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



from above, from which it will be seen that the rows of 

 scent-cups, indicated by dotted circles, are separated by 

 glandular areas without cups, whereas in niavius there were 

 no such intermediate rows. As a result of this formation, 

 a transverse section of the scent-patch taken parallel with 

 the general direction of the nervures has a quite different 

 appearance from a section taken at right angles to this 

 direction. PI. XVIII, fig. 1, shows a section parallel with 

 the nervures. On the upper wing surface the scent-cups 

 are seen, each with its gland consisting of a few more or 

 less vacuolated cells with prominent nuclei. The scales 

 are rather thick and apparently solid. It should be 

 noted that the glandular cells are short and do not reach 

 to the ventral or lower wing-surface, thus leaving a space 

 which is possibly filled with fluid during life. Alternating 

 with these cupped glands are what may be termed " blind 

 glands," formed of cells resting on a mass of material 

 which is preserved in the section and presents a horizontally 

 striated appearance. That this is really a mass of material, 

 and not a membrane, is proved by the fact that its appear- 

 ance is continuous in consecutive sections in whatever 

 direction they may be cut. It presents the same appear- 

 ance in sections cut in planes at right angles to each other. 



PI. XVIII, fig. 2, shows a section transverse to the 

 nervures and through a row of the cupped glands. The 

 difference in wing thickness is due to the section having 

 been taken nearer to the edge of the patch. Also, the 

 glands being somewhat flattened, they appear broader in 

 this section. In all other respects the structure of the 

 glands appears to be the same as in fig. 1. 



The abdominal brushes are situated as in A. niavius, 

 and are similar in so far as they consist of chitinous hairs 

 arising from the lining of a membranous bag. They are 

 everted by fluid pressure in the body cavity of the insect, 

 and withdrawn by means of a muscle attached to the 

 proximal end of the bag and to one of the ventral abdo- 

 minal plates. The structure of the brush is, however, 

 very different. When dissected out it is found to be of a 

 fairly uniform thickness up to within about 1 mm. of the 

 proximal end, where it is somewhat sharply constricted 

 and, for the remainder of its length, considerably narrowed. 

 From this constricted portion there arise structures which 

 are very difterent from the well-rounded hairs more com- 

 monly found in these brushes. They are presumably in 



