510 Prof. S. P. Langley on the Invisible 



Temporarily mounted on A R, and moving about with it, is the 

 large spectro-bolometer described in a previous memoir. The 

 centre of its graduated circle (C) lies under the point P. Its 

 two long arms are not free to move as usual, but are con- 

 strained by mechanical attachments (not here shown) to 

 occupy the positions Pp, PD. 



Two large 60° prisms of the same material (pure rock-salt 

 from the same mine), their faces worked with the greatest 

 accuracy, are placed with their equal refracting angles in 

 opposite directions, one (Pj at the obtuse, the other (p) at the 

 acute angle of the parallelogram (PpDc), the vertices of all 

 whose angles in the mechanical construction are pivoted and 

 connected by inflexible arms, so that (both prisms being kept 

 automatically in minimum deviation by the attachments, M m) 

 the angle of minimum deviation (cp P) for the first prism is 

 necessarily equal to the angle of minimum deviation (R P D, 

 or its equal PDc). 



Thus, if the pencil of solar lunar heat (reflected from a 

 large siderostat on the north, not shown here) passes from N 

 towards S, on moving the beam A R, pivoted at p (p being 

 the projection in our drawing of a vertical line passing 

 through the centre of the turn-table and the median line of 

 N S and A R) into various positions (N S remaining fixed), 

 the rays, which are refracted by the prism p in the direction 

 jp P, will emerge from P in the direction P D and fall upon 

 the bolometer B. A condensing-lens {I) forms the solar 

 image on the slit (s{) of the first spectroscope, whose train 

 (consisting of collimator, l l9 prism, p, and image-forming 

 lens, / 2 ) forms a spectrum on the slit (s 2 ) of the second spec- 

 troscope. Here a narrow pencil from the first spectrum, 

 comprising only the particular wave-lengths which fall within 

 the width of s 2 , is admitted, and, by the second train L : P h 2} 

 formed into a horizontal spectrum at and on either side of B. 

 When we move A R this spectrum moves in turn past the 

 vertical linear thread of the bolometer B, which lies in the 

 focal plane of this spectrum, and is immersed in its successive 

 absorption-lines as these defile past it. The function of the 

 first spectroscopic train (I, l\,p), l 2 ) is solely to sift out the 

 extraneous radiations, and to present at the second slit (s 2 ) 

 only those which legitimately belong to that part of the spec- 

 trum we wish to study. These pure rays pass into the second 

 slit and are analysed by the second train in the usual way, by 

 the aid of the linear bolometer at B, and of the circle (C) 

 reading to ten seconds of arc. 



The objection to this apparatus is its complexity, which, 

 however, we have been unable to advantageously diminish. 



