Ejcpcnmenfs on the TvrpeJf). 359 



each olker^s left hands, plnngea pouited piece of inetal into 

 a drop of water placed upon an insiilaliiig body. 



12. By substituting" flame instead of a dro]} of water, the 

 .comnumication is interrupted, and no sensation is expe- 

 .rienced until the two pieces of metal touc-^ each other 

 within the flame. 



'13. We must also observe that 110 shock will take plac-e 

 either in air or under water, unless v/e immedialelv touch 

 the body of ihe electric ilsh. They are unable to give their 

 ■stroke through a iaver of water, however thin; a fact 

 which is the more rem;wkable, as we know, that in Galvanic 

 experiments, wlicre the frog i-s placed under water, it is 

 sutficient to bring the filver forceps near to the muscles^ 

 and that the contraction takes place when the Jaycr of watcj 

 interposed is one or two milliuietreg in tliici:ne«s. 



Such, m\' respected friend, are the principal obaervEtion? 

 whic& we iiave n,Kide upnn the torpedo. Experiments 4th 

 .and 1 0th prove that tlie electric orgaos of these animak 

 manifest no tension or excels of charge. We should rather 

 be inclined to .compare their action to a chain of small 

 Leyden phials than to the pileof Volta. As some commu- 

 nication is always necessary for the occurrence of a s:hock, 

 and having •received strokes from the gymnolus through 

 very dry cords, I conclude that in the case where this 

 powerful animal appeared to give these violent shocks with- 

 out the existence of any conimunicatiwn, it must have 

 arisen from my impel tcct in:=ulation. If the torpedo acts 

 by poles, by an electrical e(piili'!)riuni vrbich tends to re- 

 establish itself, experiments 5th and 6th seem to prove that 

 these poles exist near one another upon the same surface of 

 the organ: for we fee] the shock by toncliing a single sur- 

 face with the finger. A plate interposed between Uhe hand 

 and the oruan .(ii.) re-establishes of itself the «|uiliiH-imTi, 

 and the hand \\ hich supports this plate fcfk nothing^ be- 

 cause it is out of the current of the electj-ic inlluence. But 

 if we suppose a number of jwles -fjf tlifl-erent descriptions 

 upon each siiTJace of jlvp organ, \\ by is it that, by cover- 

 ing these with two metallic plates the edges of whidi do 

 not touch, and pixcing the hands upon thcoc plates, the 

 eqftilibriimi is re-cf-tabTished through ihe me/]iu«i of the 

 arms ? Wherefore, it may be asked, does not the positive 

 electricity oi' the inferior surface seek, at the moment of 

 explosion, the negative electricity of the neigWxjuringpole? 

 And whtrefore does it find it only in the .superior surface of 

 the ekplnc organ ? These difiiculties arc perhaps not in- 

 Z 4 SMJUJOuntable, 



