82 
F. McAllister 
entire pyrenoid may become a single starch mass. In both of these 
latter cases the entire pyrenoid becomes starch. 
Lutman (ii) has been unable to prove with certainty the cleaving 
off of such rudiments of starch grains in Closterium. He finds a 
cleavage of the pyrenoid but is not positive that the resulting segments 
give rise to starch grains. He seems inclined to the view that they 
may give rise to new pyrenoids. His results make it clear that the 
pyrenoid in Closterium is not homogeneous but nearly always shows 
areas which absorb the stain in a varying degree. Often faintly stained 
lines in the pyrenoid are continuous with the clefts between the sur- 
rounding starch grains suggesting a relation between these lines in 
the pyrenoid and the cleavage of segments. In other cases the entire 
pyrenoid may break up into ''lens-shaped segments." 
Yamanouchi (23) has described what he regards as a new species 
of Hydrodictyon in the cells of which are to be seen numerous ovoid 
or spheroid chloroplasts. These chloroplasts "have two functions, 
one to produce characteristic pyrenoids and the other to form reserve 
starch grains." Starch formation, according to Yamanouchi, here 
has no relation to the pyrenoid but occurs at or near one side of the 
small chloroplasts, in much the same manner as the starch grains 
are formed by the leucoplast of Phajus. If these results are sub- 
stantiated by further study we shall have the most curious phenomenon 
of a pyrenoid in the green algae having no relation to starch formation 
and the starch formed by minute chloroplasts in a manner apparently 
the same as in the chloroplasts and leucoplasts in the higher plants. 
Such evidence would seem at least sufficient to exclude this alga from 
the genus Hydrodictyon. 
I have shown (12) that the pyrenoids of Tetraspora may split up 
to form several small starch grains, or the entire pyrenoid may be- 
come converted into a single starch mass much in the same manner as 
Timberlake (21) has described for Rhizoclonium and certain other 
algae. 
Due to the recent researches of Lewitsky (10), Guilliermond (8), 
and others, a voluminous literature on the subject of chondriosomes 
in plant cells has been developed. The last mentioned authors claim 
to have established the origin of plastids from chondriosomes. It is, 
however, beyond the province of this paper to deal with the origin of 
plastids as such. 
