316 THE ENTOMOLOdHST's RECORD, 



gives rise to migratory flights in insects, more particularly in lepidop- 

 tera. Some species appear to have a constant migratory instinct. 

 Certain individuals, why we do not know, set oil' and fly apparently as 

 far as they can, in a more or less definite direction, possibly because in 

 some way they recognise that the ground is already occupied, and the 

 excess individuals when the species abounds set off on a migratory 

 quesb for unoccupied ground. The presence of flying specimens of 

 their own (or other) species may be looked upon as the stimulus, and 

 so long as a swarm keeps together, the stimulus persists and the 

 swarm keeps on so long as its individual members can. When a species 

 03cupies a habitat very favourable to it {Pyraiiiri>i canlni in north Africa 

 in perhaps an instance), this abundance must occur almost every year, 

 and it is only by such migration that the species does not exterminate 

 itself by eating up every available scrap of food before the larvae are 

 half grown. Despite their well-known migratory habits, there is 

 reason to imagine that locusts do actually exterminate themselves over 

 large areas. Suppose this self-extermination does actually happen, 

 then the survivors are only a few that did happen to migrate ; their 

 progeny would have to work back to the favourable areas again, and 

 then multiply exceedingly and repeat the process, and, after a time, the 

 species would be descended from these occasional migi'ants in a number 

 of their preceding generations, and it would become an instinct of the race. 

 But with short-lived animals, such as insects, there could not possibly 

 be any of that regular movement due to knowledge, that one readily 

 understands in birds, and the migration of insects appears to us to have, 

 beyond the initial cause of the tendency to dispersal, nothing in common 

 witli that of birds. Trimen, whose observations, being made in South 

 Africa, are especially valuable, w-rites {in lift.) : " There is, in the nature 

 of things, no migration proper, i.e., like that of birds, among such short- 

 lived creatures as butterflies. The great assemblages of the latter, all 

 moving onward in the same direction — which seem to be almost 

 confined to Pierinae and Danainae — appear to be cases of emigration 

 under pressure of abnormal multiplication in specially favourable 

 seasons for the larvae. It has been sometimes stated that only males 

 composed these travelling myriads, but, although that sex was much 

 more numerous in the few instances of the kind that came under my 

 own and Colonel Bowker's observation in South Africa, we took a 

 fair proportion of females among them, and my belief is that the food- 

 plants over a large area, having been denuded of foliage by the 

 unwonted number of larvae, compelled the females to go far afield in 

 search of some district containing fit plants for oviposition, and that 

 they were naturally accompanied by the males. Of course, under normal 

 conditions, every species is doing its best to extend over larger areas, 

 and there is evidence in South Africa (and no doubt also elsewhere) of 

 conspicuous species having extended their range to the southward in 

 recent years, e.g., at Durban, in Natal, a locality well-worked for 

 l)utterflies by a good many resident collectors, it is certain that three 

 species, Crenis rosa, Godartia wakejieldli and Pieris apilleri — which 

 nobody could avoid seeing if they were about — made their first 

 appearance there and on the Natal coast generally, some few years 

 since, as stragglers from the north, but are now firmly established 

 as resident species." 



The causes of insect dispersion, then, so far as they relate to 



