pritchard's wedge photometer. 313 



of uniform substance and figure, and apart from the consideration of reflection, etc. — 

 increase in geometrical progression for equal increments of thickness, if there is no 

 selective absorption in the wedge. That this is not so, is shown in c, where we find 

 on the contrary that there is a very strongly marked progressive increase in the 

 quotients there given. The variations of b are shown graphically in Fig. 5. 



A 3'et more striking proof of the great selective absorption at least of the total 

 heat is shown otherwise in line h, where we see that the heat transmitted through 

 the thickest part of the wedge is hardly over ^V ^^^^ of the direct beam. We 

 consider this in connection with the fact that the light here transmitted has been 

 found by Professor Pickering to be something like yoVtt of the direct light. But 

 though this is evidence of great selective absorption of heat, taking all kinds indis- 

 criminately, it may still be quite possible that this absorption exists chiefly in the 

 invisible, lower part of the spectrum, — that is, in the infra-red, — and that the 

 wedge is sensibly non-selective in the visible part. 



To determine this distinct absorption in the visible part, we have the prismatic 

 measures, of which there are four or five incomplete series, obtained on three dif- 

 ferent occasions. To each of these we have also added a set of measures just 

 below the visible part (at 1'') to make the tendency more evident.* 



Extract from Original Record. 

 Statioti, Allegheny. 

 Bate, July 27, 1886. 



Temperature of Apparatus = 26° C. * 



State of Ski/, milky blue. 

 Aperture of Slit = 2 mra. (Heiglit of wedge was but 0.59, that of slit used to-day, but no heat 



entered in any case excejjt through wedge save when that was away altogether. Deflections 



with wedge away are reduced by multiplying by the factor 0.59.) 

 Prism used, Ililger No. 3. (A very large Flint prism.) 

 Spectrum thrown west. 

 Galvanometer, No. 3. 

 Tim,e of single Vibration, 9 sec. 

 bolometer, No. 16 (1 mm. aperture). 



• It may possibly not be superfluous to recall to the reader, that, wliile the wave-length 0''.4 in the following 

 tables corresponds nearly to tlie extreme violet, that at Om.7 to the extreme red, and the visible spectrum or "light 

 heat" lies almost wholly between these points, the greater part of tbc wbole solar energy is found in wave-lengths 

 greater than 0^.7 in the form of invisible radiation or "dark heat." (The radi.atiou in wave-lengths beyond the violet 

 is here quite negligible.) As we are not particularly concerned here with this "dark heat," we have confined our 

 measures to points between 0.4 and 0.7 inclusive (light heat), except one set at li» (dark heat), just below the red, 

 which we have added to make more evident the tendency of the absorption to increase a.% the wave-lengths increase, 

 though for one and the same unit thickness. If there be any linear selective absorption in the wedge, it is not evident 

 in such observations as these. Of course throughout this investigation the as.scnt of the reader is assumed to the 

 recognized principle that a change in the rate of absorption of the heat in any ray implies a like change in the rate 

 of absorption of light in the same ray. 



