in Rhodymeni ales : II. Delesseriaceae . 197 
so gradual, that no rupture of their continuity can be 
detected. 
4. The tearing away of the pericarp is further said to be 
commonly omitted immediately above the auxiliary cell, 
where a strand of filaments remains connecting that cell with 
the pore of the cystocarp above. The strand of cells referred 
to is doubtless the group of sterile derivatives of the peri- 
central cell so often described in the foregoing accounts, and 
if so, it is not a strand of cells but a bushy tuft in D. san- 
guined, reaching only a part of the distance to the pore, and 
pushed aside when the gonimoblast-filaments subsequently 
arise. In the other species of Delesseria the cells are fewer 
in number, and lie loosely imbedded in a mucilage above the 
auxiliary cell, still less resembling a strand of cells connecting 
it with the pore. In D. sinuosa and the species of N itophyllum 
the cells are more compact, and the gonimoblast-filaments 
grow round and over them, justifying the description ‘ nabel- 
formig’ to this stage in the appearance of the cystocarp. In 
all cases the pore is the gap left above these cells by the 
over-arching converging filaments. It is not accurate to 
describe the sterile group as in any way attached to the 
pericarp at the pore. 
5. As to the formation of a second chamber below, sepa- 
rated from the spore-containing cavity by the middle layer as 
a kind of diaphragm, a condition figured by Schmitz and 
Hauptfleisch for N '. punctatum, I have not been able to find it 
in the species examined. In the maturer stages of the growth 
of the cystocarp, the site of the auxiliary cell is the apex of 
a papilla projecting into the cavity, and while the luxuriant 
gonimoblast-filaments depress the base of the cavity round 
about it, they do not, as far as I have been able to see, 
enter a second cavity on the opposite side of the middle 
layer. 
6. A general fusion of the auxiliary cell with neighbouring 
cells is described as taking place at the ‘placenta.’ While 
such a confluence apparently occurs in some species of Nito- 
phyllum , and is very general in Florideae, it is strikingly 
