Latent Heat of Vaporization of Liquids. 273 



It is evident that a comparison between observed* and 

 calculated values should only be carried out at the same 

 temperature. Owing to lack of data, however, it has been 

 necessary in several cases simply to quote such results as are 

 available. When the difference between the temperature in 

 the two cases is great (e. g. xylene, cymene, acetic acid, &c), 

 one cannot expect the results to be very comparable, for latent 

 heat has in general a fairly high (negative) temperature 

 coefficient. A more important point still, however, lias 

 reference to the compressibility data. An accurate determi- 

 nation of the compressibility of a liquid involves very great 

 experimental difficulty, and it is therefore to be expected 

 that the results for the substances quoted may in some cases 

 be subject to rather large error (cf. Quincke, Annalen tier 

 Physik, xliv. p. 776 (1891)). Further, since for all the 

 substances mentioned with the exception of mercury, the 

 compressibility alters, i. e. decreases, as the pressure increases, 

 to obtain comparable results the values should be extrapolated 

 to zero pressure. This has only been done in a few cases 

 (cf. Ritzel, Zeitsch. physik. Chem. lx. p. 319 (1907)). 



The determination of the latent heat of vaporization itself 

 may also involve considerable error f . With these reser- 

 vations one is justified in considering that the relation holds 

 for liquids which are not associated either in the liquid or 

 vapour state. For these same substances one obtains the 

 " normal v value for the Trouton constant, On the other 

 hand, in the case of those liquids such as the alcohols which 

 are known to be considerably associated in the liquid state 

 into di- and tri-molecules, but are practically non-associated 

 in the state of vapour, it will be observed that the calculated 

 latent heat is only about one half of the observed. For these 

 substances the Trouton constant is abnormally high. In the 

 case of water, owing to the existence of the density maximum 



* In several ca?es the " observed " value is really calculated from the 

 Clapeyrou equation. 



t In this connexion Nernst (' Applications of Thermodynamics to 

 Chemistry,' p. 103) employs the relation 



R S,Hh 



p -, 



in which p x and ^denote the vapour pressures corresponding to T l and 

 T 2 , two^ temperatures which differ by so small an amount that their 

 geometric and arithmetic means may for practical purposes be set equal 

 to one another. This mean temperature is the one to which X refers/' 

 .... u in general the heats of vaporization calculated with its aid are 

 more accurate than those determined calorimetrically." 



Phil Mag. S. 6. Yol. 22. No. 128. Avg. 19.11. T 



