Wright and Larsen — Quartz as a Geologic Thermometer. 431 



accuracy the general course of the curves. Equations of 

 similar nature have also been written down by the writers and 

 the agreement of the calculated values with the observed data 

 is fairly close, but in no case could more than approximate 

 agreement over the entire range of the curve below 575° be 

 obtained. Many equations were solved with a view to adjust- 

 ment throughout the course of the curve, but every formula 

 adopted proved more or less unsatisfactory at certain points 

 along the curve. It can be stated, however, that the decrease 

 in birefringence of quartz from to 575° can be represented 

 approximately by a parabolic curve, while above 575° the 

 birefringence increases gradually and almost as a linear func- 

 tion of the temperature. A more detailed discussion of the 

 different formulas obtained and comparison with Mallard's 

 formula is interesting but of little direct value in the present 

 instance, and may therefore be omitted. 



The angle of circular polarization for different temperatures 

 was also measured by the use of the thermal microscope. The 

 results, however, are less concordant owing to the difficulty of 

 determining accurately the exact position of total extinction, 

 and also the difficulty of passing light waves through the plate 

 exactly parallel with the principal axis. Basal plates of quartz 

 of different thickness were cut and polished and the measure- 

 ments carried on in sodium light. 



To increase the sensitiveness of the method, the bi-quartz 

 wedge-plate* was used. The results of these measurements 

 are listed in Table II and expressed graphically by curve IV of 

 tig. l.f It will be noted that although the general agreement 

 of single determinations is here less satisfactory than in the 

 data on birefringence, the average value expressed by the 

 curve IV indicates a rise of the angle with increasing tempera- 

 ture up to 575°, where an abrupt increase is observed, while 

 above 575° the increase is slight. 



The determinations by Joubert and Le Chatelier of the 

 angle of circular polarization in sodium light for different tem- 



* This Journal, xxvi, 349, 1908. 



f The observed values of Table II and fig. 1 have been recalculated for a 

 plate l mm thick. The increase in thickness due to expansion at different 

 temperatures has also been taken into account, although this latter factor is 

 of minor influence and practically negligible. 



Table 3 contains the results of the measurements of the angle of rotation 

 of different basal plates of quartz, calculated to the uniform thickness of l mm 

 at different temperatures in sodium light. The noticeable lack of agreement 

 between individual readings on plates at the same temperature is due in 

 part to the difficulty of locating the position of total extinction accurately, 

 and in part to the different size plates used where iu large plates it is hardly 

 possibly to preserve uniform temperature throughout the plate in such a 

 small furnace. The difficulty of passing the light through the plate precisely 

 parallel with the principal axis is still a third difficulty which it is not easy 

 to eliminate under the conditions prescribed. 



