small range of individual variability; and, if so, they would naturally require a long time to exhibit any change of type under the influence of apogamy alone. But, be this as it may, Weismann himself adduces these cases merely for the sake of showing that there are cases which seem to tell against the general principle of modification as due to apogamy alone--i.e. the general principle which, under the name amixia, he is engaged in defending. And the conclusion at which he himself arrives is, that while it would be wrong to affirm that apogamy _must_ in all cases produce divergence, we are amply justified in affirming that in many cases it _may_ have done so; while there is good evidence to prove that in not a few cases it _has_ done so, and therefore should be accepted as one of the factors of organic evolution[57]. [56] Namely, _Lycaena denzelii_, _L. pheretes_, _Argynnis pales_, _Erebia mante_. [57] Since the above was written, I have heard of some cases which seem to present greater difficulties to our theory than those above quoted. These refer to some of the numerous species of land mollusca which inhabit the isolated rocks near Madeira (Dezertas). My informant is Dr. Grabham, who has himself investigated the matter, and reports as follows:-- "It is no uncommon thing to meet with examples of the same species, sub-fossil, recent, and living upon one spot, and presenting no variation in the long record of descent." Then, after naming these examples, he adds, "All seem to vary immediately on attaining new ground, assuming many aspects in different districts." Unquestionably these statements support, in a very absolute manner, Mr. Wallace's opinion, while making directly against my own. It is but fair, however, to add that the cases are not numerous (some half-dozen at the most, and all within the limits of a single genus), and that, even in the opinion of my informant himself, the facts have not hitherto been sufficiently investigated for any decisive judgement to be formed upon them.