20 -2 



REV. S. M. JOHNSTON 



was maintained. The above curves, therefore, show that the differences in observed 

 elevation of boiling points were due for the most part to unavoidable inaccuracy in 



7 Z -3 -4 -5 



ELEVATION OF BOILING TEMP. 



FlQ. 4. 



determining the boiling point of water, and only to a slight extent to error in deter- 

 mining the boiling points of solutions. It would follow that values of the elevation 

 of the boiling point of a solution above that of water may be affected by a large 



