148 The Classification of Insects 



are, as a rule, infertile. It has indeed been proposed 

 to make the infertility of hybrids the test of " specific 

 distinctness." But experiments have shown, in the 

 case of the European " Emperor" moths for instance, 

 that a male hybrid can be crossed with a female of 

 one of its parent-species so as to produce offspring 

 (85). Such occurrences must, however, be very rare 

 in nature ; and it is evident that if intercrossing be- 

 tween different species were at all general, specific 

 distinctions would soon vanish. 



There are various ways in which intercrossing can 

 be prevented (91). A variety in process of develop- 

 ment into a new species may be geographically isolated 

 from its parent stock, having migrated into a new 

 locality, where changed conditions are inducing modi- 

 fication in its form or appearance. There may be a 

 preference on the part of individuals to pair with 

 members of their own variety or race ; this factor, 

 though of undoubted value among back-boned animals, 

 is probably of little effect among insects. Intercross- 

 ing may be rendered impossible by a difference in 

 the time of year at which the two separating forms 

 become mature. A very interesting case of this form 

 of isolation is furnished by two British moths — 

 Tephrosia crepuscular la and T. hisortata, so closely related 

 that it is almost impossible to distinguish them by 

 structural characters, yet with distinct life-cycles, the 

 former appearing in the perfect state once a year 

 during May and June, the latter in March and April 

 and again in July. In many insects intercrossing 

 is rendered impossible by variation in the genital 

 armature ; the male claspers become so modified that 

 they will only fit the corresponding parts of a female 

 of their own race. The well-known North American 

 "seventeen-year Cicad" has a larger and a smaller 

 race or sub-species, which usually differ markedly in 



