[wheeler] CUBICAL EXPANSION OF VITREOUS QUARTZ 141 



less readily in an oxidizing flame, 1 it appears to be impossible to anneal 

 it perfectly and thus insure complete isotropy. 



II. MEASUREMENTS OF THERMAL EXPANSION 

 PRIOR TO 1911. 



Le Chatlier, 2 in 1889, appears to have been the first to measure 

 the approximate coefficient of expansion of fused silica. Curve I, 

 (Fig. I) 3 represents his results, the dilatations being expressed in n per 

 meter. His mean coefficient between 0° and 1,000°C. was 0-7xl0- 6 . 

 Curve II shows his results for the crystalline form of quartz parallel 

 to the axis; Curve III, perpendicular to the axis; and Curve IV is for 

 a quartzose sandstone. It will be seen that in the crystalline form 

 the expansion is quite regular and much more rapid than that of 

 the fused substance up to about 570°C, where a sudden expansion 

 takes place which is followed by a steady contraction on further heat- 

 ing. It will also be noticed that Curve IV lies just above the mean 

 position of Curves II and III. 



The work of Le Chatlier, Boys, Shenstone and others had brought 

 to the notice of physicists and chemists a substance whose many 

 desirable properties and importance were quickly recognized. Many 

 workers soon set about the more exact study of these properties 

 which was necessary if fused silica were to be used to advantage in 

 the carrying out of accurate w r ork. 



The results of the numerous investigators, who studied its thermal 

 expansion during the first decade of the present century, have been 

 so well summarized in an excellent paper by Dr. G. W. C. Kaye 4 

 that it has seemed sufficient for this period merely to mention the 

 principal workers, with a brief statement of the more important 

 results 5 wmich had been obtained at that time. 



As was seen from Fig. 1, quartz in the crystalline state has a 

 fairly large coefficient of expansion, w r hile" in the amorphous state 

 its coefficient becomes very small. For example, the approximate 

 values at moderate temperatures as given by Dr. Kaye are : — 



1 Dufour, Loc. cit. 



2 Comptes Rendus 108, pp. 1046-1049, 1889. 



130, pp. 1703-1705, 1900. 



3 Fig. 1 is taken from Shenstone's paper as it appeared in Chem. News 83, 

 p. 206, 1901, and is due to Le Chatlier. 



4 Phil. Mag. (6) 20, pp. 718-728, 1910. 



5 For a more complete account of the work of this period reference should be 

 made to Dr. Kaye's paper, to which the writer begs to acknowledge his indebted- 

 ness, as well as to the original papers. 



