138 
R. A. HARPER 
ticular stage may divide at a little later period. Since, however, as 
indicated above, the process of cleavage is rather rapid, it is quite 
possible in Didymium that only one nuclear division occurs between 
the time when the sporanges have reached their full size and the stage 
of complete ripening of the spores. 
As noted, the cleavage process is initiated both from the outer 
and inner surfaces of the spore-plasm. A stage when almost an entire 
series of peripheral blocks have been cut off is shown in figure 12. At 
this stage the cleavage next to the columella is markedly less advanced. 
In the segments which have been cut off the nuclei are found to be 
irregularly distributed in their interior. There is apparently no 
tendency now for the nuclei to take a position just next to the plasma 
membrane. In figure 15 a later stage in which cleavage has gone still 
farther is shown. Cleavage is advancing also in the region of the 
columella but is markedly behind that on the periphery. Certain of 
the outer blocks have begun to subdivide by surface furrows at this 
stage. The planes of the cleavage furrows become very irregular as 
the process advances and the clefts which marked the course of the 
capillitial threads gradually disappear. 
Tangential sections of these stages show the formation of the 
cleavage furrows very clearly (figs. 13 and 14). In the periphery of 
such sections the irregular blocks of protoplasm seem entirely free 
from each other, a little nearer the center the furrows are seen ad- 
vancing and branching to cut off blocks of all sizes. Still deeper and 
at the center we find cross sections of rounded and angular cavities, 
which indicate the first stages in the formation of the cleavage furrows 
from the surfaces of the capillitial openings. The pushing out of the 
first clefts from these tubular capillitial cavities is very well shown in 
these figures. The cleavage is seen to be strictly progressive from the 
surface inward, and further the surfaces adjacent to the capillitial 
ca\ities are in their relation to cleavage exactly equivalent to the 
external surfaces of the mass of spore-plasm as a whole. The plasma 
membranes of the capillitial openings are the source of cleavage 
furrows to even a greater degree than the original surface plasma mem- 
brane of the spore sack as a whole. Figure 16 shows a drawing of a 
multinucleated block of the spore-plasm as it is being cut up into smaller 
fragments. It is plain here, as in the photographs, that the furrows 
cut into entirely undifferentiated protoplasm with no hyaline zones 
or rows of granules to indicate the direction they will take. Even in 
