Grinnellia americana , Harv. 21 
gives additional testimony to the belief that the sterile 
filaments connecting pericarp and cortical cells in the base of 
the cystocarp (Fig. 21) are agents for conducting food-material 
from the pericarp to the egg-cells of the cystocarp. 
In Mr. T. Johnson’s work on Sphaerococcus and Gracilaria, 
he figures and describes a complicated fusion of the procarpic 
cells shortly after fertilization of the trichogyne. A very 
large cell is the result of this fusion of several cells and from 
this the ooblastema-filaments develop. Nothing of this nature 
is discovered in the growth of the cystocarp of Grinnellia. 
It seems, on the contrary, that the large protoplasmic 
companion-cells simply contribute nourishment to the large 
central cell through the pits in the walls of their connecting 
processes (Fig. 26 p ), and that this large cell gives rise to a few 
ooblastema-filaments and to a papilla-like growth of cells 
containing dense protoplasmic contents from which nearly all 
the ooblastema-filaments develop (Figs. 24, 25). In this 
way the production of these filaments is continued at different 
intervals and gives rise to chains of carpospores of various 
ages, ranging from mature to very young forms (Figs. 22, 26). 
The lack of fusion in the auxiliary cells of Grinnellia is one 
of the most notable points observed in the study of its 
development. It is especially peculiar, as fusion of basal, 
procarpic cells has been reported in the related genus 
Gracilaria. A careful examination of a very large number of 
stained and unstained preparations of median longitudinal 
and transverse sections of young and old cystocarps failed to 
reveal a single case of cell-fusion. The cell-walls of the old, 
empty, auxiliary cells in the base of the mature cystocarps 
were unabsorbed, though often irregular and distorted, because 
the cell-contents had been used in supplying food-material to 
the ooblastema-filaments and carpospores. As only a thin 
membrane is left to mark the cell-walls of the empty auxiliary 
cells, the freezing method was especially valuable in exactly 
determining whether there had been any true fusion of 
adjacent cells. Any process employed in preparation for 
sectioning, which would have produced even a slight rupture 
