﻿L 998 ] 



GIL Anisotropic Expansion of a Drawn Tube of Vitreous 

 Silica. By H. L. Callendar, M.A., LL.I)., F.R.S* 



IT has been customary to assume that the expansion of an 

 amorphous material should be isotropic, and the cubical 

 coefficient has generally been taken as three times the linear. 

 The discrepancy between the value of the cubical coefficient 

 of silica from 0° to 100° C. f namely I'OOxlO" 6 (deduced 

 from the weight thermometer observations of Harlow and 

 Eumorfopoulos by assuming the values of Callendar and 

 Moss for the absolute expansion of mercury), and the values 

 of the linear coefficient [-500x10"° (Chappuis). -465xl0~ 6 

 vScheel), -424 x 10~ 6 (Eandall)], has not unnaturally been 

 regarded by Scheel and Heuse (Phil. Mag. March 1912, 

 p. 412), and also by Eumorfopoulos (Phil. Mag. April, 1912, 

 p. 653) as conclusive evidence of an error in the determi- 

 nations of Callendar and Moss. The magnitude of the error 

 in the cubical coefficient would vary from '50xl0 -6 to 

 '21 x 10 -(i , according to the value selected for the linear 

 coefficient. 



Considering the great care taken in the testing and verifi- 

 cation of the apparatus for the absolute expansion of mercury, 

 and the close agreement of the observations taken before and 

 after the whole apparatus had been dismounted and set up 

 afresh with new tubes, it seemed to me scarcely possible that 

 an error of this order of magnitude should exist in the results. 

 In the April number of the Philosophical Magazine, p. 682, 

 I, therefore, expressed the opinion, " that the explanation must 

 be sought in the inequality of axial and radial expansion of 

 a drawn tube due to intrinsic strain." Jn order to settle the 

 question, if possible, by a direct appeal to experiment, I 

 obtained from the Silica Syndicate of Hatton Garden, who 

 had supplied the bulbs employed by Harlow and Eumorfo- 

 poulos, a drawn silica tube, 2 cm, in diameter, as nearly as 

 possible similar to those from which the bulbs had been made. 

 A tripod and a ring were cut from this tube by Messrs. 

 Hilger, and were arranged as shown in the annexed figure 

 for observing the difference of the axial and radial expansion 

 by the Fizeau method. The design was exhibited and de- 

 scribed at a meeting of the Physical Society on April 28th, 

 but I was unable to give the results of the test at the time, 

 as the worked specimens did not arrive until a day or two 

 after the meeting. 



With the assistance of Mr. A, Eagle, I have since suc- 

 ceeded in obtaining several concordant determinations of the 

 difference between the axial and radial coefficients of ex- 

 pansion of this particular tube. The exact reduction of the 

 * Communicated by the Author. 



