BROOKS AND RITTENHOUSE: ON TURRITOPSIS. 
455 
while in Turritopsis the morula stage results directly from segmenta¬ 
tion without any recognizable migration of cells. 
The formation of the endoderm in Turritopsis therefore differs 
from nearly all the methods which have previously been described, 
and which in the main conform to one or another of the stereotyped 
methods as established by Metschnikoff. The nearest approach is 
that briefly described by Hargitt for Eudendrium and Pennaria, in 
which there is also more or less of a syncytium formed prior to the 
differentiation of the germ layers. 
Cell Multiplication. 
During the early cleavage phases the cells multiply entirely by the 
process of mitosis. In the later phases, especially when the egg is 
approaching that stage in which the cell boundaries are lost, there is 
good evidence that direct cell division is also of frequent occurrence. 
In this period of development mitosis and amitosis take place simul¬ 
taneously in the different cells of the segmenting egg. Figure 40 
(pi. 34) shows a karyokinetic spindle in the metaphase; and figure 41, 
(pi. 34) one in the anaphase. The chromosomes are large and promi¬ 
nent, but are too closely crowded together to be counted with accuracy. 
The nuclei, which divide amitotically, vary in size considerably, 
and have a reticular appearance. Figure 42 of plate 34 shows a large 
nucleus of this reticular character with the chromatin scattered about 
in the linin mesliwork. Figures 43-46 (pi. 34) illustrate nuclei in 
various stages of amitotic division. Frequently in cells where amitosis 
takes place, many of the yolk granules have been digested and conse¬ 
quently are fewer than in cells where digestion is less active. It may 
be that the more active functions of digestion and the phenomena of 
direct cell division are associated with each other; or it may be that the 
view of Flemming and Ziegler, that amitosis is connected with a high 
specialization of the cell or is the forerunner of degeneration, applies 
in this ease. This latter conception seems plausible, for we find 
amitosis to be most abundant shortly before the cell boundaries dis- 
appear and the embryo is transformed into the syncytium. 
For a number of years it has been known that amitosis is common 
in follicle cells, digestive epithelial cells, supporting cells, etc.; but 
generally it was not supposed to take place in early embryonic develop- 
