WCP2762

[1]

UNIVERSITY COLLEGE BRISTOL

27^th October 1897

Dear Mr. Wallace,

I criticised Romanes'¹ application of Delboeuf's Law, in conversation when I was fuming over the proofs of Chap. 1 with him, on somewhat similar lines to yours. I said suppose we separate a given group into two moieties; and that in (1) we have 95% of normal individuals a b c d e and 5% with an abnormality widening them expressible by the formula a b c d e n, which in (2) we have 95% a b c d e as before and 5% with a different abnormality a b c d e m; then the only effect under Delboeuf's Law per se will be that n will be incorporated as a character in group (1) while m will be incorporated in group (2); but since the differences implied by n and m are respectively small, the resulting difference is proportionately small. Only on the further hypothesis that under the new conditions (if such there be) of (1) n be of selective value which those of m in (2) be likewise of selective value, with new species characterized expectantly by n. ness and m. ness both raised to higher values by natural selection be formed. I could not get him to see the matter as I did, and he poured forth a cloud of words which rather confused than enlightened me. So I concluded by asking whether it was to stand [2] or not; and he replied that he would think it over and alter it if on consideration it failed to satisfy him. It was not altered: so I let it stand. I purposely abstained form adding critical notes on his or a score of other points which we discussed. It was not for me to give a running fire of criticism. Some points which we discussed were altered in the margin of the proof: so I took it that more which were not altered were regarded as imimportant [sic].

On the question of "2 or more simultaneous divergent tines of variation" — perhaps the matter is, as you say, largely verbal. Romanes's [illeg.]sumption of Natural selection under isolation as one of its modes, is in my judgement unsatisfactory and confusing. But it does seem to me that granting continued and free intercrossing between all members of a given group divergent evolutn in that group is impossible. If the continued and free intercrossing ceases, this is a form of isolation permitting natural selection to bring about cumulative divergence.

I cannot but feel (of course I speak in confidence) that the book is not only in parts but as a whole weak. I have had a most difficult task in editing it. I can only say that the much I excluded was weaker that the little retained. There is so much that is mere imitation of previously expressed opinion, so little that is new in fact or argument. Had I not been bound by solemn injunction and given promise to do my best for the book I should long ago have given up the task as hopeless and fruitless. And he thought so much of it, poor fellow! And was so uniformly kind & patient under my criticisms with which he mostly disagreed!!

Yours very truly | C. Lloyd Morgan² [signature]