xxi 



first being easily removed, the second requiring more than the boiling-point 

 of water to remove it. This atom of water may be replaced by a salt, form- 

 ing a class of double salts. Amongst the salts with basic water he put s 

 sulphuric acid as sulphate of water, and an additional atom as equal to the 

 atom of crystallization. 



In the same year he speaks of ammonia performing the function of 

 water in compounds of copper. In 1836 he says that his "researches 

 make it probable that the correspondence between water and the magnesian 

 class of oxides extends beyond their character as bases, and that in certain 

 subsalts of the magnesian class of oxides the metallic oxide replaces the 

 water of crystallization of the neutral salt and discharges a function which 

 was thought peculiar to water." The inquiry was extended to the con- 

 stitution of the phosphoric acids, and the amount of base taken up by 

 them shown to be equivalent to the amount of water in the acid ; from this 

 he passed to the arseniates. It is quite evident that he treated water as he 

 treated metallic oxides; indeed he speaks of metallic oxides performing 

 the functions or taking the place of water. It w T as a distinct recognition of 

 hydrogen as a metal in its place in salts, whilst his latest paper in 1869 

 endeavours to establish its specific gravity when combined with palladium 

 as an alloy. 



This was one of the chains of discovery which, at an earlier period, led 

 to the doctrine of substitution. It is still sound ; and although the water 

 does not now hold the same place of honour in the phosphorus acids, 

 the place is held firmly by the hydrogen, which takes its position as a 

 metal. This idea cannot be regarded as originating with Graham ; Davy 

 seems to have hinted it, and Dulong made it distinct ; but it was Graham 

 whose careful experiments and cautious reasoning gave it consistency and 

 force, although he himself did not actually adopt it in general teaching. 

 Probably nothing tended so much to give hydrogen its present place as the 

 inquiry into the constitution of the phosphates, and his explanation of the 

 monobasic, bibasic, and tribasic acids. 



We have the first results of his experiments on diffusion in the Philoso- 

 phical Magazine for 1829. 



After giving various details of experiments he says, " It is evident that 

 the diffusiveness of gases is inversely as some function of their density, 

 apparently the square root of their density." This is the conclusion he 

 arrived at finally. 



The separation of gases by simple diffusion is shown to be practicable, 

 and is there illustrated ; he mentions it as conceivable " that imperceptible 

 pores and orifices of excessive minuteness may be altogether impassable (by 

 diffusion) by gases of low diffusive power, that is, by dense gases, and 

 passable only by gases of a certain diffusive energy." Here we observe his 

 wonderful caution : he will not say that the atoms or molecules may be 

 too large, he will not say that the gas will not pass, but he says "impass- 

 able (by diffusion)." 



