Fig. 2. Experimental tank and current-interrupting equipment. 
contact with the main body of water in the 
tank determined how much current would 
flow. As the entire electrode was covered by 
water, gases formed on its surface evenly, and 
their escape did not affect the current flow. 
With this arrangement, it was possible to low- 
er the enclosed electrode to any depth without 
changing the current flow, and a more sym- 
metrical field could be created at each end of 
the tank. 
Carbon was chosen as the most suitable 
electrode material because it is a fair con- 
ductor, inexpensive, and does not erode rapid- 
ly with electrolysis as do most metals. 
To remove the products of electrolysis as 
they were formed and to prevent contamina- 
tion of the water (chiefly by chlorine) in the 
tank, the water within each electrode was 
siphoned off continuously be means of a rub- 
ber tube. The loss of water from the tank was 
compensated for by a steady addition of new 
water. 
Tank 
A tank measuring 2 X 2 X 12 feet was 
constructed from redwood. The inner surface 
was thoroughly impregnated with heated par- 
affin to prevent leaching and to aid in water- 
proofing. Net barriers were placed about 1 
inch in front of each electrode to prevent the 
fish from injuring themselves by contact with 
the electrode. Two other net barriers were 
arranged across the center of the tank to form 
a small area in which the fish could be held 
