128 MIMICRY AND VARIATION [ch. 



the mainland, however, the enemies of Pseudacraea 

 are weU acquainted with the Planema models which are 

 here common, and discriminate against individuals 

 which are not close mimics of the Planemas. The 

 result is that on the mainland transitional forms are 

 scarcer than on the island. Natural selection main- 

 tains a high standard for the mimetic likeness on the 

 mainland owing to the abundance of the model; but 

 when the model is scarce the likeness ceases to be 

 kept up to the mark strictly, and tends to become 

 lost owing to the appearance of fresh variations which 

 are no longer weeded out. 



Here it should be stated that the various Pseud- 

 acraeas form a population in which the different forms 

 mate freely with one another. In the few breeding 

 experiments that Dr Carpenter was able to make he 

 found that ohscura could produce terra, and that 

 tirikensis was able to give ohscura, the male in each case 

 being, of course, unknown. Far too little work has as 

 yet been done on the genetics of these various forms, 

 and it would be rash to make assumptions as to the 

 nature of the intermediates until the method of experi- 

 mental breeding has been more extensively employed 

 in analysing their constitution. Possibly it is not 

 without significance that the abundance or scarcity of 

 the ohscura form runs parallel with the abundance or 

 scarcity of the intermediates. It suggests that the 

 intermediates are heterozygous in some factor for 

 which the typical ohscura is homozygous, and the 

 fact that the intermediates are more numerous than 



