456 BRITISH LEPIDOPTERA. 



species, apparently more or less continuously-brooded in its native 

 haunts f, but, in common with several other Sphingid species, it is 

 a well-known migrant, reaching, in the course of its wanderings 

 north, over the greater part of the Palaearctic area, whilst it 

 extends to the southernmost point of Africa. The allied Asiatic 

 M. styx, previously considered a variety of this species, is treated by 

 Jordan and Rothschild (Revision of Sphingidae, pp. 21 et seq.) as 

 a distinct species. In the Palaearctic area, M. atropos appears to be 

 more or less partially triple-brooded in the most suitable seasons, i.e., 

 including the early May-June imagines which are occasionally 

 immigrant here, and which, in hot summers, produce imagines in 

 August and these again in October-November, or, as a continuous 

 brood, from August to November. In other years, the earliest imagines 

 do not appear till July-August, when the only race of home-bred 

 imagines is to be found in October-November, most of the larvae 

 and pupae coming to grief unless protected. Under artificial 

 conditions, and when protected, the autumnal pupae give up their 

 imagines partly the same autumn and partly the next spring and 

 early summer, so that, however short may be the pupal stage in 

 its tropical haunts, such hybernating period as it has belongs to 

 the pupal stage. There seems to be a general opinion on the 

 continent that the July larvae normally produce imagines in the 

 autumn of the same year, whilst the September-October larvae go 

 over the winter, e.g., Siepi says (Feuilles de Jeunes Nahiralistes, 1903) 

 that, "in the Bouches-du-Rhone district, the larvae that pupate in 

 July emerge in September, those that pupate in September and 

 October, or later, go over until the following June," but there is very 

 little, if any, evidence that pupae that go over the winter in a state 

 of nature do successfully emerge even in the south of Europe, 

 although records are not wanting that, in confinement, hybernating 

 pupae successfully produce imagines without artificial heat, and with 

 only the protection of being kept in a room without a fire, i.e., 

 protected against actual frost. Thus we have Rashleigh recording 

 (Z00L, ii., p. 473) an imago bred June 23rd, 1844, from a pupa of 

 the preceding year, at Horton Kirby. Goatley tells "(Z00L, v., 

 p. 1863) of one imago emerging on October nth, 1846, and another 

 on July 7th, 1847, from pupae obtained from larvae that went down at the 

 end of July, 1846; Corbin also records (Ent., vi., p. 520) many larvae 

 pupating well, all producing imagines in the autumn, except one pupa 

 •-vhich lived throughout the winter, and produced an imago in June, 

 whilst Colthrup, from larvae of September, 1899, taken at Dover, 

 obtained imagines October 7th-i8th, 1899, and Howe from 

 pupae of August, 1899, obtained imagines on September 16th and 18th, 

 1899, without any forcing. Anderson notes {Ent., xi., p. 188) that, at 



f Mathew writes {in litt.) : " I think that, in the warm parts of the Mediterranean, 

 the species may be continuously-brooded for I have found the larva? very small in 

 November and December when I have been shooting in Asia Minor. In Britain, 

 all the pups; that I have ever had, have produced their imagines in autumn, or, if they 

 have attempted to pass the winter in the pupal state, have died before the early 

 summer, and, in 1901, when I kept a number of pupae out-of-doors in as natural a 

 state as possible, at Dovercourt, all perished. It seems probable, therefore, that ova 

 deposited in spring here arc from parents that have crossed the English Channel or 

 North !Sea , the build of the moth suggests that it must be a very rapid flier, so that 

 a passage across the sea would occupy only a very short time." 



