OF ARTS AND SCIENCES. 57 



per, and subsequently with alcohol, until no trace of free sulphuric acid 

 remained. After drying, the precipitate of sulphate of lead and gold, 

 with the filter cut into small pieces, was transferred to a beaker glass, 

 and the whole was then digested with pure bicarbonate of soda * (pre- 

 pared from the oxalate), during at least forty-eight hours. The sul- 

 phate of soda formed by this opei'ation was then filtered off from the 

 carbonate of lead and metallic gold, and the amount of sulphuric acid 

 which it contained determined by precipitation with chloride of barium 

 in the usual way. From the amounts of sulphate of baryta thus 

 obtained, recorded in column (3) of the table, the amounts of lead 

 (col. 4) in the mixed precipitates of sulphate of lead and gold were 

 calculated. 



The carbonate of lead, mixed with bits of paper and gold, above 

 mentioned, was now dissolved in dilute nitric acid, the solution evapo- 

 rated with sulphuric acid until all the nitric acid had been expelled, 

 subsequently treated with water acidified with sulphuric acid, and, after 

 standing for forty-eight hours, collected upon a tared filter, washed 

 with dilute sulphuric acid and alcohol, dried at 100°, and weighed. 

 The results may be found in column (a) of the table. 



The re-agents used in the foregoing operations were all chemically 

 pure. A special experiment was moreover made, in which portions 

 of the nitric acid, ammonia, chloride of ammonium, and sulphuric acid 

 used, twice as large as any which were actually employed in the 

 analysis of either of our samples of coin, were evaporated together, 

 and subjected to a course of treatment identical with that which the 

 solutions of coin were forced to undergo ; but not a trace of sulphate 

 of lead, or of any precipitate, other than a few light and utterly insig- 

 nificant floating flocks, probably alumina, could be detected as the 

 result of this trial. 



That our process is sufficiently accurate to exhibit all the lead which 

 the samples of coin examined really contained, we do not think pi'ob- 

 able. All of the sources of error to which the process, when properly 

 conducted, is subject, tend to cause a slight loss of lead. Thus, in the 

 first place, it is scarcely possible that the last traces of chloride of lead 

 can be removed by washing from the caseous precipitate of chloride 



* See Fresenius, Anleitung zur Quantitativen Analyse, (Bnvunschweig, 4'^ Aufl. 

 1858,) p. 286, ^ 132, II. b. /3. 

 VOL. V. 8 



