I06 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 85 



regard the terminal appendages of the Dytiscus larva as the cerci. 

 However, the evidence presented by Blunck of the fusion of the first 

 and second segments of the abdomen is not generally accepted by 

 students of Coleoptera, and is not convincing, while the fact that the 

 segment preceding the appendicular processes bears the eighth pair of 

 abdominal spiracles would ordinarily be taken as conclusive evidence 

 that this latter segment is the true eighth somite of the abdomen, 

 especially since these respiratory apertures, as described by Blunck, 

 have the structure typical of lateral abdominal spiracles. 



For the present, therefore, the nature of the terminal appendages of 

 the Dytiscus larva must be left in doubt, but if the organs do not 

 belong to the series of lateral, stylus-like appendages, fully repre- 

 sented in Coptotomus and Gyrinidae, it is possible that they are struc- 

 tures homologous with those dorsal processes often developed on the 

 ninth abdominal segment of other coleopterous larvae, which appear 

 to be mere cuticular outgrowths, though they may become movable 

 at their bases. The muscles inserted on the processes in Dytiscus are 

 evidently not specific muscles of the appendages, but groups of seg- 

 mental or intersegmental fibers that, by reason of their attachments 

 at the bases of the processes, serve to move the latter. 



The abdomen of most other coleopterous larvae consists of lo 

 distinct segments (fig. 44 B, C). The ninth segment is usually well 

 developed and frequently supports a pair of processes (B, C, ug) 

 arising from its dorsum. These processes are termed urogomphi by 

 Boving, since from their segmental position it is clear that they can 

 not be cerci. In some species the urogomphi are fixed outgrowths 

 of the posterior end of the ninth tergum (C) ; in others they arise 

 from the membrane behind the ninth tergal plate (B), and are then 

 flexible at their bases. They vary much in size and shape from 

 short, spine-like points to long, thick processes or multiarticulate fila- 

 ments, and they are sometimes distinctly jointed. The urogomphi 

 appear to be equivalent structures in all cases, and are probably but 

 cuticular outgrowths of the ninth abdominal tergum, which in 

 certain species become movable by a membra nization of the tergal 

 wall at their bases. The mobile appendages of the Dytiscus larva, 

 therefore, are perhaps urogomphi of the rudimentary ninth segment, 

 secondarily movable by some of the intersegmental muscles normally 

 extending between the eighth and ninth segments. 



The tenth abdominal segment in many coleopterous larvae is a 

 short, often tubular structure (fig. 44 B, C, X), at the extremity 

 of which is a small retractile and eversible membranous pad sur- 

 rounding the anus, known as the " pygopodium " (C, XI), which may 



