420 
fine branches ending within the capsule in flat, disc-like enlargements. 
The bodies of these cells, which are usually large, present a depression 
on their axis cylinder side, giving the cell more or less the shape of 
a bell, and, to carry the figure still further, the nerve process may 
be likened to the clapper of the bell. The axis cylinder is large and 
begins with a cone-shaped thickening. A short distance from the cell 
body the axis cylinder gives off several thin processes, which, after a 
more or less wavy or spiral course, run back (with reference to the 
axis cylinder) toward the cell body, and terminate in small, disc-like 
expansions, which resemble MER- 
KEL’s tactile discs or the discs in 
the corpuscles of GRANDRY. 
Fig. 1 shows such a cell. From 
a, the axis cylinder, are given off 
five fine branches, 0, c, two of which, 
c, are clearly traced to the axis 
cylinder. Each one of the fine 
processes ends in a disc-like ex- 
pansion, d, which, as may be seen 
from the figure, may be nearly 
round or oval or pear-shaped. In 
preparations not too deeply stained, 
these terminal discs or plates are 
more or less granular, the granules, 
Fig. 1. Spinal ganglion cell of Rana C. a, axis cylinder of cell; 5, ce, fine pro- 
cesses ending in d, end-discs; e, secondary process given off from axis cylinder some 
distance from cell body; /, capsule of cell; g nuclei within the capsule. 
which vary much in size, staining much more deeply blue than does 
the ground substance. The fibril e, undoubtedly was given off from 
the axis cylinder at a point farther away from the cell body; its con- 
nection with the axis cylinder was missed in this section. The ar- 
rangement of these secondary branches varies greatly among the dif- 
ferent cells. Often their arrangement is very complicated; now and 
then they are wound spirally about the axis cylinder, so that their 
connection with it is made out with difficulty and often cannot be 
discovered. 
This is quite well shown in Fig. 2, where only one of these 
branches, 5b, can be clearly traced to the axis cylinder. Now and then 
these fine branches divide into two or three secondary branches (see 
Fig. 2 c, and Fig. 3), each of which ends in the disc-like expansion. 
