ELECTROLYSIS 193 



cyanide. The electrolytic test cannot be applied to compounds 

 such as those mentioned, and in the case of carbon compounds 

 generally there are difficulties which attach to the view that the 

 constituent atoms are held together by electrons. It is only 

 necessary to inspect the formula of such a compound as one of 

 the sugars. 



H OH H HO HO HO 



V 



H 



H HO H H H 



Here are five hydroxyl groups, HO, and more than an equivalent 

 number of hydrogen atoms, but neither hydroxyl nor hydrogen 

 is capable of separating from the carbon so as to form an ion. 



Ionic dissociation in a liquid has been sometimes regarded as 

 similar in nature to the dissociation of many volatile compounds 

 when converted into vapour by heat. This, however, is a case 

 of imperfect analogy. When, for example, ordinary ammonium 

 chloride is heated it can easily be shown that the vapour contains 

 both ammonia and hydrogen chloride, for on passing through it 

 a pipe of porous clay, such as a tobacco pipe stem, and blowing 

 air through the pipe, the issuing gas is alkaline, owing to the 

 presence in it of an excess of ammonia. This is explained by the 

 escape through the pores of the clay of both ammonia and 

 hydrogen chloride, but in accordance with the ordinary law of 

 gaseous diffusion, the ammonia being the lighter gas, diffuses 

 most rapidly. The law states that gases diffuse at rates which 

 are inversely proportional to the square roots of their densities. 

 The density of ammonia (compared with hydrogen) is 8-5, while 

 that of hydrogen chloride is 18-25 ; hence their relative rates of 

 diffusion are \/18'25 and -\/8-5, so that ammonia gas passes 

 through a porous partition nearly 1J times faster than hydrogen 

 chloride. 



If an electrolyte, such as sodium chloride, is dissolved in 

 water, and a layer of pure water floated above it, it might be 

 expected that if the ions sodium and chlorine into which much 

 of it is supposed to be resolved are free it should be possible to 

 detect the chlorine in the watery upper layer in excess of the 

 sodium, while in the lower layer there would remain an excess 

 of sodium. But as explained in the previous chapter the ions 



