pC DISSOCIATION 527 



" For all vapours condensible above zero, what is called 

 the vapour-density is a mathematical fiction, to which it is 

 impossible to give a physical meaning. Its interpretation 

 is only possible when the vapour, at a point sufficiently 

 above the boiling-point, obeys the law of Mariotte, and 

 possesses the constant coefficient of expansion, 0*00367, of 

 the perfect gases" (Comptes rendus, 1867, 64, 242). 



The association of water. In the case of water-vapour, 

 the vapour-density agrees almost down to the boiling-point 

 with that calculated for the formula H 2 O. But Bose (Zeit. 

 Elektrochem., 1908, 14, 271) has recently found evidence of 

 polymerisation in the neighbourhood of 100, the increase of 

 density corresponding with the formation of about 9% 

 of H 4 O 2 . Liquid water at 100 appears to be polymerised 

 to such an extent that the average size of the molecules is 

 approximately that shown by the formula H 4 O 2 (Guye, 

 Trans, Faraday Soc., 1910, 6, 78). The formation of 

 unstable complex molecules during the condensation of a 

 vapour is usually described as ASSOCIATION. The phenome- 

 non does not differ essentially from those cases of dissocia- 

 tion (S 8 ^ 4S 2 , N 2 O 4 ^ 2NO 2 ), in which only a single 

 product is formed. The term is, however, usually applied 

 to cases in which the vapour has a normal density and the 

 polymerisation is, for the most part, confined to the liquid 

 and solid states. 



The evidence of association in liquid water is particularly 

 strong. Rontgen pointed out in 1891 that the anomalous 

 contraction between o and 4 might be explained by a 

 gradual dissociation of " ice-molecules " into denser " water- 

 molecules." Sutherland suggested that these might be 

 formulated as follows : 



" steam-molecules " H 2 O = " hydrol " 

 " water-molecules " H 4 O 2 - " dihydrol " 

 "ice-molecules" H 6 O 3 = trihydrol." 



Following the analogy of nitrogen peroxide, it may be 



