1B22.] Mr. Keates on the Analysis of Brass, 327 



Now had the mode of analysis under consideration been per- 

 fect, the precipitate (B) ought only to have weighed about 87*5 

 grains instead of 111*5, and that at (C), 36*5 instead of 14 ; — a 

 difference much too great to Ix; attributed to any error in the 

 inanipulation. Had the residue left by the potash been taken 

 for oxide of copper, it would have given only 11*2 per cent, of 

 zinc in the brass, which approaches pretty nearly the quantity 

 assigned by some Avriters ; and as the defect in the analysis 

 probably arises from some mechanical cause, it is, therefore,, 

 likely that no two experiments v/ould give the same results for 

 an alloy of uniform composition. 



In recommending another mode of analysis to the notice of 

 chemists, I do not pledge myself that it is one which will give 

 results of mathematical accuracy ; but which nevertheless will 

 be found sufficiently correct for any practical purposes, and may 

 serve until some one more versed in analytic research than 

 myself shall point out a better. 



(A.) 70 grains of copper and 30 grains of zinc were dissolved 

 in dilute nitric acid. The solution, with the addition of a little 

 sulphuric acid, was evaporated to dryness, and redissolved in 

 dilute sulphuric acid, adding the acid considerably in excess. A 

 little sulphate of lead which remained was separated, but not 

 weighed ; it might arise from the zinc or the acid. 



(B.) The solution being much diluted was introduced into a 

 flask, and brought to the boihng point. Three polished cylinders 

 of iron, one inch long each, were introduced, and the boiling 

 continued until the solution became nearly colourless. A portion 

 of the liquid was now tried for copper by a fresh piece of iron, 

 but the surface remained perfectly free from it ; and upon adding 

 one drop of a solution of nitrate of copper, a precipitate took 

 place, which shows that this test was sufficiently delicate. 



(C.) The mixture was filtered while hot, and the copper well 

 washed with very dilute sulphuric acid, and afterwards with 

 boiling water ; and being dried, was put in a crucible covered 

 with charcoal powder, and melted, the button weighed 69*5 

 grains. 



The filtered fluid was now proceeded with precisely in the 

 same manner as at (E) in Experiment I. The oxide of zinc 

 obtained weighed 3ti*75, equal to 2y-4 of zinc. This gives 



Copper 69-5 



Zinc 29-4 



98-9 

 Loss 11 



100-0 



The deficiency of half a grain of copper, and little more than 

 half a grain of zinc, is not greater than what often occurs in 



