on Milllerian Mimicry and Diaposemrdism. 569 



which latter class will therefore undergo a second exaction 

 of the same number. The result of course is that the two 

 original forms, A and B, each lose 50 individuals, or 10 per 

 cent. ; while the two diaposematic intermediates, Ab and 

 Ba, each lose 25 only, or 5 per cent. Needless to say that 

 in view of these considerations I have no intention of 

 impaling myself on either horn of the dilemma so care- 

 fully prepared for me on p. 100 of my friend's paper. 



It may possibly be objected that the numbers of Ab, Ba 

 are not likely at the outset to be equal to those of A 

 and B. This is true enough, but any one who is willing to 

 incur the trouble can easily convince himself that taking 

 the numbers of Ab, Ba smaller only accentuates their 

 advantage over A, B. The number of individuals experi- 

 mentally destroyed may of course be multiplied indefinitely 

 without disturbing the relation between A, B and Ab, Ba. 



But it may still be urged, is there any evidence that 

 such intermediate forms as those exemplified in Ab, Ba 

 are actually to be found in nature ? Undoubtedly there 

 is ; about this I shall have more to say later on, but 

 meanwhile M'e may take as a single example two forms of 

 Leuceronia and Nychitona that occur together in the neigh- 

 bourhood of the Victoria Nyanza. The former (Z. pharis), 

 though still unmistakably a Leuceronia, differs from its 

 nearest allies by points in which it plainly approaches the 

 Nychitona ; while the latter, without losing its general 

 resemblance to its own group, shows features of likeness 

 to the Leuceronia which are peculiar to itself among its 

 congeners. It may still be urged that there is no evidence 

 of distaste fulness in respect of these forms. This may be 

 readily allowed without damaging the argument, for if such 

 approach is possible between forms that belong to the 

 edible category, it must be at least equally possible of 

 occurrence between forms that are distasteful. And if it 

 once occurs as a variation, its perpetuation is provided for 

 in the manner already shown. 



To summarise the foregoing : — Mr. Marshall has omitted 

 to take into account the factors of (1) relative distasteful- 

 ness, and (2) relative conspicuousness and powers of display. 

 These omissions vitiate his argument as to the effect of 

 relative population. Further, he has ignored (8) the 

 effect of the possession of a double aposeme upon relative 

 mortality, and (4) the fact that a persistence of a mimetic 

 variation does not necessarily involve the disappearance of 



