igoj\ Mills — Molecular Attraction. 



Vapor Pressures of Water 



95 



Temperatures 



From Regnault's 

 curve 



From Broch's 

 formula 



A 



100° 760.00 

 120 1489.0 

 150 3572.0 

 200 11660.0 



Corrected for latitude, 



760.00 0.00 

 1544.7 55.7 

 4244.3 672.3 

 31861.0 20201.0 



etc. and normal degrees C. 



200° 11660.0 31839.0 20179.0 



Preliminary constants before employment of method of least 



squares 



200< 



11660.0 



40547.0 



28887.0 



It surely cannot be necessary to give further illustrations 

 of the fact that the trend of no empirical curve can be trust- 

 worthy near its end point. 



In conclusion therefore: — 



3. Errors of observation occurring in the vapor pressure 

 will exert fer se little effect upon the constant of formula 2. 



"4. Errors in the vapor pressure may cause greatly multi- 

 plied error in the constant by affecting Biot's formula used 



for the calculation of the -^= r . This source of error will be 



greatly more apparent at the end points for which the Biot 

 formula was calculated. 



THE EVIDENCE FOR THE THEORY. 



The measurements considered in calculating the constants, 

 of which Table 1 gives a summary, include thirty-one sub- 

 stances, and cover a range of 290° in temperature. The fore- 

 going discussion has shown that errors of observation may 



