FLUATE OP LIME, AND THE DIAMOND. 161 



4uced, but when the same fringes were held parallel to the 

 fringe b e, the sum of their effects was produced. Hence it 

 follows, that the structure which produces the fringes a d, cf y 

 is the same as that of one class of doubly refracting crystals, 

 and the structure which produces the fringe b e, the same as 

 that of the other class. An effect exactly similar to what is 

 shewn in Fig. 5. is exhibited by the sclerotic coat in the eyes 

 of fishes, and may be produced by crushing a piece of soft 

 isinglass, or by pressing a mixture of rosin and bees-wax *, 

 between two plates of glass. 



The preceding experiments entitle us to conclude, that mu- 

 riate of soda, fluor-spar and the diamond, combine in the same 

 specimen three different structures, and form a new class of 

 doubly refracting crystals. In some parts they act upon light 

 like that class of crystals in which the deviation of the ex- 

 traordinary ray is supposed to be produced by an attractive 

 force. In other parts they act upon light like the other class 

 of crystals in which the extraordinary ray deviates from the 

 axis in virtue of a repulsive force ; and in intermediate por- 

 tions they exhibit that mean structure, in which the light is 

 urged neither by attraction nor repulsion, and where there is 

 neither a polarisation nor a division of the transmitted pen- 

 cil. If the laws which regulate the crystallisation of these 

 minerals had been allowed an undisturbed operation, it is pro- 

 bable that the crystals would have had a perfect cube or an 

 octahedron for their primitive form, and would have exhibited 

 none of the phenomena of double refraction. 



The slightest irregularity, however, in the operation of these 

 laws, would produce a deviation from the perfect primitive 

 form, and the crystal would therefore deviate from the inter- 

 mediate class into the attractive and repulsive classes, and 



Vol. VIII. P. I. X would 



* See Phil. Trans. 1815, p. 32, 33. ; and 1816, p. 172, 173. 



