168 Dr. H. Eltringham's Further Observations on the 



will become flattened during expansion. The patch is 

 already flat, and being incapable of extension, becomes 

 covered by a fold. PI. XIX, fig. 24, shows a cup and a 

 socket-cell from the imaginal wing. These cells are highly 

 vacuolated, finely granular, and occasionally provided 

 with more than one nucleus, especially the socket-cells. 

 Mayer {loc. cit.) describes the nuclei of the scale-producing 

 cells in the wing of D. plexippus as occasionally undergoing 

 amitotic division, so that we may suppose this process 

 also to take place in the present species (a broken scale is 

 shown arising from the right-hand cell). The " cups " 

 have no relic of a scale-stalk as in D. lotis, but appear to 

 possess a central pore, though this is scarcely so obvious 

 as in some species of Amauris. The scales lining the 

 pocket have an appearance which suggests that they are 

 traversed by pores, though the structure is less easily 

 discernible than in lotis. There is no evidence in either 

 pocket or brush of the production of any dust material. 



A longitudinal section of the brush within the pupa 

 shows the hairs arising from highly vacuolated cells, each 

 with a large nucleus. The cytoplasm is finely granular, 

 and stains less readily towards the extremity from which 

 the hair arises. 



Danaida lotis. Cram. 



In this species there is a fold in the hind- wing resembling 

 that of D. chrysippus, and a pair of abdominal brushes, the 

 latter presenting no special features. The hairs are of one 

 kind and strongly chitinised. In general their sections 

 resemble those of chrysippus, and small spherical particles 

 of a dried secretion may be observed entangled amongst 

 them. Some hairs are flattened for a portion of their 

 length or depart in other ways from the more truly circular 

 section usually exhibited. Their functions are probably 

 merely mechanical. PI. XVI, fig. 2, is a photograph of a 

 section of the brush not far from the distal end. PI. XVI, 

 fig. 3, shows a photograph of a transverse section of the 

 wing-pocket, which is seen to be a more or less spiral fold 

 of the wing. The pocket contains no dust material, and 

 as none is found in the brush it is evidently not produced 

 in this species. A surface view of the membrane with 

 which the pocket is lined shows that alternate scale- 

 sockets are modified into structures resembhng scent-cups, 

 whilst the remaining sockets present a more or less normal 



