100 NINETEENTH KEFORT. 



These temperatures, as given in the following table, represent the aver- 

 age of three series of readings and from them the uniaxiality curve of 

 Plate IV has been draM'n. 



Table 3. 



Uniaxiality Temperatures of Glauberite. 



The uniaxiality temperatures for six important Mave lengths as 

 developed from the curve follow : 



527 M/i (E line) 29. 8°C (i56fMfx 48.6°C 



,560 35.5 670 51.6 



589 (D line) 40.6 687 (B line) 54.1 



The curve shows an unexpected break at about dOO/j-fJ. for it was 

 thought that a straight line curve should result. This change in direc- 

 tion may possibly be due to several factors. First, as was noted in the 

 case of gypsum above, the hyperbolic brushes are very much broader 

 for light above GOOfxfx than for that below. This obviously makes 

 accurate determination of the exact temperature at which uniaxiality takes 

 place extremely difficult. Second, the observations plotted in Plate IV 

 are the average of three sets of readings made from different portions 

 of the same section. As already indicated, slight variations in the 

 character of the section could easily be recognized under the microscope 

 and these may account to some extent at least for the break in the 

 curve. A third reason may be that at a temperature of about 42° C 

 some slight change in chemical composition may take place. Of the 

 three possible explanations, the first undoubtedly is to be considered as 

 the most probable. 



Mineralogical Taboratory, University of Michigan. 



LEGENDS. 



Plate II. Change in temperature of uniaxiality in gypsum for various wave- 

 lengths. 



Plate III. Change in the appaicMl optical angle in oil of glauberite for various 

 wave-lengths and temperatures. 



Plate IN'. Change in temperature of uniaxiality in glauberite for \arious waNC- 

 lengths. 



