322 peitchard's wedge photometer. 



The lower line, " Mean Value of (c)," exhibits in the constant increment an inter- 

 esting proof of the correctness of a theory elsewhere applied to the results of 

 atmospheric absorption, and these prismatic observations then, on the whole, while 

 extending and elucidating those in the direct beam, are in strict accordance with 

 them. 



Conclusion. 



It appears from all the preceding observations, and from the established principle 

 that changes in the thermal and optical effects are proportional in any one ray, that 

 there is a selective absorption of light throughout the wedge, even in the visible 

 rays ; feeble in the more luminous portion of the spectrum, but of such a character 

 that, broadly speaking, the transmissibility always increases from the violet towards 

 the red. That it increases, and very greatly, beyond the red, is shown clearly 

 by the additional measures we have given at P.O in the infra-red spectrum. 



Though the eye is incomparably more sensitive than the bolometer, the latter 

 is probably able to discriminate much smaller differences quantitatively. It is 

 therefore possible that the selective absorption so apparent to the bolometer even 

 in the luminous spectrum may not be easily sensible ocularly, particularly as this 

 absorption is seen to be less in the portions which are most effective in vision, such 

 as the yellow, than in the red. 



These observations, then, demonstrate that selective absorption of a quite definite 



character and amount exists in the submitted example of the Pritchard wedge, even 



in the visible rays ; but they are not to be understood as necessarily proving that 



it does so in any degree very prejudicial to the special object of the instrument. 



S. P. LANGLEY, 



Director of the Allegheny Observatory. 

 August, 18S6. 



The unexpected character of the results obtained by Professor Langley, and the 

 importance of the conclusions to be derived from them, render it desirable that 

 they should be checked by some wholly independent method. The wedge to the 

 eye is of a nearly perfect neutral tint, and might be expected to transmit equally 

 rays of all wave-lengths. The bolometer, however, shows that the variation is very 

 great even in the visible spectrum. The proportion transmitted corresponding to a 

 length of li inches, is 0.137 for blue light, \ = 0.5, and 0.252 for red light, X = 0.7. 

 This quantity corresponds to a difference of 0.44 of a magnitude per inch in the 



