SMITH. — EXPANSION OF ETHER AND ALCOHOL. 



439 



accurately. Moreover, the liquid meniscus did not preserve a perfectly 

 symmetrical form, and this caused some error. By taking the mean of 

 twenty observations, — ten made by increasing and ten by decreasing 

 the pressure, — the probable error was reduced to one per cent or less. 



Tables IV and V contain the results of these observations on ether and 

 on alcohol. In these tables, /• is the compressibility when the pressure 

 is expressed in atmospheres, and F is the vapor-pressure, in centimeters 

 of mercury, of the liquid at the corresponding temperature t. 



TABLE IV. 



COMPRKSSIBILITY OF EtHTL EtHER. 



Part II : Liquids under Pressures less than their 

 Vapor-pressures. 



In order to have a liquid exist under a pressure less than that corre- 

 sponding to the vapor-pressure at the respective temperature, the 

 gaseous and liquid phases must, of course, not be in contact. In the 

 dilatometer used in the work on the coefficient of expansion described 

 in Part I of this paper, the liquid under investigation was in contact 

 with mercury in such a way that its gaseous phase was not allowed to 

 form. The boundary condition for superheating being thus fulfilled, 



