His paper on the torpedo was read before the Royal Society in 1776. The experiments were undertaken in order to determine whether the phenomena observed by Mr. John Walsh in connection with the torpedo could be so far imitated by electricity as to justify the conclusion that the shock of the torpedo is an electric discharge. For this purpose Cavendish constructed a wooden torpedo with electrical organs, consisting of a pewter plate on each side, covered with leather. The plates were connected with a charged Leyden battery, by means of wires carried in glass tubes, and thus the battery was discharged through the water in which the torpedo was immersed, and which was rendered of about the same degree of saltness as the sea. Cavendish compared the shock given through the water with that given by the model fish in air, and found the difference much greater than in the case of the real torpedo, but, by increasing the capacity of the battery and diminishing the potential to which it was charged, this discrepancy was diminished, and it was found to be very much less in the case of a second model having a leather, instead of a wooden, body, so that the body of the fish itself offered less resistance to the discharge. One of the chief difficulties lay in the fact that no one had succeeded in obtaining a visible spark from the discharge of the torpedo, which will not pass through the smallest thickness of air. Cavendish accounted for this by supposing the quantity of electricity discharged to be very great, and its potential very small, and showed that the more the charge was increased and the potential diminished in his model, the more closely did it imitate the behaviour of the torpedo.