oS ag I i a a a a i at 
1896 | CYSTOCARP OF GRIFFITHSIA 43 
which grows rapidly, assumes at first an obovate, later an 
elliptical form, slightly reniform, and a length greater than 
that of the cystocarp. The cells form thin, translucent, 
expanded plates, and although they completely envelop the 
favellz, the cells within may be studied, the details, however, 
being greatly obscured by them. Seven of these involucral 
branches were counted in one case, but the number varies. 
Nageli and Janczewski call attention to the bicellular character 
of these branches in G. corallina, which, the former states, had 
been overlooked by systematists, and to which the present 
writer has found no reference in published descriptions of G. 
Bornetiana. 
Although the later stages in the development of the cysto- 
carp are made out with difficulty, the placental cell can be seen 
to stretch out and become irregular in form, and it seems 
highly probable, though not quite certain, that fusion takes place 
between it, the supporting cell, and the central cell, forming one 
large placenta, the lobes arising from the portion corresponding 
to the original placental cell. Fig. 79 is drawn to the same 
scale as the others, and thus shows not only the great irregularity 
in the form of this fusion cell, but also the great increase in size 
over any cell previously met with in the cystocarp. It is diffi- 
cult to determine the exact nature of the large cells joined to 
it, but they seem to be sterile cells unconnected with the 
Sporogenous cells, which for some reason have taken on this 
increase in size. 
No evidence was found that the peripheral cells other than 
the supporting cell of the carpogenic branch have a part in spore 
production. They gradually become inconspicuous after fertili- 
zation, and finally disappear with the carpogenic branch and the 
Sterile cell given off from the supporting cell, unless the branch 
referred to in the preceding paragraph and represented in fg. 
79 is made up of such cells which have taken on vegetative 
growth. It is easier to conceive the influence of fertilization as 
transmitted to the supporting cell through the cells of the car- 
pogenic branch than to the other two or three peripheral cells, 
