[wheeler] CUBICAL EXPANSION OF VITREOUS QUARTZ 145 



the coefficient at this temperature. As the maximum length of 

 crystalline quartz occurs near this temperature, it may be that the 

 more rapid expansion of the fused quartz is due to the fact that the 

 change from the crystalline form to the amorphous is not quite 

 complete. 



Recently Mr. A. Blackie 1 at the National Physical Laboratory 

 has compared the expansions of translucent and of transparent 

 fused silica. He finds that the translucent variety expands some- 

 what more rapidly up to 500°C. than the transparent. In other 

 words, this maximum near 500°C. is more pronounced in the trans- 

 lucent than in the transparent fused silica, as would perhaps be expected 

 if this maximum is due to the presence of a small amount of crystalline 

 material in the fused quartz. 



Dr. Kaye also notes that a well-marked recalescent point at 

 about 500°C. has been observed for quartz crystals, while Rosenhain 

 did not detect such a point in fused silica. 



For moderate changes of temperature the thermal hysteresis, 

 or residual variation of the expansion coefficient, is very small, al- 

 though it has been definitely established and measured for a tem- 

 perature range not exceeding 400 or 500°C. by Mr. L. F. Richardson 2 

 at the National Physical Laboratory. Following Dr. Kaye, we take 



«11. 



^— • - as the definition of the thermal hvsteresis, where 1 is the length 

 It ■ & 



of the specimen at 0°C. and ol is the increase in length at 0° produced 

 by heating to a temperature t° and maintaining it at this temperature 

 for a considerable time, and then allowing it to cool to zero. 



For annealed specimens of fused silica, Richardson obtained 

 the following values; over the range — 190 to400°C. between — lxlO- 9 

 and — 5xl0- 9 , and over the range to 800° about — 17xl0- 9 , while 

 over the range to temperatures above 1,000°C. a positive value 

 was obtained. The negative value of course indicates that the con- 

 traction during cooling is greater than the expansion during the 

 preceding heating. 



III. RELATION BETWEEN LINEAR AND CUBICAL 

 EXPANSION. 



In January, 1911, there appeared a detailed account of the 

 measurement of the absolute expansion of mercury by Callendar and 

 Moss 3 by means of an elaboration of Dulong and Petit's method 



1 Chem. News 104, pp. 77-79, 86-88, 1911, Faraday Soc. Trans. 8, part I, 1911; 

 Nat. Phys. Lab. Call. Res. 8, pp. 139-151, 1912. 



2 Phil. Mag. (6) 20, pp. 726-727, 1910. 



3 Phil. Trans. A. 211, pp. 1-32, 1911; Proc. Roy. Soc. London, A. 84, pp. 595-597. 



