152 



The origiiuil volume of the licniid in the flask was restored by the addi- 

 tion of 15 c. c. of water, and the slightly more dilute mixture so obtained 

 was used for the next experiment. Forty-three mixtures were investigated 

 in this way, ranging from 91 to per cent, of alcohol. 



Corrections were introduced in the temperature readings for the baro- 

 metric pressure and for the exposed column of mercury in the thermometer, 

 assuming that the barometer effect would be the same as in the case of water 

 — an assumption very nearly in accordance with the facts, as shown by the 

 tables of Regnault and Classen given in Biedermarm's Chemiker-Kalender. 



The temperature results are probably accurate within 0.2 degrees, and 

 the concentrations within 2 per cent. 



The results obtained are given in the foUowing table: 



