to limitations of the Mullerian Hypothesis of Mimicry. 97 



justify the expectation that the elimination due to the 

 JBatesian factor would be competent to produce a higher 

 degree of inter-resemblance than would the factor adduced 

 by Fritz Miiller. 



Another point worth noting is the difference in the 

 periods of incidence of these two processes of selection. 

 The Mullerian factor is, as we have seen, supplied entirely 

 by young and inexperienced birds, etc.; it will therefore 

 have its greatest effect during the summer months when 

 such young animals would be most numerous. As the 

 season advances, however, these animals would be increasing 

 in wisdom and experience, and consequently by the 

 autumn or early winter we may reasonably suppose that 

 their efficiency as producers of Mullerian associations will 

 have very notably diminished. Now from this time 

 onwards until the early months of the next summer there 

 will be no further appearance of young broods of insecti- 

 vorous animals ; and it seems clear that for a considerable 

 portion of the year the forces which make for Mullerianism 

 will be at a very low ebb, if not altogether absent. On 

 the other hand, there seem to be no grounds for assuming 

 the existence of any such period of marked diminution in 

 the factors which make for Batesian mimicry ; for in most 

 tropical countries butterflies (which are here alone being 

 considered) are fairly plentiful throughout even the winter 

 months, while there is a much greater reduction in the 

 insects of most other orders {cf. Trans. Ent. Soc. 1902, 

 p. 432). Thus, although a large number of insectivorous 

 migrants will have departed at that season, the scarcity 

 of other insects, in conjunction with the comparative 

 conspicuousness of butterflies, will doubtless lead to a 

 maintenance, or even an increase, of the percentage of 

 destruction by the remaining resident birds. The inci- 

 dence of Batesian elimination, therefore, will be com- 

 paratively continuous and persistent. 



Having thus briefly examined some aspects of Mliller's 

 selective factor, we may now endeavour to ascertain the 

 conditions which will render it most effective in fostering 

 mimetic associations. The position of affairs will be best 

 appreciated by taking some hypothetical case. Let us 

 assume therefore that within any given area there exist 

 two species of butterflies, A and B, possessing nauseous 

 qualities in about the same degree, but having diflerent 

 warning colour-patterns ; and further that within the 



TRANS. ENT. SOC. LOND. 1908. PART I. (mAY) 7 



