158 F. X. WILLIAMS 



in the crop of Blattidse (Petrunkewitsch, '99), etc. Geipel ('15) 

 however was unable to find distinct tracheal end-cells in the 

 luminous elaterid (Pyrophorus noctiluca). The end-cells have 

 their cytoplasm continuous with the tracheal epithelium on the 

 one hand, and on the other, spread far along the tracheoles, if 

 they do not entirely envelop the latter. 



In general, the tracheae (T, fig. 28) which supply the light- 

 organ, extend transversely from spiracle to spiracle and are con- 

 nected by lateral branches, well illustrated by Geipel ('15), over 

 the dorsal layer and send down numerous, usually vertical 

 branches through the organ. The epithelium which invests and 

 secretes the chitinous intima of the tracheae, is, in the non- 

 photogenic layer, as elsewhere without the organ, rather thick, 

 with numerous deeply-staining nuclei which lie parallel to the 

 length of the chitinous tube itself, already being secreted at 

 this stage. In the photogenic layer however, the tracheal epi- 

 thelium is much thicker (E, fig. 28), for here the cells are not 

 parallel to the length of the tube but have their long axis at 

 right angles to it. The nuclei are in the distal portion of the 

 cells. It is in this region that the tracheal end-cells are formed. 

 They occur on the numerous little lateral branches which radiate 

 from the vertical trunks (EC, fig. 29), and are close together or 

 contiguous, thus forming a sort of outer cylinder investing the 

 now thin tracheal epithelium on the one hand, and on the other, 

 closely applied to the surrounding photogenic cells. The tracheoles 

 most commonly bifurcate and anastomose, and may often appear- 

 in cross section as thick rings between the photogenic cells. 

 Geipel ('15) who has carefully studied the end-cells in Photinus 

 marginellatus, concludes that the capillary ends blindly in the 

 middle of the end-cell ; that there is no direct connection between 

 this capillary and the processes of the end-cell, that these pro- 

 cesses do not anastomose, and finally that the latter are inter- 

 cellular. Having no osmic acid fixations of this material I am 

 unable to add anything to these disputed points, except that I 

 have found no end-cell processes penetrating the photogenic 

 cells, but have found them to be intercellular. These processes, 

 at least in a number of cases, almost certainly do anastomose, 



