SECONDARY PRECIPITATION OF GOLD 261 
from normal up to saturation, but no precipitation occurred, even 
when the solutions were boiled for several minutes. All of these 
solutions gave acid reaction, due partly to the hydrolysis of the 
auric chloride: 
AuCl,+3H.O = Au(OH);+3HCl 
and partly to the addition product of auric chloride and water, 
H.AuCl,O, which ionizes as an acid’ as follows: 
H.AuCl,O = 2H+-+ AuCl,O= 
It is well known that the efficacy of many reducing agents is 
increased by the alkalinity as is readily explained by the electro- 
lytic theory of oxidation and reduction.* In accordance with this 
theory it was thought that neutralization of the acid solution might 
induce precipitation. N/50 potassium hydroxide was slowly added 
to 10 c.c. of a solution N/5 with respect to manganous salt and 
containing 75 mg. of gold as chloride. When about 6 c.c. had been 
added a brown precipitate began to form. The alkali was added 
until the solution was nearly neutral with respect to litmus, the 
precipitate was collected on a filter, and the manganese oxide 
was removed by digesting with a mixture of oxalic and sulphuric 
acids of known reducing strength. The residue was brought 
into a crucible, dried, and shown to be metallic Au. 
In order to determine which oxide of manganese was formed in 
this reaction, the solution in the oxalic and sulphuric acid was 
divided into two equal parts. One of these was analyzed for 
manganese, and the other titrated with standard potassium 
permanganate solution to ascertain the loss of oxalic acid due to 
oxidation by the manganese oxide. The titration showed an 
oxidation of oxalic acid corresponding to one atom of oxygen for 
every molecule of MnO as determined in the analysis for manganese. 
The precipitate was clearly MnO,, doubtless with water of hydration. 
In a check experiment the manganese oxide was taken into solution 
with a mixture of oxalic and sulphuric acids, equivalent to 45.28 c.c. 
N/to alkali. After dissolving the precipitate the solution was 
equivalent to 32.26c.c. N/1o alkali, having lost the equivalent 
« Hittorf and Salkowsky, Zeitschr. f. Phys. Ch., XXVIII, 546-55 (1899). 
2 See Stieglitz, Qualitative Analysis, Part I, 2609 ff. 
