11 Mr. W. Sutherland on Weak Electrolytes and 



If we regard the water in 50 per cent, alcohol as dihydrol 

 which is changed by more alcohol into hydrol, and if we 

 take A/p 2 — 0'0S2 to measure the amount of this change, the 

 relation for Nernst's data becomes 



(K 2 — l)//a a -72 = 106(A/jp s -0-082), 



and for Thwing's 



(Kj-l)//^- 63-8 = 375 (A/p 2 -0'082). 



An observation of Abegg's is of great interest. He found 

 his law hold for a mixture of 10 volumes of ethyl alcohol 

 with 1 volume of water down to — 83°*6 C, K for the 

 mixture being given by 145e -T ' /19 °. He found, moreover, 

 that if he applied the simple mixture formula (32) to his 

 results, he got for water in alcohol below 0° C. 



K=372<r T / 190 , 



a formula which applies excellently to the experiments of 

 Heerwagen and Drude on water up to 76° C. It is very 

 remarkable that, though water is a mixture whose com- 

 position varies with temperature, it should have K given by 

 this simple formula, which confirms the conclusion that K 

 for trihydrol is nearly equal to K for dihydrol and is pro- 

 duced by the same mechanism in both substances, the Jf \} 

 bond joining to in (H 2 0) 3 and (H 2 0) 2 . In the same 

 way Abegg's discovery "that K for a mixture of alcohol and 

 water conforms to his law, shows that in alcohol and water 

 the same mechanism confers the large abnormal K. But 

 Abegg's result for the mixture brings in a seeming difficulty. 

 In his mixture p! = 0'89. But we have seen that in Nernst's 

 experiments 90 per cent, of alcohol changes (K 2 — l)//o 2 f° r 

 water considerably, and in Thwing's much more, whereas in 

 Abegg's 89 per cent, alcohol causes no change. It appears 

 then that liquid hydrol has a K which varies much with the 

 experimental conditions, being equal to that of K for water 

 under those of Abegg, but much smaller under those of 

 Thwing. Now at the Faraday Society's discussion I brought 

 out the fact that in water of crystallization hydrol has a 

 value of K about 6. As water of crystallization is hydrol 

 in the solid state, this recalls Abegg's observation that when 

 the alcohols solidify, the large value of K for the liquid 

 changes to a value nearly equal to n 2 in accord with the law 

 of Maxwell. It appears then that the whole of this subject 

 is full of interesting clues to the functions of electrons in 



