Opalina. 245 
the nuclear membrane, one would surely see more frequent evidence 
of their doing so. It seems certain that only a very small proportion, 
if any, pass as solid bodies out of the nucleus. In the great majority 
of nuclei apparently none do so. 
It seems, however, not improbable that the chromatin spherules 
of the nucleus and the endoplasmic spherules. may be somewhat 
related. Their staining reactions suggest this.. Stained with Drena- 
FIELD'S haematoxylin the newly formed chromatin spherules are very 
dark blue; the older chromatin spherules stain less and less, and 
finally are not stained at all. The endosarc spherules are entirely 
unstained with this reagent. Similarly with safranin and light green 
the chromatin spherules, if newly formed, stain deep red; older 
chromatin spherules stain more faintly, and in nuclei in which the 
new spindle has appeared they are either very faint or have already 
disappeared. With the same dye the endoplasmic spherules are 
colored a very faint pink, resembling the almost dissolved chromatin 
spherules within the nucleus. As the chromatin spherules loose 
their staining capacity one sees that many are growing smaller; 
some, on the other hand, in the same nuclei often seem to be 
enlarged and more diffuse. Not infrequently one finds faintly staining 
irregular masses which look like dissolved chromatin spherules filling 
several alveoli of the nuclear foam, and I believe this is the proper 
interpretation (cf. Fig. 58, Pl. XVIII, in which near the centre of the 
nucleus are such faintly stained areas). Is seems well-nigh certain that 
the chromatin spherules dissolve, and it is probable that they pass in 
liquid form through the nuclear membrane into the cytoplasm. It is not 
improbable that, having reached the cytoplasm, this material reforms 
in the endoplasmic spherules. Very likely, however, material from the 
cytoplasm as well is used in the formation of the cytoplasmic spherules. 
The origin of the endoplasmic spherules from the chromatin 
spherules is by no means assured. The granules that with proper 
staining are always seen in the endoplasmic spherules, and especially 
the lines occasionally seen within them, connecting their granules, 
suggest that they are formed elements of considerable complexity. 
If they really divided, as TOnntexs (1898) and Kunsruer & GrNesTE 
(1905) describe, their interpretation as living constituents of the 
cell would seem unavoidable. I do not, however, find evidence of 
their division, the constricted portions of the frequent dumbbell- 
shaped spherules never being very slender as if about to part. 
If the two sorts of spherules be related as suggested, probably 
the material from one chromatin spherule is enough to form or aid 
17* 
