to limitations of the Milllerian Hypothesis of Mimicry. 105 



equilibrium so far as concerns the Miillerian factor, and no 

 mimetic approach will take place. Again, where A is 

 largely in excess of B, X will be large and Z small; 

 therefore the disadvantage accruing to A' will be large, 

 the advantage small, the net result being a considerable 

 disadvantage ; so that the effect of the Miillerian factor 

 will be to stamp out A' and to establish A, in other words, 

 to prevent its mimicking B. It is thus evident that so 

 far from upsetting my general argument this criticism 

 merely serves to confirm it when all the factors are taken 

 into account. 



We may now turn to an examination of the various 

 cases which have been cited as furnishing conclusive 

 evidence as to the actual occurrence of diaposematic 

 resemblances. In looking at these as a whole an interest- 

 ing fact emerges, namely, that no example has yet been 

 brought forward as occurring between any two of the 

 most dominant distasteful groups of the world, viz.: 

 Danaini, Euploeini, Lycorgeini, Ithomiinse, Heliconinse and 

 Acrseinse ; although it is among some of these groups that 

 Miillerian mimicry finds its highest development, and 

 cases of reciprocity should be evident and numerous, if 

 Diaposematism be a vera causa. Perhaps it is a mere 

 coincidence that in the great majority of instances cited 

 (5 out of 7) one of the mimetic forms belongs to a species 

 not generally considered to be unpalatable. But it has 

 been claimed that the establishment of a case of diapose- 

 matic resemblance is of itself good evidence of the 

 unpalatability of both species involved. This claim is 

 obviously justifiable only where it can be shown that it is 

 impossible to interpret the facts on any hypothesis other 

 than Diaposematism ; and such a case has not yet come 

 under my notice. It was in connection with this aspect of 

 the question that I first found myself in conflict with this 

 theory ; for its application has led to the assumption that 

 the coloration of certain South African butterflies has a 

 warning significance, indicating the existence of distasteful 

 qualities, whereas my own observation of the insects in 

 the field (supported by a few experiments) has led me to 

 an entirely opposite conclusion. These divergent results 

 have induced me to undertake a closer examination of the 

 foundations upon which the hypothesis of Diaposematism 

 has been built. 



