Structure and Polarity of Electric Motor Nerve-Cell in Torpedoes. 217 
Each of the cells in question is a typical nerve-cell of extraordinary 
size. They grow with the growth of the fish to a certain degree and 
in a well-developed fish, of say 35 cm. in length, these cells will average 
75 microns in diameter. They are compact in shape, as the various 
figures will show, and each gives off from 20 to 40 processes, one of which 
is the neuraxon or main impulse-carrier from the cell to the electro- 
plax which it supplies. The remainder are dendritic in structure and 
are thin and richly branching. Some traces of a weak polarity in the 
arrangement of the processes can be seen (plate 3, all figs.) in the fact 
that in many of the cells the neuraxon is given off at one end of a 
slightly elongated cell, while a majority of the dendrites are given off 
at the other end. In other cases this is not so apparent and the 
distribution of the dendrites seems to be very general and not restricted 
D 
\\ Toran Gach 
MU AU nr tiga oh Ed 
See (ply Waa ly 
~ ae / ns Sse ww 
~~ — Z SS ~—— 
iS eee 
hie. i Fie. 2. 
Fic. 1.—Diagram, after Cavendish, to indicate course taken through the surrounding water by 
the electric current generated at time of discharge of electric organ in Torpedo marmorata. 
Copied from ‘‘The Physiology of Nerve, Muscle and Electric Organ” by Dubois- 
Reymond, translated by J. Burdon-Sanderson. Oxford, 1887. 
Fic. 2.—Improved and corrected diagram of same conditions as in fig. 1, by Dubois-Reymond. 
Copied from same source as fig. 1. 
to any one area. Preparations made by teasing bits of electric lobe 
that have been macerated for 24 to 36 hours in one-third alcohol serve 
to show these features to advantage in plate 3, and in such isolated 
cells the processes can be seen for some distance, being seldom broken 
off less than the diameter of the cell in length, while in many cases 
they are shown for a much greater distance. There is seldom any 
difficulty in distinguishing the axis-cylinder process from the dendrites. 
It takes its origin from the efferent pole or end of the cell, which forms 
a large and fairly well-defined implantation cone and which is further 
marked by being at a somewhat thinner and tapering end of the cell 
and by being bounded by the larger mass of pigment material. The 
process is not noticeably thicker than the dendrites, but it holds its 
thickness for a long distance from the cell. Also, like the implantation 
