458 



PROCEEDINGS OF THE AMERICAN ACADEMY. 



considerable care plotted some of the high isothermals of alcohol. In 

 obtaining from these curves values of the coefficient of compressibility 

 and of the coefficient of expansion considerable uncertainty was intro- 

 duced, but at the three temperatures given in the following table it 

 seemed possible to find a value of each of these coefficients having 

 sufficient accuracy to be useful in this paper. The three values of a 

 calculated from the coefficients thus obtained are given in Table XVII. 



TABLE XVII. 

 Alcohol. 



It would appear from Tables XVI and XVII, as they here stand, 

 that the two quantities a and a' meet and cross in magnitude in the 

 neighborhood of 215° C. This is unlikely. It is more probable that 

 the data for the calculation of a and a' in the neighborhood of the 

 critical temperature and pressure are somewhat unreliable. 



Summary. 



1. The coefficient of compressibility, k, and the coefficient of expan- 

 sion, e, of ether and of alcohol have been measured at various tempera- 

 tures at pressures near the saturation pressures of the respective vapors. 

 In some cases the liquids have been examined in the superheated con- 

 dition. Within the error of my observations these coefficients have 

 been found to be constant over the range of pressure employed in this 

 paper. By reference to Figure 1 the meaning of this result may be more 

 clearly brought out. It states that the slope of any one of the isother- 

 mals examined in this paper, — e. g. the isothermal A F H J (Figure 1) 

 — has the same value at C, where the external pressure is somewhat 

 greater than the vapor-pressure, that it has at E, where the external 

 pressure is less than the vapor-pressure ; i. e., the direction of the 

 isothermal does not change when the saturation curve is crossed. It 

 has been recently stated by Amagat ^'^ that nothing is known of the 



" Comp. Eend., May 21, 190G. 



