272 On the Action of Uncrystallized Films upon Light. 



dicularly at B, and emerging at C in the direction CD. As a 

 portion of the ray B C is reflected at C, and again reflected at 

 B and transmitted at C, the pencil C D will consist of two di- 

 stinct portions, one of which has been twice refracted, and the 

 other and much feebler portion twice reflected. As neither of 

 these portions are polarized, no physical change is produced by 

 their combination, unless when the plate M N is so extremely 

 thin as to produce the colours of thin plates by the interference 

 of the reflected with the refracted portions. 



When a ray R B is incident obliquely at B, it suffers refrac- 

 tion at B and E, and the emergent pencil E F contains a portion 

 of light polarized by refraction. This ray, in passing through 

 other plates or films parallel to M N, is at last completely polar- 

 ized in one plane, having grown feebler in intensity by the 

 abstraction of the light reflected at the two surfaces of each 

 plate. 



The portion of the refracted pencil B E which is reflected at 

 E and G, and a portion of it polarized, emerges at K as a pen- 

 cil K L, partly polarized by reflexion. A portion of G K is 

 again reflected at K and H, and emerges at P as a pencil P S, 

 more polarized by reflexion than K L. Hence the principal or 

 refracted pencil E F is combined with the pencils K L, P S (and 

 others by reflexions at P, &c), polarized in an opposite plane, so 

 that with a certain number of plates, varying with the angle of 

 incidence, the emergent pencil E F, K L, and P S consists of 

 two oppositely-polarized portions of light approximately equal. 



When polarized light is incident upon a pile of these thin and 

 colourless films, and subsequently analysed, it exhibits all the 

 properties of a plate cut perpendicularly to the axis of a 

 uniaxal crystal. The line A D corresponds with the axis of the 

 crystal ; and the different azimuths in which the polarized ray 

 may be inclined to this axis correspond with the principal sec- 

 tions of a uniaxal crystal.- The polarized tints have the same 

 value in every azimuth at the same angle of incidence, and 

 therefore form rings which, when crossed with plates of sulphate 

 of lime, descend in Newton's Scale like the tints of negative 

 uniaxal crystals. 



Out of hundreds of specimens now on the table, I have found 

 a few so colourless and so perfect as to produce, at different inci- 

 dences, all the polarized tints or rings up to the blue of the 

 second order of Newton's Scale. These colours are so pure, 

 and so regularly developed by the inclination of the plate, that 

 the most skilful observer could not fail to pronounce it to be a 

 portion of a doubly refracting crystal. 



The production of the leading phenomena of doubly refracting 

 crystals, namely, two oppositely polarized pencils, and the system 



