ye) BOTANICAL GAZETTE [JANUARY 
2-celled families in this locality. They vary in form from circular- 
lenticular (fig. 15) to elliptical-lenticular (fig. 11); from those hav- 
ing plane walls (fig. 14) to those having marked polar flattenings 
with a central ridge (fig. 16); from those with thick bands and caps 
to those entirely lacking the gelatinous secretion. When seen from 
the flat side, the bands when present are cruciform. They lie well 
within the mother wall, but may extend some distance beyond the 
cells (fig. 12). In the late maturity stage the mother wall is lost 
and the cells enlarge until they rather than the bands become the 
conspicuous feature (fig. 19). The cells may then separate at once 
or divide internally before separation (fig. 23). After separation 
of the aplanospores development or rest period follows as described 
under the 2-celled families. The tetrahedral 4-celled family (fg. 
20) differs in having a close-fitting mother wall, in the necessarily 
different arrangement of the bands, and in the absence of polar 
caps. The life history is probably the same as for the preceding 
The 8-celled families are exceedingly rare. Out of the hundreds 
of specimens of the plant which I have examined from the collec- 
tions and cultures, not more than a score of this type have been 
found. Of these there have been two forms about equally abun- 
dant: (1) those having the 8 cells arranged in the form of a cube 
with an edge 35-50 » long, and (2) those with the 8 cells grouped 
more or less irregularly. These latter have in one or two instances 
resembled a cube that has been compressed so that the upper and 
lower faces are the shape of a rhombus. 
These 8-celled families have a close-fitting mother cell wall until 
late maturity, when they are held only by the gelatinous bands. 
STOCKMAYER predicted the possibility of 8-celled families on the 
basis of the caps in the 4-celled forms, but he evidently had in 
mind families with 8 cells in the same plane. 
The most important result of the cultures, however, has been 
to show that the mature colonies do not divide, that is, the 2-celled 
mature colonies do not form 4-celled colonies directly. When the 
bands are once formed they are permanent structures. Whether 
the family shall be 1-celled, 2-celled, 4-celled, or 8-celled is deter- 
mined by the number of divisions that take place within the aplano- 
