274 Dr. W. C. M«C. Lewis on the 



at 4° C. (under ordinary pressure), the coefficient of expan- 

 sion is negative from 0° 0. to 4° C. and thereafter positive, 

 thus causing a corresponding change of sign in the cal- 

 culated value of the latent heat. In the special case of 

 acetic acid, the calculated latent heat comes out approxi- 

 mately double the observed. The Trouton constant is abnor- 

 mally low. This behaviour is due to the fact that not only is 

 the liquid associated but the vapour is also polymerized con- 

 siderably and apparently to a higher degree of polymerization. 

 Ramsay and Young *, employing the Clapeyron equation, 

 found that the latent heat of vaporization of this substance 

 rose with rising temperature, reaching a maximum at about 

 110° C. (L= 92*79 cals.), thereafter steadily decreasing. 



In the case of associated liquids it is generally assumed 

 that the observed latent heat of vaporization consists of two 

 parts — one part being the heat required to isothermally 

 cause the breaking down of the complex molecules into the 

 simple form, and the second the heat required for the 

 molecular spatial distribution change which is involved in 

 vaporization. Since the "calculated" values of this paper 

 do not take any direct account of heat required to cause 

 association to disappear, one would expect the calculated 

 values to be less than the observed. Such is the case ; and if 

 we assume that the discrepancies observed may be given a 

 quantitative significance, it would appear that it requires 

 approximately the same amount of heat to cause the depoly- 

 merization of the molecules as it does to vaporize the resulting 

 simple molecules. To account for the behaviour of acetic 

 acid we must assume that the " latent heat of depolymerization 



* Journ. Chem. Soc. xlix. p. 790(1886). 



In the case of methyl alcohol — whose factor of association in the liquid 

 state at 20° C. is as large as 2*32,— Ramsay and Young- found that at this 

 temperature the density of the saturated vapour was 16*23, indicating a 

 molecular weight of 32*46. The theoretical molecular weight is 32. On 

 the other hand, in the case of acetic acid, — whose factor of association in 

 the liquid state at 20° C. is 2*13, — Ramsay and Young- found that the 

 vapour density was 59*3, indicating a molecular weight of 118 6 while 

 the theoretical is 60. This is sufficient to indicate the extraordinary 

 behaviour of the vapour, though the figures appear to indicate that the 

 degree of association in the vapour is not quite so large as in the liquid. 

 In this connexion, however, attention should be drawn to the observation 

 made by Ramsay and Young, viz. " Condensation takes place before the 

 pressure becomes nearly constant ; when a considerable amount of liquid 

 is present, however, pressure remains nearly constant with decrease of 

 volume." The acid acts in fact as though it consisted of more than one 

 substance. Further, Ramsay and Young remark that in consequence of 

 this abnormal behaviour " (he numbers [for the density of the vapour] 

 must be regarded as minimum values, although above 40° it is probable 

 they are nearly correct." 



