STUDIES ON INSECT SPERMATOGENESIS. 



393 



may be accompanied by still other peculiarities of constant occurrence 

 (see Montgomery '10). Within each lobe the cells are grouped in 

 cysts which are arranged as a rule in a double ^ series (as seen in 

 longitudinal sections like that of Text-figures I A and B) throughout 

 the length of the lobe, the various stages following one another in an 

 accurate series, beginning at the blind end of each lobe with spermato- 

 gonia. At any giAcn le\el the cysts of each lobe are in about the same 

 stage (the large spermatocyte generations introduce some discrepan- 

 cies) so that in a given testis it is usually possible to find in one lobe 

 or another every step in the spermatogenesis. When the cysts of 



D 



C 



B 



Figure 1. Structure of the pentatomid testis (xl8). Lobes numbered as 

 described in the text; r/, efferent duct. (Flemming;-hematoxylin). ^4, longi- 

 tudinal section of the testis of Euschistus scrrus; B, the same — Mtirgantia his- 

 trionicn; C, cross-section of the testis of Murgantia histrionica; D, the same — 

 Stiretnis anchorago. 



completed sperms reach the open ends of the lo]:)es, the sperms are 

 emptied into the collecting chamber, whence they pass into the efferent 

 duct. 



The occurrence of morphological differences between the lobes or 

 their contents makes some system of nomenclature desirable, whereby 

 reference may readily be made to any particular compartment in a 

 testis. For this reason I have adopted the plan ^ of numl)ering the 



2 In large-celled lobes the cysts are often arranged in a single series. 



3 This method of designation was used by Montgomery ('98) who appears 

 to have numbered the lobes quite arbitrarily, his numerical order being the 



