336 ROBERT H. BOWEN. 



in the nuclear sap. The apparent anomalies of the Hemipteran 

 centrioles are thus cleared up, and they are seen to conform with 

 the usual types of sperm formation. 



To summarize: (i) in the spermatid the customary pair of 

 centrioles is present, which (2) remain in the "neck" region and 

 give rise to the axial filament of the tail; (3) the centrioles in 

 early stages are accompanied by a chromatic pseudo-blepharo- 

 plast, which is subsequently resorbed by the nuclear sap. 



The fate of the spermatid mitochondria in Euschistus has 

 been outlined by Montgomery ('n) and the much more com- 

 plete account which I have worked out is not essential for the 

 present purpose. Suffice it to say that the threads allotted to the 

 spermatids by the second maturation division are rapidly con- 

 densed by characteristic steps into a compact sphere (Fig. 22) 

 (the spermatid nebenkern of authors), which soon divides into 

 two equal parts, and these, rapidly elongating, come to form a 

 sort of mitochondrial sheath for the axial filament which lies 

 between them. During the intermediate stages in the elongation 

 of the nebenkern halves they develop a series of paired bleb-like 

 swellings (Fig. 27) which increase in number as the mitochon- 

 drial sheaths spin out (Fig. 32) and are the most conspicuous 

 thing in Benda preparations of this period. Subsequently they 

 seem to be reabsorbed and the sheaths become mere threads run- 

 ning very nearly the entire length of the tail of the sperm. These 

 vesicular swellings on the sheaths are to be found in a number 

 of insects, as I have ascertained, but I find scarcely any direct 

 mention of them in acounts of spermatogenesis ; except in Dues- 

 berg's ('10) work on Blatta where something of a similar nature 

 occurs. In the mature sperm the two sheaths appear to retain 

 their thread-like structure showing no tendency to fuse into a 

 complete mantle for the axial filament as is the case in Blatta ac- 

 cording to Duesberg. Smears oi mature sperm show clearly 

 that the tail is flat and ribbon-like, each edge being formed of a 

 distinct thread the derivative in all probability of one of the 

 original halves of the spermatid nebenkern. 



I have, further, made some interesting observations on the 

 structure of the mitochondria as revealed especially in the divided 

 halves of the nebenkern during the early stages O'f their elonga- 



