HARPER: BINARY FISSION AND SURFACE TENSION 161 
These are the relations of the two groups of four, as described by 
Biitschli and Overton. As noted, it is with reference to the position of 
the eight equatorial cells that clear description has been lacking. Text- 
figure 2 shows an equatorial view of the model and it is obvious at 
once that a least surface configuration requires the eight equatorial 
cells to form a zigzag belt, each cell alternating with two cells of the 
polar groups of four. With this arrangement the colony is made up 
of eight groups of three cells, the most compact arrangement possible, 
each of the four cells of each polar group appearing in two of the groups 
of three. The groups of three are so placed with reference to each 
other as also to form the groups of four seen in both the polar and 
equatorial views. The whole colony would consist of eight of these 
groups of four, since each of the eight equatorial cells is a member of 
three such groups and each cell of the polar group is a member of 
two such groups. 
The whole forms as perfect an approximation to a least surface 
configuration as can be achieved by sixteen cells arranged in as nearly 
as possible a spherical group. Such a configuration may be regarded, 
it seems to me, as the type configuration and illustrates the inter- 
action of surface tension, adhesion and binary fission as morphogenetic 
factors. In the actual colonies the cells are, of course, flattened upon 
each other and, as I have pointed out, elongate during division. As the 
photograph (Fig. 7, Pl. Il) shows, they are quite variable in both form 
and size. As is shown clearly by Klein also in all figures of the so- 
called polar openings, phialopores, the posterior group of four do not 
form any such definite square group as is maintained by the anterior 
group of four, yet the tendency to the formation of the groups shown 
in the model is, it seems to me, obvious. The abundance of pentagons 
and hexagons in the cell outlines of the adult colonies is good evidence 
of a tendency to the most compact configuration possible. 
I have not attempted to make the model with any great accuracy. 
We lack data as to the relative efficiency of adhesion and surface 
tension in the cells, which would be necessary for the exact determina- 
tion of their interrelations. The actual relations of fission, adhesion 
and surface tension in the processes just described are perhaps brought 
out more clearly in the two diagrams (Text-figs. 3 and 4). 
If the protoplasm of the mother cell were a mass which grows 
merely by imbibition and swelling, and if the cutting up of the mass 
into cells were merely a secondary phenomenon following the principle 
of the rectangular intersection of the cleavage planes and surface 
tension, as. Hofmeister (’67), Sachs (’78), and other critics of the cell 
theory have assumed, we should expect a configuration of the cells like 
that shown in text-figure 4. This is the sort of configuration Magnus 
