syz Mr. Cavendish’s Experiments on the 
faites par le moyen de la Machine eleBrique Teylerienne , p. 1 82. has 
deicribed the method employed by him and M. Van Troots- 
wyk. They ufed a glafs tube, the upper end of which was 
flopped by cork, through which an iron wire was palled, and fe- 
cured by cement, and the lower end was immerfed into mercury ; 
fo that the electric fpark palled from the iron wire to the foap- 
lees. After fo much of a mixture of five parts of dephlogifti- 
cated and three of common air as was equal to twenty-one 
times the bulk of the foap-lees * was ablorbed, fome paper was 
moiftened with the alkali, which by its burning appeared to 
contain nitre, but (hewed that the alkali was not near fatu- 
rated. The experiment was then continued with the fame 
foap-lees till more of the air, equal to fifty-fix times the bulk 
of the foap-lees, was abforbed, which is near double the quan- 
tity required to faturate them ; but yet the diminution went 
on as fall as ever. It was then tried, by the burning of paper 
dipped into them, how nearly they were faturated ; but they 
Hill feemed far from being fo. 
The circumftance of ufing the iron wire appears evidently 
objectionable, on account of the danger of the iron wire being 
calcined by the eleCtric fpark, and abforbing the dephlogiffi- 
cated air ; and when I firft read the account, I thought this 
the moft probable caufe of the-" difference in the refult of our 
experiments ; but I am now inclined to think that the cafe 
was otherwife. From the manner in which M. VanMarum 
exprelfes himfelf, it feems that the only circumftance, from 
which they concluded that the alkali was not faturated, was 
the imperfefl: marks of deflagration, that the paper dipped into 
it exhibited in burning ; which, as we have feen, might pro- 
ceed as well from fome of the mercury having been diffolved 
* This is rather more than half of that requifite to faturate the foap-lees. 
as 
