328 M. C. Lea — Silver llemisulphate. 



neonsly to resolve themselves into metal and normal haloid. 

 They rarely remain as hemihaloid for more than an hour or 

 two, and often for much less time. The change is often quite 

 sudden and is easily observed by the alteration of color, the 

 black of the hemihaloid passing into the metallic gray color 

 belonging to a mixture of normal haloid with metallic silver. 

 The hemibromide seems to be a little less unstable than the 

 hemichloride. 



This instability does not render an analysis impossible since 

 both the products of the change are insoluble ; but renders it 

 somewhat more difficult, as the freshly formed silver haloid 

 tends to run through a filter. Sometimes indeed it seems as 

 if traces of the silver chloride were for a few moments soluble 

 in water with a yellow coloration. The appearance of this 

 yellow color in the water is apt to be the first indication of 

 the splitting up of the hemichloride. 



Two analyses were made, one of material obtained by acting 

 on the brown salt with dilute hydrochloric acid ; this con- 

 tained 81*79 per cent of silver. One by decomposing it with 

 sodium chloride ; this gave 81-93 per cent. .A substance hav- 

 ing the formula already given should, by conversion into 

 chloride, give a mixture in which two-thirds of the silver 

 should exist as hemichloride, and one-third as normal chloride. 

 We have then — 





, — Found.—, 









1. 2. 



Mean. 



Calculated. 



Ag per cent.. 



. 81-79 81-93 



81-86 



82-35 



a result sufficiently close to afford a confirmation of the con- 

 stitution assigned. 



"When the brown salt is decomposed with dilute hydro- 

 bromic acid or an alkaline bromide, a corresponding result is 

 obtained. By treatment with hydrobromic acid a mixed 

 bromide resulted which proved to contain 66 - 06 per cent of 

 silver. 



A general consideration of all the reactions which I have 

 obtained seems to indicate that the action of sulphuric acid 

 and sodium hypophosphite on silver carbonate does not 

 lead directly to the production of the double salt which I 

 have described, but that the hemisalt is produced in excess, 

 often in large excess ; that the nitric acid oxidizes this excess, 

 being able to attack the free hemisalt, but not that portion 

 which is combined with protosalt and so rendered stable. It 

 follows that whatever has been the original relative propor- 

 tion between the two salts the nitric treatment leaves always 

 one molecule of each. If it were possible to control the 

 formation it is not improbable that a pure hemisulphate might 



