LORD RAYLEIGH ON THE COLOURS OF THIN PLATES. 165 



to spheres, or masses of other shape, small in all their dimensions relatively 

 to the wave-lengths of light. According to this view, the colour is that 

 found by taking the components of white light proportionally to X -4 , instead 

 of X~ 2 .* 



The curve, starting thus from a definite point, takes a nearly straight course 

 in the direction of white (W), which it passes a little upon the green side. The 

 white of the first order on Newton's scale is thus somewhat greenish, as must 

 obviously be the case when we consider that it arises when the maximum 

 reflection is in the green or yellow portion of the spectrum, so that the red and 

 blue must be relatively deficient ; but the deviation from white is very small, 

 and is not usually recognised. After leaving white the curve passes through 

 the yellow, and approaches pretty close to the side of the triangle at a point 

 representing the D-line in the orange.t The retardation is here 1688. The 

 colour then reddens, but makes no approach to the spectrum reds lying near 

 the corner of the triangle. Passing rapidly through the purple "transition- 

 tint," it becomes bluer, until it attains the magnificent blue or violet of the 

 second order, in the neighbourhood of V = 2328. At this.point there is a good 

 approach to the corresponding spectrum colour, although the latter lies here a 

 little outside the triangle. Leaving blue the colour rapidly deteriorates, 

 becoming greener, but nowhere attaining a good green. The best yellow of the 

 second order at 3400 is nearly as pure as the best of the first order, but inclines 

 less to orange. The reds of the second order are even less pure than those of 

 the first, but the inferiority diminishes as we approach the second transition- 

 tint in the purple. The blue of the third order at 4200 is much inferior to the 

 corresponding colour of the second order, but gradually acquires a superiority 

 as it becomes greener near 4400. The blue-greens which follow, and the full 

 greens from 4800 to 5000, are splendid colours, beyond comparison superior to 

 the corresponding colours of the second order, but yet falling far short of the 

 spectrum colours near (44). On the other hand, in the third order the yellows 

 are not so pure as in the first and second orders, and there is even less 

 approach to red, although a better show is made in the purple at 6000. In the 

 transition from this purple to green, the blue falls short even of the blue of the 

 first order, but the green at 6800 is very fine, sensibly equal to one of the 

 greens of the third order. It will be remarked that in the fourth order greens 

 there is little variety, the direction both on the outward and on the backward 

 course being nearly in a line through white. On the return to white, which is 



* See several papers by the Author, published in the Philosophical Magazine, " On the Light from 

 the Sky, its Polarisation and Colour," Feb. 1871, April 1871 ; "On the Scattering of Light by small 

 Particles," June 1871 ; " On the Electro-Magnetic Theory of Light," August 1881, &c. 



t The points 20, 24, 28, .... on the diagram, represent the spectrum colours as determined by 

 Maxwell. 



VOL. XXXIII. PART I. Y 



