178 BULLETIN OF THE 
Since in the ripe individual the mass of sexual products whch fills 
the body cavity would exert a dangerous or even fatal pressure on cells 
so delicate as these, it is evident that the partition is absolutely neces- 
sary for their development, and may be so for the protection of the brain. 
While the presence of the partition is essential to the existence of these 
cells as they are, it is impossible from the evidence at hand to form any 
idea of the cause which led to its development. However, the partition 
being formed, I believe it is possible to understand how the large cells 
may have attained their size and position. 
It has already been mentioned that the dorsal cells vary considerably 
in size; an early sketch of a living animal, made before the structure 
was well understood, shows in a dorsal view, not two, but three succes- 
sive pairs of large cells. There were, however, in this case, actually 
only two pairs of dorsal cells, and the supposed third pair was the most 
posterior pair of large ganglion cells in the brain. It has already been 
said of these that they are located, not zz the mass of the brain, but in 
great part above it. Now, given two pairs of lateral cells located on the 
upper surface of the brain anterior to the fifth pair of ganglion cells 
near the place where the processes of the dorsal cells enter the brain 
mass, it is easy to conceive how they may have become larger and larger, 
and finally may have risen entirely above the brain into the free space 
dorsal to the cesophagus, where no obstacle is offered to their further 
increase in size. At the same time, the stalk would be produced as a 
mere mechanical result of the lengthening of the cell process to accom- 
pany this migration. Once free in this cavity there is every reason to 
believe that the cell might continue to develop in size until, with its 
companions, it should occupy the entire space, which is approximately the 
present condition of affairs. The fifth pair of large ganglion cells in the 
brain, half projecting, as they do, above its surface, would then repre- 
sent the first stage in the migration which the dorsal cells have already 
accomplished. Attention must be called to the fact that the three pairs 
of cells which have been compared in this hypothetical statement of the 
case are not similar in one important particular ; for whereas the pro- 
cesses of the fifth pair of ganglion cells are unbranched and may be 
traced far posteriad through the ventral nerve cord, the processes of the 
dorsal cells are branched and can be traced only a comparatively short 
distance. I do not see, however, that this difference greatly affects my 
explanation, — which is purely mechanical, — since the cell and not 
the process is directly concerned. The nuclei of both kinds of cells are 
nearly equal in size and alike in structure. The great difference be- 
