THE DISTRIBUTION OF VARIANTS IN NATURE 113 



are differences in size, shape, colour, tarsal claws and width 

 of penis. In the south of England only elegans is found, while 

 in the north of Scotland only depressus and in Ireland only 

 depressus or approximating intermediates. But in N. England 

 and S. Scotland a completely intergrading series is found. 

 There is a similar relation between Gyrinus natator and G. sub- 

 striatus, except that the overlap appears to be wider. 



3. Geographical variation in butterflies. 



In this group geographical races have been more studied 

 than in any other order of insects. Frequently, however, 

 races have been described from too little material and their 

 geographical limits are often very uncertain. A well-studied 

 example is described by Waterhouse (1914, 1922), who deals 

 with the races of an Australian butterfly, Tisiphone abeona. 

 This species is found on the S.E. coast to the seaward of the 

 main dividing range. Five races follow one another in 

 succession down the coast. Two of the races have been proved 

 to be interfertile with a third, which is not in direct contact 

 with either of them. Another two races appear to interbreed 

 and produce peculiar forms not known elsewhere. A similar 

 outburst of peculiar forms where two races meet is recorded 

 by Harrison and Carter (1924) in Aricia medon in England. 

 Doubtless a variety of genetic conditions will determine 

 whether the recombinations resulting from an interracial 

 cross shall produce an intergrading series or an unexpected 

 new type. 



4. Geographical variation in fleas. 



Jordan (1931) gives an interesting example in the variation 

 of the common mouse flea Ctenophthalmus agyrtes. This species 

 is represented in Western Europe by five races — one in England 

 and N.W. France, one in E. France, Germany and Switzer- 

 land, three in Switzerland and N. Italy (separated by various 

 mountain ranges). There are several peculiarities in this 

 distribution. First, the presence of the English Channel has 

 not led to the formation of a peculiar English race. Secondly, 

 the environment of fleas is unusually constant and wide 

 differences in external conditions do not appear to affect them 

 (e.g. in the Alps they occur without modification right up to 

 the tree limit). It is difficult, therefore, to see why races 



