Sir DxyiD BREVfSTEVL on the Compensations of Polarized Light. 381 



to the common light, or to the total quantity in each beam, is the same ;* but this 

 equality is accidental, as appears from the fact already mentioned. 



The remarkable phenomena produced at this angle in glass, and at the cor- 

 responding angle in all transparent bodies, where cos. (i-\-i') zz cos.'^ (i — i') 

 require to be more minutely stated, and lead us to the construction of what may 

 be called the compensating rhomb, which is shown in Plate, Fig. 1 . It consists 

 of a well annealed rhomb of glass, or of any other uncrystallized body abcd, 

 having, in the case of glass, the angles bad, bcd = 139° 25', and abc = 40" S5', 

 when the index of refraction is 1.525. If a ray of light Rr, is incident upon ab, 

 at an angle of 82'' 44', exactly one-half of it will be reflected in the direction rm, 

 and the other half refracted in the direction rN, having each the same quantity 

 of polarized light, as already stated. But the ray rN is again reflected at n at an 

 angle of 40° 35', and it will emerge from the face ad nearly perpendicularly, 

 without suffering any perceptible refraction, in the direction nm'. If we now ex- 

 amine this ray m'n, we shall find it to be in the state of common light, although 

 the incident ray rN contained 145 polarized rays, or nearly one-half of the pencil 

 rN. In order to be satisfied of this, the compensating rhomb should be made of 

 two equal and similar rectangular prisms, abc, ado, cemented to or nearly touch- 

 ing one another. By removing adc, the ray rN emerging nearly perpendicularly 

 from the face ac, will exhibit the state of its polarization, when it falls upon the 

 face DC at the point n. 



We have now obtained by this experiment a very singular result. If the 

 pencil rN consists of 145 rays of polarized light, and 333 — 145 = 188, of com- 

 mon light, the effect of a single reflexion at n has been to unpolarize polarized 

 light ! and to produce no change at all upon common light ! a property of a re- 

 flecting surface hitherto unheard of, and incompatible with all our present know- 

 ledge of the polarization of light. After such a conclusion, it would be an un- 

 profitable task to adduce any further arguments ; and I shall therefore only state 

 that all the phenomena of polarization, by successive reflexions and refractions, 

 stand in direct contradiction of the views which I have been combating. 



The restoration of the pencil rN to common light by reflexion at n, furnishes 

 us at once with the principle of compensation, in conformity with the laws of po- 

 larization deduced in my papers of 1830. The whole of the ray rN has suffered a 

 physical change by refraction at r, consisting of a rotation of Its planes of polari- 



* This is the only angle where this equahty obtains. 



