510 



other mass or passes into its interior (Figs. 32, 33, 35). The united 

 mass thus formed is generally noticeably larger than that of spermatids 

 which have received only four chromosomes, as may be seen by com- 

 paring Figs. 35 and 34 of which Fig. 34 is the spermatid without 

 the odd chromosome. 



In each type the centrosome may be seen to have persisted and 

 to be connected with the chromatin mass by means of a filament,, 

 presumably the first indication of the axial filament of the sperma- 

 tozoon tail. In the pigeon the axial filament first appears as a 

 thread connecting two centrosomes which lie in. the cytoplasm. The 

 thread later continues backward through one of the centrosomes, which 

 is ring-like, and out of the cell, while the other centrosome which 

 marks its anterior end ultimately comes to lie inside the nuclear 

 membrane, obscured by a covering of chromatin. Seemingly this intra- 

 nuclear portion represents the middle-piece of the typical sperma- 

 tozoon as no other structure comparable to a middle-piece is discernible. 

 In the guinea on the contrary, it appears that the axial filament is in 

 direct contact with the nuclear contents from the beginning. From 

 the fact that the only visible centrosome may, in favorable preparations, 

 be seen to be ring-like in nature, it seems reasonable to infer that it 

 corresponds to the ring-like chromosome of the pigeon spermatid, and 

 that the polar centrosome of the secondary spermatocyte has divided 

 at an early period, the smaller component remaining in the chromatin 

 mass from the beginning instead of standing apart and moving to it 

 later. 



In the transformation of the nucleus of the spermatid into the 

 head of the spermatozoon, there seems to be no wholly invariable 

 course of procedure. Figs. 36 to 39 represent perhaps the sequence 

 observed as frequently as any. A dense chromatin strand (Fig. 36) 

 sprouts out along one side or splits oflf, as it were, from the main 

 mass of chromatin and then the latter gradually unrolls or untwists 

 (Figs. 37, 38) into a definite elongated head marked at one end by 

 the head-spine and at the other by the developing tail. In its process 

 of straightening out the spermatozoon usually becomes so oriented 

 that the tail lies towards the lumen of the tubule and the anterior 

 end toward the base of the tubule wall where the Sertoli or nurse 

 cells are to be found. 



The foregoing account describes the more usual method of pro- 

 cedure for spermatids which have the larger mass of chromatin ; that 

 is, for those containing the odd chromosome. Of frequent occurrence, 

 however, and particularly in the case of the smaller type of nucleus, 



