SMITH. — EXPANSION OF ETHER AND ALCOHOL. 443 



is supposed to be measured at the temperature of the liquid under 

 investigation. In the present method, however, the change of volume 

 due to a change of temperature or pressure was practically transl'erred 

 either from the bulb A B (Figure 8), where it had a temperature greater 

 than 0° C, to the tube E F (Figure 8), where its temperature was about 

 0° C, or from the tube E F back to the bulb A B. In either case it is 

 obvious that the observed change of volume is less than the true 

 change of volume. Taking an apj)roximate value of the coefficient of 

 expansion of the liquid, one at once finds what value the observed 

 change of volume would have if it had been measured at the tempera- 

 ture of the liquid in the bulb instead of at o° C. The magnitude of 

 this correction varied with the temperature of the liquid in the bulb, 

 but was never greater than seven per cent of the total change of 

 volume. 



Expansion. 



If observations were to be made on the coefficient of expansion, the 

 dilatometer was filled in the manner already described on page 429 and 

 then placed in position in the water-bath, the temperature of which 

 was raised to the desired value and allowed to become constant. 

 Enouiih of the liquid was removed from the tube E F (Figure 8) to 

 allow the free surface to stand about 2 cm. above the point E. The 

 end of the tube E F was connected to the pressure gauge described on 

 page 429 and the external pressure on the liquid was made slightly 

 less than the vapor-pressure corresponding to the temperature of the 

 li(|nid in the bulb of the dilatometer. This temperature was deter- 

 mined by means of a well-tested Baudin thermometer which passed 

 down through the cover of the water-bath and had its bulb near the 

 middle of the cylindrical bulb of the dilatometer. The pressure re- 

 maining constant, observations were then made on the change of 

 volume corresponding to a definite change of temperature. Before 

 observing the change of volume a wait of about one hour was allowed 

 for the temperature to attain equilibrium. After observations on the 

 change of volume and on the change of temperature had been made, 

 the external pressure on the liquid was increased, until the liquid was 

 no longer superheated. ^° The liquid was then allowed to cool dowji to 



" This method of procedure was adopted after the first four attempts to 

 obtain values of the coefficient of expansion of ether in tlie superheated state. 

 In each of these attempts the following peculiarity was noticed. 



I had the ether at a temperature of 35° C. and under a pressure of 75 cm. of 

 mercury. Leaving the pressure unchanged, I increased the temperature up to 

 45° C, and when the temperature had become constant, noted the change of 

 temperature and the corresponding change of volume. Without changing the 



