THE DISTRIBUTION OF VARIANTS IN NATURE 121 



structure, while Myers (1926) states that the song is the most 

 stable single character in the cicadas, though here morpho- 

 logical differences are also fairly conspicuous. In grasshoppers, 

 taken as a whole, structure would appear to be more distinctive 

 than song, though the latter is difficult to define owing to 

 environmental effects (temperature, presence of other indi- 

 viduals, etc.). Promptoff (1930) records statistical local 

 differences in the song of the Chaffinch in two different areas 

 of Russia. Again Kinsey (1930), in his valuable revision of 

 the genus Cynips {Spathegaster and Dryophanta, auctt.), finds 

 several pairs of species or races which are only to be distin- 

 guished by their galls. His actual summary for the genus 

 (p. 38) is as follows : 52 species have structure more distinctive 

 than the galls ; 24 species have galls more distinctive than 

 structure ; 17 species have the two equally distinctive. The 

 formation of the galls is known to be due to the action of the 

 larval gall-wasp. 



Thorpe's account (1929, 1931 ) of the races of the Small 

 Ermine Moth (Hyponomeuta padella) shows that structural 

 and physiological differences are about evenly balanced, 

 neither being very great. There is a distinct food preference, 

 indicated by oviposition-response of the female and even more 

 by larval choice ; members of one race cross with one another 

 more easily than they do with members of the other, and there 

 are slight overlapping colour-differences between the adult 

 moths ; the larvae construct different types of cocoons. The 

 two forms of the Human Louse (Pediculus) are somewhat more 

 distinct and crossing is liable to lead to abnormalities in the 

 hybrids. 



Unfortunately data as to selective mating between races are 

 very scanty. If we knew more it might be possible to regard 

 species differing only in the male genitalia as a special type 

 of biological race. In a number of forms (Lucilia, the blow- 

 fly ; Chironomus-midgcs, etc.) the females are morphologically 

 indistinguishable and the maintenance of the species must 

 depend on the reactions of the male, perhaps to a scent emitted 

 by his mate. 



In connection with biological races it is interesting to 

 consider the differences which may be found in the develop- 

 mental stages of animals, especially in larval forms. If we 

 eliminate species which are still imperfectly known, it is probable 



