Structure of Scent Organs in Male Danaine Butterflies. 171 



opens into the base of the brush-bag proper, in which latter 

 a few of the hairs can be seen cut across. Fig, 4 

 shows that the vesicle consists of an outer layer of cells 

 with conspicuous nuclei. From these cells arise narrow 

 funnel-like structures somewhat chitinised and homo- 

 logous with hair-sockets. From these, again, spring very 

 delicate tubes of irregular section which more or less fill the 

 cavity. PI. XX, fig. 31, shows one of these funnel-sockets 

 arising from the deep layer of vacuolated cells, the latter 

 being partly seen in section. The delicate tube arising 

 from the socket bulges out and overhangs the latter. The 

 socket shows a core of cytoplasm which, judging from 

 stained preparations, appears to divide at the distal end. 

 The whole structure strongly suggests that some secretion 

 is discharged from the funnel-sockets into the vesicle, 

 through the delicate tubes, and passes thence into the 

 cavity of the brush-bag proper. PL XVI, fig. 6, represents 

 a photograph of a longitudinal, and PL XVII, fig. 2, a trans- 

 verse, section of the brush-bag. The latter is distinctly 

 glandular, many more cells being present than are found 

 in species of Amauris. The hair-sockets are long and more 

 or less funnel-shaped as in the vesicle, though somewhat 

 less constricted at the middle of their length. PL XX, 

 fig. 30, shows one of the hair-sockets under a higher power. 

 Several cells are seen to form a basal gland in connection 

 with a single hair, and at one side appears the lumen of a 

 trachea, many of which are found in intimate connection 

 with the brush-bag in this species. PL XX, fig. 32, shows 

 a transverse section of one of the brush-hairs at a point 

 towards its extremity, whilst fig. 33 represents the appear- 

 ance of a longitudinal view of the same portion of the same 

 hair. Near their origin and for a considerable portion of 

 their length these hairs are either comparatively smooth 

 or only faintly striated, but towards the distal extremity 

 they develop more of less regular lines of projections. I 

 have been unable actually to see pores in these hairs, but 

 that there are openings to the exterior seems a reasonable 

 conclusion from the glandular nature of the brush-bag. 

 One other point should be noticed. The wall of the bag 

 is glandular and gives rise to hairs along nearly its whole 

 length, so that the appearance on e version must be that 

 of a test-tube-brush instead of a spherical tuft as in 

 Amauris. I have found no evidence of any dust-apparatus 

 in this species. 



