456 BRITISH LEPIDOPTERA. 



ing of the ist oblique stripes. July 17th. One is now nearly full- 

 grown, in 4th skin, while the majority of the remainder are about 

 to undergo a 4th moult. Fullfed larva in 4th skin (July 17th): 

 Head : not so large or tall as that of S. ocellata in 5th skin, and 

 the yellow bands down the sides of the face are stronger and broader. 

 The resemblance of the apical points and their position on the crown of 

 the head are remarkably close to S. ocellata in everything except colour. 

 This is, however, partly due to the fact that, with S. ocellata in the 

 adult stage (5th skin), the tubercles or processes have greatly dwindled 

 in size. Body : in colour it now almost exactly resembles a typical 

 larva of S. ocellata ; the oblique stripes are much stronger than 

 when the larva first changed into this skin; the subdorsal is very 

 weak and faint ; the spiracles bright red, whilst in 5. ocellata they are 

 deep crimson, and much larger ; anal flap outlined with yellow, in ,S. 

 ocellata with white. [In general shape and habits the larva is still very 

 similar to that of A. populi, and is restless under examination, as is the 

 larva of the latter species. The larva of S. ocellata, on the other hand, 

 will usually retain its stiff and rigid resting-position during the process 

 unless greatly provoked.] The first larva went down on July 21st, 

 having only undergone three, moults. Fifth stadium : The same as 

 in the fourth. By the 24th the majority were fullfed, or nearly so, 

 and the remainder, with one or two exceptions, were in the final 

 stage. So far as I was able to follow them, the majority had four 

 moults, a few only attaining their full growth in the 4th skin (Bacot). 

 House notes that in the first stage there was no difference observ- 

 able between the larva of the hybrid and that of A. populi, very 

 little in the second stage, more difference in that of the third, 

 whilst finally they were more like larvae of S. ocellata than those 

 of A. populi. Birchall observes that the larvae he had were barely 

 distinguishable from those of A. populi, and appeared healthy, but 

 there must have been constitutional weakness, for of 1 6 which 

 assumed the pupal stage, only six produced moths. 



Variation of larva. — Bacot observes ( Ent. Record, x., p. 

 190) that the hybrid larvae of S. hybr. hybridus, bred by House and 

 Birchall, differed in some particulars from those he reared. These 

 latter resembled more closely those of ^. ocellata than those 

 of A. populi, whilst theirs resembled the larvae of A. populi more nearly 

 than those of S. ocellata, although House records that after the 3rd 

 moult his larvae began to lose their A. populi traits, and finally were 

 more like S. ocellata. Different broods may, of course, vary 

 considerably in their tendency to follow either parent species. 

 Frings observes (Soc. Ent., xv., p. 164) that, when young, the 

 larvae had the characteristic form, narrowed anteriorly, as well as 

 the colour and markings, of A. populi of like age, so that had he 

 not known their history he would unhesitatingly have called them 

 that species. But after the third moult, the greater part of them al- 

 ready approached the larva of A. ocellata ; the ground-colour assumed a 

 bluish-white colour in place of yellow-green as heretofore. Neverthe- 

 less, nearly all retained the form of populi larvae. Not till after the 

 4th moult did his specimens assume the form, colour and markings 

 of the ocellata type, but then so completely that they were scarcely 

 distinguishable from true ocellata larva 1 . Only a few specimens had 

 retained the short, pointed horn of populi larvae, but this was 



