THE PYRENOID OF ANTHOCEROS 
91 
The starch in the sporogenous cells and in other cells of the sporo- 
phyte of Anthoceros which lack pyrenoids, seems however to be formed 
in much the same manner as in Reboulia and the other Marchantia 
forms. This is probably also the case with those Anthoceros species 
with multiple chloroplasts which lack pyrenoids, although in these 
plastids photosynthesis takes place while in the above mentioned 
plastids of the sporophyte the starch is probably deposited from 
soluble carbohydrate formed elsewhere. 
Campbells work (3, 4) on certain tropical Anthoceros forms would 
seem to indicate that some relation exists between the size of the 
chloroplasts and the presence of pyrenoids in them. The conditions 
in Megaceros Salakensis are especially suggestive, where the single 
chloroplast of the surface cells contains a poorly defined pyrenoid, 
while the four to six plastids of the deeper lying cells contain no pyre- 
noids. It is, of course, to be expected that small chloroplasts would 
contain correspondingly smaller pyrenoids, but their entire disappear- 
ance is unexplainable. The pyrenoid is to be seen in every cell of the 
gametophyte of Anthoceros laevis whether in the small chloroplasts 
in the region of the growing point or in those relatively large chloro- 
plasts of the surface cells of active, mature thalli. Here the size of 
the chloroplast merely determines the size of the pyrenoid. On the 
other hand pyrenoids are lacking in all of the cells of the foot and the 
embryonic tissue of the sporophyte. Although they later become 
evident in the assimilative tissue they are never to be identified in the 
cells of the sporogenous layer. The cause of the absence of pyrenoids 
in the chloroplasts of the embryonic cells of the sporophyte and of the 
sporogenous layer may be due to a failure to form starch directly by 
photosynthesis in these cells. 
Since Campbell's references to the pyrenoids and starch in the 
chloroplasts of Megaceros Tjibodensis and M. Salakensis were only 
incidental, his attention having been directed mainly to the grosser 
morphological details, it seems very much to be desired that these 
forms be reinvestigated to determine if possible the relation existing 
between the size and the number of the chloroplasts and the presence 
or absence of pyrenoids. 
Denniston (6) has shown that in the transformation of the soluble 
carbohydrates to starch in the leucoplasts of Canna and Dieffenbachia 
and in the chloroplasts of Pellionia Daveauana there is a visible inter- 
mediate product which stains orange with Flemming's triple stain. 
