THE DEATH'S HEAD MOTH 241 



difficult problem to solve), by supposing that its numbers 

 are at such times largely reinforced by immigrants from 

 the Continent. 



This subject of migration of Lepidoptera is beginning 

 to attract the attention of entomologists, and it is 

 expected that investigation will show that the abun- 

 dance or scarcity of many species, besides the Death's 

 Head Moth, is largely regulated by immigration. It 

 remains for the future to disclose what are all the 

 influences which cause the migration. 



Somewhat unreliable in its times of appearance with 

 us, usually the moth emerges in the autumn, in Sep- 

 tember or October — September appearing to be pre- 

 eminently the month when it is on the wing — and 

 probably hybernates, or else it lies in pupa through the 

 winter ; and hybernated specimens, or specimens from 

 hybernated pupse, are sometimes found in June. In a 

 state of nature, in the opinion of some entomologists, 

 the moths always emerge in the autumn, but when the 

 pupa is kept under artificial conditions, the moth often 

 appears in spring. Thus, Atropos seems taken from 

 June to October inclusive, and it occasionally appears in 

 November, though the latter date is exceptional. As 

 for the larva, examples almost full-fed have been 

 obtained at the end of June ; the latest recorded date 

 is October; but July and August, as already implied, 

 appear to be the great months for this stage. It is 



