376 University of California Publications in Zoology [Vol. 1G 



toward one of the poles to a considerable extent aliead of the others, 

 the plane of sectioning might come between it and the equatorial plate. 

 This would leave nineteen chromosomes in one section as seen in polar 

 view. If, on the other hand, the plane of sectioning did not come 

 between the equatorial plate and the advancing chromosome, a polar 

 view would show nineteen chromosomes in one plane and a single one 

 at another level. 



In making counts of the number of .secondary spernuitocytes having 

 twenty or nineteen chromosomes, it was found that more had twenty 

 than had nineteen. This is to be expected, since it is not at all likely 

 that one-half of the sections would be made separating the advancing 

 chromosomes from the equatorial plate. 



A side view of a secondary spindle shows that the chromosomes 

 are very different in shape from those of the first spindle. Instead of 

 having the quadripartite form, they have assumed a dumbbell shape, 

 with their long axes lying parallel to the long axis of the spindle 

 (pi. 30, fig. 7). 



With the excepton of one, all of the chromosomes divide at about 

 the same time. As they migrate they become so massed together that 

 it is extremely difficult to count them, and by the time they have 

 I'eached the poles few of the chromosomes can be distinguished in the 

 mass of chromatin. Here too it seems that the period of migration is 

 rapid, for anajjhases are very rare. 



Guyer (1910), Wodsedalek (1913), Jordan (1911), and others 

 working on various vertebrates have claimed a second pairing of the 

 chromosomes previous to the secondary division. This fusion I have 

 failed to find in the mouse. Many metaphase plate stages of the sec- 

 ondary spermatocytes have been studied and all showed the number 

 of chromosomes to be nineteen or twenty, and in figure 7 (pi. 30) 

 seventeen chromosomes may be counted. "Wodsedalek (1913) finds 

 that in this secondary pairing the accessory chromosome does not pair. 

 This process of pairing could not take place in the mouse unless 

 one of the autosomes should pair with the allosome, since there is an 

 uneven number of autosomes. 



A chromosome has been mentioned which behaves dil?erently from 

 the others and which in several ways resembles an accessory chromo- 

 some. Throughout the development of the early spermatocytes, a 

 mass of chromatin is to be seen lying against the nuclear membrane, 

 and it is particularly conspicuous during the late spireme stages. 

 This appears similar to a chromatin nucleolus found in some insects. 



