Wright and Larsen — Quartz as a Geologic Thermometer. 429 



there is a sharp decrease, while above that temperature the 

 change is very slight, the birefringence increasing gradually. 

 The birefringence for Li light appears to be slightly lower than 

 that for Na light, but otherwise its rate of change is approxi- 

 mately the same throughout the scale. — Above 1000° the light 

 from the furnace itself increases rapidly in intensity, and the 

 light from the sodium or lithium flame becomes rapidly in- 

 sufficient in relative intensity for satisfactory determinations, 

 and intense white light (electric arc) should be employed. 



In 1890 Mallard and Le Cliatelier* measured the birefrin- 

 gence of quartz at different temperatures by means of interfer- 

 ence fringes after the method of Fizeau and Foucault, and 

 observed a sudden decrease in the birefringence at about 570°. 

 For yellow light the following measurements (Table II, column 

 I) are indicated on the curves of fig. 1 of Mallard's article. 







Tab 



LE 



II 











I 







II 





Temperature 





Mallard 



Wright and Larsen 



Difference 



15° 



= 



•00917 







•00910 



•00007 



100° 



= 



•009045 







•00902 



•00003 



220° 



= 



•(•08865 







•00882 



•00005 



535° 



zrr 



•008145 







•00811 



•00004 



570° 



— 



•00804 







•00797 



•00007 



590° 



= 



•007765 







•00760 



•00016 



665° 



— 



•00777 







•00762 



•00015 



1060° 



:zz 



•00800 







•00787 



•00013 



Table 2. Comparison of the measurements of birefringence of quartz at 

 different temperatures by E. Mallard after the method of interference 

 fringes of Fizeau and Foucault with those of Wright and Larsen using 

 Babinet compensator. 



In column II, the results of the measurements at the same 

 temperatures in sodium light by the writers are included for 

 comparison ; while in column III the differences between I and 

 II are given and show the relatively close agreement between 

 the two sets of observations, Mallard's readings being slightly 

 higher, especially for the /3-quartz. In his paper Mallard 

 gives a formula representing the birefringence-temperature 

 curve below 575° and also a second formula representing the 

 curve above 575°. His equations are of the second degree and 

 parabolic in nature, and represent with a fair degree of 

 * Bull. Soc. Min., xiii, 123-129, 1890. 



withstanding the relatively large scale. Curve III represents the observed 

 birefringence data of quartz in lithium light ; curve IV, which is not so 

 exact in its definition as II and III, indicates the change in the angle of 

 rotation of a basal plate of quartz l ram thick, with rise in temperature. The 

 measurements of the angles of rotation were less satisfactory owing to thick- 

 ness of some of the plates used and also the difficulty of determining the 

 exact angle of rotation in each case. 



