Vapour-pressure and the Volumes of Vapour and Liquid. 137 



§ 3. A substance specially suited for the comparison of the 

 theoretically determined with the observed vapour-tensions is 

 ether. For this we possess the series of tensions determined 

 by Kegnault*, extending from —20° to 120°. of which the 

 reliabilitv cannot be questioned, and a series of tensions by 

 Sajotschewskyf. which extends from 100° to the critical tem- 

 perature 19 C : 



Of these observed tensions I have employed three for the 

 determination of the constants occurring in equation (7), and 

 found the following numbers: — 



a = 2665: h = 076786: n=l 19233. 



Bv employing these numbers, for every value of T the corre- 

 sponding values of C can be calculated from (7); and then 

 the value belonging to it of the fraction II II c can be taken 

 from the Table. From this fraction, which can also be written 

 PT C P C T, the value of P is immediately obtained, since P c and 

 T c are known. In this way I have calculated the tensions for 

 a series of temperatures advancing by increments of 20 c each, 

 and arranged them in the following Table under the designa- 

 tion P calc. For comparison I have added, under P obs., the 

 observed values — below 100° those observed by Pegnault, 

 above 100° those observed by Sajotschewsky, and for 100° 

 and 120° the mean values taken from the data of both ob- 



t 



-20= 



0° 



20 3 



40° ; 



60° 



80° 



Pcalc. ... 



Pobs 



A 



00881 



0-0907 

 -00026 



0-2427 



02426 



4-0-0001 



0.-72 



0569 



+0003 



1-195 



1193 



4-<HXE 



2-265 



2 270 



-0-005 



3-973 



3-977 



4-0-001 





t 



100° 



120 3 



140° 



160° 



180° 



190° 



P calc. . . . 



Pobs 



A 



6-557 



6-549 



4-0-003 



10-27 

 10-23 



- o-oi 



1541 



15-42 

 - 001 



22-33 



22-34 



- 001 



31-41 



31-90 



- 0-49 



36-90 

 36-90 







The accordance between the calculated and the observed 

 tensions is seen to be almost complete. Only at 180° does a 

 difference of inadmissible magnitude occur, which, in compa- 

 rison with the rest of the differences, is very striking. This, 

 however, is doubtless occasioned by an inaccuracv of the 

 observation-value, as is most clearly to be inferred from the 



* Eegnault. Relation des Experiences, Sec. ii. p. 393. 

 t Sajotscb-ewskv, Beiblatter, iii. p. 741 (1579j. 



