PHYSARELLA MIRABILIS PECK AND STEMONITIS FUSCA ROTH 285 
the capillitial cavities were started to be formed some time after the 
columella wall deposit had attained some degree of thickness. If the 
capillitial tubules should begin to form as early as columella deposit, 
the lumen of the two should undoubtedly be continuous. 
The formation of the reticulated, capillitial sac just within the 
exterior protoplasm of Stemonitis is at first glance a puzzling phe- 
nomenon. This would obviously result in a layer of spores being 
formed outside the main capillitial network; in fact, a careful examina- 
tion of a fresh, mature sporangium shows that such is indeed the case. 
It is evident that this rich development of a layer of capillitial threads 
just within the peripheral protoplasm must result from a very rapid 
secretion of wall substances from the rapidly maturing external proto- 
plasm. Only comparatively few of the very delicate threads which 
connect this anastomosing network with the external wall apparently 
have an opportunity to form, so quick are these changes in the periph- 
eral portion of the sporangium. 
Contrasting the sporangial walls of the two species under discussion, 
it seems clear that the thickness of wall deposit must depend to some 
extent upon the length of time the peripheral protoplasm remains in 
a relatively quiescent state. In Physarella, external secretion ap- 
parently continues during a considerable portion of the time of develop- 
ment of the sporangia. In Stemonitis, on the other hand, cleavage 
begins at the exterior very shortly after the heaping protoplasm attains 
its final shape, and in consequence the sporangial wall reaches only a 
meager development. 
Some mature capillitial threads in Stemonitis fusca are somewhat 
flattened; some are somewhat irregular in shape, especially at the 
meshes; some present a wrinkled or crumpled appearance, as before 
mentioned. In some cases the occurrence of solidity is hard to dis- 
prove; but studies of many sections and of mature capillitia indicate 
without any doubt that hollowness of the threads is in reality very 
common. 
The smoothness of the exterior of the capillitial threads in both 
species studied is in striking contrast to the spirally thickened and 
otherwise marked, distinctly hollow threads of certain other Myxomy- 
cetes, such, for example, as Trichia. But in the essential features of 
the process, the phenomena connected with the formation of the two 
types of threads are regarded as entirely similar, consisting in both 
cases in the deposition of hollow threads by plasma membranes 
