24 Mr. Tomlmson on [July 



placing them severally on the reflecting surface of the 

 mercury, and proceeding in the same manner as with the 

 solutions, I obtained complete and distinct accidental or 

 rather complementary colours. But one great difficulty 

 still remained, the emj^loyment of inconvenient bulks of 

 mercury. 



14. To obviate this objection I had recourse to a plane 

 mirror which I formed at first out of a piece of common 

 window glass, * the size of the stained disks, and obtained 

 the desired results by placing the disks successively thereon. 



r5. In furtherance of this idea I have constructed a little 

 apparatus for the purpose of observing the accidental 

 colours. It consists of a plane circular mirror, three 

 inches in diameter, on which a coloured disk of the same 

 size is placed, and the two glasses are thus enclosed in a 

 circular frame of wood. On holding the instrument 

 obliquely inclining to the light,t and bringing the periphery 

 close to the eye, and looking downwards into the glasses, 

 all reflected objects are doubled, such as the window frame, 

 chimneys, the distant cathedral spire, &:c. The window 

 frame being marked with broad bands of complementary 

 colours, while the second reflection of the spire, &:c., is the 

 complementary one. It will be evident that the frame 

 may be constructed in the form of a box, the inner side of 

 the cover containing simply a plane mirror, and the box 

 from three to seven or nine disks of stained glass. If three 

 only are employed, they of course will be red, yellow, and 

 blue; if seven, they will be the primitives. In employing 

 the instrument it will only be necessary to place one of the 

 disks on the plane mirror within the cover, and after the 

 observation as directed above, the stained disk may be re- 

 moved and another substituted. 



16. It must not be forgotten that the stained glass must 

 either be very thin, or the stain sufficiently light to allow 



* It is of course known that a plane mirror may be speedily formed by the re- 

 duction of mercury from the fulminate, the powder being equally and evenly 

 spread over the surface of the glass, and then ignited. A less expeditious, but an 

 infinitely better mode, is that recommended by Faraday, Chemical Manipulation, 

 page 574. 



t I should have observed before, that the experiments in this and the preceding 

 paper must be performed near the window, except, of course, where flame is em- 

 ployed ; but the instrument mav be employed in all cases where direct natural 

 light is present. 



