190I-2.] Photography in Natural Colours. 375 



in any desired proportions, to form a patch upon the screen, visible to 

 the entire audience. The h'ght from the crater of the electric arc is 

 converged, by the condensers of the lantern, upon the slit at one end of 

 a collimating tube which contains at the other end a lens whose 

 principal focus coincides with the slit. The beam of light from the 

 lantern hence emerges from the collimating tube parallel, and will form 

 a distinct image of the slit at the principal focus of any lens inserted in 

 its path. The interposition, between this and the collimating lens, of 

 direct vision prisms, constructed to give dispersion without deviation, 

 breaks up the single uncoloured image of the slit into a number of over- 

 lapping coloured images forming a pure spectrum \show7i~\ in the plane 

 containing the principal focus of the lens. If a large condensing lens be 

 placed beyond this plane, its function will be to collect all these coloured 

 images, and form an image of the last surface of the prisms. This image 

 is, however, too small to be visible at any distance, and an enlarged image 

 of this image may, by means of another large lens, be formed upon a 

 screen beyond, and of such a size as to be plainly visible to all. By 

 causing this image to fall upon a small square of white card on a black 

 velvet background, the effect of coloured edges can be eliminated ; and 

 the colourless nature of the image formed in this case is evidence that 

 the union or recomposition of all the spectum colours gives white light 

 \shown\. If a card containing a narrow slit be moved along in the 

 plane of the spectrum, the patch on the screen will assume each 

 spectrum colour in turn, isolated, of course, from its fellows and hence 

 uninfluenced by contrast. The substitution for this single slit of a 

 brass frame containing three slits, whose relative positions and apertures 

 can be varied at will, and which was specially constructed for this 

 experiment, enables us to determine the resultant colour produced by 

 the mixture of any two or three spectrum colours, in any desired 

 proportions. If the positions of these slits be made to coincide with the 

 primary colours, as determined by Abney, which can easily be done 

 by burning lithium and magnesium salts in the arc, thus "scaling" the 

 spectrum, the resultant colours produced by the mixture of these 

 primary colours can be at once determined. 



The union of the three primaries, red, green, and blue-violet, in 

 certain definite proportions, easily determined by trial, and measured by 

 the relative apertures of the slits, produces an uncoloured patch on the 

 screen \shown\ ; and this white light, although not optically equivalent, 

 being produced by the mixture of three narrow isolated bands instead 

 of the whole range of the spectrum, can not be distinguished from 

 ordinary white light. Nor can colours, produced by the mixture of 



