268 KANSAS UNIVERSITY SCIENCE BULLETIN. 



the group, and in its fully extended state measures about eight 

 times the length of the smallest. This chromosome stands out 

 prominently through all the generations from spermatogonia 

 to spermatids and will serve excellently for comparison in 

 other species. Chromosome No. 10 is characterized by the 

 form of a ring and is easily distinguished in the first division, 

 both in a. lateral and a polar view. The single ring seems to be 

 constant for this species. Occasionally the largest two chromo- 

 somes assume the ring shape (fig. 9), and on the other hand 

 cells are found which contain no rings (fig. 6). The formation 

 of the ring is exactly like the other chromosomes of this group, 

 a bivalent chromosome possessing, when it comes into the 

 spindle, a longitudinal split along which it divides in the first 

 spermatocyte. The chromosomes numbered 9, 8, 7, show a 

 gradual diminution in size but have no peculiarity of form to 

 distinguish them. Chromosome No. 6 does not differ in form, 

 but, as is characteristic of the accessory, can be recognized by 

 its eccentric position in the spindle. The next four chromo- 

 somes are not distinguishable by shape, but No. 1 is always 

 spherical. No. 2 is usually similar to No. 1 in form but 

 slightly larger, and is sometimes found in the form of a cross 

 (pi. XVII, cell F, row 2). 



There are, then, five chromosomes of this group that can be 

 recognized with certainty in the spermatocyte generations : 

 No. 1, the small spherical; No. 6, the accessory; No. 10, the 

 ring; No. 11, the long chromosome, always fully extended and 

 next largest, and No. 12, the largest chromosome of the com- 

 plex. 



Despite the crowding in the spermatogonia, two pairs of 

 chromosomes. No. 12 and No. 1 (fig. 1), can be easily dis- 

 tinguished ; the others are doubtful. 



The chromosomes divide longitudinally in the first division, 

 with the exception of the accessory, which passes undivided to 

 one pole. As some dispute has arisen concerning this division 

 it might not be amiss to follow out briefly the behavior of the 

 chromosomes from the spermatogonial stage. Here we find the 

 chromosomes often showing the longitudinal split (fig. 1, 

 chr. 9). In the early growth stages the spireme is distinctly 

 split, and even after segmenting and much condensing this 

 division is still visible. Very soon after synapsis the bivalent 

 chromosome begins to separate into halves along the longi- 



