WENRICH: spermatogenesis of PHRYNOTETTIX MAGNUS. 87 



with other chromosomes, especially in the growth-period. Here its 

 polar granule may unite with those of the other chromosomes to form 

 a composite granule. (2) Its behavior, while unique in many respects, 

 differs from that of the autosomes in the degree and the chronology, 

 rather than in the kind, of its changes. The autosomes form vesicles 

 in the telophase of the spermatogonia, as Sutton ('00) long ago pointed 

 out, just as does the accessory, but they are not quite so large or 

 persistent as with the latter. In the growth-period the accessory 

 forms a looped spireme, just as the autosomes do (see fig. 71 and 72, 

 Plate 6), but its thread is much more dense and heavily stained than 

 the others. Although it fails to find a mate in synapsis, its behavior 

 is very like that of the autosomes and its spireme loop may occupy the 

 entire circumference of the nucleus. The process of shortening and 

 thickening, which all the chromosomes undergo, occurs very early in 

 the case of the accessory and it passes through most of the growth- 

 period as a rather compact mass of chromatin. In the postspireme 

 stages, at the time when the chromatids separate from each other by 

 the formation of the secondary longitudinal split, the accessory forms 

 a more or less bent or twisted rod, which often shows a longitudinal 

 split. This split must be homologous to the secondary split seen in 

 the autosomes, which divides longitudinally each of the chromosomes 

 united in synapsis. In the anaphase of the first spermatocyte division 

 its halves separate at the distal end, so that it forms a dyad similar in 

 all respects to those of the autosomes, except for its more roughened 

 condition. In fact, the accessory dyad cannot always be distinguished 

 from the others in the late anaphase. In the metaphase of the second- 

 ary spermatocytes it divides along with the autosomes and usually 

 is indistinguishable from them. Its behavior may therefore be more 

 nearly parallel to that of the whole series of chromosomes than we are 

 sometimes led to suppose. 



c. The Spermatogonial Divisions. 



Let us now consider the subject of persistent chromosomal organiza- 

 tion from the standpoint of the spermatogonial divisions. Figures 

 1-20 (Plates 1 and 2) are intended to represent the most important 

 stages included in the cycle of changes from one cell division to the 

 next. In this description no reference will be made to the selected 

 chromosomes, but the general behavior of the chromatin material 

 will be considered. We shall also leave out of account the mechanics 



