GIBBS. — COMPLEX INORGANIC ACIDS. 267 



Auramin Pyrofliospliales, 



When a solution of double chloride of gold and sodium AuCl4Na is 

 made as nearly neutral as possible and then boiled with sodic pyro- 

 phosphate P207Na4 , a very pale yellow solution is formed, which con- 

 tains the auro-sodic salt discovered by Persoz, the formula of which 

 we should now write PaOyAu'^'Na. The solution of this salt gives 

 with ammonia a white precipitate which quickly becomes yellow and 

 crystalline. With argentic nitrate it gives a pale yellow flocky crys- 

 talline precii^itate, which is very insoluble and does not blacken readily 

 in the light. With mercurous nitrate it gives a greenish gray flocky 

 crystalline precipitate. A nearly white curdy precipitate is formed 

 with sulphate of luteo-cobalt ; none with sulphate of croceo-cobalt. 

 With Pt (NH3)4Cl2 gold is reduced. 



The crystalline precipitate formed by ammonia in the auro-sodic 

 pyrophosphate solution was well washed, dried at 150°, and analyzed. 



0.9665 gram gave 0.6756 gram gold = 78.44% AugOs. 

 0.9031 gram gave 0.7003 gram gold = 78.34% AuoO,.- 

 0.9665 gram gave 0.1621 gram P-AMgo = 10.73% PA- 

 1.0031 grams gave 0.1703 gram PAMga = 10.86% PA- 

 0.7375 gram gave 0.02286 gram NII3 = 8.10%. 

 0.7691 gram lost, at 150° C, 0.0420 gram water = 5.46%. 



The analyses lead to the formula 



14 AU.O3 . 6 PA • 14 NII3 . 3 Na^O -f 24 Aq., 



7886 100.00 



The analysis is due to Dr. Morris Loeb, who assisted me in this 

 part of my work most efficiently. 



The salt does not explode at low temperatures, but on heating to a 

 little above 170° C. a violent explosion takes place. This shows 

 clearly that part at least of the gold is in the form of an auramin, or 

 compound with ammonia. 



