278 THE VARIATION OF ANIMALS IN NATURE 



In species such as Psammochares plumbeus, which burrows in 

 loose sand, the well-developed comb makes the front legs a 

 much more efficient organ for scraping away the soil. In 

 other species with a rudimentary comb its value is doubtful, 

 and a considerable number of species without a comb seem to 

 be able to burrow equally well. It may be mentioned that a 

 similar comb is developed in a number of species of sand- 

 nesting wasps belonging to other families. 



One other modification which appears to be of some value 

 is the peculiar flattened head and thorax and thick fore femora 

 of the species of Aporus which, preying on spiders living in 

 burrows (Ferton, 1901, p. 121), are much more fossorial 

 than the other species. 



If we consider the habits of the species the problem is 

 equally perplexing. The species of Pseudagenia build mud cells ; 

 those of Deuteragenia use ready-made crevices, old nests of 

 other insects, or snail-shells ; most Psammocharids dig burrows 

 in sand or earth ; Parrqferreola lays its egg on a spider, which 

 runs about with it in the open ; Ceropales and a few species of 

 Psammochares are parasitic on their allies. Apart from the 

 exceptions already mentioned it is impossible to seize on any 

 point in their structure which specially fits them for their mode 

 of life. Further, these variations in habit themselves do not 

 seem of much use to the species : all types of nest seem equally 

 good, as far as we can see. There is a certain amount of 

 specialisation in the nature of the prey, though further work 

 would probably show that many species are more polyphagous 

 than is at present known. How far such food differences can 

 be considered adaptive is considered later (p. 301). 



The following are the main conclusions to be drawn from 

 this example, which could be reduplicated again and again 

 from other divisions of the animal kingdom. 



1. The majority of characters, separating either sub- 

 families, genera or species, have no known use to the species 

 and have no known relation to the special habits. In the actual 

 example, there is no case of a useful character separating 

 closely allied species : the characters which are useful (and 

 possibly adaptive) are generic, or they separate distinct groups 

 of species within the genus. 



2. The differences in habits also do not appear to be 

 definitely adaptive. We can see that, if a wasp decides to 



