New Analysis of Solar Light. 409 



the case may be, and thus exhibit all the degrees of colour which 

 are observed through different thicknesses of the blue glass. 



According to the method of Brewster, all parts of the spec- 

 trum, weakened and unweakened, are before the eye of the ob- 

 server at once, and it is therefore impossible to prevent the irre- 

 gidarly dispersed portion of the brighter coloui-s from entering the 

 eye. Hence the problem reduces itself to thefinding out of another 

 method of repeating these experiments, by which the disturbing 

 colours shall be totally, or almost totally, excluded from the field 

 of view. If the spectrum be observed through a telescope, it is 

 possible to procure any desired colour isolated from the others, 

 but the irregular refraction and reflexion of the light without 

 the eye will be increased by the glasses of the telescope. The 

 changes of colour of the yellow stripe I found certainly less when 

 they were viewed thus singly, but nevertheless they still existed. 

 Another method, however, gave me perfectly satisfactory results. 

 The method is derived immediately from that of Brewster, if 

 instead of permitting the unchanged sunlight to enter through 

 the aperture, we transmit it first through a prism, and then per- 

 mit those portions only to pass through the aperture whose 

 changes of colour arc to be investigated. My manner of pro- 

 ceeding is as follows : — Solar rays reflected from a mirror are per- 

 mitted to enter through a narrow slit into a dark room and to 

 fall upon a vertical prism. Immecbately behind the latter is a 

 lens which casts the spectrum formed by the prism upon a screen. 

 In the latter is a second very fine vertical slit. The hght of that 

 band of the spectrum which falls exactly upon the slit passes 

 through, while the rest is cut off. The observer stands behind 

 this second screen, the back of which is well blackened, best 

 covered with black velvet, and looks at the slit through a second 

 prism of the best possible quality. If in the first prism, or in 

 the lens, no light was dispersed, then woidd homogeneous light 

 alone of a determinate colour arrive at the second slit and pass 

 through it j and this, on account of its homogeneity, would, when 

 looked at through the second prism, form no spectrum, but 

 remain a narrow band, just as if it were seen by the naked eye. 

 But as a small portion of white irregularly-refracted light enters, 

 a very feebly luminous spectrum is formed by the latter, in which 

 a single coloured band, that of the regularly refracted light, 

 comes very brilliantly forward. The light dispersed in the 

 second prism and in the eye, belonging, as it does, for the most 

 part to that of the bright band, cannot when mixed with the 

 latter change its colour, for it is homogeneous with it. Of the 

 other colours, those only which arc irregularly refracted in the 

 first prism pass through the slit; and this cpiantity of light is 

 so small, that the portion of it dispersed in the second jnism and 

 in the eye of the observer cannot be further perceived. 



