236 PROCEEDINGS OF THE AMERICAN ACADEMY 



Investigations on Light and Heat, made and published wholly or in part with 

 appropriation from the Romfokd Fund. 



CONTRIBUTIONS FROM THE PHYSICAL LABORATORY OF THE 

 MASSACHUSETTS INSTITUTE OF TECHNOLOGY. 



XIL — PHOTOMETRIC RESEARCHES. 

 By William H. Pickering. 



Presented Feb. 11, 1880. 



Although many forms of Photometer have been devised within the 

 last hundred years, little has been done towards measuring the bril- 

 liancy of our brighter lights, and nothing, so far as I am aware, 

 towards determining the relative intensity of their component colors. 

 With these two objects, and especially the latter one, in view, the 

 following determinations have been made. 



The first difficulty encountered was to obtain a steady light, of 

 which the ratio of the component colors should be constant, for use aa 

 a standard. Experiments were made using a platinum crucible con- 

 taining a salt having a high point of fusion. This was to be kept 

 just at the melting point, and thus a constant temperature would be 

 obtained, and consequently a constant light. Several difficulties were 

 encountered, however ; among others the cracking of the crucible, 

 owing to the alternate melting and solidification. And moreover the 

 lio-ht was not sufficiently brilliant to be wholly satisfactory, so that 

 the idea had to be given up. One interesting result, however, was 

 obtained. Theoretically, the same amount of light would be given 

 out in any direction by a curved surface, as by a flat one ; but it 

 was found that at the very edges, the curved surface was noticeably 

 more brilliant than in the middle. That this effect was not due 

 to contrast was shown by placing a brilliant background behind 

 the crucible, when, if contrast produced the effect, it would now 

 have been reversed. But no change in the result was noticeable. 

 Should this result be confirmed, it has an interesting application. 

 The opacity of the atmosphere surrounding the sun has been calcu- 

 lated on the supposition that the theoretical rule is correct; heuce, if 



