362 ON LIGHT. 



which, as we have seen, depends on the proportion of 

 velocity of the hght in and out of a medium. In the 

 case of one set of vibrations, again, the propagation of 

 the medium may be equally impeded or influenced in 

 all directions of the ray in which case a wave starting 

 from any point in the- surface would run out spherically 

 within the crystal, while in that of the other the amount 

 of obstruction might vary with the direction of the ray, 

 and thus give rise to a wave running out with different 

 velocities in diffei'ent directions from its centre of propaga- 

 tion, and therefore not spherically. 



(142.) To this conclusion, but without passing through 

 the intermediate considerations which have led us up to 

 it, Huyghens (who certainly had formed no conception 

 of transverse vibrations) appears to have jumped^ by one 

 of the happiest divinations on record in the history of 

 science, viz., that in the double refraction of Iceland 

 spar, while the ordniary ray is propagated in a spherical^ 

 that of the extraordinary spreads from its point of origin 

 at the surface of the crystal in an elliptical w^lvq., the 

 form being that of an oblate spheroid of revolution, 

 having its polar axis parallel to the axis of the rhom- 

 boid, and bearing to its equatorial diameter a defnrite 

 numerical proportion, viz., that of eight to nine (very 

 nearly). Making this assumption, and laying it down 

 as a principle (capable of demonstration), that the 

 direction of a ray of liglit in such a mode of propaga- 

 tion is not that of a perpendicular to the surfoce of the 

 wave at any point, but that of a line drawn from the 

 centre of the wave to its point of contact with a plane, 



