512 BBITISH LEPIDOPTERA. 



two pupae squeezed together and with no partition between them. 

 Foddy mentions, in 1890, three cocoons with two pupae in each. 

 Pearson also records an instance of a cocoon with two pupae. Vaughan 

 found among many cocoons made by this species, two much larger 

 than the rest, and each of these was found on being broken open to 

 contain two dead larvae. He further notes that Bond had observed a 

 cocoon with three inmates. Thouless records two similar cocoons each 

 containing two pupae. 



Variation in colour of cocoons. — Poulton found that the cocoons 

 of this species were of various shades of colour, and supposed this to 

 be due to the influence of the colour of the environment, among which 

 the cocoons were spun, upon the larva, and he further assumed that the 

 latter had the power of adjusting the colour of the cocoons to their en- 

 vironment as a means of concealment from their enemies. Newman 

 reported that cocoons spun on white paper were creamy-white in 

 colour, whilst those spun among leaves were dark brown in colour, the 

 leaves becoming when dead of the same tint as the cocoon. From 

 this also it was deduced that the dark brown colour was highly pro- 

 tective. It was, however, discovered that the normal brown colour of 

 the cocoons of this species was due to a red-brown viscous fluid, which 

 is voided on the silk by the spinning larva, and which becomes dark 

 brown on exposure. This colour has, undoubtedly, a great protective 

 value in nature, and pale cocoons are very rarely observed except when 

 spun by larvae reared under artificial conditions. The explanation of 

 the coloration of these pale cocoons now became simple. It was 

 observed that if a larva were disturbed just as it was about to spin, it 

 voided the greater part of the red-brown fluid owing to the disturbance, 

 and then, whatever the colour of the environment, the larva spun a 

 pale cocoon, having already lost the colouring matter, with which it 

 would, normally have stained the cocoon. It was also noticed that if 

 larvae were removed from their food and starved for three or four days 

 previous to spinning, they also spun pale cocoons ; and so, also, did 

 larvae whose constitutions had been undermined by parasites. Some 

 of these j)ale cocoons were, indeed, quite white. Bateson considers 

 that the colouring matter is probably a chlorophyll product, and this 

 is the reason that the hue of the normally dark coloured cocoons, bears 

 such a striking resemblance to decayed or decaying leaves. He further 

 considers that it comes from the alimentary canal, but is not certain 

 whether it be voided from the mouth or anus. It would appear, 

 therefore, that the substance with which the silk is stained, is entirely 

 different from that with which the cocoon is hardened. The variation 

 in the colour of tho cocoons extends from dark black-brown (the tint 

 of black coffee), through dark brown, pale brown, cream-colour, to 

 those that are absolutely white. The pale cocoons, as might be 

 supposed are, as a rule, thin and papery, whilst the dark cocoons are 

 stout, stiff and shiny. Bussell notes of a batch of cocoons that he had 

 from larvae reared in confinement that the majority were of a whitish 

 tint, a few were brownish, and one or two pale greenish. 



Pupa. — The pupa is large for the size of the cocoon, the segments 

 increasing in width from the head to the fourth abdominal. It is pale 

 brown in colour, the ventral appendages and wings paler, and some- 

 what transparent. The dorsal area is rather darker, and there is a 

 slender mediodorsal, blackish line. Ventrally: The head is prominent, 



