Accidental and Complementary Colours, 209 



^ placed before the eye, the latter will be unable to distin- 

 guish any thing for a second or two. The order, therefore, 

 will be as follows: — green — red — green — black. This 

 can be repeated any number of times, and with any coloured 

 disk, provided the colour be sufficiently deep. 



2. The rationale of this experiment, I assume to be as 

 follows : — White light is composed of red, yellow and blue ; 

 green is composed of the two latter, which the green disk 

 transmits easily and readily, while it opposes the trans- 

 mission of the red rays : the surface of the glass, there- 

 fore, next the eye, transmits green light, and the opposite 

 side is receiving and reflecting red ; if the glass be suddenly 

 removed, the red rays, whose path was obstructed by the 

 disc enter the eye ; and, during this process, if the green 

 disk be restored, we view the three colours red, yellow and 

 blue by super-position ; hence, the result is black. 



3. In order to prove that the red rays are reflected, we 

 have only to reverse the experiment. Take a disk of green 

 glass, sufficiently transparent to allow reflection from its 

 under surface, — let the observer's back be opposite to the 

 window, so as to get the window frame reflected ; th,e bars 

 will appear of a strong and decided green ; this reflection 

 is obtained from the under surface of the glass — a faint 

 image of the bars, &c., will be obtained from the first sur- 

 face, which will be red. When the two surfaces are not 

 quite parallel, this effect is most apparent, as the two 

 images are separated to distances varying from each other, 

 thus indicating, very clearly, the surface from which either 

 colour is produced ; accordingly, the Perichromascope, 

 (Records, vol. ii. p. 24,) which depends wholly on this 

 principle, is more satisfactory, when the coloured glass 

 and the plane mirror form an angle with each other. The 

 reason why the complementary colours are so seldom seen 

 in a piece of common stained glass is, that its small thick- 

 ness prevents a sufficient separation of the images from the 

 two surfaces. 



4. The rationale of the experiments (1, 3,) derive ad- 

 ditional support from the following experiment. When 

 the sun is shining on the window, place a prism so as to 

 throw a well defined spectrum upon the ceiling of the 

 room. Arrange the Perichromascrope so that the coloured 



VOL. IV. p 



