406 HARRY H. CHARLTON 



While no opinion as to the origin of the acrosome is advanced, 

 the fact that it arises in relation to the granules from which the 

 axial filament originally developed would point to the probabil- 

 ity of a centrosomal origin. The granular thread which he 

 considers to be the drawn-out middle-piece is, in Lepisma 

 domestica, simply the proximal end of the axial filament carried 

 to the apex of the cell along with the centrosome, and it would 

 seem as though such an interpretation could be made of 

 Goldsmith's results. 



The middle-piece 



Wilson ('06, p. 337), in speaking of the essential structures 

 in a spermatozoon, lists the middle-piece as a body Which " either 

 contains a formed centrosome or a pair of centrosomes, or is 

 itself a metamorphosed centrosome. " 



It is altogether possible that the middle-piece in Lepisma 

 domestica arises from one or both of the granules which we 

 found to have their origin from the centrosome and later traced 

 them to where they were closely applied to the nuclear wall. 

 In fact, we would naturally expect something of the kind, but 

 unfortunately the later history of the granules could not be 

 followed. The question is further confused by the presence of 

 a body which condenses from the nebenkern ring and which, 

 from its size, staining power, and position, would point to its 

 becoming the middle-piece. Furthermore this body can be 

 found practically in all the stages up to the apposition of the 

 middle-piece to the nucleus, but here again we are stopped, for 

 we have not seen the actual formation of this body into the 

 middle-piece. It is possible that a further study will solve this 

 difficulty. 



Comparison with Orthoptera 



At the first glance there seems to be a similarity between the 

 spermatogenesis in Thysanura as described above and that in 

 Orthoptera as previously described and particularly given by 

 Payne ('16) for Gryllotalpa borealis, but there are also differences 

 which cause one to question whether the resemblance is not 

 more superficial than real. 



