1*0 T^olta on Armnal Eit-clrichv, 



©r eontrariwife by means of a conduftor of the firft clafl 

 placed between two of the fecond clafs, alfo different. ; thcr^ 

 isftill a third method of exciting the ele6tric fluid, though in 

 a degree fo much weaker that it is fcarcely capable of cauling 

 convulfions in a pcrfeclly prepared frog, in which there is 

 fiill a ftrong degree of vitality. This new method confifts in 

 formingr the circle of three different condutlors, all of the 

 fecond clafs, without the intervention of one of the fitft or 

 a metal one, Some think they iind in this method a ftrong 

 objeftion agaiuft my prineipie, 



Fig. 20. repfefents this third method compared with the 

 other two*. In the experiments of Profeilbr \^alli, refpeft- 

 in*' which fo much noife has been made without any realon, 

 t reprcfents the leg of the frog, and particularly the hard ten- 

 dinous part of the ]\IuJ'atlus Q^alh-ocncmhis ; ;» the rump or 

 the nuifcles of the back, or the ifchialii' nerves, to which the 

 faid tendinous parts are applied; and a tlic blood, or the viU 

 cous faponaceous or falinc fluid, applied to the point of 

 conta(5l. 



1 have fully defcribed this new method, where no metal is 

 ufed, in my third and fourth letter to Profeffor Vaflali, writ-, 

 ten in the autunui and winter of the year 1795. I have there 

 fiiewn, that thefe new fa6ts, far from altering my ideas an4 

 principles, ferve rather to eflablifli them ; and that they ren-, 

 der more general the principle that the conductors, by hete-r 

 rosencous contaft, that is of two different from each other, 

 become exciters of eleftricity, and confirm the beautiful law 

 arifinty from <t, that to produce an eleclric ftream the cir- 

 cle muft neceffarily be formed of three difl'crent conduiSlors. 

 You now fee in what the whole fecret, the whole magic con- 

 fifts; and that it depends not merely on metals, as might 

 have been believctl, but on all tlie different conduftors. Aa 

 Ion"- as we adhere to thefe principles, it will be eafy to ex- 

 plain all the before-mentioned experiments without being 

 feduccd to the neceffity of having rccourfe to any imaginary 



* Sec riite ]. i\\\ ihe preceding >Juinbcr,_) 



pr^nci|ol?, 



