26 



SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 65 



, We call attention to the decided difference between the behavior 

 of nearly homogeneous rays, as observed by the bolometer, and of 

 the total radiation, as observed by the pyrheliometer. The logar- 

 ithms of the pyrheliometer readings of September 20 are plotted 

 against Bemporad air-masses in the upper curve of figure 3, and 

 the reader will readily perceive the pronounced and steady change 

 of curvature of the resulting plots. This is in sharp distinction to 

 the close approximation to straight lines shown in the logarithmic 

 plots of the bolometric observations at single wave lengths. Forbes, 

 Radau, Langley, and many others have discussed this relation 

 between total radiation and air-masses, and have shown why such 





