500 BRITISH LEPIDOPTERA. 



exceptions, they may be described as being generally distributed 

 throughout the superfamily, except among the Eumorphids. 

 Whether these hairs should not be considered a generalised rather 

 than a specialised character is still open to doubt. They are 

 certainly found in quite separate and distant superfamilies. One 

 is further inclined to look upon the tumid thoracic segments 

 exhibited in Spliinginae and the swollen ist and 2nd abdominals 

 of the Eumorphidi as specialised characters, for there can be little 

 doubt, one suspects, that the more uniformly cylindrical-shaped 

 larvae have retained, in this particular, and as a necessity of their 

 environment, what one supposes must have been the ancestral 

 form. The loss of the caudal horn, within the phylum itself, 

 must also, it appears to us, be looked upon as a specialisation. 

 It is really remarkable how few useful structural larval characters 

 one appears to find in the Sphingids, i.e., characters sufficiently 

 marked to be of service in allowing one to feel comparatively safe 

 in discussing the phylogeny. The tumid thoracic segments of the 

 Sphingids and Manducids, for example, suggest to us a distinct 

 alliance, as also do the swollen ist and 2nd abdominals of the 

 various Eumorphid branches, in spite of the possible advantages 

 of these structures for the purpose of protection. 



Chapman's notes on the pupa (antea, pp. 371 — 373) suggest that 

 in this stage the most marked modifications of the superfamily are 

 written. Viewed broadly the Sphingids must be looked upon as one 

 of the most successful groups of large lepidoptera in the struggle for 

 existence ; their ability to feed upon the wing and their capacity for 

 amazingly rapid and long-continued flight are their marked 

 characteristics, and hence the modification of tongue and wings 

 have been two of the most marked features leading to success in 

 the struggle for existence, and the modification of the pupa to 

 meet the requirements with regard to these organs forms a fruitful 

 basis for study. As compared with the Amorphid pupa, the Sphingid 

 pupa is longer, more attenuated towards the anal end, has a much 

 smoother skin-surface, and exhibits, in some cases, considerable 

 power of movement. In spite of the smooth appearance of the 

 Sphingid pupal skin, Chapman has discovered the presence of 

 unquestionable hairs or bristles of microscopic size on various pupne, 

 viz., Sesia stellatarum, Hippotion celerio, Eumorpha elpeno?', Man due a 

 atropos, Smerinthus ocellata, &c. These bristles are placed in a 

 circular opening in the chitin, are there jointed, and then diminish 

 regularly to a sharp point, like typical tactile bristles. They are 

 in considerable numbers on the pupae of H. celerio and E. elpenor, 

 occupying exactly the same positions as in S. stellatarum, viz., 

 the anterior borders of the segments ventrally, and especially on 

 the modified portions close to the intersegmental membrane. In 

 M. atropos and iS". ocellata they are smaller and less abundant, 

 but here they occur on the central zone of the segments and are 

 not found near either the anterior or posterior borders. The hairs 

 are not to be seen unless a piece of the pupa be prepared and 

 mounted ; sufficient material has been examined to suggest that 

 probably all Sphingid pupaa have these hairs, and that their 

 distribution is almost identical in Sesia ( Macroglossum) and Eianorpha, 

 different in Amorplia and Sphinx. 



