4 
Davis . — Spore Formation in Derbesia. 
The larger nuclei are distributed rather uniformly throughout the 
protoplasm within the sporangium, and they lie generally at some distance 
from one another. The protoplasm between is vacuolate and filled with 
plastids among which the smaller nuclei lie singly or in groups (Fig. 7). 
These nuclei at this time are of about the same size as when the 
protoplasm entered the developing sporangium, that is, about the size 
of the plastids, but some of them are a little larger and some smaller. They 
show a marked tendency to become massed in groups (Fig. 10) in regions of 
granular cytoplasm free from plastids. 
There is no evidence that the smaller nuclei fuse with one another 
to form the larger. This point was studied with great care. These nuclei 
decrease in size, and shortly become much smaller than the plastids (Fig. 10). 
Although they may lie so close together as to be in actual contact, they 
never unite. The small nuclei are not easily found when they have become 
half the size of the plastids, but careful search will reveal them in larger and 
smaller groups and also singly, and the fully mature sporangium, shortly 
before spore formation, has probably just as many nuclei as when first 
formed. The small nuclei finally break down and are generally completely 
lost at the time when the spores are formed, except for deeply-staining 
globules, which are probably the remains of the nucleoli. 
It is difficult to understand the reasons for this rapid differen- 
tiation of the nuclei in size. I was not able to discover any cytoplasmic 
structures, such as coenocentra, which could be interpreted as centres of 
dynamic or metabolic activity affecting the nourishment of nuclei in their 
vicinity, as appears to be the case during oogenesis in the Perisporiales and 
Saprolegniales. However, the arrangement of the protoplasm around the 
larger nuclei is very suggestive of important dynamic or metabolic relations 
between them and the cytoplasm in the immediate vicinity. The envelope 
of granular cytoplasm, which has the appearance of kinoplasm, and more 
especially the radiating protoplasmic strands, give to the nuclei the appear- 
ance of being themselves important centres of dynamic and metabolic 
activity. Their distribution also at relatively wide intervals throughout the 
protoplasm indicates that each is really the centre of a definite region 
of protoplasm in the sporangium. It is probable that the radiating strands 
are the paths of protoplasmic streams which pass to and fro between the 
large nuclei and the surrounding protoplasm. They are, therefore, probably 
the paths of dynamic and metabolic activities between the large nuclei and 
the regions of the protoplasm which they dominate. 
It would seem as though there were an actual struggle for existence 
among the nuclei in the sporangium to control a limited amount of proto- 
plasm with limited metabolic and dynamic possibilities, and that relatively 
few of the nuclei were successful while the majority perished. I have 
already suggested this explanation to account for the degeneration of the 
