WCP2197

[1]¹

6 Brunswick Square, W. C.

April 10/[18]88

My dear Wallace,

I have carefully read & considered the MSS. of your VII Chap[ter]. & think the best service I can render you is to treat it ultra-critically so that if my objections have any weight you may reconsider the points which require strengthening before appearing before the scientific public.

Your arguments appear to be à priori perfectly sound, but the groundwork wants strengthening because you start with certain assumptions which must be proved before the hypothesis can take rank as part & parcel of the theory of descent (which I hope to see in due time). You suppose that a sp[ecies]. is giving rise to two var[iant]s. adapted to different modes of life in the same area & that it is to the advantage of the sp[ecies]. that these two forms should remain ^distinct & become enhanced in distinctness. Granted. This is tantamount to admitting that among all the variations that occur two are of real value & would (under the old theory) be accumulatively differentiated by nat[ural]. selec[tion]. But in order that these two forms sh[ou]ld. remain distinct it is absolutely essential that there should [2] be some degree of sterility between the two forms in some portion of their area to start with. In other words you have to assume that some degree of sterility is correlated with those particular characters which give the speci two forms an advantage. Now if it is once granted that sterility may arise between the var[iant]s. of a sp[ecies]. & that this sterility is at the same time associated with useful characters, we are admitting the assumption that two fundamental modes of variation (utility in character plus sterility) must be correlated whenever & wherever a species is undergoing differentiation. Our enemies are likely to say that this is putting an undue strain upon the theory & that the probabilities are against us — that it is unlikely that sterility should be associated with just those particular characters wh[ich]. are required for the survival of the two forms. Take a hypothetical case. Suppose a land bird to be subject to such persecution by terrestrial foes that it became advantageous to it to acquire aquatic habits. To meet those conditions webbed feet & a capacity for diving would be the most essential requirements. Now in the usual course of development by nat[ural]. selec[tion]. it could not be assumed that the result (aquatic habits) had been brought about by the [1 word crossed out] survival of specimens ^varieties in which a tendency [3] to a webbing of the feet & the capacity for remaining long under the water had been correlated at first, but we should have to believe that first one character & then the other character had been acquired in the usual way & that the correlation now seen is the result of the addition of these two distinct kinds of variation. The chances against a tendency to webbing & a faculty of diving being correlated at first must obviously be enormous.

The assumption you have to start with is equivalent to that correlation shown to be improbable in the last paragraph — the only difference is that you suppose the correlation to exist in one part of the area occupied by a sp[ecies]. instead of all over the area. This seems to me to be the weakest point of the argument & it is this point that requires the strongest support in the way of facts that you can bring together. If you can do this all the rest of the reasoning follows naturally & you will (in my opinion) have added the strongest factor to the Darwinian theory since Weismann's non-transmissibility of acquired characters. DO strengthen this basis & wipe out physiological selec[tion].!! Your refutation of Romanes is splendid.

I have been hyper critical because I do not want the enemy to have ^find a vulnerable point in the discussion.

[4] I have just returned from a 4 days' walking tour with J. P. Thompson. We walked from Lyndurst Road thro[ugh]' New For[est]. to Christchurch, Bournemouth, Swanage, & Corfe Castle.

Kind regards from wife & mother to you & Mrs. Wallace, | Yours very truly, | R. Meldola [signature]