OF ARTS AND SCIENCES. 9 



same reaction. In one experiment the reaction was still preceptible in 

 the fourteenth wash-water. lint under the action of the boiling water, 

 the precipitate becomes crystalline or granular and the action lessens, 

 until at last the water does not dissolve sufhcient chloride of silver to 

 cause even a cloudiness on the addition of nitrate of silver, as just 

 described. Mr. G. M. Hyams, a student in this laboratory, washed 

 two different portions of chloride of silver with boiling water until the 

 action ceased, and then weighed and examined the residue. In the 

 first experiment 1.4561 grammes of chloride of silver were washed with 

 66 litres of water. The chloride of silver was then collected, and 

 found to weigh 1.2320 grammes. Hence, 0.2241 gramme, correspond- 

 ing to 15.39 per cent, had passed into solution. In the second experi- 

 ment 60 litres of water were used, and 16.03 per cent of the cliloride 

 of silver originally precipitated were dissolved. These numbers, how- 

 ever, are only approximately accurate ; for, as the precipitate becomes 

 granular, it settles with less readiness, and there was necessarily some 

 loss in filtering off so large a volume of liquid. 



In the experiments above described the boiling water produced only 

 a very slight decomposition of the chloride of silver. The precipitate, 

 granulated by the washing, readily dissolved in aqua ammonia, leaving 

 less than a milligramme of a black powder, which was proved to be 

 metallic silver. 



The solvent power of water on freshly precipitated chloride of sil- 

 •ver did not appear to be influenced by the presence of free nitric acid, 

 even in large quantities. We tried the effect both of dropping the 

 nitric acid on the precipitate before pouring on hot water, and also of 

 previously adding nitric acid to the boiling wash-water. We used 

 amounts of nitric acid {8= 1.355) varying from five to two hundred 

 cubic centimetres to the litre of water, but without finding any marked 

 difference in the result. 



The presence of a small amount of nitrate of silver in the water 

 entirely prevented its solvent action, so far as we could discover. In 

 order to determine the limit of the action, we added different quantities 

 of nitrate of silver to the boiling water before pouring it on to the 

 precipitated chloride of silver. With one centigramme of nitrate of 

 silver to the litre of water, there was a marked turbidness on subse- 

 quently adding an excess of the same reagent to the filtrate. With 

 two, three, or even four centigrammes to the litre, an opalescence could 

 still be distinguished, although constantly diminishing with the in- 

 creasing amount of the salt. With five centigrammes, there was 

 no opalescence, and we concluded that one decigramme of nitrate of 



