ON THE NATURE OF LIGHT AND COLOURS. 469 



wineglass, and placed in a vertical position, its upper edge becomes ex- 

 tremely thin, and appears nearly black, while the parts below are divided by 

 horizontal lines into a series of coloured bands; and when two glasses, one 

 of which is slightly convex, are pressed together with some force, the plate 

 of air between them exhibits the appearance of coloured rings, beginning 

 from a black spot at the centre, and becoming narrower and narrower, as the 

 curved figure of the glass causes the thickness of the plate of air to increase 

 more and more rapidly. The black is succeeded by a violet, so faint as to 

 be scarcely perceptible; next to this is an orange yellow, and then crimson 

 and blue. When water, or any other fluid, is substituted for the air between 

 the glasses, the rings appear where the thickness is as much less than that 

 of the plate of air, as the refractive density of the fluid is greater; a circum- 

 stance which necessarily follows from the proportion of the velocities with 

 which light must, upon the lluygenian hypothesis, be supposed to move in 

 different mediums. It is also a consequence equally necessary in this theory, 

 and equally inconsistent with all others, that when the direction of the light 

 is oblique, the effect of a thicker plate must be the same as that of a thinner 

 plate, when the light falls perpendicularly upon it; the difference of the 

 paths described by the different portions of light precisely corresponding 

 with the observed phenomena. (Plate XXX, Fig. 447 . . 449.) 



Sir Isaac Newton supposes the colours of natural bodies in general to be 

 similar to these colours of thin plates, and to be governed by the magnitude 

 of their particles. If this opinion were universally true, we might always 

 separate the colours of natural bodies by refraction into a number of different 

 portions, with dark spaces intervening; for every part of a thin plate, which 

 exhibits the appearance of colour, affords such a divided spectrum, when 

 viewed through a prism. There are accordingly many natural colours in 

 which such a separation may be observed; one of the most remark- 

 able of them is that of blue glass, probably coloured with cobalt, which 

 becomes divided into seven distinct portions. It seems, however, im- 

 possible to suppose the production of natural colours perfectly identical 

 with those of thin plates, on account of the known minuteness of the 

 particles of colouring bodies, unless the refractive density of these par- 

 ticles be at least 20 or 30 times as great as that of glass or water; which is 

 indeed not at all improbable with respect to the ultimate atoms of bodies^ 



