SPERMATOGENESIS OF EUSCHISTUS Us 
B. The idiozome 
The idiozome of the spermatogonia (id, figs. 1, 3) may lie 
against any region of the nucleus, and fig. 3 illustrates how vari- 
able its position may be. But there is here an interesting rela- 
tion not noted in my previous studies. No matter where the 
idiozome is situated, it always touches the nucleus at that wall 
where the chromatin plate is; the nature of this plate was men- 
tioned with the description of the autosomes. Therefore this 
idiozome probably stands in some intimate chemical relation 
with the chromatin plate. The idiozome is irregularly granular 
and becomes browned after osmic acid fixation but doesnot stain to 
any degree with any of the stains employed. It disappears 
during the prophases of division by breaking into smaller granules, 
that seem to be identical with the small bodies lying outside of 
the spindle (figs. 2,4). In resting cells are found smaller scattered 
granules in addition to the large idiozome (figs. 1, 3), and these 
smaller granules may be either the remains of a previous idiozome 
or else mitochondria. 
The young spermatocytes at the end of the last spermatogonial 
division (fig. 5) show no large idiozome, but merely scattered 
granules that may be the remains of the previous one. Very 
early, however, an idiozome develops in them, and throughout 
its existence it constantly maintains a position at the distal pole 
of the nucleus just where the chromatin plate lies. Fig. 6 shows 
the first appearance of the idiozome in a spermatocyte, and the 
succeeding drawings to fig. 36 represent its appearance up to the 
time of its beginning disintegration. Frequently it seems dis- 
tinctly paired, and then each of its portions seems to be related 
to a particular one of those autosomes that make the chromatin 
plate; this relation is seen in figs. 8, 11, 13, 21, 25, 28, 34. Per- 
haps it is always paired but appears so only when seen from a par- 
ticular pole. The figures also show that the spatial relations of 
the idiozome are rather closely related to the spatial extent of the 
chromatin plate, which would indicate growth of idiozome sub- 
stance within the cytoplasm corresponding to the area of the 
nuclear wall involved in the nuclear chromatin plate. But there 
