SMITH. — EXPANSION OF ETHER AND ALCOHOL. 



437 



TABLE III. 

 Expansion of Ethyl Alcohol — 99.5 Per Cknt. 



Compressib ility. 



In compression the liquid is heated. One must therefore proceed 

 in one of two ways. One must either wait for the heat of compression 

 to disappear, or read the change of volume at once and correct for the 

 error due to the heat of compression. The first method depends upon 

 the operation being as nearly as possible isothermal ; the latter, on the 

 condition that the compression be very nearly adiabatic. Since it was 

 practicable to read the change of volume in less than 30 seconds after 

 the change of pressure had been made, it seemed possible to get very 

 nearly adiabatic compression. This method had the further advan- 

 tage that, by reading the change of volume rapidly, any drift of tem- 

 perature, to which we have already referred as a source of error, would 

 be much less serious than if we were obliged to wait for the heat of 

 compression to equalize. To prevent as far as possible any loss of heat 

 during compression, the brass jacket was filled with the same kind of 

 liquid which was in the piezometer. The temperature of the liquid 

 which surrounded the piezometer was therefore raised the same amount 



