Absorption and the Colours of Thin Plates. 215 



the particles had sizes corresponding to these pure tints, and 

 that the light which composed them should not pass through 

 different thicknesses of these particles. As I have demon- 

 strated, however, in a paper already referred to, that the tint 

 which Newton reckoned one of the third order, has no con- 

 nexion whatever with that or with any other order, and that 

 all other tints of absorbent media are in the same predica- 

 men;, we are not only free from the difficulty which embar- 

 rassed Newton ; but it is actually necessary to have recourse 

 to particles of an ordinary refractive power, and having such 

 forms and occupying such positions as will permit lateral 

 transmissions and thus produce compound tints, such as we 

 actually observe in natural bodies, and as we have shown to 

 be produced by thin plates. 



Now if we suppose the colouring particles to be spherical, 

 or to have the form of plates or cubes, or other solids dissemi- 

 nated through the fluid or solid bodies which they colour, 

 the tints would be permanent and compound as we find them 

 in nature. 



The second point of difference to which I have referred, 

 namely, the absolute disappearance of the reflected tints in 

 several coloured solids, fluids, and gases, is one of great mag- 

 nitude. Newton has evaded this difficulty in his theory ; but 

 from the manner in which he gets rid of the intromitted light 

 in black bodies, it is obvious that he would ascribe the dis- 

 appearance of the reflected tints to their being " variously 

 reflected to and fro until they happened to be stifled and 

 lost." 



As I shall have occasion to discuss this subject experiment- 

 ally in a paper on the permanent colours of natural bodies, I 

 shall only state at present that I have succeeded by particular 

 methods in rendering reflected tints visible in many coloured 

 fluids and glasses, but 1 cannot consider them as equivalent 

 to the reflections of thin plates. 



I have endeavoured to corroborate the views contained in 

 the. preceding pages by a series of collateral experiments on 

 the periodical colours of polarized light. When we divide 

 the spectrum into bands by doubly refracting plates, the phae- 

 nomena are beautiful beyond all description. If we dissect 

 or subdivide the luminous bands in the spectrum, as seen by 

 one analysing prism, by means of successive plates and prisms, 

 the result is very remarkable; and if the doubly refracting, 

 plates are inclined to each other or to the incident beam, the 

 black bands will also be inclined to each other, and the lu- 

 minous spaces have the form of a triangle either complete or 

 truncated at its apex. By using plates of the same or of va- 



