58 K. D. Kraevitch on an Approximate Law of the 



the temperatures 55°, 60°, 65°, and 70°. I find 6-9207. The 

 arithmetical mean of these two results is 6*9227 ; the last 

 two figures may contain an error. I take the differential 



coefficient -^-for 60° as equal to 6*923, where an error maybe 



suspected in the last figure ; the result thus obtained is placed 

 in the third column opposite to 60°. In exactly the same way 



I calculated the differential coefficient^- for 65°, regarding it 



as — - 3 and as ~- 2 ' in the first instance I obtained 8*4044, and 

 dt dt ' 



in the second 8*4002*; the mean 8*402 is placed in the table 

 opposite 65°. The other differential coefficients for tempera- 

 tures 70°, 75°, and 80° were determined in the same manner ; 

 each of them is the arithmetical mean of two figures. 



By placing the resultant values of the differential coefficient 

 in equation (10), I calculated the heat of vaporization for 

 five temperatures t ; these results are placed in the second line. 



t 60° 65° 70° 75° 80° 



r 57291 570-04 567*13 564-18 561-36 



2-87 2-91 2-95 2-82 



o-c 1 0-574 0-582 0-590 0-564 



The figure 567*36 was found for the heat of vaporization at 

 70° according to the first method ; it may be considered as 

 coinciding with that now found by another method, 567*13. 

 From the tables it is seen that the heat of vaporization falls 

 with a rise of temperature. The third line gives the difference 

 between the two adjacent values of r. These differences are 

 near to each other, but are not equal, as would be expected ; 

 but on following out the calculation it is easily seen that the 

 second decimals in the heat of vaporization, and hence also 

 those figures in the differences, may contain errors depending 

 upon errors in the pressures. The fourth line of the table con- 

 tains c — Ci ; these values are obtained by dividing the differences 

 (figures in the third line) by 5. They and their arithmetical 

 mean, 0*578, may be said, taking into consideration the pos- 

 sible errors, not to differ from that found above by more 

 than 0*58831. 



When the coefficients of *• and x are known, the pressures 



* The fig-ures for one and the same coefficient differ constantly by the 

 two last figures (3rd and 4th). This is due to the inaccuracy of the 

 figures of the pressures ; hence the labour expended by Broch in their 

 calculation may be counted as fruitless. 



