NO. 13 SCALE OF SOLAR RADIATION ABBOT AND ALDRICH 3 



resistances of the same coils in 1932. The readings of current on the 

 niilliammeter ranged between 50.0 and 63.0 on its scale, and could 



usually be read to somewhat better than -; — , possibly to . 



^ 600 -^ looo 



The milliammeter. with its i^osition and surroundings unchanged, 

 was calibrated against two standard cells which were obligingly checked 

 against standards between two of our comparisons by a member of the 

 stall of the California Institute of Technology, and found highly ac- 

 curate. We believe that we have measured the pyrheliometer currents 



by these methods to better than , possibly t(^ — — . As the current 



500 1000 



enters the formula in its square, the error due to inaccuracy of cur- 

 rent measurements may reach 0.5 percent but is believed to be less. 

 Accidental errors of the standardization of silver-disk pyrheliometer 

 S.I.5bis against the water-flow pyrheliometer are to a great extent 

 eliminated by the numerous repetitions of the comparisons. 



We now give in the following table a summary of the results of 

 the standardization. 



This mean value is 2.^ percent below the value 0.3715, which, as 

 stated in our former paper, represents the Smithsonian scale of 1913. 

 The defect 2.3 percent is in close agreement with the defect 2.5 per- 

 cent indicated by our results of 1932. But on account of the numerous 

 observations and highly satisfactory conditions prevailing in 1934, 

 we prefer to attach full weight to the correction 2.3 percent. The 

 causes which may have contributed to the higher scale found in 1913 

 are fully discussed in our former paper. 



