by Sodium Thiosulphate 53 



The errors in experiments 1, 3, 4, 5, 6, 10, 11, 12 and 17 are, 

 proportionally to the amount of material handled, practically 

 the same and this fact caused me to make a careful revision of 

 all standards ; but no mistake could be found. The reaction 

 upon which the process depends requires the formation of 

 Hg 2 S, but this mercurous sulphide breaks down immediately 

 into mercuric sulphide and mercury. The latter is probably 

 acted upon by the free nitric acid present to form mercuric 

 nitrate, which in turn is transformed into the compound 

 2HgS.Hg(]N"0 3 ) 2 by the action of the thiosulphate. At any 

 rate, with an error so large, whatever its source may be, the 

 process is plainly impracticable. 



The third step in Scherer's process deals with the action of 

 sodium thiosulphate on mercuric nitrate. In the following 

 experiments a solution of mercuric nitrate was prepared either 

 by dissolving as far as possible 20 grams of mercuric nitrate in 

 about 200 cm3 of cold water, filtering off the supernatant liquid 

 and diluting to a definite volume (exps. 1-8), or by dissolving 

 the salt in strong nitric acid and diluting (exps. 9-19). The 

 standard of the solution was obtained by precipitation as 

 metallic mercury by means of the electric current. The yel- 

 low precipitate, formed according to Scherer's reaction, 



3Hg(N 3 ) 2 + 2Na 2 S 2 3 = 2HgS. Hg(N0 3 ) 2 + 2Na 2 S0 4 + 2N 2 5 , 



on adding the sodium thiosulphate settles much better than in 

 the case of either mercurie chloride or mercurous nitrate ; but, 

 as the supernatant liquid takes on a permanent yellow color 

 towards the end of the reaction, it is impossible to see when 

 the thiosulphate produces no further precipitation. On this 

 account, therefore, the same method of procedure was adopted 

 as in the case of the mercuric chloride and mercurous nitrate, 

 i. e., filtration and titration of the excess of sodium thiosul- 

 phate with iodine and starch solution. The result of the 

 experiments is shown in the following table. 



These results seem to show the impossibility of obtaining 

 accurate results according to Scherer's method for the determi- 

 nation of mercuric nitrate by direct titration with sodium 

 thiosulphate. The constant plus error cannot be accounted for 

 on the hypothesis that the nitric acid present decomposes the 

 sodium thiosulphate, for in that case the error would lie in the 

 other direction. It is more probable that the constitution of 

 the compound 2HgS, Hg(N0 3 )„ is not definite enough to make 

 it the basis for an analytical process. 



