88 
F. McAllister 
those of the gametophyte. And as is to be expected, the structure 
of the pyrenoid, the mode of starch formation from it, as well as the 
distribution of the starch, are also the same. 
Thus it will be seen that all chloroplasts actively engaged in photo- 
synthesis have the same structure. On the other hand those chloro- 
plasts in which photosynthesis has not yet begun, as in the embryonic 
area of the sporophyte, and in those cells having other functions than 
photosynthesis as the sporogenous cells and those of the foot, the 
pyrenoid is lacking. In these latter cells starch is deposited in con- 
siderable amounts but not through the agency of a visible pyrenoid. 
The protoplasm of the less active plastids, both of the sporophyte 
and the gametophyte, is usually of a compact texture showing very 
fine and uniform reticulation (figure 3). Plastids in which starch 
is being formed rapidly have a texture less uniform and regular. The 
reticulations are coarse and elongated in the direction of the long 
axes of the starch grains (figure 7). The presence earlier of starch 
grains in this protoplasm is no doubt responsible for this peculiarity 
in its texture. 
The chloroplasts of Anthoceros multiply by fission as is usual with 
the plastids of other plants. Seen in cross section the plastid seems 
to elongate, and the pyrenoid to elongate with it (figure 18). The 
pyrenoid finally separates into two parts and we have a much elongated 
chloroplast with a pyrenoid in each end (figure 19). This plastid 
pinches in two and forms the new plastids. The chloroplast shown in 
figure 19 is an unusually large one found in the older part of the thallus. 
As a. type of the mode of starch formation in the Marchantia 
group, I will here refer briefly to this process in Reboulia hemispherica. 
I hope later to give a detailed account of starch formation in some of 
the members of this group. 
The chloroplasts of the cells of this liverwort are spheroid or ovoid 
bodies having an average diameter of about 6 microns. The average 
number in each cell seems to be from 8 to 16. 
The protoplasm of those plastids which lack starch is irregularly 
distributed. Scattered vacuole-like areas are present in which no 
granular protoplasm is to be detected (figure 20). In the chloroplasts 
which are partly filled with starch grains these clearer, homogeneous 
areas are less conspicuous or lacking (figure 21). No deeply stained 
bodies other than the starch bodies are present, either aggregated or 
scattered. The smallest stainable body to be detected in these 
plastids seems to be starch. 
