HARPER: BINARY FISSION AND SURFACE TENSION 163 
Biitschli regards it as an open question whether all the cells divide in 
the later stages. It is, however, obvious that successive bipartition 
of all the cells is the natural method of maintaining the globular form 
already achieved. Any excess or deficiency of the number of divi- 
sions in any considerable group of cells would manifest itself at once 
as a bulge or depression in the surface of the expanding sphere unless 
compensatory divisions elsewhere and far-reaching, gliding move- 
ments of the cells among one another were possible. There is no 
evidence either of the occurrence or the possibility of such movements. 
As has been many times observed, the daughters of the original group 
of four can be recognized late in the life of the colony in their original 
positions with respect to each other and the colony as a whole. 
Oltmanns (04) follows Goebel (’82) and Goroschankin (’75) in 
asserting that this original polar group of four cells does not divide 
after the sixteen-cell stage but does not give any very positive evidence 
on the point. Overton (’89, Taf. III, Fig. 18) in a colony of about 
200 cells shows a group of eight cells, six of which are dividing. It is, 
of course, obvious that a failure of the original group of four to con- 
tinue dividing after the sixteen-celled stage would not prevent the 
maintenance of the rounded form of the colony in case there were 
compensating divisions in the adjacent cells. 
It is interesting to note that Kirchner (’79) finds the development 
of the colony from the fertilized egg of V. aureus essentially like that 
of the asexual germ cell. His figures give some indication of the 
elongation of the cells before division and he describes the cup-shaped 
form of the colony in the eight-cell stage. 
It seems to me that we are justified in concluding that Volvox, 
though showing deep-seated specialization of somatic and germ cells 
in which it contrasts markedly with Eudorina, Pandorina and Gonium, 
still like them shows vegetative totipotency and equivalence of its 
cells in the growth of the colony. This is an important consideration 
in view of the question as to the origin of the differentiation of germ 
and somatic cells which is so conspicuous in the adult colony of Vol- 
vox, and entirely lacking in the simpler members of the series. This 
differentiation is such that the germ cells are distributed solely in a 
single half or three fourths of the colony, the remaining portions re- 
maining persistently somatic-sterile. The fertile area of the colony is 
regularly the posterior half or three fourths as the colony swims. 
It would be of great theoretical interest in this earliest appearance of 
the differentiation of soma and germ plasm if it could be shown that 
the cells bearing the germ plasm were different in cell lineage, age 
since last division, relative maturity as indicated by total number of 
divisions undergone, or in any other way, from the remaining cells of 
