156 



ANNUAL REPORT SMITHSONIAN INSTITUTION, 1912. 



correctly measure the relative intensities of rays of all wave lengths, 

 whether visible or not. 



The indications of the bolometer may be automatically observed 

 by photography, and thereby the solar spectrmn may be exhibited, 

 as in figure 1, as a sinuous line whose elevation above the base 

 line of zero radiation gives the relative intensity of the different 

 colored or invisil)le ra5^s. The two curves shown are taken independ- 

 ently about an hour apart, for the purpose of studying the increase 

 of intensity of the sun's rays of different wave lengths as their 

 path in air dmiinishcs in length in consequence of the approach of 



090 



osc 



0.6C 



0.50 



C-Wi 



Fig. 2.— Vertical atmospheeic transmission for different wave lengths. 

 Upper curve for Mount Wilson; lower curve for Washington. 



the sun to the meridian. From such studies the results shown in 

 figure 2 are found. The upper curve represents the percentage 

 transmission of a vertical column of air above Mount Wilson (ele- 

 vation 1,750 meters), while the lower curve shows the less transmis- 

 sion for vertical rays at Washington (30 meters). The wave lengths 

 are indicated as abscissae. From this we see how much more loss the 

 violet rays of wave length 0.40/( suffer than do the red rays of wave 

 length 0.70/jL in traversing the air. 



In order to determine the quantity of the solar radiation, we 

 must fix our conditions independent of the variable losses in the 



