ii2 THE VARIATION OF ANIMALS IN NATURE 



Thirdly, we are a little doubtful if the data for various groups 

 are really comparable and whether samples of populations 

 consisting of a few individuals, such as are used in mammals 

 and birds, afford a sound basis for distinguishing local races. 

 Sumner (191 8, p. 292) seems to express this doubt concerning 

 the races of small mammals. We do not as a matter of fact 

 think this vitiates the general principle, for there are groups 

 (e.g. the Cephalopoda) in which the numbers used are equally 

 low and yet few races are recorded. What we feel is that 

 comparable data are required and that some modifications 

 of the alleged incidence of race-formation might result, if 

 large numbers were regularly used. 



In conclusion, it seems likely that geographical variation 

 will ultimately be found to be as frequent in groups like 

 terrestrial arthropods and molluscs as in vertebrates. Where 

 all the present evidence is against the likelihood of such races 

 being discovered, it will be usually found that special habits 

 and other factors that prevent isolation and colony-formation 

 are mainly responsible. Again, in some species we must 

 look to the inherent capacity for variation as a cause. It must 

 always be recalled that our knowledge of variation is at present 

 very unequal in its incidence in the various groups and is less 

 easily obtainable in some than in others. 



Examples may now be given of the occurrence and dis- 

 tribution of geographical variation in various groups. 



1. Geographical variation in Lygaeus kalmii (Hemiptera, Hetero- 



ptera) . 



Parshley (1923) has shown that there is a clearly marked 

 eastern and western race in the United States. These meet 

 at a line joining Winnipeg to Brownsville, Texas. Along this 

 line intermediates occur, which cannot be referred to either 

 race. Since the species is, in addition, highly variable in 

 colour, it is only possible to recognise the geographically 

 significant characters by careful study. 



2. Geographical variation in water beetles. 



Omer Cooper (1931) summarises the evidence for two 

 examples. The extremes in each case are treated as species, 

 but they correspond to what are called geographical races in 

 other groups. Thus in Deronectes depressus and D. elegans there 



