200 Mr. S. P. Langley's Experimental Determination of 



Figure 1 illustrates the means finally adopted, and the 

 course of the rays through the apparatus; although, for the 

 sake of distinctness, the mechanical devices used to maintain 

 the proper arrangements of the parts are omitted. The rays 

 of light, coming from the 300 millim. flat mirror of the large 

 siderostat, pass across the apparatus, and fall upon a concave 

 speculum of 180 millim. aperture at M, by which at a distance 

 of about 1*5 metre they are converged to a focus at Si. At this 

 point is a vertical slit, adjustable to any desired width by a 

 double screw, which moves both jaws at once, so as to keep 

 the centre always in the same place. This slit is protected 

 from the great heat by a plate of iron pierced with an aper- 

 ture only a little larger than the slit when open to the usual 

 width. Beyond S x the rays diverge and fall upon the concave 

 grating, Gr. Directly opposite the grating is a second slit, S 2 , 

 also double-acting ; and the apparatus is so arranged that the 

 two slits, S 1? S 2 , and the grating, G, always lie upon the cir- 

 cumference of a circle whose diameter is 1*63 metre; and 

 therefore in whatever manner the slits may be placed, the 

 light coming through Si forms a sharp spectrum upon S 2 . 

 A very massive arm, carrying the grating, the slit S 2 , and the 

 heavy spectro-bolometer, is pivoted at the centre of the circle, 

 so that the relative positions of these parts are unchanged. 

 The slit S 2 is automatically kept diametrically opposite the 

 grating, and on the normal to its centre. 



The slit S 2 is the slit of the spectro-bolometer, provided 

 with the same attachments as when used for mapping the 

 visible spectrum (except that it is now fitted with simple 

 collimating and objective lenses of the same special kind of 

 diathermanous glass as the prism, instead of its own concave 

 mirror). Its eyepiece and the bolometer are interchangeable. 



By means of the eyepiece and graduated circle the devia- 

 tion, and consequently the refractive index, of the rays passing 

 through the slit can be determined, if they are visible. If 

 they are invisible, their exact wave-length is known by a 

 simple ocular observation of the visible ones, on which they are 

 superposed by the action of the grating, while their subse- 

 quent deviation is determinable by the bolometer placed at B, 

 provided they retain sufficient energy to affect the instrument. 

 It will be seen that, according to this method, all those in- 

 visible rays which arc n times the definitely known length of 

 some visible ray are caused to pass together through a slit, 

 and then through a prism, which sorts out the ray of the first 

 spectrum from that of the second, that of the second from 

 that of the third, and so on, so that the corresponding index 

 of refraction may be determined by obseivation — with the 



