574 



teen certainly and possibly eighteen, the uncertainty of the count 

 arising from the difficulty in determining whether I was dealing in 

 one case with two chromosomes or with a slightly cleft one. Perhaps 

 it would be safer to say that the spermatogonial number of chromo- 

 somes is not less than fifteen and not more than nineteen with the 

 probability that in most instances, at least, it is seventeen. 



Fig. 2 represents an arrangement of the spermatogonial chromo- 

 somes that was observed occasionally. Lying to one side of the 

 general equatorial plate is a closely united set of three small chromo- 

 somes or perhaps what might rather be regarded as a tripartite 

 chromosome. From the fact that what has been identified as the odd 

 or accessory chromosome in the next division shows more or less of 

 a tripartite structure at times, I believe it probable that this outlying 

 spermatogonial mass is the same body or group of bodies. 



At the division of the spermatogonial chromosomes one small 

 chromosome is often the first to divide so that not infrequently its 

 two small, round daughter chromosomes lead off for the poles of the 

 spindle somewhat in advance of the others. 



In the following description of the dividing spermatocytes I have 

 confined myself chiefly to the numerical and topographical relations 

 of the chromosomes, inasmuch as the chief interest of the study centers 

 in them. As preliminary to the first division of the spermatocyte the 

 usual growth period occurs together with a well defined synizesis and 

 subsequent spireme formation. In the spireme stage chromatin nu- 

 cleoli sometimes were (Fig. 4), but often were not observed (Fig. 3). 

 The two figures in question are disproportionately large when com- 

 pared with the others of the plate, due mainly to the fact that they 

 were drawn from a very thin smear preparation in which the nuclei 

 were considerably flattened. In Fig. 4 the condensation to form 

 chromosomes has apparently progressed a little farther than in Fig. 3. 

 When chromatin nucleoli were present they frequently comprised a 

 larger pair and from three to five smaller individuals although no 

 definite rule to this effect could be established. 



Little could be determined concerning the early prophase in the 

 division of the primary spermatocyte but in late prophase nine chromo- 

 somes generally appear. Fig. 5 represents a smear preparation of a 

 late prophase or equatorial plate stage showing the nine individual 

 chromosomes. The odd is recognizable as a large elongated body 

 slightly curved and enlarged at one end. There is considerable range 

 in the size of the other chromosomes. In the division of the primary 

 spermatocyte the odd chromosome passes undivided to one pole. 



