218 Hazen: Lire HISTORY OF SPHAERELLA LACUSTRIS 
In cloudy weather one is likely to find cases of division at al- 
most any hour of the day if the temperature is favorable. This 
fact indicates that the light which promotes photosynthesis hinders 
the resorption of materials in preparation for division. 
The time required for the formation of megazooids after the ac- 
tual beginning of division varies with different conditions of light 
and temperature. In all cases that I have observed continuously 
under the microscope, the process has occupied at least an hour, 
often several hours. But there is good reason to believe that light 
retards the process of division, even as it is unfavorable to the 
preparatory process, for in completely darkened cultures I have 
repeatedly found that nearly all division is completed within an hour 
after its beginning is observed, while in cultures left in ordinary 
light dividing cells will be fourid for several hours. 
Alexander Braun ('51) says that even before the beginning of 
division the red color begins to be replaced by a yellowish-green 
layer from the periphery inwards, but I have not found this to be 
the fact except when division is delayed, and in case of cells which 
have not gained the entirely red color of typical resting-cells. 
When division is accomplished promptly red cells produce red 
zooids. 
After the first cleavage ik is transverse, the division follows 
no constant rule; the second cleavage, though in a plane perpen- 
dicular to the first, may begin at both poles simultaneously or 
division may be completed in one hemisphere before it begins in the 
other. Often when four daughter-portions appear to be completely 
formed they divide again so as to form eight zooids, but more fre- 
quently when eight or sixteen zooids are to be produced the di- 
vision is so irregular that it cannot be followed easily, or the mass 
may even appear to break up all at once into numerous daughter- 
cells, 
Dangeard ('88) suggests as an explanation for such cases that 
possibly the resting cells contain several nuclei, and he also thinks 
that when more than four daughter-cells are formed, the reason 
may be that the mother-cells have been for some cause prevented 
from dividing at the normal point and have continued to increase 
in size. This explanation does not appear to me quite sufficient, 
for not infrequently eight or sixteen zooids are produced by cells 
no larger than those which produce four or eight. 
