VERTICAL DISTRIBUTION OF THE LEPIDOPTERA. 347 



well known to those who rear exotic forms in this country. This 

 is possibly the principal reason why Acherontia atrojws and 

 several other species of Sphingidae do not occur permanently 

 more plentifully in the British Isles than is the case, although 

 certain plants which the larvaG are known to feed upon are found 

 in sufficient abundance. 



Deficiency of moisture seems in a similar way to circumscribe 

 the distribution of certain species in other countries. These are 

 also apparently instances of the direct influence of the physical 

 conditions. 



The distribution of the Lepidoptera both vertically and hori- 

 zontally is not by any means always co-extensive with that of 

 their pabula. In some instances it may be more extensive than 

 that of their usual food-plant, when they select a closely-allied 

 species in its stead ; but in the majority of cases the plant 

 possesses the widest distribution of the two. In the latter the 

 struggle for existence with their natural enemies proves too 

 great for the insects, notwithstanding the abundance of their 

 food supply. 



We have good grounds for supposing that many species of 

 animals and plants formerly existed in situations and at eleva- 

 tions where the climatal conditions were very different from 

 those which they occupy at the present day. There are numer- 

 ous examples of this to be found in the Alps, the correct causes 

 of which we know. The chamois, for instance, which formerly 

 frequented the pine forests, are now seldom seen there, except 

 in the winter time, having been forced to flee upwards before 

 their chief enemy, man. The bear, the lynx, and the ibex have 

 similarly had to seek refuge among the higher and more in- 

 accessible peaks. Many insects are also now found less com- 

 monly than formerly at the lower elevations, owing to the 

 extension of cultivation. Numerous other similar instances 

 might be adduced all in support of the superior power of the 

 organic environment. 



From the discontinuity in the present distribution of many 

 genera and species of fauna and flora, their former more exten- 

 sive area which they occupied may be estimated. Take, for 

 instance, the Pyrenean desman or water mole {Mygale pyreiiaica), 

 which possesses its nearest relative in the steppes of South 

 Kussia. Then we have two species of butterflies in Europe, 

 namely, Zegris eiq^heme, which occurs in Andalusia, South 

 Russia, and Asia Minor; and also Satyrus hippolyte, which is 

 likewise found in the Iberian Peninsula, and again in the South 

 of Russia, but not in any intermediate locality. All of these 

 must once have possessed a continuous distribution in order to 

 account for their present dispersion. 



The genus CEnis, or its immediate stirps, undoubtedly pos- 

 sessed in former times an almost complete cosmopolite distribu- 



