190] LYON—THE SPORE COATS OF SELAGINELLA 289 
time, which render it more impervious to the entrance of solutions 
from the sporangium, this undifferentiated portion of the membrane 
in its turn has probably served for nutriment or has been taken into 
the protoplast. Often, however, a little of the original membrane 
may be detected crowded back against the exospore, forming a 
layer that-easily may be misinterpreted as an additional coat, if 
intervening stages have been overlooked (jig. 8, h). 
As the spores now nearly fill the sporangium cavity and the tapetal 
célls are disorganized, all future growth must be at the expense of 
the central vacuole. The principal changes in the maturing of the 
spore are the increase in the amount of protoplasm, with marked 
“§rowth of the nucleus, and the thickening of the endospore by suc- 
cessive laminae upon its inner surface, which is in contact with the 
protoplast at first only in the region of the nucleus (fig. 9). Even- 
tually, however, it becomes a layer of equal thickness throughout 
(fig. 10). The exospore is elaborately bossed and sculptured, and 
the protuberances form a layer—the so-called perinium—that stains 
(ute differently from the part in contact with the endospore (jigs. 
7-10). It seems to me from its formation that it should not be 
Tegarded as a distinct coat, however, but as the outer region of the 
exOspore. The central vacuole throughout the spore development 
1S filled with a transparent fluid, which in early stages disappears in 
Preparing the slide, but which becomes a coagulable mass as the 
female gametophyte forms (fig. 11). 
€ most conspicuous differences between S. Emmeliana or S. 
“pus (taken as a type) and S. rupestris, described above, are due I 
- to the simultaneous transformation of the spore membrane of 
g former fro coats, and to certain phenomena due to mechanical 
Tains that arise between these two coats and the protoplast. In 
oa , = i — gnc 
<) teatare p AA: eee ‘ rt 
ca Z § Species about determining the origin of that port 
as = membrane on the contiguous surfaces (fig. 14). Earlier 
. the formation of nuclear plates, and there is no such 
pe ce of lobing of the spore mother cell as in S. rupestris. At 
be more membrane is homogeneous and translucent (jig. 15); 
*n its thickness is about half the diameter of the protoplast 
