TRANSCRIBER’S NOTES:
—Obvious print and punctuation errors were corrected.
—Bold text has been rendered as =bold text=.
AN EXAMINATION
OF
WEISMANNISM
Oxford
HORACE HART, PRINTER TO THE UNIVERSITY
[Illustration: AUGUST WEISMANN]
AN EXAMINATION
OF
WEISMANNISM
BY
GEORGE JOHN ROMANES
M.A., LL.D., F.R.S.
HONORARY FELLOW OF GONVILLE AND CAIUS COLLEGE, CAMBRIDGE
London
LONGMANS, GREEN, AND CO.
1893
PREFACE
AS already stated in the Preface to the second edition of _Darwin and
after Darwin_, Part I, severe and protracted illness has hitherto
prevented me from proceeding to the publication of Part II. It is
now more than a year since I had to suspend work of every kind, and
therefore, although at that time Part II was almost ready for press,
I have not yet been able to write its concluding chapters. Shortly
before and during this interval Professor Weismann has produced his
essays on _Amphimixis_ and _The Germ-plasm_. These works present
extensive additions to, and considerable modifications of, his previous
theories as collected together in the English translation, under
the title _Essays on Heredity, Vol. I_. Consequently, it has become
necessary for me either to re-write the examination of his system which
I had prepared for Part II of my own treatise, or else to leave that
examination as it stood, and to add a further chapter dealing with
those later developments of his system to which I have just alluded.
After due reflection I have decided upon the latter course, because
in this way we are most likely to obtain a clear view of the growth of
Weismann’s elaborate structure of theories—a view which it is almost
necessary, for the purposes of criticism, that we should obtain.
Having decided upon this point, it occurred to me that certain
advantages would be gained by removing the whole criticism from the
position which it was originally intended to occupy as a section of my
forthcoming volume on the Post-Darwinian period. For, in consequence
of the criticism having been written at successive intervals during
the last six or eight years as Professor Weismann’s works successively
appeared, it has now swelled to a bulk which would unduly encumber
the volume just mentioned. Again, the growth of Professor Weismann’s
system has of late become so rapid, that if the criticism is to keep
pace with it in future, the best plan will doubtless be the one which
it is now my intention to adopt—viz., to publish the criticism in a
separate form, and in comparatively small editions, so that further
chapters may be added with as much celerity as Professor Weismann
may hereafter produce his successive works. Lastly, where so much
elaborate speculation and so many changes of doctrine are concerned, it
is inevitable that some misunderstandings on the part of a critic are
likely to have arisen; and therefore, should Professor Weismann deem it
worth his while to correct any such failings on my part, the plan of
publication just alluded to will furnish me with the best opportunity
of dealing with whatever he may have to say.
It must be understood, however, that under the term “Weismannism” I
do not include any reference to the important question with which the
name of Weismann has been mainly associated—i.e., the inheritance or
non-inheritance of acquired characters. This is a question of fact,
which stands to be answered by the inductive methods of observation
and experiment: not by the deductive methods of general reasoning.
Of course Professor Weismann is fully entitled to assume a negative
answer as a basis whereon to construct his theory of the continuity of
germ-plasm; but no amount of speculation as to what the mechanism of
heredity is likely to be if once this assumption is granted, can even
so much as tend to prove that the assumption itself is true. Therefore,
in this “examination of Weismannism” I intend to restrict our attention
to the elaborate system of theories which Weismann has reared upon his
fundamental postulate of the non-inheritance of acquired characters,
reserving for my next volume our consideration of this postulate itself.
Lest, however, it should be felt that “an examination of Weismannism”
in which the question of the transmission of acquired characters is
omitted must indeed prove a case of Hamlet without the Prince of
Denmark, I may be allowed to make two observations. In the first
place, this great question of fact is clearly quite distinct from
that of any theories which may be framed upon either side of it.
And, in the second place, the question was not raised by Weismann. It
appears, indeed, from what he says, that he never caught a glimpse of
it till about ten years ago, and that he then did so as a result of
his own independent thought. Moreover, it is perfectly true that to
him belongs the great merit of having been the first to call general
attention to the subject, and so to arouse a world-wide interest with
reference to it. But to suppose that the question was first propounded
by Weismann is merely to display a want of acquaintance with the
course of Darwinian thought in this country. As far back as 1874 I had
long conversations with Darwin himself upon the matter, and under his
guidance performed what I suppose are the only systematic experiments
which have ever been undertaken with regard to it. These occupied
more than five years of almost exclusive devotion; but, as they all
proved failures, they were never published. Therefore I here mention
them merely for the purpose of showing that the idea of what is now
called a “continuity of germ-plasm” was present to Darwin’s mind as
a logically possible alternative to the one which he adopted in his
theory of pangenesis—an alternative, therefore, which he was anxious
to exclude by way of experimental disproof. If it be said that no one
could have been aware of this in the absence of publication, I reply
that I think it may be perceived by any one who reads attentively his
chapter on Pangenesis. Moreover, early in the seventies his cousin,
Mr. Francis Galton, published a “Theory of Heredity,” which, as we
shall see in the course of the following pages, presented as distinctly
as could possibly be presented the question of the transmission of
acquired characters, and answered it in almost exactly the same manner
as Weismann did about ten years later. Lastly, as Weismann has himself
been careful to point out, he was likewise anticipated in this matter
by Jäger (1878), and Nussbaum and Rauber (1880).
For these reasons, then, I exclude this question from the following
examination of what I think we ought to understand as distinctively
“Weismannism.”
G. J. R.
CHRIST CHURCH, OXFORD,
_July, 1893_.
CONTENTS
CHAP. PAGE
I. STATEMENT OF WEISMANN’S SYSTEM UP TO THE YEAR 1886 1
II. LATER ADDITIONS TO WEISMANN’S SYSTEM UP TO THE YEAR 1892 28
III. WEISMANN’S THEORY OF HEREDITY (1891) 48
IV. EXAMINATION OF WEISMANN’S THEORY OF EVOLUTION (1891) 86
V. WEISMANNISM UP TO DATE (1893) 117
APPENDIX I.—ON GERM-PLASM 173
” II.—ON TELEGONY 191
AN
EXAMINATION OF WEISMANNISM.
CHAPTER I.
STATEMENT OF WEISMANN’S SYSTEM UP TO THE YEAR 1886[1].
SEEING that Professor Weismann’s theory of heredity, besides being
somewhat elaborate in itself, is presented in a series of disconnected
essays, originally published at different times, it is a matter of
no small difficulty to gather from the present collection of them a
complete view of the system as a whole. Therefore I propose to give
a brief sketch of his several cognate theories, arranged in a manner
calculated to show their logical connexion one with another. And,
in order also to show the relation in which his resulting theory
of heredity stands to what has hitherto been the more usual way of
regarding the facts, I will begin by furnishing a similarly condensed
account of Mr. Darwin’s theory upon the subject. It will be observed
that these two theories constitute the logical extremes of explanatory
thought; and therefore it may be said, in a general way, that all
other modern theories of heredity—such as those of Spencer, Häckel,
Elsberg, Galton, Nägeli, His, Brooks, Hertwig, and De Vries—occupy
positions more or less intermediate between these two extremes.
Therefore, also, we need not wait to consider these intermediate
theories[2].
“When closely analyzed, Mr. Darwin’s theory—or the “provisional
hypothesis of Pangenesis”—will be found to embody altogether seven
assumptions, namely:—
1. That all the component cells of a multicellular organism throw off
inconceivably minute germs, or “gemmules,” which are then dispersed
throughout the whole system.
2. That these gemmules, when so dispersed and supplied with proper
nutriment, multiply by self-division, and, under suitable conditions,
are capable of developing into physiological cells like those from
which they were originally and severally derived.
3. That, while still in this gemmular condition, these cell-seeds
have for one another a mutual affinity, which leads to their being
collected from all parts of the system by the reproductive glands of
the organism; and that, when so collected, they go to constitute the
essential material of the sexual elements—ova and spermatozoa being
thus aggregated packets of gemmules, which have emanated from all the
cells of all the tissues of the organism.
4. That the development of a new organism, out of the fusion of two
such packets of gemmules, is due to a summation of all the developments
of some of the gemmules which these two packets contain.
5. That a large proportional number of the gemmules in each packet,
however, fail to develop, and are then transmitted in a dormant
state to future generations, in any of which they may be developed
subsequently—thus giving rise to the phenomena of reversion or atavism.
6. That in all cases the development of gemmules into the form of their
parent cells depends on their suitable union with other partially
developed gemmules, which precede them in the regular course of growth.
7. That gemmules are thrown off by all physiological cells, not only
during the adult state of the organism, but during all stages of
its development. Or, in other words, that the production of these
cell-seeds depends upon the adult condition of parent cells: not upon
that of the multicellular organism as a whole.
At first sight it may well appear that we have here a very formidable
array of assumptions. But Darwin ably argues in favour of each of them
by pointing to well-known analogies, drawn from the vital processes
of living cells both in the protozoa and metazoa. For example, it
is already a well-recognized doctrine of physiology that each cell
of a metazoon, or multicellular organism, though to a large extent
dependent on others, is likewise to a certain extent independent
or autonomous, and has the power of multiplying by self-division.
Therefore, as it is certain that the sexual elements (and also buds
of all descriptions) include formative material of some kind, the
first assumption—or that which supposes such formative matter to be
participate—is certainly not a gratuitous assumption.
Again, the second assumption—viz., that this particulate and formative
material is dispersed throughout all the tissues of the organism—is
sustained by the fact that, both in certain plants and in certain
invertebrated animals, a severed portion of the organism will develop
into an entire organism similar to that from which it was derived, as,
for example, is the case with a leaf of Begonia, and with portions cut
from certain invertebrated animals, such as sea-anemones, jelly-fish,
&c. This well-known fact in itself seems enough to prove that the
formative material in question must certainly admit, at all events in
many cases, of being distributed throughout all the tissues of living
organisms.
The third assumption—or that which supposes the formative material
to be especially aggregated in the sexual elements—is not so much an
assumption as a statement of obvious fact; while the fourth, fifth,
sixth, and seventh assumptions all follow deductively from their
predecessors. In other words, if the first and second assumptions be
granted, and if the theory is to comprise all the facts of heredity,
then the remaining five assumptions are bound to follow.
To the probable objection that the supposed gemmules must be of a
size impossibly minute—seeing that thousands of millions of them
would have to be packed into a single ovum or spermatozoon—Darwin
opposes a calculation that a cube of glass or water, having only one
ten-thousandth of an inch to a side, contains somewhere between sixteen
and a hundred and thirty-one billions of molecules. Again, as touching
the supposed power of multiplication on the part of his gemmules, he
alludes to the fact that infectious material of all kinds exhibits
a ratio of increase quite as great as any that his theory requires
to attribute to gemmules. Furthermore, with respect to the elective
affinity of gemmules, he remarks that “in all ordinary cases of sexual
reproduction, the male and female elements certainly have an elective
affinity for each other”: of the ten thousand species of _Compositae_,
for example, “there can be no doubt that if the pollen of all these
species could be simultaneously placed on the stigma of any one
species, this one would elect, with unerring certainty, its own pollen.”
Such, in brief outline, is Mr. Darwin’s theory of Pangenesis.
Professor Weismann’s theory of Germ-plasm is fundamentally based upon
the great distinction, in respect of their transmissibility, between
characters that are congenital and characters that are acquired. By
a congenital character is meant any individual peculiarity, whether
structural or mental, with which the individual is born. By an
acquired character is meant any peculiarity which the individual may
subsequently develop in consequence of its own individual experience.
For example, a man may be born with some malformation of one of his
fingers; or he may subsequently acquire such a malformation as the
result of accident or disease. Now, in the former case—i.e., in that
where the malformation is congenital—it is extremely probable that the
peculiarity will be transmitted to his children; while in the latter
case—i.e., where the malformation is subsequently acquired—it is
virtually certain that it will not be transmitted to his children. And
this great difference between the transmissibility of characters that
are congenital and characters that are acquired extends universally as
a general law throughout the vegetable as well as the animal kingdom,
and in the province of mental as in that of bodily organization. Of
course this general law has always been well known, and more or less
fully recognized by all modern physiologists and medical men. But
before the subject was taken up by Professor Weismann, it was generally
supposed that the difference in question was one of degree, not one of
kind. In other words, it was assumed that acquired characters, although
not so fully—and therefore not so certainly—inherited as congenital
characters, nevertheless were inherited in some lesser degree; so that
if the same character continued to be developed successively in a
number of sequent generations, what was at first only a slight tendency
to be inherited would become by summation a more and more pronounced
tendency, till eventually the acquired character might be as strongly
inherited as any other character which was _ab initio_ congenital. Now
it is the validity of this assumption that is challenged by Professor
Weismann. He says there is no evidence of any acquired characters
being in any degree inherited; and, therefore, that in this important
respect they may be held to differ from congenital characters in kind.
On the supposition that they do thus differ in kind, he furnishes a
very attractive theory of heredity, which serves at once to explain
the difference, and to represent it as a matter of physiological
impossibility that any acquired character can, under any circumstances
whatsoever, be transmitted to progeny.
But, in order fully to comprehend this theory, it is desirable first
of all to explain Professor Weismann’s views upon certain other topics
which are intimately connected with—and, indeed, logically sequent
upon—the use to which he puts the distinction just mentioned.
Starting from the fact that unicellular organisms multiply by fission
and gemmation, he argues that, aboriginally and potentially, life
is immortal. For when a protozoon divides itself into two more or
less equal parts by fission, and each of the two halves thereupon
grows into another protozoon, it does not appear that there has been
any death on the part of the living material involved; and inasmuch
as this process of fission goes on continuously from generation to
generation, there is seemingly never any death on the part of such
protoplasmic material, although there is a continuous addition to
it as the numbers of individuals increase. Similarly, in the case
of gemmation, when a protozoon parts with a small portion of its
living material in the form of a bud, this portion does not die, but
develops into a new individual; and, therefore, the process is exactly
analogous to that of fission, save that a small instead of a large part
of the parent substance is involved. Now, if life be thus immortal
in the case of unicellular organisms, why should it have ceased to
be so in the case of multicellular? Weismann’s answer is, that all
the multicellular organisms propagate themselves, not exclusively by
fission or gemmation, but by sexual fertilization, where the condition
to a new organism arising is that minute and specialized portions of
two parent organisms should fuse together. Now, it is evident that with
this change in the method of propagation, serious disadvantage would
accrue to any species if its sexual individuals were to continue to be
immortal; for in that case every species which multiplies by sexual
methods would in time become composed of individuals broken down and
decrepit through the results of accident and disease—always operating
and ever accumulating throughout the course of their immortal lives.
Consequently, as soon as sexual methods of propagation superseded
the more primitive a-sexual methods, it became desirable in the
interests of the sexually-propagating species that their constituent
individuals should cease to be immortal, so that the species should
always be recuperated by fresh, young, and well-formed representatives.
Consequently, also, natural selection would speedily see to it that all
sexually-propagating species should become deprived of the aboriginal
endowment of immortality, with the result that death is now universal
among all the individuals of such species—that is to say, among all
the metazoa and metaphyta. Nevertheless, it is to be remembered that
this destiny extends only to the parts of the individual other than the
contents of those specialized cells which constitute the reproductive
elements. For although in each individual metazoon or metaphyton
an innumerable number of these specialized cells are destined to
perish during the life, or with the death, of the organism to which
they belong, this is only due to the accident, so to speak, of their
contents not having met with their complements in the opposite sex:
it does not belong to their essential nature that they should perish,
seeing that those which do happen to meet with their complements in the
opposite sex help to form a new living individual, and so on through
successive generations _ad infinitum_. Therefore the reproductive
elements of the metazoa and metaphyta are in this respect precisely
analogous to the protozoa: potentially, or in their own nature, they
are immortal; and, like the protozoa, if they die, their death is an
accident due to unfavourable circumstances. But the case is quite
different with all the other parts of a multicellular organism. Here,
no matter how favourable the circumstances may be, every cell contains
within itself, or in its very nature, the eventual doom of death. Thus,
of the metazoa and metaphyta it is the “germ-plasms” alone that retain
their primitive endowment of everlasting life, passed on continuously
through generation after generation of successively perishing organisms.
So far, it is contended, we are dealing with matters of fact. It must
be taken as true that the protoplasm of the unicellular organisms, and
the germ-plasm of the multicellular organisms, has been continuous
through the time since life first appeared upon this earth; and
although large quantities of each are perpetually dying through
being exposed to conditions unfavourable to life, this, as Weismann
presents the matter, is quite a different case from that of all the
other constituent parts of multicellular organisms, which contain
within themselves the doom of death. Furthermore, it appears extremely
probable that this doom of death has been brought about by natural
selection for the reasons assigned by Weismann—namely, because it is
for the benefit of all species which perpetuate themselves by sexual
methods, that their constituent individuals should not live longer
than is necessary for the sake of originating the next generation, and
fairly starting it in its own struggle for existence. For Weismann has
shown, by a somewhat laborious though still largely imperfect research,
that there is throughout all the metazoa a general correlation between
the natural lifetime of individuals composing any given species and
the age at which they reach maturity, or first become capable of
procreation. This general correlation, however, is somewhat modified
by the time during which progeny are dependent upon their parents for
support and protection. Nevertheless, it is evident that this fact
tends rather to confirm the view that expectation of life on the part
of individuals has in all cases been determined with strict reference
to the requirements of propagation, if under propagation we include
the rearing as well as the production of offspring. I may observe in
passing that I do not think this general law can be found to apply to
plants in nearly so close a manner as Weismann has shown it to apply to
animals; but, leaving this consideration aside, I think that Weismann
has made out a good case in favour of such a general law with regard to
animals[3].
We have come, then, to these results. Protoplasm was originally
immortal, barring accidents; and it still continues to be immortal in
the case of unicellular organisms which propagate a-sexually. But in
the case of all multicellular organisms, which propagate sexually,
natural selection has reduced the term of life within the smallest
limits that in each given case are compatible with the performance
of the sexual act and the subsequent rearing of progeny—reserving,
however, the original endowment of immortality for the germinal
elements, whereby a _continuum_ of life has been secured from the
earliest appearance of life until the present day.
Now, in view of these results the question arises,—Why should the
sexual methods of propagation have become so general, if their effect
has been that of determining the necessary death of all individuals
presenting them? Why, in the course of organic evolution, should these
newer methods have been imposed on all the higher organisms, when
the consequence is that all these higher organisms must pay for the
innovation with their lives? Weismann’s answer to this question is as
interesting and ingenious as all that has gone before. Seeing that
sexual propagation is so general as to be practically universal among
multicellular organisms, it is obvious that in some way or another it
must have had a most important part to play in the general scheme of
organic evolution. What, then, is the part that it does play? What is
its _raison d’être_? Briefly, according to Weismann, its function is
that of furnishing congenital variations to the ever-watchful agency
of natural selection, in order that natural selection may always
preserve the most favourable, and pass them on to the next generation
by heredity. That sexual propagation is well calculated to furnish
congenital variations may easily be rendered apparent. We have only to
remember that at each union there is a mixture of two sets of germinal
elements; that each of these was in turn the product of two other sets
in the preceding generation, and so backwards _ad infinitum_ in an ever
doubling ratio. Remembering this, it follows that the germinal elements
of no one member of a species can ever be the same as those of any
other member born of different parents; on the contrary, while both
are enormously complex products, each has had a different ancestral
history, such that while one presents the congenital admixtures of
thousands of individuals in one line of descent, the other presents
similar admixtures of thousands of other individuals in a different
line of descent. Consequently, when in any sexual union two of these
enormously complex germinal elements fuse together, and constitute a
new individual out of their joint endowments, it is perfectly certain
that that individual cannot be exactly like any other individual of
the same species which has been born of different parents. The chances
must be infinity to one against any single mass of germ-plasm being
exactly like any other mass of germ-plasm; while any amount of latitude
as to difference is allowed, up to the point at which the difference
becomes too pronounced to satisfy the conditions of fertilization—in
which case, of course, no new individual is born. Hence, theoretically,
we have here a sufficient cause for all individual variations of
a congenital kind that can possibly occur within the limits of
fertility, and, therefore, that can ever become actual in living
organisms. In point of fact, Weismann believes—or, at any rate,
provisionally maintains—that this is the sole and only cause of
variations that are congenital, and therefore (according to his views)
transmissible by heredity. Now, whether or not he is right as regards
these latter points, I think there can be no question that sexual
propagation is, at all events, one of the main causes of congenital
variation; and seeing of what enormous importance congenital variation
must always have been in supplying material for the operation of
natural selection, we appear to have found a most satisfactory answer
to our question,—Why has sexual propagation become so universal among
all the higher plants and animals? It has become so because it is thus
shown to have been the condition to producing congenital variations,
which in turn constitute one of the primary conditions to the working
of natural selection.
Having got thus far, I should like to make two or three subsidiary
remarks. In the first place, it ought to be observed that this theory
touching the causes of congenital variations was not originally
propounded by Professor Weismann, but occurs in the writings of
several previous authors, and is expressly alluded to by Darwin[4].
Nevertheless, it occupies so prominent a place in Weismann’s system of
theories, and has by him been wrought up so much more elaborately than
by any of his predecessors, that we are entitled to regard it as, _par
excellence_, the Weismannian theory of variation. In the next place,
it ought to be observed that Weismann is careful to guard against the
seductive fallacy of attributing the origin of sexual propagation to
the agency of natural selection. Great as the benefit of this newer
mode of propagation must have been to the species presenting it, the
benefit cannot have been conferred by natural selection, seeing that
the benefit arose from the fact of the new method furnishing material
to the operation of natural selection, and therefore constituting
the condition to the agency of natural selection having been called
into existence at all. Or, in other words, we cannot attribute to
natural selection the origin of sexual reproduction without involving
ourselves in the absurdity of supposing natural selection to have
originated the conditions of its own activity[5]. What the causes may
have been which originally led to sexual reproduction is at present
a matter that awaits suggestion by way of hypothesis; and, therefore,
it now only remains to add that the general structure of Professor
Weismann’s system of hypotheses leads to this curious result—namely,
that the otherwise ubiquitous and (as he supposes) exclusive dominion
of natural selection stops short at the protozoa, over which it cannot
exercise any influence at all. For if natural selection depends for its
activity on the occurrence of congenital variations, and if congenital
variations depend for their occurrence on sexual modes of reproduction,
it follows that no organisms which propagate by any other modes can
present congenital variations, or thus become subject to the sway of
natural selection. And inasmuch as Weismann believes that such is
the case with all the protozoa, as well as with all parthenogenetic
organisms, he does not hesitate to accept the necessary conclusion that
in these cases natural selection is without any jurisdiction. How,
then, does he account for individual variations in the protozoa? And,
still more, how does he account for the origin of their innumerable
species? He accounts for both these things by the direct action of
external conditions of life. In other words, so far as the unicellular
organisms are concerned, Weismann is rigidly and unconditionally an
advocate of the theory of Lamarck—just as much as in the case of
all the multicellular organisms he is rigidly and unconditionally an
opponent of that theory. Nevertheless, there is here no inconsistency:
on the contrary, it is consistency with the logical requirements of
his theory that leads to this sharp partitioning of the unicellular
from the multicellular organisms with respect to the causes of their
evolution. For, according to his view, the conditions of propagation
among the unicellular organisms are such that parent and offspring
are one and the same thing; “the child is a part, and usually a half,
of its parent.” Therefore, if the parent has been in any way modified
by the action of external conditions, it is inevitable that the child
should, from the moment of its birth (i.e., fissiparous separation),
be similarly modified; and if the modifying influences continue in the
same lines for a sufficient length of time, the resulting change of
type may become sufficiently pronounced to constitute a new species,
genus, &c. But in the case of the multicellular or sexual organisms,
the child is not thus merely a severed moiety of its parent; it is the
result of the fusion of two highly specialized and extremely minute
particles of each of two parents. Therefore, whatever may be thought
touching the validity of Weismann’s deduction that in no case can
any modification induced by external conditions on these parents be
transmitted to their progeny, at least we must recognize the validity
of the distinction which he draws between the facility with which such
transmission must take place in the unicellular organisms, as compared
with the difficulty—or, as he believes, the impossibility—of its
doing so in the multicellular.
We are now in a position fully to understand Professor Weismann’s
theory of heredity in all its bearings. Briefly stated, it is as
follows. The whole organization of any multicellular organism is
composed of two entirely different kinds of cells—namely, the
germ-cells, or those which have to do with reproduction, and the
somatic-cells, or those which go to constitute all the other parts of
the organism. Now, the somatic-cells, in their aggregations as tissues
and organs, may be modified in numberless ways by the direct action
of the environment, as well as by special habits formed during the
individual lifetime of the organism. But although the modifications
thus induced may be, and generally are, adaptive—such as the increased
muscularity caused by the use of muscles, “practice making perfect”
where neural adjustments are concerned, and so on,—in no case can
these so-called acquired, or “somatogenetic,” characters exercise any
influence upon the germ-cells, such that they should reappear in the
next generation as congenital, or “blastogenetic,” characters. For,
according to the theory, the germ-cells as to their germinal contents
differ in kind from the somatic-cells, and have no other connexion or
dependence upon them than that of deriving from them their food and
lodging. So much for the somatic-cells.
Turning now to the germ-cells, these are the receptacles of what
Weismann calls the germ-plasm; and this it is that he supposes
to differ in kind from all the other constituent elements of the
organism. For the germ-plasm he believes to have had its origin in the
unicellular organisms, and to have been handed down from them in one
continuous stream through all successive generations of multicellular
organisms. Thus, for example, suppose that we take a certain _quantum_
of germ-plasm as this occurs in any individual organism of to-day. A
minute portion of this germ-plasm, when mixed with a similarly minute
portion from another individual, goes to form a new individual. But,
in doing so, only a portion of this minute portion is consumed; the
residue is stored up in the germinal cells of the new individual,
in order to secure that continuity of the germ-plasm which Weismann
assumes as the necessary basis of his whole theory. Furthermore, he
assumes that this overplus portion of germ-plasm, which is so handed
over to the custody of the new individual, is there capable of growth
or multiplication at the expense of the nutrient materials which are
supplied to it by the new soma in which it finds itself located; while
in thus growing, or multiplying, it faithfully retains its highly
complex structure, so that in no one minute particular does any part of
a many thousand-fold increase differ, as to its ancestral characters,
from that inconceivably small overplus which was first of all entrusted
to the embryo by its parents. Therefore one might represent the
germ-plasm by the metaphor of a yeast-plant, a single particle of which
may be put into a vat of nutrient fluid: there it lives and grows upon
the nutriment supplied, so that a new particle may next be taken to
impregnate another vat, and so on _ad infinitum_. Here the successive
vats would represent successive generations of progeny; but, to make
the metaphor complete, one would have to suppose that in each case the
yeast-cell was required to begin by making its own vat of nutrient
material, and that it was only the residual portion of the cell which
was afterwards able to grow and multiply. But although the metaphor
is thus necessarily a clumsy one, it may serve to emphasize the
all-important feature of Weismann’s theory—namely, the almost absolute
independence of the germ-plasm. For, just as the properties of the
yeast-plant would be in no way affected by anything that might happen
to the vat, short of its being broken up or having its malt impaired,
so, according to Weismann, the properties of the germ-plasm cannot be
affected by anything that may happen to its containing soma, short of
the soma being destroyed or having its nutritive functions disordered.
Such being the relations that are supposed to obtain between the soma
and its germ-plasm, we have next to observe what is supposed to happen
when, in the course of evolution, some modification of the ancestral
form of the soma is required in order to adapt it to some change on the
part of its environment. In other words, we have to consider Weismann’s
views on the _modus operandi_ of adaptive development, with its result
in the origination of new species.
Seeing that, according to the theory, it is only congenital variations
which can be inherited, all variations subsequently acquired by the
intercourse of individuals with their environment, however beneficial
such variations may be to these individuals, are ruled out as regards
the species. Not falling within the province of heredity, they
are blocked off in the first generation, and therefore present no
significance at all in the process of organic evolution. No matter how
many generations of eagles, for instance, may have used their wings for
purposes of flight; and no matter how great an increase of muscularity,
of endurance, and of skill, may thus have been secured to each
generation of eagles as the result of individual exercise; all these
advantages are entirely lost to progeny, and young eagles have ever to
begin their lives with no more benefit bequeathed by the activity of
their ancestors than if those ancestors had all been barn-door fowls.
The only material which is of any count as regards the species, or with
reference to the process of evolution, are fortuitous variations of the
congenital kind. Among all the numberless congenital variations, within
narrow limits, which are perpetually occurring in each generation of
eagles, some will have reference to the wings; and although these
will be fortuitous, or occurring indiscriminately in all directions,
a few of them will now and then be in the direction of increased
muscularity, others in the direction of increased endurance, others
in the direction of increased skill, and so on. Now each of these
fortuitous variations, which happens also to be a beneficial variation,
will be favoured by natural selection; and, because it likewise happens
to be a congenital variation, will be perpetuated by heredity. In the
course of time, other congenital variations will happen to arise in
the same directions; these will be added by natural selection to the
advantage already gained, and so on, till, after hundreds and thousands
of generations, the wings of eagles have become evolved into the
marvellous structures which they now present.
Such being the theory of natural selection when stripped of all
remnants of so-called Lamarckian principles, we have next to consider
what the theory means in its relation to germ-plasm. For, as before
explained, congenital variations are supposed by Weismann to be due
to new combinations taking place in the germ-plasm as a result of
the union in every act of fertilisation of two complex hereditary
histories. Well, if congenital variations are thus nothing more
than variations of germ-plasm “writ large” in the organism which is
developed out of the plasm, it follows that natural selection is
really at work upon these variations of the plasm. For, although it is
proximately at work on the congenital variations of organisms after
birth, it is ultimately, and through them, at work upon the variations
of germ-plasm out of which the organisms arise. In other words,
natural selection, in picking out of each generation those individual
organisms which are by their congenital characters best suited to their
surrounding conditions of life, is thereby picking out those peculiar
combinations or variations of germ-plasm, which, when expanded into a
resulting organism, give that organism the best chance in its struggle
for existence. And, inasmuch as a certain overplus of this peculiar
combination of germ-plasm is entrusted to that organism for bequeathing
to the next generation, this to the next, and so on, it follows that
natural selection is all the while conserving that originally peculiar
combination of germ-plasm, until it happens to meet with some other
mass of germ-plasm by mixing with which it may still further improve
upon its original peculiarity, when, other things equal, natural
selection will seize upon this improvement to perpetuate, as in the
previous case. So that, on the whole, we may say that natural selection
is ever waiting and watching for such combinations of germ-plasm as
will give the resulting organisms the best possible chance in their
struggle for existence; while, at the same time, it is remorselessly
destroying all those combinations of germ-plasm which are handed over
to the custody of organisms not so well fitted to their conditions of
life.
It only remains to add that, according to Weismann’s theory in its
strictly logical form, combinations of germ-plasm when once effected
are so stable that they would never alter except as a result of
entering into new combinations. In other words, no external influences
or internal processes can ever change the hereditary nature of any
particular mixture of germ-plasm, save and except its admixture with
some other germ-plasm, which, being of a nature equally stable, goes
to unite with the first in equal proportions as regards hereditary
character. So that really it would be more correct to say that any
given mass of germ-plasm does not change even when it is mixed with
some other mass—any more, for instance, than a handful of sand can be
said to change when it is mixed with a handful of clay.
Consequently, we arrive at this curious result. No matter how many
generations of organisms there may have been, and therefore no matter
how many combinations of germ-plasm may have taken place to give rise
to an existing population, each existing unit of germ-plasm must
have remained of the same essential nature or constitution as when
it was first started in its immortal career millions of years ago.
Or, reverting to our illustration of sand and clay, the particles
of each must always remain the same, no matter how many admixtures
they may undergo with particles of other materials, such as chalk,
slate, &c. Now, inasmuch as it is an essential—because a logically
necessary—part of Weismann’s theory to assume such absolute stability
or unchangeableness on the part of germ-plasm, the question arises, and
has to be met, What was the origin of those differences of character
in the different germ-plasms of multicellular organisms which first
gave rise, and still continue to give rise, to congenital variations
by their mixture one with another? This important question Weismann
answers by supposing that these differences originally arose out of the
differences in the unicellular organisms, which were the ancestors of
the primitive multicellular organisms. Now, as before stated, different
forms of unicellular organisms are supposed to have originated as so
many results of differences in the direct action of the environment.
Consequently, according to the theory, all congenital variations which
now occur in multicellular organisms, are really the distant results
of variations that were aboriginally induced in their unicellular
ancestors by the direct action of surrounding conditions of life.
I think it will be well to conclude by briefly summarising the main
features of this elaborate theory.
Living material is essentially, or of its own nature, imperishable;
and it still continues to be so in the case of unicellular organisms
which propagate by fission or gemmation. But as soon as these primitive
methods of propagation became, from whatever cause, superseded
by sexual, it ceased to be for the benefit of species that their
constituent individuals should be immortal; seeing that, if they
continued to be so, all species of sexually-reproducing organisms
would sooner or later have come to be composed of broken-down and
decrepit individuals. Consequently, in all sexually-reproducing or
multicellular organisms, natural selection set to work to reduce the
term of individual lifetimes within the narrowest limits that in the
case of each species were compatible with the procreation and the
rearing of progeny. Nevertheless, in all these sexually-reproducing
organisms the primitive endowment of immortality has been retained
with respect to their germ-plasm, which has thus been continuous,
through numberless generations of perishing organisms, from the first
origin of sexual reproduction till the present time. Now, it is the
union of germ-plasms which is required to reproduce new individuals
of multicellular organisms that determines congenital variations on
the part of such organisms, and thus furnishes natural selection with
the material for its work in the way of organic evolution—work,
therefore, which is impossible in the case of unicellular organisms,
where variation can never be congenital, but always determined by
the direct action of surrounding conditions of life. Again, as the
germ-plasm of multicellular organisms is continuous from generation to
generation, and at each impregnation gives rise to a more or less novel
set of congenital characters, natural selection, in picking out of
each generation the congenital characters which are of most service to
the organisms presenting them, is really or fundamentally at work upon
those variations of the germ-plasm which in turn give origin to these
variations of organisms that we recognize as congenital. Therefore,
natural selection has always to wait and to watch for such variations
of germ-plasm as will eventually prove beneficial to the individuals
developed therefrom, who will then transmit this peculiar quality of
germ-plasm to their progeny, and so on. Therefore also—and this is
most important to remember—natural selection as thus working becomes
the one and only cause of organic evolution in all the multicellular
organisms, just as the direct action of the environment is the one and
only cause of it in the case of all the unicellular organisms. But
inasmuch as the multicellular organisms were all in the first instance
derived from the unicellular, and inasmuch as their germ-plasm is
of so stable a nature that it can never be altered by any agencies
internal or external to the organisms presenting it, it follows that
all congenital variations are the remote consequences of aboriginal
differences on the part of unicellular ancestors. And, lastly, it
follows also that these congenital variations—although now so entirely
independent of external conditions of life, and even of activities
internal to organisms themselves—were originally and exclusively
due to the direct action of such conditions on the lives of their
unicellular ancestors; while even at the present day no one congenital
variation can arise which is not ultimately due to differences
impressed upon the protoplasmic substance of the germinal elements,
when the parts of which these are now composed constituted integral
parts of the protozoa, which were directly and differentially affected
by their converse with their several environments.
Again, if for the sake of distinctness we neglect all these
far-reaching deductions from his theory of _heredity_ whereby Weismann
constructs this elaborate theory of _organic evolution_, and fasten
our attention only upon the former, we may briefly summarize the
fundamental difference between his theory of _heredity_ and Darwin’s
theory of _heredity_ thus.
Darwin’s theory of heredity is the theory of _Pangenesis_: it
supposes that _all_ parts of the organism _generate_ anew in every
individual the formative material which, when collected together in the
germ-cells, constitutes the potentiality of a new organism; and that
this new organism, when developed, resembles its parents simply because
_all_ the formative material in each of the parents has been thus
_generated_ by, and collected from, _all_ parts of their respective
bodies. Weismann’s theory of heredity, on the other hand, is the
theory of the _Continuity of Germ-plasm_: it supposes that _no_ part
of the parent organism generates _any_ of the formative material which
is to constitute the new organism; but that, on the contrary, this
material stands to all the rest of the body in much the same relation
as a parasite to its host, showing a life independent of the body,
save in so far as the body supplies to it appropriate lodgement and
nutrition; that in each generation a small portion of this substance is
told off to develop a new body to lodge and nourish the ever-growing
and never-dying germ-plasm—this new body, therefore, resembling its
so-called parent body simply because it has been developed from one and
the same mass of formative material; and, lastly, that this formative
material, or germ-plasm, has been continuous through all generations of
successively perishing bodies, which therefore stand to it in much the
same relation as annual shoots to a perennial stem: the shoots resemble
one another simply because they are all grown from one and the same
stock.