1815.] in the Physical Sciences. 23 



was rendered opalescent, both by the pronitrate and pernitrate of 

 mercury, but not by the third mercurial salt; but this third salt 

 produced a slight opalescence in a solution of one part of muriatic 

 acid in 3,552,000 parts of water. 



3. Arsenic. — Some discussions have taken place in Germany 

 respecting the best test for white arsenic. Gartner observed that 

 the formation of Scheele's green, by pouring cuprated ammonia 

 into the liquid containing the white arsenic, does not in certain 

 circumstances take place. Schweigger recommends sulphurated 

 hydrogen, which is certainly a most delicate test of arsenic, in 

 consequence of tiie fine yellow j)recipitate which it forms. The 

 employment of nitrate of silver as a test for arsenic does not seem 

 to be known in Germany. This test, first pointed out by Mr. Hume, 

 but much simplified and improved by Dr. Marcct, is certainly very 

 delicate ; and, when the precautions suggested by Dr. Marcet are 

 attended to, does not seem liable to ambiguity. 



4. Manganese. — Pfaff has published a set of experiments on the 

 method of separating manganese from iron, and has shown that all 

 the methods hitherto proposed on the Continent are imperfect. 

 Bergman's process has been long given up by all chemists. Vau- 

 quelin's, by means of bicarbonate of potash, he found likewise 

 unsuccessful. The method by means of tartrate of potash is in- 

 complete, because the tartrate of potash-and-manganese is soluble 

 in water. Dr. John's method, by oxalate of ammonia, was like- 

 wise unsuccessful. Nor did he succeed better by means of the 

 succinates, benzoates, or phosphates. Mr. Hatchett's method of 

 separating the iron, by means of ammonia, was not tried. I find 

 that by means of it I can very easily procure perfectly pure oxide of 

 manganese. Whether it would succeed equally in forming an 

 accurate separation of manganese and iron for the purposes of 

 analysis, I have not tried ; though I think it probable that it might 

 be used for that purpose with advantage, supposing iron and man- 

 ganese to be the only substances in the solution upon which we 

 operate. 



5. Iron. — Mr. Porrett has recommended triple prussiate of 

 potash, or ferrureted chijazate of potash, as he calls it, as the best 

 method of throwing down iron, and ascertaining its quantity. The 

 precautions necessary to be attended to are the following. The 

 ferrureted ciiyazite must be pure"; There must be no excess of acid 

 in the solution, or as little as possible. All other substances preci- 

 pitated by the ferrureted chyazate must be previously removed. The 

 ferrureted chyazate must not be added in excess, or as little so as 

 possible. The solution being boiled, and the Prussian blue sepa- 

 rated and weighed, every hundred parts of it must be considered as 

 equivalent to 3f235 of peroxide of iron in the solution. 



I have no doubt that when no manganese is present the ferrureted 

 chyazate may be employed with considerable accuracy to detect the 

 quantity of iron in any substance under analysis ; but as it most 

 frequently happens in mineral an.ilysis that iron and manganese ar& 



