262 Pirsson and Robinson — Determination of Minerals 



In a rock slide containing a number of sections of the same 

 mineral, the ellipsoid of elasticity will be cut in various direc- 

 tions and some of them must approach the maximum of bire- 

 fringence, the probability increasing with the number of the 

 sections. "While it will rarely happen that one is thus oriented 

 exactly parallel to the axial plane or optic axis, sensible diver- 

 gences may exist without perceptibly altering the birefringence 

 or vitiating the method to be presently described. In a num- 

 ber of haphazard sections of a mineral then, the one showing 

 the maximum of birefringence may be assumed to be nearly 

 parallel to the optic plane or axis thus giving approximately 

 the highest birefringence of the mineral. This determination 

 may be confirmed by the fact that such sections should not 

 yield any loci of optic axes or axial bars or bisectrices in con- 

 vergent light. The color in plain light, the pleochroism with 

 one nicol, the cleavage and crystal form often help, in addition, 

 to confirm the orientation. 



If now we consider only those sections of minerals which 

 give the maximum of birefringence, we may eliminate the 

 factor of the orientation of the section in considering the 

 strength of the interference colors and this will then depend on 

 the other two factors, the thickness of the plate and the value 

 of birefringence. There is thus a constant relation between 

 them and if we know the thickness of the plate and the value 

 of the maximum birefringence, we may readily calculate the 

 interference color ; or, given the thickness and the color, the 

 maximum birefringence may be found ; or, finally, with this 

 and the color, the thickness of the plate. 



These relations have been put into graphic form by Michel- 

 Levy in the beautiful colored plate in " Les Mineraux des 

 roches " by Michel-Levy and Lacroix. We have not found 

 them expressed in graphic form in any other handbook with 

 which we are acquainted.* The advantage of having these 

 relations shown in a diagram where they can be readily seen, 

 followed and used is very great. It is probable that the uncer- 

 tainty, difficulty and cost of the colored plate may also have 

 deterred some authors from using it. It is, however, by no 

 means necessary that the plate should be colored, the diagram 

 is the essential part and this we have thought would be of ser- 

 vice to petrographers and students in this country, many of 

 whom are unacquainted with the colored plate of Michel-Levy, 

 or are deterred from using it by its cost and inconvenient form. 

 We have put it into compact shape and have simplified the 

 color-names, by indicating the place of the distinct and well- 



* Such as Rosenbusch, Rosenbusch-Iddings, Zirkel, Luquer or Harker. The 

 last named, however, gives some very useful practical hints in this direction. 



