21,1 Shaw: Merrillosphaera 105 
least as much as 192 », at which time the gonidia of the 
daughters are about 32 to 42 » in diameter; and, according 
to the figure, while the gonidia in different daughters of the 
same parent vary, all the gonidia in any one daughter are about 
equal in size. An asexual coenobium in which the gonidia had 
not yet segmented, as indicated by the statement that the coeno- 
bium contained only one generation, shown in his fig. 3 (Plate 
7, fig. 45), measures by the figure 637 by 643 pe 
The number and size of the somatic cells were not given. 
These cells were described ag globular and biciliated. Carter’s 
fig. 1 a (Plate 7, fig. 43) of such a cell shows no intercellular 
protoplasmic filaments, though this point loses significance 
when we note that his figure of a somatic cell of a true Volvor 
on the same plate (fig. 2 a of his plate) shows a stellate proto- 
plast without depicting the intercellular filaments which are 
characteristic of that genus. 
The number of the gonidia was given as generally eight, 
and emphasis was laid on the statement that they are distinctly 
visible in the daughters before birth. The distribution of these 
gonidia in the coenobia was not indicated more particularly 
than by the specification: “Daughters confined to the posterior 
three fourths of the sphere, the anterior fourth being empty.” 
The “almost mathematically regular arrangement” of the goni- 
dia, which was emphasized by Powers (’08, p. 153) in his de- 
scription of the American variety, is masked in Carter’s figures 
by a degree of artificiality. This artificiality of the figures is 
evident from the fact that, of the sixty-four gonidia in his 
fig. 1 and the eight gonidia in his fig. 8, no two overlap. Even 
the overlapping of the daughters in his fig. 1 was kept down as 
much as possible. With this in mind, we can appreciate the 
probability that the arrangement of the gonidia in Carter’s 
material was as regular as has been found in any variety of this 
species. 
Carter’s statement that each “daughter” (that is, gonidium) 
consists of an enlargement of a peripheral (somatic) cell re- 
ferred, for illustration, to his fig. 11, which he listed as repre- 
Senting another species, described in the same paper, and that 
happens to be a true’ Volvox, wherefore we may discount its 
strict applicability to the species under consideration. Acord- 
ing to his account (Carter ’59, p. 3), as the gonidium— 
enlarges, the chlorophyll and protoplasm together are seen to form an 
areolar structure around the internal periphery of the cells (fig. 4), which 
