34^ Experiments in Galvanic EleSiricily, 



afcended from this wire ; but when the machine wa? in full 

 force, a conficlcrai)!e ftream could be obferved. 



*' The gas was collcfted, and found to be a mixture of hy- 

 drogen and oxygen, in the proportion of three parts of the 

 former to one of the latter. No jxreat dependence however, 

 was placed upon this in point of accuracy. The zinc wire 

 was found to be much corroded, and looked as if a confider- 

 able portion of it had been dilfolved. As the cloud which 

 was formed around this wire became purple on expofure to 

 the light, Mr. Cruickfliank Aifpe£led it might be luna cor- 

 nea, or muriat of filver proceeding from the filver, which 

 had been fomchow diffolved, and afterwards precipitated in 

 this ftate, by the muriatic falts in the common water." 



Diftilled water, to which a little tinfture of litmus was 

 added, was next employed in the tube. Gas arofe from both 

 wires, but in greateft quantity from the filver wire. In a 

 fI)ort time, the whole fluid below the point of the zinc wire 

 became red, and the fluid above the filver wire looked of a 

 deeper blue than before, the flight tinge of purple being de- 

 ilroyed. 



Diftilled water, tinged with Brazil wood, foon became of 

 as deep a purple as could be produced by ammonia, while 

 the portion of the fluid round the zinc wire became very 

 pale. From thcfe experiments it appears to Mr. Cruick- 

 Ihank, that an acid, probably the nitrous, is produced at the 

 wire conneiSted with the zinc, and an alkali, probably am- 

 pionia, at that connected with the filver, end of the pile. 



When lime-water was employed, the wire was likewife 

 a5;ed upon, but in a lefs degree. The cloud at firfl: had ati 

 olive colour, exaelly refembling the precipitate of filver by 

 )ime water. 



In thefe experiments the quantity of filver diflfolved was 

 confiderable, and, where water was employed, a portion of 

 it remained in folution, which was proved by adding muri- 

 atic acid. More would probably have been fufpended, but 

 that an evident precipitation near the upper extremity of the 

 sine wire, vi'as occafioned by the alkali generated by the 

 procefs. 



;\sT^y4rQgen gas, when heated^ or in its nafcent ftate, 



reduce^! 



