402 Mr. Talbot's Facts relating to Optical Science. 



let them be fixed firmly in this position. Then if the observer 

 looks through the cube at a bright 

 white object, or at the sky, he will 

 see a number of coloured parallel 

 bands, the direction of vision be- 

 ing supposed to be perpendicu- 

 larly through two opposite sides, 

 ns A B, C D. If instead of this 

 he looks through A B at a light 

 coming from tlie direction X X 

 and then reflected internally on 

 the face A C, he will again see 

 numerous coloured bands upon 

 A C, but these will be comple- 

 mentary in their tints to the former 



ones. These coloured bands are analogous in their nature to 

 Newton's rings, differing only in being formed between two 

 plane surfaces either jiarallel or very nearly so, and viewed 

 by the observer at an incidence of 45°. 



But the beauty of the appearances may be surprisingly in- 

 creased by transporting the apparatus into a dark chamber, 

 and suffering a single pencil of the brightest solar light to pass 

 through the prism, or to be reflected from the face A C. If 

 then a sheet of white paper be held up, at any distance from 

 the prism, the coloured bands are depicted upon it with the 

 greatest vivacity and distinctness. The transmitted bands 

 have altogether a different character from the reflected ones, 

 so that it is impossible to mistake one for the other, even with- 

 out reference to tlie path of the ray. 



This experiment, easily tried, is one of the most beautiful 

 in optical science; I shall not, however, dwell upon it, be- 

 cause I believe it is sufficiently well known, and that it has 

 been exhibited in some public lectures. 



Now, in making this experiment with care, I have observed 

 some remarkable circumstances. 



The coloured bands are not, as has been supposed, isochro- 

 matic lines. The deviation is sometimes very marked, so that 

 a band in the course of its progress acquires very different 

 tints from those which it possessed originally. This fact may 

 be considered of some importance with respect to the theory. 

 It takes place when the prisms are in close contact, and the 

 bands few in numljer. But the following is still more deserv- 

 ing of attention. When the contact of the prisms is diminished 

 by interposing a hair between them, (still pressing them to- 

 gether,) the coloured bands depicted upon the paper, become 

 more numerous, narrow and crowded. Frequently they alter- 



