DOUBLE REFRACTION IN ISOMORPHOUS SUBSTANCES. 265 



observe in this manner the centre of the rings, it is advisable to 

 place the plane of polarization in the first instance parallel to the 

 plane of the optical axes, then at an angle of 45° to this plane. 



If the apparent inclination of the optical axes to the normal 

 of the plane of emergence is once measured, it is easy to calcu- 

 late their true interior inclination, provided that the correspond- 

 ing index of refraction, that is to say the index of a ray situated 

 and polarized in the plane of one of the optical axes, has been 

 previously determined. 



If the true inclination of the optical axes to the normals of 

 one or two faces of known position is known, the direction of 

 their bisecting line in relation to these normals follows from it, 

 and the position of the axes of optical elasticity in the interior 

 of the crystal is thus completely determined. 



This second method requires therefore a tolerably large num- 

 ber of measurements ; each of these involves an error ; and con- 

 sequently the former method, which is moreover more general, 

 appears to me susceptible of a higher degree of accuracy. 



In experimenting with biaxial crystals, I have always deter- 

 mined the optical character of the bisecting line in artificial 

 plates, either very nearly normal to that line or parallel to it, and 

 at the same time normal or parallel to the plane of the optical axes.- 



In the former case the plate includes within its plane. the axis 

 of mean optical elasticity B, and one of the axes A or C. If 

 the maximum elasticity is found in the direction of B, the axis 

 of minimum elasticity can be no other than C. The bisecting 

 line will therefore coincide with the axis of maximum elasticity A. 

 When, on the contrary, the minimum elasticity is found in the 

 direction of B, the axis of maximum elasticity can be no other 

 than A, and the bisecting line will coincide with the axis of 

 minimum elasticity C. 



This bisecting line is included within the plane of the two other 

 kinds of plates ; its character may therefore be determined at once. 



In order to determine with these plates the character of the 

 axes of optical elasticity which they contain, I placed them upon 

 laminae of quartz, which were likewise parallel to their axis, and 

 brought them in the way already described in a pencil of parallel 

 polarized light. The thickness which is necessary to enable 

 them to develope colours by compensation may always be found, 



SCIEN. UEM.^Nat. Phil. Vol. I. Part III. U 



