216 Professors Redtenbacher and Liebig 



ning the atomic weight of a metal, that it might be supposed 

 the results obtained would enable us to settle the question, 

 whether or not the atomic weight of a metal is a multiple of 

 that of hydrogen; but the nearer we approach to absolute 

 accuracy, the greater are the obstacles we have to overcome 

 in arriving at such a point. Besides, the circumstance that 

 very few bodies submitted to analysis are absolutely free from 

 all impurity, or from the substances from the compounds of 

 which they are separated, introduces a difficulty which is often 

 not less than that of accurately conducting the analysis itself. 



" I was of opinion that crystallized nitrate of lead, which 

 is ignited in a platinum crucible until the nitric acid is entirely 

 decomposed, would give perfectly pure oxide of lead ; but 

 when this oxide is reduced by means of hydrogen gas, and 

 the metallic lead dissolved in nitric acid, lead-coloured scales, 

 which prove to be metallic platinum, remain behind. It is 

 true the quantity of this residue is very small, but if the result 

 be depended upon up to the last figure, then even the smallest 

 impurity should be avoided. Crucibles of gold and of silver 

 were also employed, but these are oxidized, and combine with 

 the oxide of lead, even when the salt is introduced in small 

 quantities into the crucible previously heated to redness ; 

 indeed oxide of lead thus obtained, is impregnated with the 

 foreign metal to a greater extent than when the calcination 

 takes place in a crucible of platinum. 



" This induced me to employ carbonate of lead prepared 

 by precipitation, partly from the acetate and partly from the 

 nitrate of lead, by means of the carbonate of soda ; and in 

 order to avoid all admixture of the carbonate of soda with 

 the precipitate, an excess of the salt of lead was employed ; 

 but notwithstanding this precaution, and the perfect washing 

 of the precipitate, the atomic weight obtained by means of 

 reduction, oscillated in both cases between 1303*5 and 1306'0 ; 

 and when the metallic lead was heated with pure water, the 

 latter was found to contain carbonate of soda. Carbonate of 

 ammonia, employed as the precipitant, is not objectionable on 

 this ground; but, partly because, if special care be not taken, it 

 is difficult to obtain that reagent perfectly free from all traces 

 of the hydrochlorate and of the sulphate of ammonia ; and 

 also because, during the calcination, the hydrogen of the am- 

 monia may reduce a small quantity of the oxide to the con- 

 dition of suboxide, which, though not perceptible, nevertheless 

 notably alters the result. 



" By the following method I think I succeeded in obtaining 

 perfectly pure oxide of lead. Nitrate of lead is heated to 

 redness in a platinum crucible, until nothing but oxide of lead 



