Chapter III

CHARACTERS AS HEREDITARY AND ACQUIRED (_continued_).

(A.) _Indirect Evidence in favour of the Inheritance of Acquired Characters._

Starting with the evidence in favour of the so-called Lamarckian factors, we have to begin with the Indirect--and this without any special reference to the theories, either of Weismann or of others.

It has already been shown, while setting forth in the preceding chapter the antecedent standing of the issue, that in this respect the _prima facie_ presumption is wholly on the side of the transmission, in greater degree or less, of acquired characters. Even Weismann allows that all "_appearances_" point in this direction, while there is no inductive evidence of the action of natural selection in any one case, either as regards germs or somas, and therefore, _a fortiori_, of the "all-sufficiency" of this cause[35]. It is true that in some of his earlier essays he has argued that there is no small weight of _prima facie_ evidence in favour of his own views as to the non-inheritance of acquired characters. This, however, will have to be considered in its proper place further on. Meanwhile I shall say merely in general terms that it arises almost entirely from a confusion of the doctrine of Continuity as absolute with that of Continuity as partial, and therefore, as admitting of degrees in different cases--which, as already explained, are doctrines wide as the poles asunder. But, leaving aside for the present such _prima facie_ evidence as Weismann has adduced on his side of the issue, I may quote him as a hostile witness to the weight of this kind of evidence _per contra_, in so far as it has already been presented in the foregoing chapter. Indeed, Weismann is much too logical a thinker not to perceive the cogency of the "appearances" which lie against his view of Continuity as absolute--although he has not been sufficiently careful in distinguishing between such Continuity and that which admits of degrees.

[35] See, especially, his excellent remarks on this point, _Contemp.
Rev._ Sept. 1893.

We may take it, then, as agreed on all hands that whatever weight merely _prima facie_ evidence may in this matter be entitled to, is on the side of what I have termed moderated Lamarckianism: first sight "appearances" are against the Neo-Darwinian doctrine of the absolute non-inheritance of acquired characters.

* * * * *

Let us now turn to another and much more important line of indirect evidence in favour of moderated Lamarckianism.

The difficulty of _excluding the possibility_ of natural selection having been at work in the case of wild plants and animals has already been noticed. Therefore we may now appreciate the importance of all facts or arguments which _attenuate the probability_ of natural selection having been at work. This may be done by searching for cases in nature where a congenital structure, although unquestionably adaptive, nevertheless presents so small an amount of adaptation, that we can scarcely suppose it to have been arrived at by natural selection in the struggle for existence, as distinguished from the inheritance of functionally-produced modifications. For if functionally-produced modifications are ever transmitted at all, there is no limit to the minuteness of adaptive values which may thus become congenital; whereas, in order that any adaptive structure or instinct should be seized upon and accumulated by natural selection, it must from the very first have had an adaptive value sufficiently great to have constituted its presence a matter of life and death in the struggle for existence. Such structures or instincts must not only have always presented some measure of adaptive value, but this must always have been sufficiently great to reach what I have elsewhere called a selection-value. Hence, if we meet with cases in nature where adaptive structures or instincts present so low a degree of adaptive value that it is difficult to conceive how they could ever have exercised any appreciable influence in the battle for life, such cases may fairly be adduced in favour of the Lamarckian theory. For example, the Neo-Lamarckian school of the United States is chiefly composed of palaeontologists; and the reason of this seems to be that the study of fossil forms--or of species in process of formation--reveals so many instances of adaptations which in their nascent condition present such exceedingly minute degrees of adaptive value, that it seems unreasonable to attribute their development to a survival of the fittest in the complex struggle for existence. But as this argument is in my opinion of greatest force when it is applied to certain facts of physiology with which I am about to deal, I will not occupy space by considering any of the numberless cases to which the Neo-Lamarckians apply it within the region of palaeontology[36].

[36] There is now an extensive literature within this region. The
principal writers are Cope, Scott and Osborn. Unfortunately,
however, the facts adduced are not crucial as test-cases
between the rival theories--nearly all of them, in fact, being
equally susceptible of explanation by either.

Turning then to inherited actions, it is here that we might antecedently expect to find our best evidence of the Lamarckian principles, if these principles have really had any share in the process of adaptive evolution. For we know that in the life-time of individuals it is action, and the cessation of action, which produce nearly all the phenomena of acquired adaptation--use and disuse in animals being merely other names for action and the cessation of action. Again, we know that it is where neuro-muscular machinery is concerned that we meet with the most conclusive evidence of the remarkable extent to which action is capable of co-ordinating structures for the ready performance of particular functions; so that even during the years of childhood "practice makes perfect" to the extent of organizing neuro-muscular adjustments, so elaborate and complete as to be indistinguishable from those which in natural species we recognized as reflex actions on the one hand, and instinctive actions on the other. Hence, if there be any such thing as "use-inheritance" at all, it is in the domain of reflex actions and instinctive actions that we may expect to find our best evidence of the fact. Therefore I will restrict the present line of evidence--(A)--to these two classes of phenomena, as together yielding the best evidence obtainable within this line of argument.

* * * * *

The evidence in favour of the Lamarckian factors which may be derived from the phenomena of reflex action has never, I believe, been pointed out before; but it appears to me of a more cogent nature than perhaps any other. In order to do it justice, I will begin by re-stating an argument in favour of these factors which has already been adduced by previous writers, and discussed by myself in published correspondence with several leaders of the ultra-Darwinian school.

Long ago Professor Broca and Mr. Herbert Spencer pointed to the facts of co-adaptation, or co-ordination within the limits of the same organism, as presenting good evidence of Lamarckian principles, working in association with natural selection. Thus, taking one of Lamarck's own illustrations, Mr. Spencer argued that there must be numberless changes--extending to all the organs, and even to all the tissues, of the animal--which in the course of many generations have conspired to convert an antelope into a giraffe. Now the point is, that throughout the entire history of these changes their utility must always have been dependent on their association. It would be useless that an incipient giraffe should present the peculiar form of the hind-quarters which we now perceive, unless at the same time it presented the correspondingly peculiar form of the fore-quarters; and as each of these great modifications entails innumerable subordinate modifications throughout both halves of the creature concerned, the chances must have been infinitely great against the required association of so many changes happening to have arisen congenitally in the same individuals by way of merely fortuitous variation. Yet, if we exclude the Lamarckian interpretation, which gives an intelligible _cause_ of co-ordination, we are required to suppose that such a happy concurrence of innumerable independent variations must have occurred by mere accident--and this on innumerable different occasions in the bodies of as many successive ancestors of the existing species. For at each successive stage of the improvement natural selection (if working alone) must have needed all, or at any rate most, of the co-ordinated parts to occur in the same individual organisms[37].

[37] For another and better illustration more recently published by
Mr. Spencer, see _The Inadequacy of Natural Selection_, p. 22.

In alluding to what I have already published upon the difficulty which thus appears to be presented to his theory, Weismann says, "At no distant time I hope to be able to consider this objection, and to show that the apparent support given to the old idea [i. e. of the transmission of functionally-produced modifications] is really insecure, and breaks down as soon as it is critically examined[38]."

[38] _Essays on Heredity_, vol. i. p. 389.

[For further treatment of the subject under discussion _see_
Weismann, _The All-sufficiency of Natural Selection_ (Contemp.
Rev. Sept. and Oct. 1893), and _The Effect of External
Influences upon Development_. "Romanes Lecture" 1894, and
Spencer, _Weismannism once more_ (Cont. Rev. Oct. 1894). C. Ll.
M.]

So much for what Weismann has said touching this matter. But the matter has also been dealt with both by Darwin and by Wallace. Darwin very properly distinguishes between the fallacy that "with animals such as the giraffe, of which the whole structure is admirably co-ordinated for certain purposes, it has been supposed that all the parts must have been simultaneously modified[39]," and the sound argument that the co-ordination itself cannot have been due to natural selection alone. This important distinction may be rendered more clear as follows.

[39] _Variation_, &c., vol. ii. p. 206.

The facts of artificial selection prove that immense modifications of structure may be caused by a cumulative blending in the same individuals of characters which were originally distributed among different individuals. Now, in the parallel case of natural selection the characters thus blended will usually--if not invariably--be of an adaptive kind; and their eventual blending together in the same individuals will be due to free intercrossing of the most fit. But this _blending of adaptations_ is quite a different matter from the _occurrence of co-ordination_. For it belongs to the essence of co-ordination that each of the co-ordinated parts should be destitute of adaptive value _per se_: the adaptation only begins to arise if all the parts in question occur associated together in the same individuals _from the very first_. In this case it is obvious that the analogy of artificial selection can be of no avail in explaining the facts, since the difficulty presented has nothing to do with the blending in single individuals of adaptations previously distributed among different individuals; it has to do with the simultaneous appearance in single individuals of a co-adaptation of parts, none of which could ever have been of any adaptive value had it been previously distributed among different individuals. Consequently, where Darwin comes to consider this particular case (or the case of co-adaptation as distinguished from the blending of adaptations), he freely invokes the aid of the Lamarckian principles[40].

[40] E. g. _Origin of Species_, p. 178.

Wallace, on the other hand, refuses to do this, and says that "the best answer to the difficulty" of supposing natural selection to have been the only cause of co-adaptation may be "found in the fact that the very thing said to be impossible by variation and natural selection, has been again and again affected by variation and artificial selection[41]." This analogy (which Darwin had already and very properly adduced with regard to the _blending of adaptations_) he enforces by special illustrations; but he does not appear to perceive that it misses the whole and only point of the "difficulty" against which it is brought. For the case which his analogy sustains is not that which Darwin, Spencer, Broca and others, mean by _co-adaptation_: it is the case of a blending of _adaptations_. It is not the case where adaptation is _first initiated in spite of intercrossing_, by a fortuitous concurrence of variations each in itself being without adaptive value: it is the case where adaptation is _afterwards increased by means of intercrossing_, through the blending of variations each of which has always been in itself of adaptive value.

[41] _Darwinism_, p. 418.

From this I hope it will be apparent that the only way in which the "difficulty" from co-adaptation can be logically met by the ultra-Darwinian school, is by denying that the phenomenon of co-adaptation (as distinguished from the blending of adaptations) is ever to be really met with in organic nature. It may be argued that in all cases where co-adaptation _appears_ to occur, closer examination will show that the facts are really due to a blending of adaptations. The characters A + B + C + D, which are now found united in the same organism, and, as thus united, all conspiring to a common end, may originally have been distributed among different organisms, where they _severally_ subserved some other ends--or possibly the same end, though in a less efficient manner. Obviously, however, in this case their subsequent combination in the same organism would not be an instance of co-adaptation, but merely of an advantageous blending together of already existing adaptations. This argument, or rejoinder, has in point of fact been adopted by Professor Meldola, he believes that all cases of seeming co-adaptation are thus due to a mere blending of adaptations[42]. Of course, if this position can be maintained, the whole difficulty from co-adaptation would lapse. But even then it would lapse on the ground of _fact_. It would not have been overturned, or in any way affected, by Wallace's _argument_ from artificial selection. For, in that event, no such argument would be required, and, if adduced, would be irrelevant, since no one has ever alleged that there is any difficulty in understanding the mere confluence of adaptations by free-intercrossing of the best adapted.

[42] _Nature_, vol. xliii. pp. 410, 557; vol. xliv. pp. 7, 29. I say
"adopted," because I had objected to his quoting the analogy of
artificial selection, and stated, as above, that the only way
to meet Mr. Spencer's "difficulty" was to deny the fact of
co-adaptation as ever occurring in any case. It then appeared
that Professor Meldola agreed with me as to this. But I do not
yet understand why, if such were his view, he began by
endorsing Mr. Wallace's analogy from artificial selection--i.
e. confusing the case of co-adaptation with that of the
blending of adaptations. If any one denies the fact of
co-adaptation, he cannot assist his denial by arguing the
totally different fact that adaptations may be blended by free
intercrossing; for this latter fact has never been questioned,
and has nothing to do with the one which he engaged in
disputing.

Now, if we are agreed that the only question in debate is the question of fact whether or not co-adaptation ever occurs in nature, it appears to me that the best field for debating the question is furnished by the phenomena of reflex action. I can well perceive that the instances adduced by Broca and Spencer in support of their common argument--such as the giraffe, the elk, &c.--are equivocal. But I think that many instances which may be adduced of reflex action are much more to the point. _For it belongs to the very nature of reflex action that it cannot work unless all parts of the machinery concerned are already present, and already co-ordinated, in the same organism._ It would be useless, in so far as such action is concerned if the afferent and efferent nerves, the nerve-centre, and the muscles organically grouped together, were not all present from the very first in the same individuals, and from the very first were not co-ordinated as a definite piece of organic machinery.

With respect to reflex actions, therefore, it is desirable to begin by pointing out how widely the adaptations which they involve differ from those where no manufacture, so to speak, of special machinery is required. Thus, it is easy to understand how natural selection alone is capable of gradually accumulating congenital variations in the direction of protective colouring; of mimicry; of general size, form, mutual correlation of parts as connected with superior strength, fleetness, agility, &c.; of greater or less development of particular parts, such as legs, wings, tails, &c. For in all such cases the adaptation which is in process of accumulation is from its very commencement and throughout each of its subsequent stages, of _use_ in the struggle for existence. And inasmuch as all the individuals of each successive generation vary round the specific mean which characterized the preceding generation, there will always be a sufficient number of individuals which present congenital variations of the kind required for natural selection to seize upon, without danger of their being swamped by free intercrossing--as Mr. Wallace has very ably shown in his _Darwinism_. But this law of averages can apply only to cases where single structures--or a single group of correlated structures--are already present, and already varying round a specific mean. The case is quite different where a _co-ordination_ of structures is required for the performance of a _previously non-existent_ reflex action. For some, at least, of these structures must be _new_, as must also be the function which all of them first conspire to perform. Therefore, neither the new elements of structure, nor the new combination of structures, can have been previously given as varying round a specific mean. On the contrary, a very definite piece of machinery, consisting of many co-ordinated parts, must somehow or other be originated in a high degree of working efficiency, before it can be capable of answering its purpose in the prompt performance of a particular action under particular circumstances of stimulation. Lastly, such pieces of machinery are always of a highly delicate character, and usually involve so immensely complex a co-ordination of mutually dependent parts, that it is only a physiologist who can fully appreciate the magnitude of the distinction between "adaptations" of this kind, and "adaptations" of the kind which arise through natural selection seizing upon congenital variations as these oscillate round a specific mean.

Or the whole argument may be presented in another form, under three different headings, thus:--

In the first place, it will be evident from what has just been said, that such a piece of machinery as is concerned in even the simplest reflex action cannot have occurred in any considerable number of individuals of a species, _when it first began to be constructed_. On the contrary, if its _origin_ were dependent on congenital variations alone, the needful co-adaptation of parts which it requires can scarcely have happened to occur in more than a very small percentage of cases--even if it be held conceivable that by such means alone it should ever have occurred at all. Hence, instead of preservation and subsequent improvement having taken place _in consequence of_ free intercrossing among all individuals of the species (as in the cases of protective colouring, &c., where adaptation has no reference to any mechanical co-adaptation of parts), they must have taken place _in spite of_ such intercrossing.

In the second place, adaptations due to organic machineries of this kind differ in another all-important respect from those due to a summation of adaptive characters which are already present and already varying round a specific mean. The latter depend for their summation upon the fact--not merely, as just stated, that they are already present, already varying round a specific mean, and therefore owe their progressive evolution to free intercrossing, but also--_that they admit of very different degrees of adaptation_. It is only because the degree of adaptation in generation B is superior to that in generation A that _gradual improvement_ in respect of adaptation is here possible. In the case of protective resemblance, for example, a very imperfect and merely accidental resemblance to a leaf, to another insect, &c., may at the first start have conferred a sufficient degree of adaptive imitation to count for something in the struggle for life; and, if so, the basis would be given for a progressive building up by natural selection of structures and colours in ever-advancing degrees of adaptive resemblance. There is here no necessity to suppose--nor in point of fact is it ever supposed, since the supposition would involve nothing short of a miracle--that such extreme perfection in this respect as we now so frequently admire has originated suddenly in a single generation, as a collective variation of a congenital kind affecting simultaneously a large proportional number of individuals. But in the case of a reflex mechanism--which may involve even greater marvels of adaptive adjustment, and _all_ the parts of which must occur in the same _individuals_ to be of any use--it _is_ necessary to suppose some such sudden and collective origin in some very high degree of efficiency, if natural selection has been the only principle concerned in afterwards perfecting the mechanism. For it is self-evident that a reflex action, from its very nature, cannot admit of any great differences in its degrees of adaptation: if it is to work at all, so as to count for anything in the struggle for life, it must already be given in a state of working efficiency. So that, unless we invoke either the doctrine of "prophetic types" or the theory of sudden creations, I confess I do not see how we are to explain either the origin, or the development, of a reflex mechanism by means of natural selection alone.

Lastly, in the third place, _even when reflex mechanisms have been fully formed_, it is often beyond the power of sober credence to believe that they now are, or ever can have been, of selective value in the struggle for existence, as I will show further on. And such cases go to fortify the preceding argument. For if not conceivably of selective value even when completely evolved, much less can they conceivably have been so through all the stages of their complex evolution back to their very origin. Therefore, supposing for the present that there are such cases of reflex action in nature, neither their origin nor their development can conceivably have been due to natural selection alone. The Lamarckian factors, however, have no reference to degrees of adaptation, any more than they have to degrees of complexity. No question of value, as selective or otherwise, can obtain in their case: neither in their case does any difficulty obtain as regards the co-adaptation of severally useless parts.

Now, if all these distinctions between the Darwinian and Lamarckian principles are valid--and I cannot see any possibility of doubt upon this point--strong evidence in favour of the latter would be furnished by cases (if any occur) where structures, actions, instincts, &c., although of some adaptive value, are nevertheless plainly not of selective value. According to the ultra-Darwinian theory, no such cases ought ever to occur: according to the theory of Darwin himself, they ought frequently to occur. Therefore a good test, or criterion, as between these different theories of organic evolution is furnished by putting the simple question of fact--Can we, or can we not, show that there are cases of adaptation where the degree of adaptation is so small as to be incompatible with the supposition of its presenting a selective value? And if we put the wider question--Are there any cases where the co-adaptation of severally useless parts has been brought about, when even the resulting whole does not present a selective value?--then, of course, we impose a still more rigid test.

Well, notwithstanding the difficulty of proving such a negative as the absence of natural selection where adaptive development is concerned, I believe that there are cases which conform to both these tests simultaneously; and, moreover, that they are to be found in most abundance where the theory of use-inheritance would most expect them to occur--namely, in the province of reflex action. For the very essence of this theory is the doctrine, that constantly associated use of the same parts for the performance of the same action will progressively organize those parts into a reflex mechanism--no matter how high a degree of co-adaptation may thus be reached on the one hand, or how low a degree of utilitarian value on the other.

Having now stated the general or abstract principles which I regard as constituting a defence of the Lamarckian factors, so far as this admits of being raised on grounds of physiology, we will now consider a few concrete cases by way of illustration. It is needless to multiply such cases for the mere purpose of illustration. For, on reading those here given, every physiologist will at once perceive that they might be added to indefinitely. The point to observe is, the relation in which these samples of reflex action stand to the general principles in question; for there is nothing unusual in the samples themselves. On the contrary, they are chosen because they are fairly typical of the phenomena of reflex action in general.

In our own organization there is a reflex mechanism which ensures the prompt withdrawal of the legs from any source of irritation supplied to the feet. For instance, even after a man has broken his spine in such a manner as totally to interrupt the functional continuity of his spinal cord and brain, the reflex mechanism in question will continue to retract his legs when his feet are stimulated by a touch, a burn, &c. This responsive action is clearly an adaptive action, and, as the man neither feels the stimulation nor the resulting movement, it is as clearly a reflex action. The question now is as to the mode of its origin and development.

I will not here dwell upon the argument from co-adaptation, because this may be done more effectually in the case of more complicated reflex actions, but will ask whether we can reasonably hold that this particular reflex action--comparatively simple though it is--has ever been of selective value to the human species, or to the ancestors thereof? Even in its present fully-formed condition it is fairly questionable whether it is of any adaptive _value_ at all. The movement performed is no doubt an adaptive _movement_; but is there any occasion upon which the reflex mechanism concerned therein can ever have been of adaptive _use_? Until a man's legs have been paralyzed as to their voluntary motion, he will always promptly withdraw his feet from any injurious source of irritation by means of his conscious intelligence. True, the reflex mechanism secures an almost inappreciable saving in the time of response to a stimulus, as compared with the time required for response by an act of will; but the difference is so exceedingly small, that we can hardly suppose the saving of it in this particular case to be a matter of any adaptive--much less selective--importance. Nor is it more easy to suppose that the reflex mechanism has been developed by natural selection for the purpose of replacing voluntary action when the latter has been destroyed or suspended by grave spinal injury, paralysis, coma, or even ordinary sleep. In short, even if for the sake of argument we allow it to be conceivable that any single human being, ape, or still more distant ancestor, has ever owed its life to the possession of this mechanism, we may still be certain that not one in a million can have done so. And, if this is the case with regard to the mechanism as now fully constructed, still more must it have been the case with regard to all the previous stages of construction. For here, without elaborating the point, it would appear that a process of construction by survival of the fittest alone is incomprehensible.

On the other hand, of course, the theory of use-inheritance furnishes a fully intelligible--whether or not a true--explanation. For those nerve-centres in the spinal cord which co-ordinate the muscles required for retracting the feet are the centres used by the will for this purpose. And, by hypothesis, the frequent use of them for this purpose under circumstances of stimulation which render the muscular response appropriate, will eventually establish an organic connexion between such response and the kind of stimulation to which it is appropriate--even though there be no utilitarian reason for its establishment[43]. To invert a phrase of Aristotle, we do not frequently use this mechanism because we have it (seeing that in our normal condition there is no necessity for such use); but, by hypothesis, we have it because we have frequently used its several elements in appropriate combination.

[43] It may be said, with regard to this particular reflex, that it
may perhaps be, so to speak, a mechanical accident, arising
from the contiguity of the sensory and motor roots in the cord.
But as this suggestion cannot apply to other reflexes presently
to be adduced, it need not be considered.

I will adduce but one further example in illustration of these general principles--passing at once from the foregoing case of comparative simplicity to one of extreme complexity.

There is a well-known experiment on a brainless frog, which reveals a beautiful reflex mechanism in the animal, whereby the whole body is enabled continually to readjust its balance on a book (or any other plane surface), as this is slowly rotated on a horizontal axis. So long as the book is lying flat, the frog remains motionless; but as soon as the book is tilted a little, so that the frog is in danger of slipping off, all the four feet begin to crawl up the hill; and the steeper the hill becomes, the faster they crawl. When the book is vertical, the frog has reached the now horizontal back, and so on. Such being the facts, the question is--How can the complicated piece of machinery thus implied have been developed by natural selection? Obviously it cannot have been so by any of the parts concerned having been originally distributed among different individuals, and afterwards united in single individuals by survival (i.e. free intercrossing) of the fittest. In other words, the case is obviously one of co-adaptation, and not one of the blending of adaptations. Again, and no less obviously, it is impossible that the co-adaptation can have been _gradually developed_ by natural selection, because, in order to have been so, it must by hypothesis have been of some degree of use in every one of its stages; yet it plainly cannot have been until it had been fully perfected in all its astonishing complexity[44].

[44] Of course it will be observed that the question is not with
regard to the development of all the nerves and muscles
concerned in this particular process. It is as to the
development of the co-ordinating centres, which thus so
delicately respond to the special stimuli furnished by
variations of angle to the horizon. And it is as inconceivable
in this case of reflex action, as it is in almost every other
case of reflex action, that the highly specialized machinery
required for performing the adaptive function can ever have had
its origin in the performance of any other function. Indeed, a
noticeable peculiarity of reflex mechanisms as a class is the
highly specialized character of the functions which their
highly organized structures subserve.

Lastly, not only does it thus appear impossible that during all stages of its development--or while as yet incapable of performing its intricate function--this nascent mechanism can have had any adaptive value; but even as now fully developed, who will venture to maintain that it presents any selective value? As long as the animal preserves its brain, it will likewise preserve its balance, by the exercise of its intelligent volition. And, if the brain were in some way destroyed, the animal would be unable to breed, or even to feed; so that natural selection can never have had any _opportunity_, so to speak, of developing this reflex mechanism in brainless frogs. On the other hand, as we have just seen, we cannot perceive how there can ever have been any _raison d'etre_ for its development in normal frogs--even if its development were conceivably possible by means of this agency. But if practice makes perfect in the race, as it does in the individual, we can immediately perceive that the constant habit of correctly adjusting its balance may have gradually developed, in the batrachian organization, this non-necessary reflex[45].

[45] We meet with a closely analogous reflex mechanism in brainless
vertebrata of other kinds; but these do not furnish such good
test cases, because the possibility of natural selection cannot
be so efficiently attenuated. The perching of brainless birds,
for instance, at once refers us to the roosting of sleeping
birds, where the reflex mechanism concerned is clearly of high
adaptive value. Therefore such a case is not available as a
test, although the probability is that birds have inherited
their balancing mechanisms from their sauropsidian ancestors,
where it would have been of no such adaptive importance.

And, of course, this example--like that of withdrawing the feet from a source of stimulation, which a frog will do as well as a man--does not stand alone. Without going further a-field than this same animal, any one who reads, from our present point of view, Goltz's work on the reflex actions of the frog, will find that the great majority of them--complex and refined though most of them are--cannot conceivably have ever been of any use to any frog that was in undisturbed possession of its brain.

Hence, not to occupy space with a reiteration of facts all more or less of the same general kind, and therefore all presenting identical difficulties to ultra-Darwinian theory, I shall proceed to give two others which appear to me of particular interest in the present connexion, because they furnish illustrations of reflex actions in a state of only partial development, and are therefore at the present moment demonstrably useless to the animal which displays them.

Many of our domesticated dogs, when we gently scratch their sides and certain other parts of the body, will themselves perform scratching movements with the hind leg of the same side as that upon which the irritation is being supplied. According to Goltz[46], this action is a true reflex; for he found that it is performed equally well in a dog which has been deprived of its cerebral hemispheres, and therefore of its normal volition. Again, according to Haycraft[47], this reflex is congenital, or not acquired during the life-time of each individual dog. Now, although the action of scratching is doubtless adaptive, it appears to me incredible that it could ever have become organized into a congenital reflex by natural selection. For, in order that it should, the scratching away fleas would require to have been a function of selective value. Yet, even if the irritation caused by fleas were supposed to be so far fatal in the struggle for existence, it is certain that they would always be scratched away by the conscious intelligence of each individual dog; and, therefore, that no advantage could be gained by organizing the action into a reflex. On the other hand, if acquired characters are ever in any degree transmitted, it is easy to understand how so frequently repeated an action should have become, in numberless generations of dogs, congenitally automatic.

[46] _Pflueger's Archiv_, Bd. xx. s. 23 (1879).

[47] _Brain_, part xlviii, pp. 516-19 (1889).--There is still better
proof of this in the case of certain rodents. For instance,
observing that rats and mice are under the necessity of very
frequently scratching themselves with their hind-feet, I tried
the experiment of removing the latter from newly-born
individuals--i.e. before the animals were able to co-ordinate
their movements, and therefore before they had ever even
attempted to scratch themselves. Notwithstanding that they were
thus destitute of individual experience with regard to the
benefit of scratching, they began their scratching movements
with their stumps as soon as they were capable of executing
co-ordinated movements, and afterwards continued to do so till
the end of their lives with as much vigour and frequency as
unmutilated animals. Although the stumps could not reach the
seats of irritation which were bent towards them, they used to
move rapidly in the air for a time sufficient to have given the
itching part a good scratch, had the feet been present--after
which the animals would resume their sundry other avocations
with apparent satisfaction. These facts showed the hereditary
response to irritation by parasites to be so strong, that even
a whole life-time's experience of its futility made no
difference in the frequency or the vigour thereof.

So much for the general principle of selective value as applied to this particular case. And similarly, of course, we might here repeat the application of all the other general principles, which have just been applied in the two preceding cases. But it is only one of these other general principles which I desire in the present case specially to consider, for the purpose of considering more closely than hitherto the difficulty which this principle presents to ultra-Darwinian theory.

The difficulty to which I allude is that of understanding how all the stages in the _development_ of a reflex action can have been due to natural selection, seeing that, before the reflex mechanism has been sufficiently elaborated to perform its function, it cannot have presented any degree of utility. Now the particular force of the present example, the action of scratching--as also of the one to follow--consists in the fact that it is a case where a reflex action is not yet completely organized. It appears to be only in course of construction, so that it is neither invariably present, nor, when it is present, is it ever fully adapted to the performance of its function.

That it is not invariably present (when the brain is so) may be proved by trying the simple experiment on a number of puppies--and also of full-grown dogs. Again, that even when it is present it is far from being fully adapted to the performance of its function, may be proved by observing that only in rare instances does the scratching leg succeed in scratching the place which is being irritated. The movements are made more or less at random, and as often as not the foot fails to touch the body at any place at all. Hence, although we have a "prophecy" of a reflex action well designed for the discharge of a particular function, at present the machinery is not sufficiently perfected for the adequate discharge of that function. In this important respect it differs from the otherwise closely analogous reflex action of the frog, whereby the foot of the hind leg is enabled to localize with precision a seat of irritation on the side of the body. But this beautiful mechanism in the frog cannot have sprung into existence ready formed at any historical moment in the past history of the phyla. It must have been the subject of a more or less prolonged evolution, in some stage of which it must presumably have resembled the now nascent scratching reflex of the dog, in making merely abortive attempts at localizing the seat of irritation--supposing, of course, that some physiologist had been there to try the experiment by first removing the brain. Now, even if one could imagine it to be, either in the frog or in the dog, a matter of selective importance that so exceedingly refined a mechanism should have been developed for the sole purpose of inhibiting the bites of parasites--which in every normal animal would certainly be discharged by an _intentional_ performance of the movements in question,--even if, in order to save an hypothesis at all costs, we make so violent a supposition as this, still we should do so in vain. For it would still remain undeniably certain that the reflex mechanism is _not_ of any selective value. Even now the mechanism in the dog is not sufficiently precise to subserve the only function which occasionally and abortively it attempts to perform. Thus it has all the appearance of being but an imitating shadow of certain neuro-muscular adjustments, which have been habitually performed in the canine phyla by a volitional response to cutaneous irritation. Were it necessary, this argument might be strengthened by observing that the reflex action is positively _improved_ by removal of the brain.

The second example of a nascent reflex in dogs which I have to mention is as follows.

Goltz found that his brainless dogs, when wetted with water, would shake themselves as dry as possible, in just the same way as normal dogs will do under similar circumstances. This, of course, proves that the shaking movements may be performed by a reflex mechanism, which can have no other function to perform in the organization of a dog, and which, besides being of a highly elaborate character, will respond only to a very special kind of stimulation. Now, here also I find that the mechanism is congenital, or not acquired by individual experience. For the puppies on which I experimented were kept indoors from the time of their birth--so as never to have had any experience of being wetted by rain, &c.--till they were old enough to run about with a full power of co-ordinating their general movements. If these young animals were suddenly plunged into water, the shock proved too great: they would merely lie and shiver. But if their feet alone were wetted, by being dipped in a basin of water, the puppies would soon afterwards shake their heads in the peculiar manner which is required for shaking water off the ears, and which in adult dogs constitutes the first phase of a general shaking of the whole body.

Here, then, we seem to have good evidence of all the same facts which were presented in the case of the scratching reflex. In the first place, co-adaptation is present in a very high degree, because this shaking reflex in the dog, unlike the skin-twitching reflex in the horse, does not involve only a single muscle, or even a single group of muscles; it involves more or less the co-ordinated activity of many voluntary muscles all over the body. Such, at any rate, is the case when the action is performed by the intelligent volition of an adult dog; and if a brainless dog, or a young puppy, does not perform it so extensively or so vigorously, this only goes to prove that the reflex has not yet been sufficiently developed to serve as a substitute for intelligent volition--i.e. that it is _useless_, or a mere organic shadow of the really adaptive substance. Again, even if this nascent reflex had been so far developed as to have been capable of superseding voluntary action, still we may fairly doubt whether it could have proved of selective value. For it is questionable whether the immediate riddance of water after a wetting is a matter of life and death to dogs in a state of nature. Moreover, even if it were, every individual dog would always have got rid of the irritation, and so of the danger, by means of a _voluntary_ shake--with the double result that natural selection has never had any opportunity of gradually building up a special reflex mechanism for the purpose of securing a shake, and that the canine race have not had to wait for any such unnecessary process. Lastly, such a process, besides being unnecessary, must surely have been, under any circumstances, impossible. For even if we were to suppose--again for the sake of saving an hypothesis at any cost--that the presence of a fully-formed shaking reflex is of selective value in the struggle for existence, it is perfectly certain that all the stages through which the construction of so elaborate a mechanism must have passed could not have been, under any circumstances, of any such value.

But, it is needless to repeat, according to the hypothesis of use-inheritance, there is no necessity to suppose that these incipient reflex mechanisms _are_ of any value. If function produces structure in the race as it does in the individual, the voluntary and frequently repeated actions of scratching and shaking may very well have led to an organic integration of the neuro-muscular mechanisms concerned. Their various parts having been always co-ordinated for the performance of these actions by the intelligence of innumerable dogs in the past, their co-adapted activity in their now automatic responses to appropriate stimuli presents no difficulty. And the consideration that neither in their prospectively more fully developed condition, nor, _a fortiori_, in their present and all previous stages of evolution, can these reflex mechanisms be regarded as presenting any selective--or even so much as any adaptive--value, is neither more nor less than the theory of use-inheritance would expect.

Thus, with regard to the phenomena of reflex action in general, all the facts are such as this theory requires, while many of the facts are such as the theory of natural selection alone cannot conceivably explain. Indeed, it is scarcely too much to say, that most of the facts are such as directly contradict the latter theory in its application to them. But, be this as it may, at present there are only two hypotheses in the field whereby to account for the facts of adaptive evolution. One of these hypotheses is universally accepted, and the only question is whether we are to regard it as _alone_ sufficient to explain _all_ the facts. The other hypothesis having been questioned, we can test its validity only by finding cases which it is fully capable of explaining, and which do not admit of being explained by its companion hypothesis. I have endeavoured to show that we have a large class of such cases in the domain of reflex action, and shall next endeavour to show that there is another large class in the domain of instinct.

* * * * *

If instinct be, as Professor Hering, Mr. Samuel Butler, and others have argued, "hereditary habit"--i. e. if it comprises an element of transmitted experience--we at once find a complete explanation of many cases of the display of instinct which otherwise remain inexplicable. For although a large number--or even, as I believe, a large majority--of instincts are explicable by the theory of natural selection alone, or by supposing that they were gradually developed by the survival of fortuitous variations in the way of advantageous psychological peculiarities, this only applies to comparatively simple instincts, such as that of a protectively coloured animal exhibiting a preference for the surroundings which it resembles, or even adopting attitudes in imitation of objects which occur in such surroundings. But in all cases where instincts become complex and refined, we seem almost compelled to accept Darwin's view that their origin is to be sought in consciously intelligent adjustments on the part of ancestors.

Thus, to give only one example, a species of Sphex preys upon caterpillars, which it stings in their nerve-centres for the purpose of paralyzing, without killing them. The victims, when thus rendered motionless, are then buried with the eggs of the Sphex, in order to serve as food for her larvae which subsequently develop from these eggs. Now, in order thus to paralyze a caterpillar, the Sphex has to sting it successively in nine minute and particular points along the ventral surface of the animal--and this the Sphex unerringly does, to the exclusion of all other points of the caterpillar's anatomy. Well, such being the facts--according to M. Fabre, who appears to have observed them carefully--it is conceivable enough, as Darwin supposed[48], that the ancestors of the Sphex, being like many other hymenopterous insects highly intelligent, should have observed that on stinging caterpillars in these particular spots a greater amount of effect was produced than could be produced by stinging them anywhere else; and, therefore, that they habitually stung the caterpillars in these places only, till, in course of time, this originally intelligent habit became by heredity instinctive. But now, on the other hand, if we exclude the possibility of this explanation, it appears to me incredible that such an instinct should ever have been evolved at all; for it appears to me incredible that natural selection, unaided by originally intelligent action, could ever have developed such an instinct out of merely fortuitous variations--there being, by hypothesis, nothing to _determine_ variations of an insect's mind in the direction of stinging caterpillars only in these nine intensely localized spots[49].

[48] For details of his explanation of this particular case, for
which I particularly inquired, see _Mental Evolution in
Animals_, pp. 301-2.

[49] Note B.

Again, there are not a few instincts which appear to be wholly useless to their possessors, and others again which appear to be even deleterious. The dusting over of their excrement by certain freely-roaming carnivora; the choice by certain herbivora of particular places on which to void their urine, or in which to die; the howling of wolves at the moon; purring of cats, &c., under pleasurable emotion; and sundry other hereditary actions of the same apparently unmeaning kind, all admit of being readily accounted for as useless habits originally acquired in various ways, and afterwards perpetuated by heredity, because not sufficiently deleterious to have been stamped out by natural selection[50]. But it does not seem possible to explain them by survival of the fittest in the struggle for existence.

[50] For fuller treatment see _Mental Evolution in Animals_, pp.
274-285, 378-379, 381-383.

Finally, in the case of our own species, it is self-evident that the aesthetic, moral, and religious instincts admit of a natural and easy explanation on the hypothesis of use-inheritance, while such is by no means the case if that hypothesis is rejected. Our emotions of the ludicrous, of the beautiful, and of the sublime, appear to be of the nature of hereditary instincts; and be this as it may, it would further appear that, whatever else they may be, they are certainly not of a life-preserving character. And although this cannot be said of the moral sense when the theory of natural selection is extended from the individual to the tribe, still, when we remember the extraordinary complexity and refinement to which they have attained in civilized man, we may well doubt whether they can have been due to natural selection alone. But space forbids discussion of this large and important question on the present occasion. Suffice it therefore to say, that I doubt not Weismann himself would be the first to allow that his theory of heredity encounters greater difficulties in the domain of ethics than in any other--unless, indeed, it be that of religion[51].

[51] For an excellent essay on the deleterious character of early
forms of religion from a biological point of view, see the Hon.
Lady Welby, _An Apparent Paradox in Mental Evolution_ (Journ.
Anthrop. Inst. May 1891).

* * * * *

I have now given a brief sketch of the indirect evidence in favour of the so-called Lamarckian factors, in so far as this appears fairly deducible from the facts of reflex action and of instinct. It will now be my endeavour to present as briefly what has to be said against this evidence.

As previously observed, the facts of reflex action have not been hitherto adduced in the present connexion. This has led me to occupy considerably more space in the treatment of them than those of instinct. On this account, also, there is here nothing to quote, or to consider, _per contra_. On the other hand, however, Weismann has himself dealt with the phenomena of instinct in animals, though not, I think, in man--if we except his brilliant essay on music. Therefore let us now begin this division of our subject by briefly stating, and considering, what he has said upon the subject.

The answer of Weismann to difficulties which arise against the ultra-Darwinian theory in the domain of instinct, is as follows:--

"The necessity for extreme caution in appealing to the supposed
hereditary effects of use, is well shown in the case of those
numerous instincts which only come into play once in a life-time,
and which do not therefore admit of improvement by practice. The
queen-bee takes her nuptial flight only once, and yet how many and
complex are the instincts and the reflex mechanisms which come into
play on that occasion. Again, in many insects the deposition of
eggs occurs but once in a life-time, and yet such insects always
fulfil the necessary conditions with unfailing accuracy[52]."

[52] _Essays_, i. p. 93.

But in this rejoinder the possibility is forgotten, that although such actions are _now_ performed only once in the individual life-time, _originally_--i.e. when the instincts were being developed in a remote ancestry--they may have been performed on many frequent and successive occasions during the individual life-time. In all the cases quoted by Weismann, instincts of the kind in question bear independent evidence of high antiquity, by occurring in whole genera (or even families), by being associated with peculiar and often highly evolved structures required for their performance, and so on. Consequently, in these cases ample time has been allowed for subsequent changes of habit, and of seasonal alterations with respect to propagation--both these things being of frequent and facile occurrence among animals of all kinds, even within periods which fall under actual observation. Nevertheless, I do not question that there are instinctive activities which, as far as we are able to see, can never have been performed more than once in each individual life-time[53]. The fact, however, only goes to show what is fully admitted--that some instincts (and even highly complex instincts) have apparently been developed by natural selection alone. Which, of course, is not equivalent to showing that all instincts must have been developed by natural selection alone. The issue is not to be debated on general grounds like this, but on those of particular cases. Even if it were satisfactorily proved that the instincts of a queen-bee have been developed by natural selection, it would not thereby be proved that such has been the case with the instincts of a Sphex wasp. One can very well understand how the nuptial flight of the former, with all its associated actions, may have been brought about by natural selection alone; but this does not help us to understand how the peculiar instincts of the latter can have been thus caused.

[53] See _Mental Evolution in Animals_, pp. 377-8.

Strong evidence in favour of Weismann's views does, however, at first sight seem to be furnished by social hymenoptera in other respects. For not only does the queen present highly specialized and altogether remarkable instincts; but the neuters present totally different and even still more remarkable instincts--which, moreover, are often divided into two or more classes, corresponding with the different "castes." Yet the neuters, being barren females, never have an opportunity of bequeathing their instincts to progeny. Thus it appears necessary to suppose that the instincts of all the different castes of neuters are latent in the queen and drones, together with the other instincts which are patent in both. Lastly, it seems necessary to suppose that all this wonderful organization of complex and segregated instincts must have been built up by natural selection acting exclusively on the queens and drones--seeing that these exercise their own instincts only once in a life-time, while, as just observed, the neuters cannot possibly bequeath their individual experience to progeny. Obviously, however, natural selection must here be supposed to be operating at an immense disadvantage; for it must have built up the often diverse and always complex instincts of neuters, not directly, but indirectly through the queens and drones, which never manifest any of these instincts themselves.

Now Darwin fully acknowledged the difficulty of attributing these results to the unaided influence of natural selection; but the fact of neuter insects being unable to propagate seemed to him to leave no alternative. And so it seems to Weismann, who accordingly quotes these instincts in support of his views. And so it seemed to me, until my work on _Animal Intelligence_ was translated into French, and an able Preface was supplied to that translation by M. Perrier. In this Preface it is argued that we are not necessarily obliged to exclude the possibility of Lamarckian principles having operated in the original formation of these instincts. On the contrary, if such principles ever operate at all, Perrier shows that here we have a case where it is virtually certain that they must have operated. For although neuter insects are now unable to propagate, their organization indicates--if it does not actually prove--that they are descended from working insects which were able to propagate. Thus, in all probability, what we now call a "hive" was originally a society of sexually mature insects, all presenting the same instincts, both as to propagation and to co-operation. When these instincts, thus common to all individuals composing the hive, had been highly perfected, it became of advantage in the struggle for existence (between different hives or communities) that the functions of reproduction should devolve more upon some individuals, while those of co-operation should devolve more upon others. Consequently, this division of labour began, and gradually became complete, as we now find it in bees and ants. Perrier sustains the hypothesis thus briefly sketched by pointing to certain species of social hymenoptera where we may actually observe different stages of the process--from cases where all the females of the hive are at the same time workers and breeders, up to the cases where the severance between these functions has become complete. Therefore, it seems to me, it is no longer necessary to suppose that in these latter cases all the instincts of the (now) barren females can only have been due to the unaided influence of natural selection.

Nevertheless, although I think that Perrier has made good his position thus far, that his hypothesis fails to account for some of the instincts which are manifested by neuter insects, such as those which, so far as I can see, must necessarily be supposed to have originated after the breeding and working functions had become separated--seeing that they appear to have exclusive reference to this peculiar state of matters. Possibly, however, Perrier might be able to meet each of these particular instincts, by showing how they could have arisen out of simpler beginnings, prior to the separation of the two functions in question. There is no space to consider such possibilities in detail; but, until this shall have been done, I do not think we are entitled to conclude that the phenomena of instinct as presented by neuter insects are demonstrably incompatible with the doctrines of Lamarck--or, that these phenomena are available as a logical proof of the unassisted agency of natural selection in the case of instincts in general[54].

[54] [See H. Spencer, _The Inadequacy of Natural Selection, A
Rejoinder to Professor Weismann_, Contemp. Rev. 1893; and
_Weismannism once more_, Ibid. Oct. 1894; Weismann, _The
All-sufficiency of Natural Selection_, Ibid. 1893; and _The
Effect of External Influences upon Development_, "Romanes
Lecture" 1894: also _Neuter Insects and Lamarckism_, W. Platt
Ball, Natural Science, Feb. 1894, and _Neuter Insects and
Darwinism_, J. T. Cunningham, Ibid. April 1894. C. Ll. M.]

(B.) _Inherited Effects of Use and of Disuse._

There is no doubt that Darwin everywhere attaches great weight to this line of evidence. Nevertheless, in my opinion, there is equally little doubt that, taken by itself, it is of immeasurably less weight than Darwin supposed. Indeed, I quite agree with Weismann that the whole of this line of evidence is practically worthless; and for the following reasons.

The evidence on which Darwin relied to prove the inherited effects of use and disuse was derived from his careful measurements of the increase or decrease which certain bones of our domesticated animals have undergone, as compared with the corresponding bones of ancestral stocks in a state of nature. He chose domesticated animals for these investigations, because, while yielding unquestionable cases of increased or diminished use of certain organs over a large number of sequent generations, the results were not complicated by the possible interference of natural selection on the one hand, or by that of the economy of nutrition on the other. For "with highly-fed domesticated animals there seems to be no economy of growth, or any tendency to the elimination of superfluous details[55];" seeing that, among other considerations pointing in the same direction, "structures which are rudimentary in the parent species, sometimes become partially re-developed in our domesticated productions[56]."

[55] _Variation of Plants and Animals_, vol. ii. p. 289.

[56] _Ibid._ p. 346.

The method of Darwin's researches in this connexion was as follows. Taking, for example, the case of ducks, he carefully weighed and measured the wing-bones and leg-bones of wild and tame ducks; and he found that the wing-bones were smaller, while the leg-bones were larger, in the tame than in the wild specimens. These facts he attributed to many generations of tame ducks using their wings less, and their legs more, than was the case with their wild ancestry. Similarly he compared the leg-bones of wild rabbits with those of tame ones, and so forth--in all cases finding that where domestication had led to increased use of a part, that part was larger than in the wild parent stock; while the reverse was the case with parts less used. Now, although at first sight these facts certainly do seem to yield good evidence of the inherited effects of use and disuse, they are really open to the following very weighty objections.

First of all, there is no means of knowing how far the observed effects may have been due to increased or diminished use during only the individual life-time of each domesticated animal. Again, and this is a more important point, in all Darwin's investigations the increase or decrease of a part was estimated, not by directly comparing, say the wing-bones of a domesticated duck with the wing-bones of a wild duck, but by comparing the _ratio_ between the wing and leg bones of a tame duck with the _ratio_ between the wing and leg bones of a wild duck. Consequently, if there be any reason to doubt the supposition that a really inherited decrease in the size of a part thus estimated is due to the inherited effects of disuse, such a doubt will also extend to the evidence of increased size being due to the inherited effects of use. Now there is the gravest possible doubt lying against the supposition that any really inherited decrease in the size of a part is due to the inherited effects of disuse. For it may be--and, at any rate to some extent, must be--due to another principle, which it is strange that Darwin should have overlooked. This is the principle which Weismann has called Panmixia, and which cannot be better expressed than in his own words:--

"A goose or a duck must possess strong powers of flight in the
natural state, but such powers are no longer necessary for
obtaining food when it is brought into the poultry-yard; so that a
rigid selection of individuals with well-developed wings at once
ceases among its descendants. Hence, in the course of generations,
a deterioration of the organs of flight must necessarily
ensue[57]."

[57] _Essays_, i. p. 90.

Or, to state the case in another way: if any structure which was originally built up by natural selection on account of its use, ceases any longer to be of so much use, in whatever degree it ceases to be of use, in that degree will the premium before set upon it by natural selection be withdrawn. And the consequence of this withdrawal of selection as regards that particular part will be to allow the part to degenerate in successive generations. Such is the principle which Weismann calls Panmixia, because, by the withdrawal of selection from any particular part, promiscuous breeding ensues with regard to that part. And it is easy to see that this principle must be one of very great importance in nature; because it must necessarily come into operation in all cases where any structure or any instinct has, through any change in the environment or in the habits of a species, ceased to be useful. It is likewise easy to see that its effect must be the same as that which was attributed by Darwin to the inherited effect of disuse; and, therefore, that the evidence on which he relied in proof of the inherited effects both of use and of disuse is vitiated by the fact that the idea of Panmixia did not occur to him.

Here, however, it may be said that the idea first occurred to me[58] just after the publication of the last edition of the _Origin of Species_. I called the principle the Cessation of Selection--which I still think a better, because a more descriptive, term than Panmixia; and at that time it appeared to me, as it now appears to Weismann, entirely to supersede the necessity of supposing that the effect of disuse is ever inherited in any degree at all. Thus it raised the whole question as to the admissibility of Lamarckian principles in general; or the question on which we are now engaged touching the possible inheritance of acquired, as distinguished from congenital, characters. But on discussing the matter with Mr. Darwin, he satisfied me that the larger question was not to be so easily closed. That is to say, although he fully accepted the principle of the Cessation of Selection, and as fully acknowledged its obvious importance, he convinced me that there was independent evidence for the transmission of acquired characters, sufficient in amount to leave the general structure of his previous theory unaffected by what he nevertheless recognized as a factor which must necessarily be added. All this I now mention in order to show that the issue which Weismann has raised since Darwin's death was expressly contemplated during the later years of Darwin's life. For if the idea of Panmixia--in the absence of which Weismann's entire system would be impossible--had never been present to Darwin's mind, we should have been left in uncertainty how he would have regarded this subsequent revolt against what are generally called the Lamarckian principles[59].

[58] _Nature_, vol. ix. pp. 361-2, 440-1; and vol. x. p. 164.

[59] Appendix I.

Moreover, in this connexion we must take particular notice that the year after I had published these articles on the Cessation of Selection, and discussed with Mr. Darwin the bearing of this principle on the question of the transmission of acquired characters, Mr. Galton followed with his highly important essay on Heredity. For in this essay Mr. Galton fully adopted the principle of the Cessation of Selection, and was in consequence the first publicly to challenge the Lamarckian principles--pointing out that, if it were thus possible to deny the transmission of acquired characters _in toto_, "we should be relieved from all further trouble"; but that, if such characters are transmitted "in however faint a degree, a complete theory of heredity must account for them." Thus the question which, in its revived condition, is now attracting so much attention, was propounded in all its parts some fifteen or sixteen years ago; and no additional facts or new considerations of any great importance bearing upon the subject have been adduced since that time. In other words, about a year after my own conversations with Mr. Darwin, the whole matter was still more effectively brought before his notice by his own cousin. And the result was that he still retained his belief in the Lamarckian factors of organic evolution, even more strongly than it was retained either by Mr. Galton or myself[60].

[60] For a fuller statement of Mr. Galton's theory of Heredity, and
its relation to Weismann's, see _An Examination of
Weismannism_.

We have now considered the line of evidence on which Darwin chiefly relied in proof of the transmissibility of acquired characters; and it must be allowed that this line of evidence is practically worthless. What he regarded as the inherited effects of use and of disuse may be entirely due to the cessation of selection in the case of our domesticated animals, combined with an active _reversal_ of selection in the case of natural species. And in accordance with this view is the fact that the degeneration of disused parts proceeds much further in the case of wild species than it does in that of domesticated varieties. For although it may be said that in the case of wild species more time has been allowed for a greater accumulation of the inherited effects of disuse than can have been the case with domesticated varieties, the alternative explanation is at least as probable--that in the case of wild species the merely negative, or passive, influence of the _cessation_ of selection has been continuously and powerfully assisted by the positive, or active, influence of the _reversal_ of selection, through economy of growth and the general advantage to be derived from the abolition of useless parts[61].

[61] For a fuller explanation of the important difference between
the mere cessation and the actual reversal of selection, see
Appendix I.

The absence of any good evidence of this direct kind in favour of use-inheritance will be rendered strikingly apparent to any one who reads a learned and interesting work by Professor Semper[62]. His object was to show the large part which he believed to have been played by external conditions of life in directly modifying organic types--or, in other words, of proving that side of Lamarckianism which refers to the immediate action of the environment, whether with or without the co-operation of use-inheritance and natural selection. Although Semper gathered together a great array of facts, the more carefully one reads his book the more apparent does it become that no single one of the facts is in itself conclusive evidence of the transmission to progeny of characters which are acquired through use-inheritance or through direct action of the environment. Every one of the facts is susceptible of explanation on the hypothesis that the principle of natural selection has been the only principle concerned. This, however, it must be observed, is by no means equivalent to proving that characters thus acquired are not transmitted. As already pointed out, it is impracticable with species in a state of nature to dissociate the distinctively Darwinian from the possibly Lamarckian factors; so that even if the latter are largely operative, we can only hope for direct evidence of the fact from direct experiments on varieties in a state of domestication. To this branch of our subject, therefore, we will now proceed.

[62] _Animal Life_, International Scientific Series, vol. xxxi.