174 Researches in Circular Solar Spectra. 



darkened room. This disk presented two or three diffraction-rings 

 similar to those of telescopic stars, but much broader. 



The same method was attempted with the sun. Various objec- 

 tives were used to obtain a solar miniature of the sun's disk. 



Plain muTors of glass silvered at the back entirely failed. 



In order to form a pure and brilliant solar spectrum under the 

 microscope, it occurred to me to take advantage of the principle 

 of total internal reflexion from a prism. I then constructed a 

 prism-heliostat, which, acting in sunshine, presented an aerial 

 miniature of the sun of great splendour (almost as dazzling as the 

 sun itself) for half an hour, without further adjustment. 



The prism-heliostat was furnished with a crown-glass double 

 convex lens, itself being of flint, and other lenses of less focal 

 length could be attached to diminish or increase the diameter of 

 the primary image of the sun, which, as the focal length gene- 

 rally used was three inches, gave an image 3 sin 30', in diameter, 

 or one-fortieth of an inch nearly. In some cases object-glasses 

 and eye-pieces were placed in the solar rays emanating from the 

 prism. 



Keceived by an inverted object-glass of the finest quality at a 

 distance of 200 inches, the solar disk could be further miniatured 

 to any desirable degree of minuteness. A theoretical diameter 

 of sixteen millionths was found convenient. To moderate the 

 overpowering brilliance of such a spectrum directly viewed, dark 

 slides of graduated neutral tints were at first used, and smaller 

 primary disks were obtained by using deeper lenses at the prism. 



When this minute spectrum is viewed with a high power 

 (800 or 1000), the phenomena attending residuary errors, whether 

 of achromatism, spherical aberration, or mechanical construction, 

 are demonstrated with so keen a severity upon the handiworks of 

 man as to throw all other methods into the shade. 



Phenomena Observed. 



With the flame-disk formerly used * only two or three diffrac- 

 tion-rings could (as already remarked) be descried ; an extra- 

 ordinary number of richly-coloured rings of dazzling brilliance was 

 now exhibited at the instant of bringing the solar disk into the 

 plane of focal vision. 



The most striking feature, amid so much effulgence, was an 

 intensely black (jet-black) diffraction-ring encircling the central 

 disk at the clearest focal point. The appearance of the rings 

 changed every instant with the slightest change of focus, and their 

 tints indicated the nature of the " secondary spectrum." 



Upon closer inspection, my curiosity was excited by observing 



* 'Phil. Trans.,' part ii., 1870, p. 595. 



