GAMETOGENESIS AND SPOROGENESIS 



217 



collecting about it at certain stages, but the situation in insects seems to 

 show that the real unit is the Golgi body composed of chromophilic 

 "Golgi matter" and relatively chromophobic "idiosome substance," 

 a close topographic relation to the centrioles being assumed only in 

 certain cases. During the meiotic divisions of the nucleus the Golgi 

 rodlets (or network or shell) form a number of dictyosomes which pass as 

 two groups to the poles; whether each divides or not is a question. The 

 fate of the idiosome substance is difficult to determine, but it is probable 

 that each dictyosome is a minute Golgi body with its two substances, 

 which are thus distributed to the four 

 spermatids. 



Each spermatid consists of cytoplasm, 

 a nucleus with the gametic number of chro- 

 mosomes, some chondriosomes, a pair of 

 centrioles, a number (usually) of Golgi 

 bodies, and often ergastic inclusions. 

 These various elements will now be followed 

 through the metamorphosis of the spermatid 

 into a spermatozoon (Figs. 130 to 134). 



Centrioles. — In the Hemiptera the two 

 centrioles are situated near the anterior 

 end of the young spermatid, one of them 

 lying in contact with the nuclear membrane. 

 From the distal member of the pair the 

 delicate axial filament of the flagellum 

 grows out and pierces the cell membrane. 

 In many insects this filament may appear in spermiogenesis in Piethodon dn- 



1 I- IP -i. • • IX ereus. (After Bowen, 1922a.) 



much earlier, before mitosis is complete or 



even in the primary spermatocyte (moths). The centrioles now 

 migrate to a position near the posterior pole of the nucleus, where 

 they take the form of parallel rods lying against the nuclear membrane, 

 and soon form a V, with the elongating axial filament extending from its 

 angle. An intranuclear body ("pseudoblepharoplast") present during 

 these stages has often been confused with the centrioles. The centrioles 

 eventually constitute the so-called "centrosomal middle piece" at the base 

 of the sperm head, but in insects this is not a very distinct structure. 



The centrioles in certain other animals show curious modifications 

 into rings and knob-like structures which have been variously interpreted 

 (Fig. 131). In his accounts of spermatogenesis in the salamander Meves 

 (1896, 1897) reported that the distal centriole, after the appearance of the 

 axial filament, transforms itself into a ring. The filament then passes 

 through this ring and attaches itself to the proximal centriole, which 

 pushes into the base of the nucleus and enlarges to form the middle piece. 

 Meanwhile the ring divides into two parts, one of which becomes the 



Fig. 131. — Behavior of centrioles 



