360 P. C. on the Colours that enter into the [Nov 



points of the compasses fall respectively on these centres ; 

 80 that, supposing the fringes to be fixed, and the cards 

 brought together, the complementary colours would be 

 superposed. That this superposition, or intersection, actu- 

 ally takes place, we may prove by bringing the two cards 

 so much nearer to each other as to cause a superposition of 

 the red and violet fringes ; this will reduce the separation 

 of the cards to about half their former distance : if we now 

 view the cards through the prism, we shall find that this 

 superposition has produced crimson, the colour formed by 

 violet and red, and that these two colours have disappeared. 

 If, pursuing the experiment, we gradually bring the cards 

 closer, as the blue covers the red, the violet will intersect 

 the yellow, and the whole of the colours will become weaker, 

 until, upon the meeting of the edges, they disappear. 



In the closing part of this experiment, we separate the 

 violet from the red, and carry it forward to the yellow ; 

 while, with the same change in the position of the cards, 

 the blue is made to overlap the red, and the whole surface 

 becomes white. But, although in this experiment white 

 light is formed by the intersection of the usual complemen- 

 tary colours, it is worthy of remark, that the superposition 

 of the three compound colours which are here collected 

 together, viz. blue, yellow, and crimson, would also pro- 

 duce white light ; and it is desirable to notice it, not only 

 as a circumstance curious in itself, but, also, to prevent its 

 misleading those, who being disposed to consider blue, yel- 

 low, and red, the primitive colours, may suppose they have 

 formed white light by their superposition, simply by mis- 

 taking crimson for red ; a mistake which, as we have before 

 observed, may easily be made. 



If we now confine our attention to one of these cards, we 

 may observe that the fringes upon the upper edge have the 

 same complementary character to the fringes upon the lower 

 edge ; and that, in fact, the whole of these fringes arise 

 from the want of a correct superposition of the three images 

 of the card upon the retina, produced by the unequal re- 

 fraction of the prism : we see the fringes upon the edges 

 only, because, upon every other part of the card white light 

 is formed by a system of mutual compensations. 



If, instead of bringing the cards nearer to each other, we 



