Section III., 1915 [147] Trans. R.S.C. 



Effect of Strain on the Coefficient of Expansion of Quartz. 

 By Professor H. T. Barnes, F. R.S.C. 



(Read May Meeting, 1915.) 



In a paper which I had the honour to present for Mr. C. B. 

 James last year before Section III of the Royal Society, on the "Co- 

 efficient of Expansion of Mercury," it was shown that a large dis- 

 crepancy exists between the coefficient as measured by the dilato- 

 meter and by the method of balancing columns. 



At present no satisfactory explanation of this error has been 

 found. Many suggestions have been made as to the possible cause 

 and one of these which occurred to me was the efifect of strain on the 

 expansion of the quartz dilatometer. The difference between the 

 coefficient measured by the two methods disappears, if we calculate 

 the coefficient by the dilatometer method, neglecting the coefficient 

 of expansion of the quartz. 



No data are available on the effect of strain on the coefficient of 

 expansion of materials, except what we know in regard to the effect 

 of strain on certain colloids, such as india-rubber and gelatine. Here 

 we find that the coefficient is changed from a positive to a negative 

 value, when the colloid is stretched slightly. 



T4ie exact composition of fused silica is not definitely known. 

 It is probably part crystalline and partly colloidal in structure. If 

 colloidal, it is conceivable that the strain under which a dilatometer 

 bulb is placed when filled with so heavy a substance as mercury 

 might have some influence on the coefficient of expansion. The 

 expansion of quartz is small and somewhat irregular at different 

 temperaturesS and it is not altogether clear or proven what effect 

 strain would have on the coefficient. 



Experimental. 



To test this point, a quartz tube was taken, approximately 80 

 cms. long, 5-4 mms. outside diameter and 3 mm. inside diameter. 

 This rod was soldered into brass rings at each end, and placed in a 

 steam jacket S, Fig. 1. One end of the tube, R, was fastened to a 

 rigid support, and the other end connected with a fine steel wire passing 

 over a pully P, and carrying a weight W. The expansion of the tube 

 was observed by a telescope and scale reflected from a mirror M, which 



iCompare N. E. Wheeler, Trans. Roy. Soc, Can. 8, 139, 1914. 



