270 Sir David Brewster on the Action of Uncrystallized Films 



the refractive power of the plates and the intensity of the inci- 

 dent pencil. 



This law, though admitted by M. Arago in his article on 

 Polarization in the Encyclopaedia Britannica*, was called in ques- 

 tion by Dr. Young f, on the ground that no finite number of 

 plates could polarize the whole transmitted beam, as a small 

 portion of light must always remain unpolarized, or in the state 

 of natural light. This is doubtless true ; but, as Sir John Her- 

 schel has shown, it does not affect the truth of the law, which 

 involves the intensity of the incident pencil. According to the 

 law of geometrical progression, indeed, a small portion of unpo- 

 larized light exists mathematically in the transmitted beam ; but 

 a beam of light may be said to be completely polarized when the 

 unpolarized portion is invisible, vanishing entirely in certain 

 positions of the analysing prism. 



Neither M. Arago nor Dr. Young has made the slightest 

 reference to that portion of the refracted light which is reflected 

 at the surfaces of each plate and returned into the transmitted 

 beam. Sir John Herschel, however, has distinctly referred to it, 

 and remarks that "it mixes with the transmitted beam, and, 

 being in an opposite plane, destroys a part of its polarization}." 



Although the law of the tangents which I have mentioned 

 refers only to the transmitted pencil, yet, in the paper which 

 contains it, I have shown that the light reflected back into that 

 pencil is distinctly visible, not as ordinary light, as Sir John 

 Herschel maintains, but as light polarized in an opposite plane 

 to the refracted pencil. 



When the angle of incidence is considerable, this oppositely 

 polarized light appears as a nebulous mass, like the nebulous 

 image in the agate; and after examining it, I found it to have 

 the same relation to the refracted pencil " as the nebulous image 

 has to the bright image of the agate, or as the first has to the 

 second pencil of doubly refracting crystals §." 



In making the experiment with a small bright image of a 

 candle, and using plates of parallel glass, I found that the 

 reflected images a, a, a, a were distinctly separated from the 

 bright or refracted image A, and were all polarized by reflexion 

 in a plane opposite to that of A ||. 



Although these two facts, which have much theoretical import- 

 ance, were not only minutely described, but represented in dia- 

 grams, in my paper of 1814, yet they escaped the notice of the 



* Encyc. Brit. vol. xviiL part 1. sect. v. 



t Ibid., in the passage within brackets. 



X Treatise on Light, art. 868. 



§ Phil. Trans. 1814, p. 226, and plate 8. figs. 2, 3. 



II Ibid. 



