SMITH. — EXPANSION OF ETHER AND ALCOHOL. 



425 



nected in series with a relay of 1700 ohms and a battery of 8 volts. 

 The other terminals of the relay were connected in series with the 

 heating coils of the air-bath. These connections are shown diagram- 

 matically in Figure 2. The oscillations which occurred in the air-bath 

 were of the order of 0.2° C, and took place in about one minute. 



I thought it well to isolate a portion of the space inside of the air- 

 bath from the remainder by means of non-conducting materials, hoping 

 thus to get an inner region in which these oscillations of temperature 

 would have little effect. Since the apparatus to be kept at constant 

 temperature was about 70 cm. long, and was to lie in a horizontal posi- 



_J5^!?^^1_ 



.^^ 9, 



tion, I chose as the region to be thus isolated the space inside of a 

 cylindrical shell, 120 cm. in length, 15 cm. in internal and 25 cm. in 

 external diameter. The walls of this cylindrical shell were made of 

 concentric sheets of asbestos, the space between these sheets being 

 packed with asbestos wool. The openings in the ends were filled with 

 cotton wool. For the purposes of observation, it was necessary to have 

 two narrow windows in the shell. These windows were diametrically 

 opposite each other, and the glass plates closing them were at the inner 

 surface of the shell. The temperature of the space thus enclosed was 

 found to remain very nearly constant. 



An incandescent lamp was used as a source of illumination, and at 

 first the heat from it, passing into the thermostat, caused a slight drift 



