1906] OLIVE—NUCLEAR AND CELL DIVISION OF EMPUSA 243 
importance of a fibrous mechanism for the accomplishment of mitotic 
division. _ ; 
I should prefer to believe that the achromatic spindle substance, 
probably present only in small amount, is a part of, and inseparable 
from the deeply staining radiations. Should this be true, then we 
may conclude that the kinoplasmic spindle-mechanism is bound up 
closely with the radiating parts corresponding to the chromosomes. 
Possibly the chromatin is here more nearly a liquid substance than is 
usual, hence it may diffuse more readily throughout the linin basis, so 
as to be indistinguishable from the latter. At. any rate, I should 
regard the chromatic filaments radiating from the centrosomes as cor- 
responding in part to the fibers of the more differentiated spindle of 
higher organisms; and, further, since these mark the paths of the 
chromatin, they must also correspond to the mantle fibres. In the 
case of Empusa, so far as studied, there is obviously nothing which 
can correspond to the central spindle of more complicated nuclei. 
CELL- pst irra 
Cell-division in Empusa, as in many other jade plants, takes 
place in entire independence of nuclear division, and also apparently 
remote from nuclear control. There is concerned in the process no 
such fibrous structure as a cell-plate; since, in fact, no cell-plate is 
ever formed at the close of the nuclear division described above. 
Further, cell-division may not take place till long after all division of 
the nuclei has ceased; hence coenocytic hyphae result. 
The branched conidiophores of E. sciarae (figs. 16, 18), as well as 
conidia in the process of abstriction (figs. 28, 30, 31, 36) furnish 
especially fine material for the study of cell-division. Examples are 
also occasionally met with in sections of vegetative hyphae (figs. 79, 21). 
A striking feature of the process as seen in conidiophores and young 
vegetative hyphae is the fact that in the cleavage of the cell, the new 
ring-formed partition-wall grows across a wide vacuolar space. In 
the case of the abstriction of the conidia, on the -other hand, and 
probably as well in older vegetative stages, although I have not as 
yet seen the phenomenon in the latter instance, the new wall 
grows through a mass of cytoplasm. Fig. 18 shows clearly the 
method of growth progressively inward of the ring-formed septum. 
