1814.] Foreign Secretary of the Royal Society. 259 



not depolarized. When subjected to the action of a rl--omb of 

 Iceland spar, it is alivays capable of being refracted in a single ray; 

 but undcra mean incidence, and one peculiar to each metal, it 

 appears complcieiy depolarized, so that it may be always divided 

 into two rays by Iceland spar. Hence metallic surfaces, as well as 

 other bodies, have a determinate angle at which they change the 

 poles of the ray which they reflect. But since in this case the plane 

 of reflection makes an equal angle with the two poles, one liirif of 

 the light is polarized in one direction, and the other half in the 

 other ; so that the reflected ray assumes the properties of a direct 

 ray.* This experiment shows us why, if we employ with metallic 

 mirrors the same method as with diaphanous bodies, the determina- 

 tion of the proposed angle becomes impossible. In fact, when 

 natural light falU under the angle prpposed, the reflected ray con- 

 tains at once the molecules polarized in one direction and in the 

 other ; so that, when decomposed by Iceland spar, it presents the 

 same phenomena as the natural ray which is reflected without 

 polarization under the gicatest and smallest incidences, which in 

 that case renders the limit indeterminable. When we subject to 

 the reflection of the mirror a ray already polarized, we avoid that 

 inconvenience ; because, instead of observing as with transparent 

 bodies tiie angle under wliich the polarization is most complete, we 

 observe, on the contrary, that in which the depolarization is ia 

 appearance the most complete. 



Thus, for metallic substances, we must employ the reflection of 

 a ray already polarized, and taking care that the poles of the ray 

 form an angle of 45° with the plane of incidence, and must observe 

 at what angle light appears depolarized like a natural ray : for 

 diaphanous substances, on the contrary, we must employ the re- 

 flection of a natural ray, and observe the atigle at which the light 

 appears completely polarized. The angle in both cases will be 

 determined with the same accuracy. 



It is now therefore proved that all bodies in nature, both opaque 

 and diaphauous, polarize light entirely when tliey reflect it under a 

 certain angle ; and that the metals, which alone appeared to con- 

 stitute an exception to this law, on the contrary polarize a greater 

 quantity of light than other bodies, since ihcy reflect a greater 

 quantity. 



From the preceding experiments there results a remarkable fact. 

 Place before a luminous body a set of parallel glass plates, and 

 increase their number till the body ceases to i)e seen when viewed 

 perpendicularly through tliem : if you look at it obliquely through 

 the same plates you will begin to see it again. Under the angle of 



• What distinguishes meials from transparent bodies U, that these last refract 

 all the li|rli( poUrizrd in one dir.-ction, and rfU.ct all polarized in the otin r ; 

 wliilr inelallic bodies reflect wliat tliey polanze in binli direclions. Ii being 

 underitiiiid, fauwevrr, tlial they possess in part the power of all otlier opaque 

 bi'dic, of absorbing a greater proportion of ihai kind of ray wliicU di*p'liauou» 

 bodies irantmit, 



B 2 



