PROCESSES FOR PURIFYING THE SPECTRUM. 45 



same manner, therefore, that from the experiment of Sir W. HERSCHEL, in which a point 

 ef maximum heat was observed beneath the red end, the physical independence of light 

 and heat was shown, so from analogous experiments on the chemical changes exhibit- 

 ed bv the salts of silver, the existence of a distinct class of rays, invisible to the eye, 

 designated "chemical ra\s." was established. 



150. These general results lead us, therefore, to suppose that there exist in the so- 

 lar beam a variety of distinct principles, and when that beam is acted upon by a prism 

 of glass, those principles are parted out from each other. Among them some are visi- 

 ble, affecting the eye with the sensation of the various colours of light, red, yellow, 

 bine, .Sec., and others are invisible, affecting the thermometer, or producing chemical 

 decompositions. The general idea which we gather from tbese remarks is, that there 

 are three separate principles coexisting in the solar ray. light, heat, and a principle of 

 chemical action ; and when this ray is dispersed by a prism, three several spectra arise, 

 of which two are invisible, and one can be seen. Their relative position is sach as is 

 given in /?i r . 112. where A B is the luminous, and therefore visible spectrum ; C D the 

 invisible, chemical spectrum; E F the invisible spectrum of heat. 



151. These are the apparent phenomena which are exhibited when an ordinary so- 

 lar spectrum is employed. But in that spectrum the several coloured spaces are far 

 from being pure, and many interesting phenomena are therefore imperceptible. NEW- 

 TOX, who has never been surpassed by any experimenter in minute investigation, studied 

 with great attention the peculiar circumstances which conspire in the production of 

 the ordinary spectrum, with a view of isolating each one of the coloured spaces in a 

 state of purity. 



15:2. First, he shows that, when an image of the sun is formed in a dark cham- 

 ber upon a screen, as at z (Jig. Ill), that image does not terminate in a sharp circular 

 edge, as is the case with the object himself, but is surrounded by a penumbra, the light 

 gradually fading away (,fig. 113). This can be wholly removed by placing a convex 

 lens so as to receive the ray, and give an image in its focus. And just as the solar im- 

 age is surrounded by the penunibral ring, so also does the same defect accompany the 

 rays when dispersed by the action of the prism, for the faint light of the penumbra w 

 equally decomposed and dispersed with the bright light of the central image. Such an 

 effect is seen in fig. 114. 



153. Suppose, now, that by the use of a lens the penumbra is removed, and the di- 

 rect solar image (c) received on the screen is seen to be terminated by a sharp and 

 well-defined edge the image, of course, will be perfectly circular. If, now, a prism is 

 placed behind the lens, in such a position as to intercept the converging beam of light, 

 as at C (fig. 115), where A is the aperture in the shutter, B the convex lens converging 

 the rays that come in through the aperture to a focus, F, where they form a circular 

 image of the sun. the rays are bent on one side by the prism C, and a neat spectrum 

 is tormed on the screen at D E, with sharp edges, and devoid of any penunibral con- 

 tinuation. 



154. But the image of the sun is circular, and the spectrum itself arises from the sep- 

 aration and successive superposition of coloured circular images, that separation being 



