results showed certain irregularities that do not usually occur in 

 cases of reciprocal decomposition, where the mass of one of the 

 compounds is successively increased. 



A prismatic examination of the rays absorbed by these salts in 

 different states of solution revealed two very suggestive facts. The 

 one is, that hi every case (except ferric acetate) the salt in dilute 

 solution not only transmits every ray that was transmitted by it in 

 saturated solution, but also some rays which it then absorbed. The 

 other is, that strong solutions of the chlorides, bromides, and iodides 

 of copper, cobalt, nickel, and iron analogous metals exhibit not 

 only the absorption due to the respective bases, but another absorp- 

 tion which can be identified with that produced by the halogens 

 themselves when simply dissolved in water ; while, when these solu- 

 tions are diluted, they cease to produce this second absorption, and 

 give precisely the same prismatic image as any compound of the 

 same base with a colourless acid. The amount of water required to 

 effect this change depends on the temperature. That the phenomena 

 indicate some difference of arrangement among the elements of the 

 dissolved salt and the water, cannot, I think, be doubted, but they 

 fail to show in any distinct manner what that difference is. 



The action of water on double salts is a still more complicated 

 problem ; but the question as to whether water separates the two 

 components did not prove so difficult of decision. While on the one 

 hand the physical properties of many double salts, as for instance 

 the potassio-chloride or iodide of platinum, prove that they are not 

 decomposed by water, the experiments of Graham, on the other hand, 

 show that some salts, as for instance alum, suffer at least a partial 

 decomposition in diffusion. 



The iodide of mercury and potassium, and the sulphocyanide of 

 silver and potassium, dissolve in a small quantity of water, but the 

 addition of more causes the separation of the insoluble component. 

 The double sulphates of copper, nickel, or chromium with potash, 

 the sulphate of copper and ammonia, the chloride of platinum and 

 potassium, the iodides of platinum or gold with potassium, and the 

 hydrochlorate of chloride of gold, do not change in colour on the 

 dilution of their aqueous solutions ; but this does not prove that no 

 separation has taken place, for the colour of these double salts in 

 solution is precisely that of an equivalent amount of that component 



