294 THE VARIATION OF ANIMALS IN NATURE 



be modified, including the mandibles, antennae, legs or abdo- 

 men, and there is a very strong prima facie case for regarding 

 the modifications as useful, the close contact of the sexes during 

 a period long enough for successful fertilisation being an 

 evident necessity. It is further well established that the 

 detailed structure of grasping organs usually differs from 

 species to species, although it is rarely possible to show any 

 detailed correlation between the organs of different types of 

 males and the structure of the corresponding parts in their 

 respective females. Not only is there great specific diversity 

 in the male without corresponding co-adaptation in the female, 

 but the actual development of grasping organs in the males is 

 highly sporadic. Thus, besides the marked specific differences 

 in the nature of these structures, it is quite common to find them 

 developed only in a few species in a genus or in a few genera in 

 a family. Of two species, otherwise very similar in structure 

 and habits, one will have a highly specialised grasping organ, 

 the other none. We will give one example from the Hymeno- 

 ptera. The small wasps of the family Crabronidae often have 

 the fore tibia, the fore basitarsus, or both, enlarged in the 

 male (fig. 27). The enlargement varies greatly in degree, from 

 a very slight increase in width to a condition in which the whole 

 apical part of the leg forms an elaborate shield which cannot 

 be used for ordinary walking ; in almost every case the 

 details of the modification are highly specific. Kohl (1915), in 

 his monograph of the palaearctic species, divides the old genus 

 Crabro into ten species-groups (by many regarded as genera or 

 subfamilies), including in all 167 species. Only forty-two 

 species are known from female specimens or have been insuffi- 

 ciently described, and of the remaining 125 species 39 have 

 the modified foreleg ; these are distributed amongst seven of the 

 ten species-groups. Bristowe (1929^, p. 348) has reviewed the 

 structures used by male spiders for grasping the females. The 

 differences appear to be usually familial or generic, but there 

 is an interesting example in the genus Pachygnatha, in which 

 the male cheliceras grasp those of the female during mating. 

 Here a marked difference in the teeth on the male chelicerje of 

 two species corresponds to two different methods of gripping the 

 female, although her chelicerae are not actually modified. 



As far as the habits are known there is nothing to show that 

 the species with grasping forelegs have a greater need for 



