82 SIR WILLIAM RAMSAY 



coexistent) could be ascertained from the theoretical isotherm 

 (continuous passage from liquid to vapour), and we found that, 

 near the critical temperature at any rate, the theoretical values 

 for ether agreed remarkably well with those observed. 



Finally, a comparison of the isochors for nitrogen peroxide 

 and acetic acid with those for ether indicated very clearly 

 the changes N 2 4 ^ 2N0 2 and (C 2 H 4 2 ) 2 ^ 2C 2 H 4 2 . 



The difference in behaviour between acetic acid and the other 

 substances was also clearly shown by the curves representing : 



(a) The relation between the densities of saturated vapour and 

 the temperature (or pressure) ; 



(b) The relation between the heats of vaporisation and the 

 temperature. 



As regards the first relation, the density of the saturated vapour 

 of ether (taking the density of hydrogen as unity) diminishes 

 with fall of temperature until it becomes normal or very nearly 

 so. This is true also for the alcohols, but the density of the 

 saturated vapour of acetic acid reaches a minimum between 140 

 and 150, increasing again at lower temperatures ; it was 

 approaching the double value at the lowest temperature at 

 which observations were made. 



The heats of vaporisation were calculated from the formula 



T dm T 



- =~r -j- The values for ether, methyl alcohol and propyl 



alcohol were found to increase with fall of temperature and 

 showed no sign of becoming constant. With acetic acid a maxi- 

 mum was reached at about 110, the heats of vaporisation 

 falling steadily at lower temperatures. For ethyl alcohol no 

 maximum was observed, although from 20 to the values 

 appear to remain practically constant. 



The behaviour of acetic acid was explained by us by the fact 

 that there is molecular association both in the liquid state and 

 in that of vapour : the rapid increase in the density of saturated 

 vapour, due to association, at low temperatures accounts for the 



