The spermatogenesis of Peripatus (Peripatopsis) balfouri. 291 



been placed one behind the other in the spirem thread have their 

 long axes exactly in the plane of the section, we find a linin thread 

 joining one end of the one with one end of the other (Figs. 13, 14, 

 16, 28, 32, 33). 



When the chromosomes in this way are segmented off from one 

 another (the segmentation of the chromatin spirem is a gradual pro- 

 cess), they shorten and thicken until they attain the definitive form: 

 this is usually that of a more or less curved rod (Figs. 13—36). The 

 volume and length of the several chromosomes of the same cell vary 

 very considerably, as can be seen by an examination of the Figures. 



In the monaster stage the chromosomes can be best counted, 

 since then they He pretty regularly with their long axes in the plane 

 of the equator and can be well studied on pole views (Figs. 22—36). 

 If they all lay exactly in this plane, and if they were straight instead 

 of usually bent rods, the computation of their number would be more 

 readily made than it is in reality, but as it is, the count is by no 

 means an easy one. In a number of cells, which I assured myself 

 were completely contained in the plane of the section, the chromo- 

 somes were carefully drawn with the camera lucida, and then counted 

 on the drawings, with the following numbers of chromosomes as a 

 result: 27, 31, 23, 31, 31, 23, 26, 29, 32, 32, 32 or 33, 29, 33 or 34, 

 26, 26, 27, 27, 31 and 34, thus the counts varied between 23 and 34, 

 and the average count would be 29. But since it is very probable 

 that the number of chromosomes would be an even one, and since it 

 is probable that the number was over-counted rather that under- 

 counted, the number from these counts would be more probably 28. 

 Now as we shall see later, the number of the chromosomes becomes 

 exactly halved during the synapsis stage, and the number in the 

 maturation divisions is 14, so that we can say that without any doubt 

 the number of chromosomes in the spermatogonia is 28. 



The earliest stages of centrosomes found were in cells of the 

 loose spirem stage, when the segmentation of the chromatin spirem 

 was nearly completed; one of these cells is figured (Fig. 10). In 

 each of these cases there were two minute centrosomes lying near 

 one another in the cytoplasm, close to that pole of the nucleus where 

 the angles of the chromatin loops are situated ; the nuclear membrane 

 was still present but faint. In one of these cells the centrosomes 

 were connected by a delicate central spindle (Fig. 10 C.Sp). The 

 next stage found was one where the centrosomes were considerably 

 further apart and joined together by a plump central spindle, made 



19* 



