DOUBLE REFRACTION IN ISOMORPHOUS SUBSTANCES. 2?? 



The optical properties are therefore perfectly identical in 

 laevo- and dextro- tartrate of soda and ammonia on the one hand, 

 and in laevo- and dextro-tartrate of soda and potash on the other ; 

 of this I satisfied myself by repeated trials. On the contrary, 

 the ammonia salts differ considerably from those of potash. 



In the laevo-tartrate and dextro-tartrate of soda and ammonia, 

 the bisecting line is indeed parallel to the height of the rhombic 

 prism of 100° 30^; but the plane of the optical axes is parallel 

 to the major diagonal of the rhombic base, and consequently 

 perpendicular to the diagonal plane in which the optical axes of 

 the potash salt would be situated. The optical axes of the dif- 

 ferent colours present a dispersion which is at least equal to that 

 observed by Sir J. Herschel ; the general angular divergence is 

 less considerable than in the previous salt, and the isochromatic 

 curves may easily be seen through a plate normal to the bisect- 

 ing line. 



I have determined the mean index of a ray polarized and situ- 

 ated in the plane of the optical axes, and the apparent angle of 

 these axes corresponding to the tolerably homogeneous red light 

 which glass of that colour transmits, and to the violet trans- 

 mitted by a concentrated ammoniacal solution of carbonate of 

 copper. These latter observations are not however susceptible 

 of any great accuracy. 



so that there could he no mistake as to the direction in which the determina- 

 tions were to be made. 



With the first plate I determined the position {trace) of the plane of the 

 optical axes by observing the rings through rectangular glass prisms; I then 

 proved by placing the plate upon a parallel quartz plate, that it developes colours 

 in polarized light when the plane of the optical axes is perpendicular to the 

 axis of the quartz ; the other plates, both parallel to the bisecting line, de- 

 velope colours under the same conditions when the bisecting line is parallel 

 to the axis of the quartz plate. 



These latter experiments were made after a great number of others, all of 

 which led to the same result. 



