Briefly, it seems to me that your argument is perfectly clear up to a certain point, but then suddenly becomes a petitio principii. In other words, so far as your view is critical of natural selection considered as a hypothetical cause of adaptive evolution, I can well believe you have adduced a formidable array of facts. But I fail to follow, when you pass on to the constructive part of your case—or your suggested substitute for natural selection in self-adaptation. For self-adaptation, I understand, consists in results of immediate response to stimuli supplied by environment. But, if so, surely the statement that all the adaptive machinery of plant-organisation is due to self-adaptation is a mere begging of the question against natural selection unless it can be shown how self-adaptation works in each case. Now I do not find any suggestion as to this. And yet this is obviously the essential point; since, unless it can be shown how self-adaptation works—i.e. that it is a vera causa, and not a mere word serving to re-state the facts of adaptive evolution. We have got no further in the way of explanation than the physician, who said, that the reason why morphia produces sleep is because it possesses a soporific quality.
Observe, I purposely abstain from considering your criticism of natural selection, which, although perfectly lucid and possibly justifiable, yet certainly does admit of the answer that incipient variations of a fortuitous kind under nature may often be inconspicuous (while Wallace shows that in animals they are, as a matter of fact, usually considerable). But we need not go into this. The interesting point to all of us must be the constructive part of your work; and I have tried to explain my difficulty with regard to it. Why should protoplasm be able to adapt itself into the millions of diverse mechanisms of nature by converse with environment? The theory of natural selection gives a logically possible, even if it be a biologically inadequate answer. But I cannot see that the theory of self-adaptation does, unless it can be shown that there is some sufficient reason why, say a direct-environment should produce self-adaptation in the direction of hairs, a marine one in that of fleshiness, &c. &c.
I have been very frank, because I know you, and therefore that this is what you would prefer. But I am too ill to make myself clear in a letter. I wish you could stop here for a day on your way home, by which time I shall probably have read your books, and we might discuss the whole business before I publish mine on the Post-Darwinian Theories.
With very many thanks,
I remain yours very truly,
G. J. Romanes.
Hôtel Costebelle, Hyères: February 24, 1894.
Dear Mr. Henslow,—Nothing can be more clear than are all your letters, and the last one, I take it, sets at rest the only question which I had to ask. For it expressly answers that, in your own view, hypothesis of 'self-adaptation' is a statement rather than an explanation of the facts. Nevertheless, it is also to some certain extent advanced as an explanation on Lamarckian lines, for in your books (for which I much thank you) you attribute adaptive mechanism in flowers to thrusts, strains &c. caused by insects. But here, if I may say so, it does not seem to me that you sufficiently deal with an obvious criticism, viz. How is it so much as conceivable that protoplasm should always respond to insect irritation adaptively, when we look to the endless variety and often great elaboration of the mechanism? Similarly as regards the inorganic environment, Lamarck's hypothesis of use-inheritance (i.e. mere increase and decrease of parts as due to inherited efforts of greater or less development by altered flow of nutrition) was at least theoretically valid. But how can you extend this to structures which, though useful, are never active, so as to modify flow of nutrition, e.g. hard shells of nuts, soft pulp of fruits, &c.? Here it is that natural selection theory has the pull. And so of adaptive colours, odours, and secretions? I confess that, even accepting inheritance of acquired characters, I could conceive of 'self-adaptation' alone producing all such innumerable and diversified adjustments only by seeing with Newman (in his 'Apologia') an angel in every flower.
Besides, I do not see why you are shut up to this, even on your own principles. For surely, be there as much self-adaptation in Nature as ever you please, it would still be those individuals (or incipient types) which best respond to stimulation (i.e. most adaptively do so) that, other things equal, would survive in the struggle for existence, and so be naturally selected. In other words, I do not see why you should accept natural selection as regards 'vigour' of seedlings, and nowhere else.