148 THE VARIATION OF ANIMALS IN NATURE 



Jeannel (191 1 and 1926) has shown how cave species tend to 

 be confined to one or a few adjacent caves. Doubtless blind- 

 ness is not the only agency confining a species to its own cave 

 (for, just as some apterous species are widespread, some 

 blind species also occur in the open), but it probably plays 

 an important role. This subject is discussed elsewhere 

 (Chapter VII, p. 269). 



The truth is that we know extremely little about the 

 powers of dispersal of animals, apart from the more sensational 

 migrations, and it is possible that some of the anomalous 

 differences in the variability of different species might become 

 clearer if we knew more. It is especially difficult to trace the 

 minor wanderings which occur within the normal area inhabited 

 by the species. The frequency of such wanderings must largely 

 determine the homogeneity (or the reverse) of the species, and 

 this, in turn, has an important effect on the significance of 

 geographical isolation, since any isolated population has a 

 greater or less chance of differing from the norm of the 

 species. The converse, however, is equally important, viz. 

 that no true random mating can occur in any species, because 

 the chance of an encounter between individuals separated by 

 a few miles of country is relatively low. Even in long-lived 

 and wide-ranging forms this must have some effect, and in 

 small, short-lived species, unless the specific range is extremely 

 small, the results must be very significant. 



II (i). Recognition marks. 



We have very little information as to the function of 

 recognition marks (or odours) amongst animals, apart from 

 structures (or odours) specifically connected with mating. 

 Something of the sort is evidently present in most gregarious 

 animals. Thus Ward (1904) has shown that the bats in 

 certain caves in Mexico roost according to their species. 

 Feuerborn (1922) has given some evidence which suggests 

 that flies of the family Psychodidae recognise the species as 

 well as the sex of other individuals. He suggests that certain 

 glands, present in both sexes, produce odours on which this 

 faculty depends. Seitz (1894) long ago suggested that some 

 Lepidoptera may produce both specific and sexual odours. 

 Colour must also play a part in species recognition, as Eltring- 

 ham (191 9) has shown in certain butterflies in which the sexes 



