22 Decomposition of light. 



Properties of The striking agreement of my hypothesis with the p£* 



deHneateTon a cunar,ties °*" the spectrum excited me the more to apply it 

 circle. to the dial of colours. This coloured figure has such sin- 



gular properties, that the mind cannot easily bend itself to 

 them. How indeed can it conceive the existence of an infi- 

 nite number of luminous rays, all different yet equally sim- 

 ple? How is it, that taken in pairs from the extremities of 

 every diameter of the dial, that is from any two opposite 

 points, they shall always form the same white? For in- 

 stance, a certain red ray with a green gives white;' art 

 orange with a blue, the same; a violet with a yellow, still 

 the same. What a strange similitude! How again are the 

 seven distinct orders of the spectrum consistent with that 

 insens'ble gradation of the tints of the dial recommended 

 by Newton, and in fact necessary? Yet afl these are so 

 completely supported by experiment, that their reality can- 

 riot be questioned. 

 Solution of the Thus I had a problem to solve, the complicated data of 

 problem. which seemed at first not to promise a simple solution; yet, 



after. various attempts, I attained my object, as will be seen. 

 First i considered, that both the nature and quantity of 

 the red, green, and violet rays, which I suppose to be the 

 sole elements of white light, are absolutely unknown. But 

 I could iikowi.se conceive them transformed into coloured 

 maiters of such intensity, or condensation, that the mix- 

 ture of an equal quantity of each should produce exactly 

 white. 



In the second place I drew Fig. 2. This consists of three 

 curves nearly circular and alike, described round the dial in 

 the following manner. I first described three equal circles, 

 having their centres in the radii drawn through the divisions 

 of 60, 180, and 300 degrees; and the circumferences of 

 which were targents to the dial at (lie divisions of 250, 360 

 and 120 degrees respectively I then modified each circum- 

 ference by this law, that, on prolonging the diameters of 

 the dial in every possible direction, the sum of the prolon- 

 gations of every diameter to the new curve should be a con- 

 stant quantity. It is easy to understand this second con- 

 struction, by which it will appear, that the .resulting curve 

 differs in fact little from the circular cin ; o inference. 



. . ' Third I v. 



