THE SPERMATOGENESIS OF AGALENA N.KVIA. 125 



are contracted away from the membrane. At this time the 

 nucleus stains so deeply that only after long extraction can any- 

 thing be learned of its structure and even then one can merely 

 say that the chromatin appears to be in the form of a dense 

 reticulum. It is apparently at this stage that synapsis occurs, 

 reducing the number of chromosomes to half the spermatogonial 

 number. The large, chromatin nucleolus of the resting stage 

 has again resolved itself into two peripherally placed rods shown 

 in longitudinal and cross sections (Figs. 8 and 9). At the close 

 of the contraction stage the growth period is introduced by an 

 evident increase in the size of the cell and the nucleus. As the 

 nuclear elements become spread out in the enlarging cavity it is 

 comparatively easy to see in what condition they are. The 

 chromatin is now in the form of irregular granules distributed 

 along the delicate loops of the spireme. The loops are long 

 and often twisted or bent, so that their number was not de- 

 termined nor could I ascertain whether or not the spireme 

 is continuous but taking what evidence one can gather from 

 this stage, together with a study of a slightly later stage, it 

 seems probable that the loops are distinct from each other and 

 that the segments of the spireme are in the reduced number 

 (Figs. 10 and 11). Later the nucleus enlarges still more, allow- 

 ing the loops to stretch out fully and it is now clear that the 

 spireme is segmented, the free ends of the loops all being directed 

 toward one side of the nucleus (Fig. 12). The linin thread is 

 delicate, at first, and the chromomeres undivided, but later a split 

 runs throughout the length of each loop, dividing the chromo- 

 meres equally. Sometimes the latter appear as large after the split- 

 ting as before but this is readily explained when one considers 

 that meanwhile the loops have shortened and so have crowded 

 the chromomeres into larger groups. At first glance the acces- 

 sory chromosomes, on account of their great staining capacity, 

 appear to have maintained their peripheral position but careful 

 focusing reveals the fact that they are now completely sur- 

 rounded by the spireme loops (Fig. 13). A cross-section of the 

 same stage is shown in Fig. 14. Although the accessory chro- 

 mosomes have moved to a more central position in the nuclear 

 cavity, their outer ends are still near the membrane and are 



