62 K. D. Kraevitch on an Approximate Laio of the 



On placing the known values of y and T in this formula, 

 we shall have 



y(f-70j 



p =p e(353)2[l+(*-70)/353] ? 



or 



f-70 



p = acxl+™{t-w), (5 2 ) 



where log a = 2*36793, 



log a = 0-018849, 

 log m = 3-46471. 



In Regnault's formula* these figures have another value, 

 because he gives his formula a different aspect, somewhat 

 diverging from theory : 



f+20 

 p =zaOLl + ™(t +20) ? 



and he endeavours to assort such values for a, a, and m as 

 would best satisfy the results of observation. 



Equation (5 2 ) contains no arbitrary coefficients ; it passes 

 through three points common to curve (5), whose abscissae 

 are 338°, 343°, and 348°. The pressures calculated from 

 this formula are given in the 4th column of the preceding 

 table. Although they agree less with the results of experi- 

 ment than the figures in the 3rd column, nevertheless it is clear 

 Roche's formula roughly and approximately expresses a law 

 of nature. 



Ether, C 4 H 10 O. 



10. We will now apply our theory to various liquids and 

 commence with ether. 



1st Method. 



It is seen from the table given below, where the first line 

 represents the temperatures {t) and the second the values of 

 x calculated after formula (8), that the maximum value of x 

 corresponds to 60°. 



t 



10° 



35° 



55° 



60° 



65° 



85° 



100° 





. 0-2821 



1-8727 



2-6274 



2-77986 



2-7258 



2-0899 



0-20548 



Above and below this temperature x becomes less, and 

 below 10° it is even negative. 

 For ether, 



AD— — 



A1J -p-74- 



* Memoires de VAcatemk des Sciences, t. xxi. p. 617 » 



