ISOLATION 145 



In longer-lived animals with a definite breeding season a 

 comparable state of affairs exists, but isolation appears to be 

 much more partial, except sometimes between races of one 

 species, e.g. Rana esculenta (Cuenot, 1921), Sepia (Cuenot, 1917)5 

 Crangon and Orchestia (Plate, 191 3). In addition long-lived 

 animals appear to be largely those which also evolve mainly 

 through geographical races, in which the breeding period is not 

 likely to be an important factor in isolation. When the races 

 have evolved so far that their ranges overlap, and we find two 

 species living side by side, other factors often override any 

 original differences in the breeding period. It may be sus- 

 pected that any environmental pressure tending to reassimilate 

 two rhythms would sooner or later be effective and, if the two 

 forms were not by that time intersterile or isolated in other 

 ways, they would be reunited. We might, therefore, expect 

 that differences in seasonal occurrence in the breeding season 

 would usually be found as specific only between forms still 

 quite closely allied. 



I [c] . General habitat. 



It must be very rarely that two closely allied species have 

 so sharply different habitats that no crossing could occur. In 

 a country like England, where no one habitat covers an exten- 

 sive area without interruption, this is obvious, but in some 

 continental areas habitat-differences may be much more 

 important, though no clear distinction can be drawn in this 

 case between restriction to one habitat and to one geographical 

 area. Even on a much smaller scale, however, habitat- 

 differences will lead to some degree of selective mating, 

 especially with forms with low powers of dispersal. This small 

 contribution towards the establishment of isolation is important 

 because some degree of differentiation in habitat preference 

 must be regarded as one of the commonest of specific characters. 

 As the general facts are well known to most zoologists, we will 

 give a few instances, confining our attention mainly to pairs 

 of closely allied forms. 



von Lengerken (191 7) and Macgillavry (1927) record that 

 the tiger-beetle Cicindela hybrida L. is restricted to the part of 

 sand-dunes which is fixed by vegetation. The subspecies (or 

 species) C. maritima Latr. occurs only on stony places on the 

 actual strand. In Holland, however, a darker race of maritima 



