436 BRITISH LEPIDOPTERA. 



1886, on Salix triandra, had the red spots more developed than in 

 any other individual of the species hitherto described, and he figured 

 the most extreme form floe, cit., pi. x., fig. 1) although the other only 

 differed in the absence of the minute dot of the upper row on the 1st 

 thoracic segment. The small size of the upper dot on the 7th ab- 

 dominal segment is noteworthy, and was the same in both individuals. 

 Comparing these larvae with that previously figured (loc. cit., 1884, 

 pp. 27 et seq.), it will be observed that, in these larvae here described, 

 the spots of the two upper rows are larger and extend on to more 

 segments, while those upon the claspers are much larger and more 

 distinct. In breeding large numbers of the species, the spots were 

 found in various degrees of development, upon many of the whitish 

 larvae, although more frequently upon the yellowish varieties. 

 This observation confirms the single instance of a whitish larva 

 with red spots recorded in 1886. With regard to the red-spotted 

 form of larva, Poulton further observes that larvae of S. ocellata, 

 reared from imagines that had themselves come from larvae of 

 the red-spotted form*, were similarly marked. Many other 

 observers note the red-spotted form. Bartel states (Palaeark. 

 Gross- Sehmett.j ii., p. 172) that there are aberrations of the 

 larva in which there are laterally two rows of red or red- 

 brown spots, which are often tolerably large, but which vary much 

 in number and size. This form is often confounded with the 

 corresponding form of the larva of Amorpha populi, from which, 

 however, it may be readily distinguished by the quite different 

 resting-position of the larva and by the colour of the caudal horn. 

 Boscher notes that larvae feeding on sallow and apple showed a 

 variation in colour that caused them to resemble their food ; those 

 on sallow, too, were ornamented with red-brown spots absent in 

 those on apple. Steuart observes that larvae found at Bedford on 

 apple and willow contrasted strongly in colour, the former being 

 of a rich apple-green tint, the latter of a very pale silvery- 

 white and green. Bouskell obtained a white form of the larva 

 at Wicken Fen from white aspen, and the difference in tint of larvae 

 feeding on Salix viminalis and ^. triandra respectively has already 

 been observed. Poulton treats this matter of the variation of the 

 larvae of this species when feeding on different foodplants at length 

 (see antea, vol. i., pp. 85-86). Allchin records (Etit. Wk. Int., ii., 

 p. 164) a brood (30) of larvae in which every individual lost the 

 caudal horn, which gradually disappeared from apex to base, leaving 

 only a slight smooth tubercular prominence, the peculiarity con- 

 tinuing through more than one change of skin. 



Cocoon. — The larvae go underground for pupation, and make a 

 cavity for themselves two inches down in the loose soil, but with 

 little or no silk spinning (Hellins) ; the larva makes a frail cell jn the 

 earth, only very few silk threads are used, although sometimes the 

 larva burrows to the depth of several inches (Bacot) ; the larvae pupate 

 * Peck observes {Can. Ent., viii., p. 239) that the larvae of Smerinthus myops 

 lias red blotches that are not uniform and are more prevalent on larvae of the late brood 

 (although some are entirely green), and correspond in colour to similar spots found 

 on the leaves of the wild cherry at that season. White's notes (Proc. E>it. Soc. 

 J. ond., 1 <S cS 7 , pp. xxiv — xxvii) should be referred to, as also those of Cameron 

 ('Dans-. Etit. Soc. Loud., 1880, p. 69), and those of Miss Gould (loc. cit., 1892, 

 pp. 242—243). 



