44 BOTANICAL GAZETTE [JULY 
either through the central cell or by separate conjugation with 
the carpogonium, 
The development of the cystocarp in this species thus rr 
bles very closely that of G. corallina as described by Janczewskt, 
with the single exception of the one carpogenic branch instead 
of two. He does not speak, either, of the fusion forming L. a. 
large central cell, though one of his figures (7) suggests that it + 
may take place. : 
The process has its more or less perfect analogy in related 4 
genera also. In Ceramium decurrens, according to Janczewski, 
the supporting cell, which, after cutting off a sterile cell, gi 
rise to two trichophores, becomes the “ carpogenic cell,” cutting 
off an upper placental cell after fertilization. Schmitz concludes 
that here conjugation probably takes place between the carpogo- 
nium and the Supporting cell, but does not claim to have observed 
it. Janczewski represents the two cells as lying adjacent, and ae 
this fusion would be more easily accomplished than in Gyriffithsia 
Bornetiana, where they are not always close together, though 
they are in some cases, as is seen in fig. ro. 
2 
sterile branches, 
becomes the auxiliary cell, cutting off, after fertilization, an 
upper sporogenous cell. In these cases the carpogenic branch is 
the process. According to Phillj 
the sporogenous cell, the auxiliary cell, and its sterile branches 
in P. nigrescens and P. Jastigiata, while in P. violacea, as appat- 
ently in Griffthsia Bornetiana, the central cell is included in the 
fusion. 
Ps, fusion takes place between 
*° On the development of the cystocarp in the Rhodomelaceze, Ann. Bot. 9: 289 
1895. 
