560 Dr. F. A. Dixey's^% to Mr. G. A. K. Marshall 



application of this latter principle to butterflies . . . seems 

 open to some serious obJL'ctions" (ihid.), I can still give my 

 assent. I liave always held that there was room for both 

 theories, which are complementary rather than contradict- 

 ory. But having said so much, he proceeds to impose 

 very serious limitations on the scope of Miillerian assimila- 

 tion, and in especial to disallow the conception of what has 

 been called Diaposematism or Reciprocal Mimicry, " even 

 as a mere working hypothesis." 



Here he no doubt expects me to join issue with him, 

 and I shall not disappoint his expectation. I maintain, 

 on the contrary, that the operation of the Miillerian factor, 

 though not universal, is a guod deal wider tlian he is dis- 

 posed to admit; and that the ^jrinciple of Diaposematism, 

 which, as he rightly says, is a corollary of the Miillerian 

 theory, affords the best explanation that can at present be 

 given of certain interesting cases of mimetic grouping. 

 This, I think, is a fair statement of the issue between us. 



The General Ao-gument. 



The opening paragraphs of Mr. Marshall's paper contain 

 a fair and lucid presentment of the Miillerian theory. On 

 these passages I have naturally no criticism to offer, though 

 it may be worthy of notice, in passing, that while the fact 

 that young insectivorous animals have to undergo an 

 education in the matter of suitable provender is, as Mr. 

 Marshall says, " sufficiently well established by now " (pp. 

 94, 95), we cannot eliminate the operation of inherited 

 instinct from the general relation of animals to their food. 

 The avoidance of poisonous fruits, for instance, must, it 

 would seem, be due to an instinct Avhich has grown up 

 under the influence of natural selection. This point, how- 

 ever, though it is well to bear it in mind, is immaterial for 

 present purposes. 



The first of Mr. Marshall's assertions that I should ques- 

 tion is his statement on p. 95 that the initial mimetic varia- 

 tion must gradually replace the original form. It is hard to 

 see why this must necessarily be the case. The original 

 form may quite conceivably continue to be able to main- 

 tain itself, even after it has given rise to a variation which 

 is also capable of a separate existence. Innumerable in- 

 stances of this persistence of an ancestral form are known 

 throughout organic nature, and indeed they are common 

 enough among the special subjects of our present study. 



