A Comparative Study of the Cliroinosoines etc. 
265 
of the granules best, for the portions of the threads between the graniiles 
graclually lose their staining capacity. As the chromosomes of the early 
strepsistene stage separate, they begin to twist about each other; but they 
remain joined together at one or both eiids (Plate XX, Figs. 87, 89, 90). 
The outlines of the threads become hazy and only the granules stain 
(Fig. 91). Soon the granules also become smaller and less distinct and 
finaUy only ill defined threads of loose texture still in connection with 
the chromatic nucleoli can be seen (Plate XX, Figs. 92, 93). Then, 
judging from the scarcity of cells in this stage, for a short period, 
during which the chromatic nucleoli disappear and the x-chromosome 
condenses, the substance of the autosomes is very indistinct and can 
not be traced as individual chromosomes (Plate XX, Figs. 94, 95). The 
cells at this stage have attained their maximum size and the transition 
to tetrad formation is rapid, for all the successive stages may be fol- 
lowed in a single cyst. 
As the chromatic matter condenses and regains its affinity for basic 
stains, it is seen to take on a variety of forms. The members of a pair 
may be twisted about each other several times (Plate XX, Figs. 98, 100), 
they may lie parallel with each other, being connected at one or both 
ends (Fig. 96), or they may simply form U’s or V’s. In follow- 
ing the tetrad formation, I have taken the macrochromosome pah- as 
an example, since they are largest, clearest, and can not be confounded 
with the smaller tetrads, even in early prophases, on account of their size 
(Plate XX, Figs. 98, 99). Figures 103 to 118 illustrate sonie of the 
various phases which this tetrad may assume. The two chromosomes 
which form it may be closely united along nearly their whole length 
(Plate XX, Fig. 103), or they may be twisted about each other while 
still United at both ends (Fig. 104), or, thircUy, they may be connected 
at one end only (Fig. 105). This Variation doubtless accounts for the 
different forms which are shown in Figures 106 to 115. The twisted chro- 
mosomes contract and split longitudinally. This split usually begins 
to be noticeable at that end which will lie in the equatorial plane of the 
spindle at the first spermatocyte division, which is likewise the end of 
attachment of the two chromosomes to each other in late prophase. 
The partial Separation of the halves of both chromosomes of the tetrad 
(Plate XX, Figs. 106, 107) leads doubtless to cross-shaped forms, such 
as are seen in Figures 102 and 114 (Plate XX). The force of the 
longitudinal contraction of the linin groundwork may cause the 
production of the short arms of the cross from the partially separated 
halves, such as seen in Figures 99 and 106 (Plate XX). As the 
