1898-99. } ON THE CYTOLOGY OF NON-NUCLEATED ORGANISMS. 471 
of them, nor any change so far as can be determined in their chemical 
reaction. It may happen, as it often does in MWZzcrocoleus terrestris, that 
while the granules of the second type are equally divided between the 
daughter cells, one of the latter may receive all the granules of the first 
type. When, however, these granules are very large in proportion to the 
cell which contains them, such granules are divided when the cell 
divides. Instances of this were seen in some preparations of Oscillaria 
natans found in a rapidly growing condition and treated with artificial 
gastric juice for from eight to ten hours. In these there were very large 
granules, or rather spherules, of the first type, which occupied a large 
portion of the cell space and which stained deeply in hematoxylin. In 
the undividing cell they were spherical, or nearly so, but in those 
dividing as well as in those which had divided, they exhibited a con- 
striction in the plane of the transverse septum formed but not completed, 
while in other cases where the septum was fully formed, each daughter 
cell had its half of the original spherule or granule. (Fig. 19). It is 
obvious that in such cases the division of the spherules was not 
physiological but mechanical and that it followed the division of the 
cytoplasm. 
Such instances of division of the granules are rare and would not, it 
must be believed, occur in the large-celled Osccllarie if the granules or 
spherules were not correspondingly large, a condition which does not 
occur. J/¢ 7s, however, interesting to note that in one form of the Cyano; 
phycee, at least, the substance in the cells which represents chromatin in 
these organisms, when present in abundance ts divided between the 
daughter cells in a mechanical way and after the daughter cells are 
Sormed. 
