80 



REPORT OF THE SECRETARY. 



ferent days. This aii)iti'ary method of i-ediiction is necessary, for differences 

 of atmospheric transparency and differences of solar distance make a direct 

 comparison of one ohservational curve with another complicated and unsatis- 

 factory. 



Plate VI shows two of the solar curves plotted with ahscissjie as air-masses 

 derived on the above assumptions, and ordiiiates as the logarithmic deflections. 



The general absorption of the solar envelope appears to be like that of the 

 earth's atmosphere, greatest at the violet end of the spectrum. This is shown, 

 and also some of the changes which have been noted in the apparent solar trans- 

 mission, in Table o. 



In this table the numbers purport to represent the percentage transmission of 

 the solar envelope for vertical rays, on different days, and for various rays 

 between wave lengths 0.4w in the violet, and 2.0/^ in the infra-red. The results 

 are based on the assumptions stated above, and a considerable difference will be 

 noted from the corresponding table of last year, owing to a change in the 

 assumed thickness of the absorbing envelope. 



