53 



interior. It will be said it never has been looked for, and that may 

 be the explanation ; but one would expect in any case in which it 

 occurred to have some deformity of the thoracic segment affected 

 that would have suggested closer examination. 



The other theory of the origin of these parts that would explain 

 this specimen is that they have a truly internal origin, that they are 

 developed as a portion of the ducts, etc., and have no relations with 

 the surface, till at the pupal stage they reach it. The ditificulty in 

 this explanation is the dermal structure of the surfaces of the 

 clasps. 



Whatever the true explanation may be, it is obvious that more 

 accurate observations on the development of these parts require to 

 be undertaken. 



The male tubercles of the ninth abdominal segment of the pupa 

 present a feature that is probably of importance in dealing with the 

 questions raised by Mr. Burrows' specimen. They are separated from 

 the rest of the ninth segment by a suture, and we may compare this 

 suture with the lines of delimitation of the pupal wings, which are 

 certainly dermal structures (superficially, at least), which have returned 

 to the surface after an internal existence due to invagination. Where 

 the wings impinge against the antennae or abdominal segments there 

 is a very marked suture, very marked because we have here two 

 structures in contact, which are not, however, continuous, as two 

 adjacent segments are, but merely in apposition ; but if we examine 

 the place when they are in structural continuity with the mesothorax 

 (or metathorax) we can only approximately say where the junction is, 

 but find that there is no trace whatever of a suture. The continuous 

 suture round the male tubercles appears to prove that they are not 

 portions of the segment returned to the surface. 



I have assumed that the male tubercles represent the clasps. I 

 think this is generally agreed, but a certain amount of further proof 

 is afforded by the specimen before us, as we see in it one of the 

 clasps just escaped from, or rather, perhaps, still slightly involved in, 

 one of the hemispherical cups. 



I have been assuming, as it is very usual to do, that these 

 tubercles are in the ninth abdominal segment, but it is more than 

 doubtful whether this is so; their posterior borders are in contact 

 with the tenth segment, and their real position is not in the ninth 

 segment, but between it and the tenth — a more probable place for an 

 interior structure to reach the surface than actually through the con- 

 tinuous surface of the segment itself. It fully accords also with 

 the fact that the structures it represents are in the imago along the 

 posterior border of the ninth segment, /. e. between the two seg- 

 ments. 



