159 



THE CLASSIFICATION OF GRACILARIA AND ALLIED 



GENEEA. 



By T. A. Chapman, M.D., F.E.S. 



(Concluded from p. 142.) 



In viewing the classifications that have been made of the 

 Gracilariads, we have seen how they have always been placed 

 close to the Lyonetiads, and usually more or less mixed up 

 with them ; in accordance, no doubt, with the fact that there 

 is probably no very profound difference in any important cha- 

 racter between them in the imago state. The superficial re- 

 semblance of Leucoptera (Cemiostoma) to Phyllocnistis is very 

 close indeed. 



I am not prepared to advance any larval characters that 

 suffice to distinguish the Lyonetiadae, but there is a very definite 

 pupal character that is, I think, both inclusive and exclusive. 

 This is that the pupa is entirely immobile, and in a special 

 manner that distinguishes it from other immobile pupao, such as 

 Perittia, Thyris, &c. It has never passed through an ordinary 

 obtect stage, in which the wings, &c., usually are attached down 

 to the fourth abdominal segment only. Here the wings, &c., are 

 attached for their whole length to the abdominal segments. In 

 Lyonetia the wings and abdominal segments form one mass that 

 tapers to a point — a point to which the wings, antennae, and 

 third legs reach, as well as the abdominal extremity. This mass 

 is not, however, as solid as it looks. As in Pupae Incompletse, 

 all the appendages separate from each other, and with rather the 

 facility one finds in Nepticida than with the difficulty that one 

 meets with, say, in Tortrix. When the appendages are pushed 

 aside, one inclines to doubt whether they were really at all 

 adherent to the abdominal segments, and these again are found 

 to be quite movable. Still, it is tolerably certain that no move- 

 ment whatever takes place in the living pupa, hardly even on 

 dehiscence. 



In Leucoptera the appendages do not come so far down, and 

 the pupa itself is comparatively short and dumpy ; still, it agrees 

 with Lyonetia in essential structure. We have here, then, a 

 pupa very different indeed from that of Gracilariads. But is it 

 after all very far off from them, if its probable evolution is con- 

 sidered ? 



If we confine the name Obtect to those pupae that have reached 

 that character by the same route as, say, Noctuae have, or by 

 some very similar one, then the Lyonetiads are certainly not 

 Obtectae. They are a separate modification of the Pupae Incom- 

 pletae. They are consolidated so far that the segments have lost 

 mobility, but are still very primitive as regards the soldering 



