Section III., 1914. [51] Trans. R.S.C. 



"A Determination of the Coefficient of Expansion of Mercury at Low 

 Temperatures ." 



By C. B. James, B.A., Demonstrator of Physics, McGill University. 



Presented by Dr. H. T. Barnes, F.R.S.C. 



(Read May 27, 1914.) 



In 1817, Dulong and Petit made a measurement of the absolute 

 expansion of mercury from 0°C. to 300°C. using the method of equili- 

 brating columns. This method is based on the principle that if two 

 colums of a U-tube containing a liquid are under the same conditions 

 of pressure, but at different temperatures, then the density of the 

 liquid in the columns will be inversely proportional to the heights. 

 By this method no correction is necessary for the expansion of the 

 containing vessel. One column was enclosed in an ice mixture, and 

 the other was encased in an oil bath heated by a furnace. The tem- 

 perature was obtained by an air thermometer extending almost the 

 whole length of the bath. No method of stirring was employed. The 

 mean value for the expansion from 0° to 100° as found by Dulong and 



Petit was - - = -00018018, In 1857, Regnault repeated this work 

 5,550 



with approved apparatus. As is well known he inverted the U-tube hav- 

 ing the horizontal connecting tube at the top and brought the lower 

 ends together. These columns were connected to an air drum and 

 were also kept at constant temperature. Regnault used a hand stir- 

 ring device to obtain a uniform temperature in the oil bath. The 

 air thermometer was used. With this improved method Regnault 



found the coefficient of expansion to be — — = -00018028 which does 



5,547 



not differ greatly from that given by Dulong and Petit. 



In 1910, Callender again improved the apparatus and repeated 

 the work. He used an electrical heater for the oil bath and obtained 

 the temperature by means of the platinum thermometer. Centrif- 

 ugal circulation driven by an electric motor kept the oil in circula- 

 tion so that uniformity of temperature was maintained. In place 

 of having a single pair of hot and cold columns as Regnault used, 

 six pairs were connected in series which gave about eight times the 

 expansion obtainable by Regnault. The expansion from -10°C. to 



