Harper . — Cell- Division in Sporangia and Asci. 517 
contain all the protoplasm of the mother cell, a considerable 
mass remaining as the so-called epiplasm. This is typical 
free cell-formation, as I have pointed out before. In all the 
sporangia studied, the cleavage is from the surface of the 
protoplasm, or from the surface of vacuoles of the mother 
cell. The daughter cells are thus separated by cleavage- 
furrows, and the nature of the division from the surface 
inwards precludes the possibility of the formation of an 
epiplasm. 
That cleavage-furrows may start either from the superficial 
surface of the mother cell or from the surface of vacuoles 
imbedded in its protoplasm, is evidence that the vacuolar 
membranes and the external plasma-membrane of the cell are 
closely related structures, as has already been maintained by 
De Vries and Pfeffer. In both ascus and sporangium the 
daughter cells first formed are naked bits of protoplasm 
bounded only by plasma-membranes and without cell-walls. 
In the sporangia studied, the cleavage does not form uninu- 
cleate cells at once, though this is probably the case in the 
sporangia of the Saprolegniaceae. In the large sporangia of 
Synchitrium and Pilobolus the cleavage is progressive, first 
dividing the mother cell into multinucleate masses, which are 
gradually split up into the uninucleate protospores. This 
progressive segmentation has no parallel in the asci, where 
from the start a single nucleus forms the centre for the forma- 
tion of each daughter cell. This progressive cleavage is by no 
means a series of bipartitions of a multinucleate cell, such as 
was suggested by Biisgen’s work on the sporangium of Mucor, 
but is entirely irregular, forming at the first segments of 
varied size which contain a varying number of nuclei. I know 
of no process which is analogous to it in this respect either in 
plant or animal cells. The growth of the protospores in 
Synchitrium and Pilobolus to form sporangia in the one case 
and by division the typically binucleated sporangiospores in the 
other, is perhaps analogous to the growth in certain asci 
whereby the typically one-celled uninucleate ascospore be- 
comes several or many celled or in some cases multinucleated 
Mm2 
