156 THE VARIATION OF ANIMALS IN NATURE 



variation in the Magpie Moth, Abraxas grossulariata, recorded 

 by Kosminsky, 191 2), it is not usually of a type to make 

 species overlap. 



If small differences in the genitalia are not in themselves 

 enough to isolate species, it becomes a matter of importance 

 to decide whether the degree of difference commonly found 

 between species is likely to have been built up in several stages. 

 One series of observations made by Foot and Strobell (1914) 

 suggests that the specific differences must be due to the action 

 of several independent hereditary factors. In crossing two 

 bugs of the genus Euschistus they found that the length of the 

 penis (a specific character) was intermediate in F x and only 

 rarely reached either parental type in F 2 . This suggests that 

 more than one factor for penis-length is involved, and we 

 may suspect that the first stages in this divergence cannot 

 have been very important as means of isolation. 



Apart from the genital armature, difference in size in itself 

 might be expected to play some part in isolation. This would 

 be more important if really closely allied species did more 

 commonly differ markedly in size. We have little very definite 

 information on this subject. Mickel (1924) has shown that 

 the Mutillid wasp Dasymutilla bioculata Cress, has a bimodal 

 variation in size owing to its having two main hosts. A male, 

 however, could mate with a female which was only half his 

 size, so that there was not much chance of the size difference 

 leading to isolation. In insects generally size-variation does 

 not appear to be very important. In a sea slug, however, 

 Crozier (191 8) has shown that mating individuals tend to be 

 of about the same size. But even in the molluscs this is not 

 universal (Robson, 1928). 



II (2). Prevention of effective fertilisation. 



Some degree of sterility on crossing is well known to be 

 a common type of difference between species. The term 

 ' sterility ' is in fact employed to describe a number of dis- 

 tinct phenomena. Only exceptionally do we know exactly 

 what occurs in a particular case. After an apparently 

 effective pairing, we may distinguish between the following 

 possibilities : 



1 . The sperm fails to reach the egg. 



2. The egg is fertilised, but development ceases at an 



early stage. 



