94 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1917. 



recent studies now enable us to construct. The one obstacle to com- 

 plete success which now seems insuperable is the lack of any means 

 to form an intense unabsorbed spectrum free from stray light, ex- 

 tending from 15 to 50 microns in wave length. 



2. AT MOUNT WILSON. 



The expedition of 1916 continued "solar-constant" and other ob- 

 servations at Mount Wilson until late in October. The expedition 

 was renewed late in June, 1917. Improvements in the supply of elec- 

 tricity and water to the station were completed in June, 1917. 



In 1916 many observations of the sky by day and by night were 

 made at Mount Wilson with the pyranometer. The plan was fol- 

 lowed from August to October of measuring with this instrument the 

 total solar radiation at a fixed zenith distance of the sun, and almost 

 simultaneously the total sky radiation over a fixed small area imme- 

 diately surrounding the sun. It seems probable that as the bright- 

 ness of the sky depends on the prevailing humidity and dust, and as 

 the radiation of the sun is diminished by presence of humidity and 

 dust, a method of combination of the two measurements may be 

 found, adapted to give approximately the " solar constant." When 

 computations are further advanced the matter will be tested. 



Restandardization of secondary pyrheliometers in 1916 against our 

 standard water-flow pyrheliometer indicated no change in their 

 constants. 



A vacuum bolometer was employed during a large part of the ob- 

 serving season. The sensitiveness was so much greater that con- 

 siderable improvement in the work on the investigation of the dis- 

 tribution of radiation over the sun's disk was possible. 



Redeterminations were made with great care on the form of dis- 

 tribution of the solar energy curve outside the atmosphere. New mir- 

 rors of stellite, a very hard nontarnishing alloy, were substituted 

 for the silvered mirrors of the spectrobolometer. It is hoped that 

 the work of 1916 will indicate conclusively how the sun's variations 

 affect the distribution of energy in the solar spectrum. 



SUMMARY. 



Preparation of apparatus and equipment for a new " solar-con- 

 stant " station of the Smithsonian Institution, now located at Hump 

 Mountain, N. C, led to valuable improvements in the bolometer and 

 the pyranometer, and to the invention and construction of a new in- 

 strument for avoiding computation in reduction of spectro-bolometric 

 observations. 



A long research on the transmission of long- wave rays by atmos- 

 pheric columns of known humidity and carbon-dioxide contents, 



