2?0 APPENDIX. 



dined to each other, and therefore refract the light in 

 the manner of a prism, the refractive indices of the rays 

 may differ so much, that while one of them passes freely 

 through such a prism, the other cannot pass at all, but 

 suffers total internal reflexion, and is thereby dispersed ; 

 just as if the prism had a larger refracting angle with 

 respect to that ray than to the other. Therefore if two 

 oppositely polarized rays are presented successively to 

 such a crystal, as in our experiment, one of them will be 

 transmitted, and the other not. That this is the true 

 explanation appears from this, that when the oblique 

 planes are well formed and clearly defined by the micro- 

 scope, the colour also is accurately limited by the same 

 boundary : so that while this part analyzes the tints of 

 a plate of sulphate of lime, the rest of the crystal is 

 inactive. 



It may be inferred by analogy, that the same cause 

 produces the analyzing power of striated or fibrous sur- 

 faces, and of those in which the striae are too minute to 

 be discernible (as in No. 1, supra, page 225) : for it is not 

 the property of all crystals with striated surfaces to have 

 the analytic power, but only of such as are doubly 

 refractive in a high degree, 



I have said that the capillary crystals (No. 3) possess 

 the analytic property, although their diameter is often 

 evanescent even with a microscope. An important infer- 

 ence may be drawn from this, viz. that a ray of light imme- 

 diately on entering one of these crystals, subdivides itself 



