204 



Mr. R. T. Glazebrook. 



[June 19, 



Yalues of principal refractive indices in Iceland spar for the three 

 hydrogen lines. 





C. 



F. 



9- 





1-65436 



1-66779 



1-67553 





1-48456 



1-49074 



1-49430 



The values of 6 are determined from the known position of the 

 wave in question, with reference to the normals to the faces of the 

 prism, and of these normals with reference to the optic axis. 



The results of the comparison are as follows : — 



For the red line C, just in the neighbourhood of the major axis of 

 the elliptic section, the curve given by experiment lies inside that 

 given by theory, the two coincide at about 15° from this axis, hence- 

 forth onwards the experimental curve lies outside, and the two coincide 

 again near the minor axis. 



The greatest difference amounts to rather over '0001, about half way 

 between the axes. If we decrease slightly the major axis of the ellipse 

 as given by theory, we may say that the result of experiment is ex- 

 pressed by an oval curve coinciding with Huyghens' ellipse at the 

 extremities of its axes, and lying outside of it for the rest of its course, 

 the greatest difference between the radii vectores being about midway 

 between the axes, and there amounting to about T ooo~o* n P ar ^ °^ 

 either. 



The experiments for the ray F lead to similar results, the differ- 

 ences, however, between theory aud experiment being markedly 

 greater, and with the assumed values for negative near the 



minor axis, so that the experimental curve is on the inside there. 



For g, taking the values of yu, 1? fi 2 , given above, the differences are 

 negative, and the experimental curve o, the interior, near the axes, 

 positive in the middle part of the arc, and the experimental curve o, 

 lies outside, they amount to rather over '0002 when a maximum. 



Thus, the results for each of the three rays point to the same con- 

 clusions, and would lead us to infer that a section through the axis of 

 the surface of wave slowness in Iceland spar would not be an ellipse, 

 but an oval curve, lying slightly outside the ellipse, and that the 

 difference between the two becomes greater as the wave-length de- 

 creases. 



These differences, however, are very small, amounting to only about 

 •0002 for the violet ray. 



I have, moreover, been able to prove that no variations in the 

 values of the constants of theory or the position of the plane of the 

 prism, with reference to the optic axis, will produce the effect re- 



