and Agency of Ele£lndty . 209 



lary glafs tube; and, after heating the tube fo as to make it 

 adhere to the point, _and cmer it in every part, I gradually 

 ground it down, till, with a pocket lens, I could dilceni that 

 the point of the cold was cxpoied. 



The fuccefs of this nieiiiud exceeding my expeftations, I 

 coated feveral wires in the fame manner, and found, that 

 wlicn iparks from the conductors before mentioned were 

 made to pafs through water, by means of a point fo guarded, 

 a fpark paffing to the diftance of i of an inch would decom- 

 pofe water, when the poiut expofed did not exceed ji, of an 

 inch in diameter. With another point, which I eftimated 

 at tt'ocj 3 fucceflion of fparks, ~ of an inch in length, af- 

 forded a current of I'niall bubbles of air. 



I have fince found that tiie fame apparatus will decompofe 

 water with a wire — - of an inch diameter, coated in the 

 manner before dcfcribed, if the fpark from the prime con- 

 dutlor palfes to the diftance of ,\ of an inch of air. 



Eixj.cr. 7. In orlfer to try how far the Itrength of the elec- 

 tric fpark miiiht be reduced by proportional diminution of 

 the extremity of tiie wire, 1 palfed a folution of gold in aqua 

 rcgia through a capillary tube, and, bv heating the tube, 

 expelled the acid. There remained a thin film of gold lining 

 the inner furface of the tube, which, by melting the tube, 

 was converted into a very fine thread of gold, through the 

 fubftanco of the glafs. 



When the extrcmitv of this thread was made the medium 

 of communication through water, I found that the men" 

 current of electricity would occafion .i ft ream of very fmal I 

 bubbles to rife from the extremity of the gold, although the 

 wire, bv which it communicated with the pofitive or nega- 

 tive conductor, was placed in abfolute contact with them. 

 Hence it appears that decompofiiion of water may take place 

 by common cledtricity as well as by t!ie electric pile, although 

 no difccrniblc fparks are produced. 



The appearance of two currents of air may alfo be imitated 

 by occafioning the electricity to pafs by fine points of com- 

 munication on both fides of the water; but, in hR, the re- 

 femblance is not complete; for, in every way in which f 

 have tried it, 1 obfcrved that each wire gave both oxygen 

 and hydrogen gas, inilcad of their being formed feparately, 

 as by the eleCtric pile. ' 



I am inclined to attribute the difiercncc in this refpeCl to 

 the greater intenfity with wJiich it \i ncceflary to en. ploy 

 common electricitv; for, that pofilive and negative electricity, 

 fo excited, have each the fame chemical power as thcv arc 



VojL. XI. O vblMvid 



