Ohfervations of Galvanic Proctjes. 397 



beginning to oxidate In many of their points, though no gas was produced upon them. 

 No gas had formed upon the filver plates, but the furface of the water in conta£l with 

 them became covered with an opaque white pellicle. 



2. The filver plates ufed In this experiment were not perfe£lly poliflied. This might 

 have influenced the refults: and fome gas might have efcaped my obfervatlon. That the 

 operation might go on in clofed vcflels, I cut off the bottoms of fome bottles with a file, 

 fo that they could be eafily joined again by cement. Into one of thefe bottles I introduced 

 a plate of polifhed zinc *, and Into another a plate of poliflied filver. The plates were 

 connetled by a wire attached to their upper angles, which protruded Into the atmofphere 

 through orifices made at the places of jundlion of the bottoms of the phials with their 

 fides ; thefe places of juniftlon being rendered perfeftly water-tight by refinous cement. 

 Four apparatufles of this kind were conftrufted. They were filled with pump water, in- 

 verted in the galvanic order In glafles containing that fluid, and made part of a connected 

 ferles of twenty glafles. 



After more than twelve hours, the zinc plates had become tarnllhed, but had given out 

 no gas. In two of the bottles with the filver plates, globules of gas, too fmall to be 

 analyfed, had coliedled : thefe plates examined in the atmofphere, as well as all the 

 other filver plates, were covered in fome points with a film of white fubftance, 

 which was foluble with flight efi^ervefcence, and without producing cloudlnefs in 

 muriatic acid. 



3. Unable to account for the non-appearance of hydrogen during the oxidation of the 

 zinc, I could not but conclude that It was condenfed or abforbed In fome new compound 

 on the furface of the filver or the zinc. Guefling that the quantity of furface might be 

 connefted with the phenomenon of its non-appearance, I fubftituted in three of the phials 

 for the fquare filver plates oblong ones of the fame length, and about ,3 Inches wide. 

 Thefe had not been long connected with the feties before gas began to form upon them j 

 and In five hours fufiicient was collefted to be examined : from the coarfe teft of inflam- 

 mability it appeared to be hydrogen. 



Thirteen pairs of a conne£ted feries of twenty-five glafles were now compofed of 

 fquare zinc plates, and oblong filver plates of dlfl^erent fizes; fome of them being about 

 ,3, and others not more than ,1 inches wide. Gas was almoft immediately given out from 

 the greater number of the oblong plates, and in largeft quantities from the fmalleft : 

 from tlie flips of ,1 wide indeed a conftant ftream of globules afcendcd through the 

 water. 



Small oval, circular, and fquare plates of nearly equal furfaces, with the flips con- 

 nected in the feries In the places of fome of them, produced precifely the fame efFefts. 

 In fhort, whenever the furfaces of the filver plates did not exceed one-fourth of the quan- 



* Of I.I inches fquare, the fize ufed in all the former experiments. 



tity 



