31 6 Analyjls of the Arfeniates 



nitric acid, and evaporated, wa;; precipitated by nitrate of 



lead. Upon filtration, 82 were left, which indicate 27 of 



arfcnic acid. Therefore this arfeniate of copper contained, 



. Copper - - - 50 



Arfenic acid - - 27 



Water - - - 22 



.99 . , 



The fecond artificial compound, which was mentionefi 

 above, was evidently more foluble than the latter ; and ana- 

 logy might lead us to fufpe61:, in arguing from the generality 

 offidts the balis of which is fuperfaturated, that it contained 

 an exccfs of acid. It was analyfed in the fame manner as 

 the laft, and aflbrdcd. 



Oxide of copper - ~ 35 



Arfenic acid _ - - 39,^ 

 Water - - - - 24 



98,5 

 Thus then have we two artificial arfeniates of copper, one 

 of which intimately corrcfponds with one of thofe which we 

 have recognized among the produAions of nature. The other 

 poffibly will be found, but we are not yet in polTeffion of it; 

 for I fhal! prefently mention the reafon why No. [., although 

 containing 39 of acid, cannot fairly be efteemed as fuch. 

 I have not yet been fo fortunate as to form the other com- 

 binations, but do not doubt that art may one day fucceed in 

 obtaining them. 



T.EMARKS. 



Before I conclude this feclion, which hitherto has had for 

 its objetl a particular account ot certain kinds of copper ore, 

 as well as of their matrices, it may not be fuperfluous to offer 

 a few remarks, not foreign to the prefcnt fubjeft, upon fome 

 methods generally ufcd in the docimallic art. To prove the 

 prefence of different fubilances in foflils, is an objed: of deli- 

 cate refcarch : but, to determine proportions with accuracy, 

 is the moft difficult operation of analytic chemiflry, and 

 often eludes in'.'edigatlcn. It is rather a pleafing refleftioii to 

 think that we are in the infancy of chemical exaclnefs ; and 

 that wc may fee the day of improvenient, when the errors 

 which we now conmiit will rcrpiire all the aid of felf-compla- 

 cency to be in the leaft excufed. And it may be of more real 

 utility to flale with frankncfs, although we cannot account 

 for them, thofe anomalous appearances which fo frequently 

 oLOur, tiiaa to coari the phamom of rigid accuracy, the re- 

 ality 



