dissimilar."

This is quite true, and we cannot wonder at it when
we consider how many different forms heredity is
able to assume. Accepting Cornevin's recent classi-
fication, we recognise the following modes :

1. Predominant or Unilateral Heredity ', where the
hereditary tendencies of one of the two mates appear
predominantly or exclusively in the progeny.

2. Bilateral Heredity. Here the progeny has
characters of both progenitors ; and it seldom, if ever,
happens that the progeny has both sorts of characters
to the same degree, or in the same proportion ; one of
the progenitors has more influence than the other.
Both sorts of characters may fuse and combine
together; white and black begetting gray for instance ;
or may coexist, remaining separate, the progeny
having some traits of the father, and others of the
mother. In this sort of heredity four cases are possible :
Heredity is Direct or Crossed, Equal or Unequal.
These terms require no explanation.

3. Atavistic Heredity, or Retrogression, which may
also be direct, crossed, or collateral. In this case the
predominant characters in the progeny are charac-



248 EXPERIMENTAL EVOLUTION LECT.



ters belonging, not to the progenitors but to their
ancestry, near or remote. Many anomalies in organ-
isation are but reversions to former ancestral types ;
and reversions exist to a less degree in the many
cases when a child, for instance, resembles his
grandfather or great-grandmother by some marked
peculiarity instead of possessing the traits of one of
his direct progenitors.

4. Indirect atavistic Heredity ', or Heredity tliroiigJi
Influence. An instance will explain this form. Lord
Morton caused a mare to be mated with a quagga ; she
gave birth to a striped hybrid. Mated the following
year with a thoroughbred, her progeny was again
striped, although not hybrid ; and this occurred three
following years, although she was but once mated with
the Quagga. The influence of the latter and we do
not know what we exactly mean by influence in this
case had incorporated itself as it were, in the mare's
organism so deeply as to make itself felt even three
years after. Such cases are met in human marriage.
They are, as yet, unexplained, but they are occasion-
ally met with, and positively ascertained.

5. Homochronoiis Heredity. This term applies to
cases where psychical or physical peculiarities put in
their appearance among the progeny, at the age when
they appeared among the progenitors. Many patho-
logical tendencies are transmitted after this mode, in



v FORMS OF HEREDITY 249

man as well as in animals, and this form of heredity
may be direct as well as crossed.

6. Re inverted Heredity. The progeny may, at first,
and during some time, closely resemble one of the
progenitors, and later on, resemble 'the other.

7. Homotopic Heredity. In this case, a given
peculiarity of one of the progenitors is met with in
the progeny exactly in the same part of the body as
in the former ; for instance, a differently coloured
lock of hair, or a naevus maternus similarly placed in
progenitors and progeny. This form may be also
direct or crossed.

8. Heterotopic, or Homo/list Heredity. Peculiarities
in one part of a progenitor may be transmitted to
other parts (made of similar tissue) in the progeny ;
the same disease, for instance, may assume one form
in progenitors, and another in the descendants.

The natural sequel to experiments in- crossing will
be experiments in hybridisation, that is, crossing
between distinct species. New hybrids must be
obtained, and the number of those which are fertile
inter se must be largely increased. And then, will the
latter hybrids form permanent varieties or species ?
Experiment only can decide. And while experiment
and artificial impregnation may go a long way in
creating entirely new and unexpected forms of life,
which may be of great interest, particular attention



250 EXPERIMENTAL EVOLUTION LECT.

must be paid to the production of hybrids under natural
conditions. This study of hybridisation, scientifically
conducted, will certainly yield very useful results in
being conducive to experiments concerning that very
marked variability in the reproductive function which
in so many cases determines sterility where fertility
might have been expected, while fertility occurs where
sterility would have seemed natural. It seems that
the process of impregnation is of the most delicate
sort, and very slight circumstances slight they seem
at least determine its success or failure. Investi-
gation of these circumstances cannot but prove
profitable.

Many experiments may also be performed in regard
to sexual and physiological selection, which are con-
nected with those on crossing and hybridisation, and
Mr. Romanes has already suggested some.

Such are, briefly stated, the methods of experimental
transformism. I sincerely hope and expect that some
years hence, perhaps when all of us shall have become
things of the past, the lecturer who shall have
assumed the pleasant task which I here fulfil, will
have much more to say on this topic, and especially
will be able to say : " this and that have been done,"
instead of the " this should be done " which is the
somewhat monotonous conclusion of each of these
lectures.



v AIMS OF EXPERIMENTAL EVOLUTION 251

And now, what are the aims of experimental
transformism ?

As stated, from the beginning, we wish to test the
theory of evolution, as applied to living beings. This
is the main object : but I wish to call your attention
also to the practical and utilitarian results which may
be expected and attained. As you all know, most, if
not all, of our garden vegetables and plants are the
result of man's industry, and in the cases where the
original wild form still exists, we can well measure the
distance between Nature's product and the perfected
form shaped by man, and best adapted to his needs.
And if man had not undertaken the task of bettering
bettering meaning here simply adapting more ade-
quately to his needs Nature's work, many of our im-
portant foods would not have existed. 1 The same is
true of animals. Compared with our domestic animals
the wild forms sink in insignificancy as concerns useful-
ness, and here again man has been a most potent
factor in creating out of the raw material those useful

1 Cf. Asa Gray : Were the Fruits made for Man or did Man make the
Fniits? (American Natural, vol. viii., p. 116.) The veteran botanist
here showed that while some fruits and vegetables have hardly departed
from their original wild type (huckleberries, cranberries, persimmons,
etc. ), others have been bettered and rendered more useful by cultivation
(currants, gooseberries, raspberries, blackberries, chestnuts, straw-
berries), while others have been so much perfected as to represent new
fruits (apple, pear, peach, etc.). He adds a list of fruits which man
should endeavour to render more perfect for his own needs.



252 EXPERIMENTAL EVOLUTION LECT.



allies which provide him with meat, milk, wool, or
energy. But, since man has been able to do what he
has done, while devoid of knowledge and unconscious
of methods, what may he not expect to perform now,
with larger means, more varied resources, many
centuries of accumulated science, and the certainty of
success if he is persevering enough ? The past is a
promise for the future, and the results already obtained
well show what we may attain to ; the whole thing
rests in our hands. Are we to believe that among
the unnumbered species of animals and plants which
are yet living in the wild state, none remain which may
be cultivated or domesticated ? Has all been done that
was possible ? No naturalist would venture to answer
yes, and to say that man has reached his Pillars of
Hercules. We are then entitled to expect many
useful discoveries if we only set to work, and the field
which lies open to us is infinite. Both organic king-
doms may be made to yield any number of new forms
whose use we cannot even foresee, so short-sighted
are we ; and even if we were only to increase the
number of the useful animals or plants, without varying
the use to which either may be put, a great deal would
doubtless have been done for the benefit of mankind.

But this cannot happen if we do not purposely set
to work. In former times, when man lived in small
or large tribes, widely separated, without easy means



v AIMS OF EXPERIMENTAL EVOLUTION 253

of intercourse, it was necessary for him to attend at
all costs to his cultures and herds ; it was a matter of
life or death. Now, our most civilised nations are
unable to produce all they need ; some work in one
line, some in others, and by cooperation and exchange
every one is provided with the things he needs, and
none feels the necessity of creating new resources.
But civilised man must be made to understand that he
can considerably increase the latter if he chooses. Such
is the practical interest of experimental transformism,
and it can in no possible manner interfere with its
scientific aim ; both are bound together.

As to the latter purpose, I must be content with a
few words on the principal lines of investigation. Our
main aim must be the study of evolution, that is of
the derivation of living forms from each other, and
the study of natural influences on the process of this
derivation. If the evolution hypothesis is true, we
must find that new forms of life may be evolved out of
pre-existing forms, by means of influences actually
operative in Nature, without man's or any other agent's
interference. Of course, in experiments on evolution,
man must interfere, but he does so only in order to
test the real efficiency of what he considers to be the
factors of evolution.

This general line of investigation is not a simple one ;
a large number of questions are intimately connected



254 EXPERIMENTAL EVOLUTION LECT.

with it, and cannot be separated, concerning which
information will be most valuable, from the practical
as well as from the scientific point of view. Some of
these questions may be briefly indicated.

First the question of variability in species, or other
groups, and how they are established. External and
internal or physiological variability must both be in-
vestigated anew, with the most delicate tests and
methods, and particular attention must be directed to
their causes. Of course, this study implicates that of
species. What is a species ; what are specific charac-
ters ? If we consider many of the recognised species,
we see that these characters are often of the most
insignificant sort, and that many of the so-called
specific characters are even of less value than those
which are used to distinguish varieties. It may be
predicted that terrific discussions will arise concerning
this vexatam questionem, which seems to become more
intricate every day, and that much sorrow will befall
that numerous and well-disposed class of systematists,
whose self-assumed task in life seems to be to in-
crease the list of specific forms. Perhaps we shall
thus understand what really makes a species ; for
while we talk much about them, we really do not
understand what they are, and no thorough definition
has yet been given.

The problem of heredity will also be investigated,



v AIMS OF EXPERIMENTAL EVOLUTION 255

What is heredity ? how does it operate ? what is
transmitted ? We hardly know anything on the
matter ; we all have read a number of anecdotes of
more or less unscientific character, and remain in the
dark. Weismann's solid and heavy essays are cer-
tainly valuable, but facts must be added to reasonings,
and the facts we need are mainly experimental. The
whole problem requires a thorough investigation, al-
though many facts are already ascertained, a great
deal remains to be done to explain heredity, and
to ascertain its limits and power. One important
question is that of the heredity of Abnormalities and
Mutilations. Many abnormalities, when not opposed
to the continuation of life, are hereditary. 1 A good
instance is provided in the case of the cats, mentioned
by E. B. Poulton (^^^,1883-6), and in the case of the
Fodli tribe, in Arabia, where all the individuals, since
very ancient times, have been born with twenty-four
digits instead of twenty, and where they all marry
within the tribe. 2 Many diseases are hereditary, under

1 Paul Bert must be included among those who have somewhat
investigated the subject, after Philipeaux, Bronn, and many others.