422 N. E. MCINDOO. 



them. They may be inverted flask-shaped as found In the legs 

 of Epilachna (Figs. i6 and 17) and in the wings of Passalus 

 (Fig. 20). Some have the shape of a flask without the neck 

 (Figs. 9, 10 and 12). Some are long and slender like fingers or 

 test tubes (Figs. 11, 18, 19 and 21). 



Their sizes also vary much. The length of a pore always 

 depends on the thickness of the chitin. The diameters of the 

 pores of a small beetle (Fig. 25) may be as large, or even larger 

 (Figs. 9 and 10) than the diameters of the pores of a large beetle 

 (Figs. 13-15)- 



A chitinous cone is always present, although it may sometimes 

 be almost indiscernible. It invariably has the same shade of 

 coloration (Fig. 17, Con) as the remaining chitin (Fig. 17, Ch'^ 

 which is formed after the insect has emerged into the imago stage. 

 This is the first time that the writer has been able to determine 

 definitely the formation of the cones. In all the hymenopterous 

 insects studied by the writer, the chitinous integument is prac- 

 tically developed when the insects emerge, but in most beetles 

 only about one third of the chitin is formed when the insects 

 emerge. Since this is true the hypodermal cells are still large 

 and they are rapidly secreting a substance which forms new chitin. 

 Their external ends stand in contact with the chitin, and when 

 no chitin is present they send processes into all holes or cavities 

 in the chitin. Thus the hypodermal cell (Fig. 23, HypC) at 

 the mouth of each olfactory pore sends a process into the pore. 

 Since the sense fiber has entered the pore aperture before the 

 cone is formed, the latter is formed at the external end of the 

 pore around the sense fiber. When the chitinous integument 

 (Fig. 17) is fully developed no hypodermal processes run into 

 the pores and the hypodermal cells are very small. 



The sense cells are always spindle-shaped (Figs. 12, 13, 16-19 

 and 23, SC). Only occasionally is an entire sense cell seen in a 

 cross section, because the entire cell seldom lies in the same plane 

 as that of the section. More entire sense cells may be seen in 

 longitudinal sections, but even in these the cells are usually eut 

 in two. Entire sense cells were best seen in the oblique sections 

 through the peduncles of the elytra of Passalus and Epilachna. 

 The nucleus (Figs. 13 and 23, SCNuc) of the sense cell is always 



