402 On Metallic Reflexion. 



room, and allowed to fall on a crystal. The reflected light was 

 limited by a slit, placed at the distance of two or three feet from 

 the crystal. This precaution was taken to ensure making the 

 observation on the regularly reflected light. Had no slit been 

 used, or else a slit placed close to the crystal, it might have been 

 supposed that the light observed was not regularly reflected, but 

 merely scattered, as it would be by a coloured powder. The 

 appearance of a spot of green light on a screen held at the 

 place of the slit showed that the light was really regularly 

 reflected. The slit was also traversed by the light scattered by 

 the support of the crystal, &c. The slit was viewed through a 

 prism and small telescope ; and the position of the dark bands, 

 or minima of brilliancy, in the reflected light could thus be 

 compared with the fixed lines, which were seen by means of the 

 scattered light in the uninterrupted spectrum corresponding to 

 that portion of the slit through which the light reflected from 

 the crystal did not pass. The minimum situated on the positive 

 side of the first bright band lay at something more than a band- 

 interval on the positive side of the fixed line D ; the minimum 

 beyond the fourth bright band lay at the distance of about half 

 a band-interval on the negative side of F. It thus appears that 

 the minima in the light reflected by the crystal were interme- 

 diate in position between the minima seen in the light trans- 

 mitted through the solution, so that the maxima of the former 

 corresponded to the minima of the latter. 



It might have been considered satisfactory to compare the 

 reflected light with the light transmitted, not by the solution, 

 but by the crystals themselves. But the crystals absorb light 

 with such energy as to be opake; and even when they are 

 spread out on glass, the film thus obtained is too deeply coloured 

 for the purpose. For to show the bands well, the solution must 

 be so dilute, or else seen through so small a thickness, as to be 

 merely pink. As M. Haidinger states that the phenomena of 

 the reflected light are the same for all the faces in all azimuths, 

 and for the polished surface of a mass of crushed crystals, it may 

 be presumed that the absorption is not much affected by the 

 crystalline arrangement, and that the composition of the light 

 transmitted by the solution is sensibly the same as that which 

 would be observed across a crystalline plate, were it possible to 

 obtain one of sufficient thinness. 



The first bright band in the reflected light does not usually 

 appear to be very distinctly separated from the continuous light 

 of lower refrangibility ; but the latter may be got rid of by 

 observing the light reflected about the polarizing angle, and 

 analysing it so as to retain only the portion polarized perpendi- 

 cularly to the plane of incidence. As the surface of the crystals 



