118 ANNXTAL REPORT SMITHSONIAN INSTITUTION, 1931 



As stated in previous reports, much effort had already been ex- 

 pended in reducing the observations made at Table Mountain, Calif., 

 but without satisfactory results. The atmosphere above Table Moun- 

 tain, though to the eye appearing very fine and clear, contains va- 

 riable amounts of ozone, water vapor, and dust, which produce 

 embarrassing difficulties in computing the solar constant of radiation. 

 Daily measurements of the amount of atmospheric ozone by the 

 method of Dobson had been in progress at Table Mountain, since 

 August, 1928, but they require fully as much time for reduction as 

 does the solar constant itself. Fortunately, as described in last year's 

 report, we were able to devise a simple method based on our holo- 

 graphic work whereby corrections can be made easily for the absorp- 

 tion of ozone on all days when solar-constant measures are made at 

 Table Mountain. All the Table Mountain solar-constant values from 

 the beginning there in 1925 have now been corrected for ozone 

 absorption. 



The changes of haziness and of absorption associated with varia- 

 tions of atmospheric water vapor make a difficulty of a more serious 

 nature. After several unsuccessful attempts to vary the Montezuma 

 procedure to suit Table Mountain conditions, the process of reduction 

 of the short-method solar-constant determinations at Table Moun- 

 tain was radically changed. It will be recalled that the essence of 

 the short method consists in employing pyranometer measurements 

 of the brightness of the sky near the sun as an index of the prevailing 

 atmospheric transparency. 



If the brightness of the sky were unaffected by varying quantities 

 of smoke or dust, we should expect the normal change of its bright- 

 ness from day to day to be exactly determined by the quantity of 

 atmospheric water vapor prevailing. In other words, there would be 

 a normal relation between pyranometry, precipitable atmospheric 

 water vapor, and atmospheric transparency, for the different wave 

 lengths. But if unusual degrees of dustiness or smokiness prevail, 

 then the pyranometer will record a positive or negative excess from 

 the normal value proper to the prevailing quantity of precipitable 

 water. This " excess " will be associated with changes in the atmos- 

 pheric transmission coefficients for all wave lengths. 



On these lines we have worked out new varieties of the short 

 method of determining the solar constant of radiation applicable to 

 conditions at Table Mountain and Mount Brulvkaros. We have re- 

 reduced all the observations made at these stations according to these 

 new methods. Great improvement in their solar-constant deter- 

 minations resulted, although it must be confessed that neither of 

 these two stations yields results as generally satisfactory as does 

 Montezuma. 



