68 Rhodora [APRIL 
branched appearance might be due to the presence of detritus not 
unusual about the bases of the rhizoids. Аз efforts to dislodge these 
branch-like processes from the rhizoids failed, I concluded that in 
some cases, at least, the branching was real. ‘he rhizoids penetrate 
the mucus on the surface of the appendages, and are well adapted to 
retain a hold on the swimming Branchipus. At first glance the 
rhizoids, particularly the short ones, might easily give the impression 
that the base consists of a dise seen in optical section, but careful 
observation with ordinary high powers, as well as with oil-immersion 
lens proves the fallacy of this supposition. ‘This suggests a possibility 
that some of the structures, hitherto described in other species as discs, 
may have been, in some instances, rhizoids which, on account of in- 
sufficient magnification, have escaped detection. 
The single parietal chromatophore shows a single conspicuous 
pyrenoid on the convex side of the cell, a position which the pyrenoids 
continue to occupy in the later stages during the entire process of 
transverse segmentation of the protoplast. ‘The pyrenoid appears as 
a highly refractile body, spherical or ovoid in outline, separated from 
the chlorophyll-bearing part of the chromatophore by a narrow 
hyaline zone; always visible, it can be followed very easily in all the 
phases of its division and movement during the process of spore pro- 
duction. See figs. 3, 4, 5, 6 and 9 (pl. 79). The division of the 
pyrenoid is accompanied by a simultaneous division of the chromato- 
phore into two equal parts. ‘The pyrenoid, in dividing, first elongates, 
then by a median constriction, assumes a dumb-bell shape, and 
finally divides into two parts, each of which assumes the original 
rounded form. ‘The division of the pyrenoid is the first visible evi- 
dence of spore development. 
The usual method of spore formation is by repeated transverse 
division of the protoplast, which gives 32 cells arranged in a single 
series. Next, by longitudinal division, each of the 30 cells in the 
middle of the plant divides into two cells; at the same time, the basal 
and the terminal cells divide transversely, thus giving 64 cells, the 
maximum observed. See figs. 6 and 9 (pl. 79), the latter representing 
the basal (z) and terminal (x) portions of a plant containing 64 spores. 
The general plane of longitudinal division throughout the mother cell 
is usually continuous. Although this is the usual method of division, 
occasionally the longitudinal division occurs in an earlier stage. 
Fig. 5 (pl. 79) shows the longitudinal division appearing in the 4 cell 
