148 



THE ROYAL SOCIETY OF CANADA 



form of a transparent silica tube 1 a meter long, by the comparator 

 method. The fact that his results agree very closely with the Values 

 obtained by the interference method, removes any serious doubt as to 

 the approximate value of the linear coefficient for silica. 



About a year after the appearance of Callendar and Moss's 

 paper, on the absolute expansion of mercury, Scheel and Heuse 2 

 published a criticism of their results. Table I., taken from their paper, 



Table I. 



shows the variation of the different results for the expansion of mercury 

 from the mean of the values obtained by Chappuis and by Thiesan, 

 Scheel and Sell. 



Callendar and Moss had raised the objection against Chappuis' 

 values, that it is not allowable to calculate the cubical expansion 

 from the linear. Scheel and Heuse urged in favor of Chappuis' 

 result the fact that not only had he measured the axial expansion of 

 the actual verre dur dilatometer bulb used (which was one meter in 

 length and 4 cms. in diameter), but that he had used the same verre 

 dur bulb and also one of platinum-iridium to measure the expansion 

 of water, with practically identical results. As Scheel and Heuse said, 

 it was not likely that the two substances would be to the same extent 

 aeolotropic. In conclusion they referred to Harlow's determination 

 of the cubical expansion of mercury by the weight thermometer 

 method, and urged that Harlow's observations were in accord with 

 Chappuis' value, and did "not force the assumption of the aeolot- 

 ropy of fused silica, but on the other hand" indicated "the presence 

 of error in the measurements of Callendar and Moss." 



'Proc. Roy. Soc. A. 85, pp. 430-447, 1911; Nat. Phys. Lab. Coll. Res. 8. pp. 

 87-105, 1912. 



2 Phil. Mag. (6) 23, pp.412-417, 1912; Verh. d. D. Phys. Ges. 14, pp. 139-144, 

 1012. 



