ON LIGHT. 345 



of the incident light. In this case, the ordinary ray has 

 been refracted ordinarily, and preserves its situation ; 

 the extraordinary extraordinarily, but its displacement 

 by the second refraction being exactly equal and oppo- 

 site (in consequence of the now reversed position of the 

 refracting rhomb) to that by the first, it is brought to 

 coincidence with the other, and the two united form one 

 image. 



(123.) The opposite sides of a rhomboid being par- 

 allel, both the ordinary and extraordinary rays after 

 transmission emerge parallel to the incident ray, by a 

 necessary consequence of that general law of retro-ver- 

 sion, in virtue of which a ray of light, wliatever path it 

 may have pursued from one point to another, can always 

 retrace that path ; the opposite faces being symmetri- 

 cally situated with respect to the axis. And the same 

 is true for a parallel plate of this or any other crystalHzed 

 substance artificially cut and polished, whatever be the 

 position which such plate may have held in the interior 

 of the crystal from which it is cut. Now it is found, by 

 cutting from rhombs of Iceland spar parallel plates in 

 various directions, that there is one through which a ray 

 of ordinary light can be transmitted perpendicularly 

 without being divided into two. This is the case when 

 the faces of the plate are at right angles to the line above 

 designated as the axis of the rhomboid. And generally 

 that a ray wliich within the crystal pursues a path par- 

 allel to this axis, will emerge from it single, whatever be 

 the situation of the surface of emergence. The axis, 

 then, is a lifie of no double refraction^ and in the case of 



