Mr . cavendish on the Torpedo. 209 
* 
parated to a given diftance ; becaufe in moil of the expe- 
riments the electricity was fo weak, that a pair of line 
pith balls fufpended from the battery would feparate 
only to a very fmall diftance; and counting the number 
of revolutions of the electrical machine is a very falla- 
cious method. 
1 found, upon trial, that though a fhock might be 
procured from this artificial torpedo, while held under 
water, yet there was too great a difproportion between 
its ftrength, when received this way, and in air; for if I 
placed one hand on the upper, and the other on the 
lower furface of the electric organs, and gave fuch a 
charge to the battery, that the fhock, when received in 
air, was as ftrong as, I believe, that of the real torpedo 
commonly is ; it was but juft perceptible when received 
under water. By increafing the charge, indeed, it be- 
came confiderable ; but then this charge would have 
given a much greater fhock out of water than the tor- 
pedo commonly does. The water ufed in this experi- 
ment was of about the fame degree of faltnefs as that of 
the fea; that being the natural element of the torpedo, 
and what Mr. walsh made his experiments with. It was 
compofed of one part of common fait diffolved in 30 of 
water, which is the proportion of fait ufually faid to be 
contained in fea water. It ap red alfo, 011 examination, 
to conduct electricity not ft fibly better or worfe than 
fome fea water procured from a nit: era! water ware- 
houfe. It is remarkable, that if I ufed frefli water in- 
ftead of fait, the fhock feemed very little weaker, when 
V o l. LXVI. E e received 
