[james] COEFFICIENT OF EXPANSION OF MERCURY 55 



It remains to select a suitable value for g in the case of fused 

 quartz. 



Much work has been done recently on the coefficient of absolute 

 expansion of quartz. A careful summary of this has been given by 

 N. E. Wheeler in a special communication to the Royal Society of 

 Canada. It is, therefore, unnecessary to go into a complete discussion 

 of the conclusions in the present paper. 



The most accurate values given by Wheeler may be taken as 



0-488 X 1(T 6 between 0° and 100°C. 



0-325 X 10 -6 # " -20°andO°C. 



0-307 X 10 -6 * " -30°and0°C. 



0-295 X 10 -6 " -37° and 0°C. 



for the linear coefficient of quartz. We have, therefore, assumed 

 these values in correcting for the expansion of our quartz dilatometer 

 and have taken the cubical as three times the linear experimental 

 values. 



The overflow from the dilatometer was determined by the or- 

 dinary method. The constancy of the total weight at 0°C. gives us 

 an indication of the fact that no permanent change in the volume of 

 the dilatometer resulted from successive heating and cooling. These 

 measurements are as follow: — 



Feb. 5th. 1527-16 gram. 

 Feb. 10th. 1527-11 « 

 May 7th. 1527-18 " 

 May 13th. 1527-15 " 



Preliminary experiments were made to determine the mean 

 coefficient of expansion of mercury between 0° and 100°C. The 

 dilatometer was immersed in a carefully prepared ice mixture with 

 the stem dipping into mercury held in the overflow receptacle. After 

 one hour, when a steady temperature was reached, the dilatometer 

 was suspended in a hypsometer and supported by a wire holder in 

 such a way as to swing free of the sides. Before placing in steam the 

 overflow receptacle was replaced by a carefully weighed empty vessel. 

 The time for immersion in steam was one hour. Careful readings of 

 a standard Fortin barometer were made during this time. The 

 following values were obtained: 



