86 K. D. Kraevitch on an Approximate Laiv of the 



considerable magnitude compared to that of the gas in a 

 state of saturation. The application of this theory to such 

 liquids is, however, interesting, because it enables us to cal- 

 culate their heat of vaporization, the determination of which 

 by experiment is very difficult. For this reason I allowed 

 myself to apply the theory to certain gases, although it was 

 impossible to expect reliable results. It appeared that the 

 existing experimental data on liquefied gases in a state of 

 saturation were exceedingly rough, and did not permit making 

 any even approximately exact conclusions. Besides which 

 certain gases give a constant increase of rate of decrease on 

 the value of x with a rise of temperature. Only sulphuretted 

 hydrogen (H 2 S) gave the possibility of obtaining x and y, 

 although only very approximatively. I here give the results 

 only, without the calculations. The temperature of the maxi- 

 mum was 65°. 



= 3-3987, y =2045*3, 

 c-C!=0'2000, r = 120*3. 



I calculated the vapour-pressure of sulphuretted hydrogen in a 

 state of saturation, at various temperatures, from these data. 



t. 



Pressures. 



t. 



Pressures. 



Calculated. 



Observed. 



Calculated. 



Observed. 



o 



-20 



4135-0 



4438-5 



o 

 50 



27809 



27815 







80516 



8206-3 



60 



33740 



33740 



20 



14071 



14152 



70 



40352 



40353 



40 ., . 



22563 



22582 









26. The second mode of determining the value of x and the 

 latent heat of vaporization requires an exact knowledge of 



the differential coefficient of the pressure i -r- J. This quantity 



may, as was shown above, be calculated from four different 

 formulae (11) : I only determined two values in each instance. 

 If the latter were equal, within the limits of logarithmic 

 errors, then the results of observation might be relied upon : 

 in the reverse case the data for the pressures contained errors, 

 made by the composers of the tables, in their calculations of 

 the interpolation formulae, or else due to misprints. The 



