946 M. M. Metcaur 
in the formation of many endoplasmic spherules; otherwise it would 
be difficult to explain the great number of the latter present in the 
body at all times. The whole body divides during each mitosis, so 
that the number of endoplasmic spherules is reduced to half. The 
chromatin spherules are formed during each mitotic cycle, but there 
are rarely, if ever, more than one hundred and twenty in one 
nucleus, while there are many hundreds of the endoplasmic spherules 
in an ordinary sized Opalina intestinalis. KunstnER & GINESTE 
estimate eight thousand for an ordinary sized O. dimidiata, a much 
larger species than O. intestinalis and having many more endoplasmic 
spherules. If the endoplasmic spherules are in any way derived 
from the chromatin spherules, and if they do not increase by division, 
there seems no escape from the conclusion that the material of one 
chromatin spherule suffices for many endoplasmic spherules. The 
endoplasmic spherules are larger than the chromatin spherules. New 
chromatin spherules do not continue to form in the nucleus while 
the earlier formed ones are dissolving, for one does not find them 
showing all varieties of staining in the same nucleus. They stain 
all about alike. The period during which the chromatin spherules 
form is indeed a rather brief one, extending from the end of the 
spireme (chromatin ribbon) stage to the beginning of the formation 
of the spindle. 
It is, of course, possible that the chromatin spherules are but 
by products of the metabolism of the chromosomes and that they 
have little significance. Yet their staining reactions, with all stains 
used, seem to indicate that they are composed of chromatin, and the 
gradations in their staining, as they dissolve, seem to connect them 
with the refractive spherules of the endosare. 
In O. caudata the condition of the chromatin spherules and _ 
endoplasmic spherules is like that in O. ¢ntestinalis. The multi- 
nucleated species, O. dimidiata, O. zelleri, O. ranarum and O. obtri- 
gona, have endoplasmic spherules of the same character. Their nuclei 
are so small that it is not easy to study in them the formation of 
the chromatin spherules, and I have not attempted it. 
Origin of the ectosare spherules. 
I have seen nothing to indicate any genetic connection between 
the spherules of the endosare and those of the ectosarc, except that 
with iron-haematoxylin, when very thoroughly extracted, one some- 
times finds, in the ectosarc, spherules showing an internally granular 
