l82 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 70 



2. The variations in terrestrial temperature depend upon varia- 

 tions in the evaporation of the ocean and corresponding cloud 

 formation and precipitation over the land. 



3. Variations in terrestrial temperatures depend upon volcanic 

 dust in the higher layers of the earth's atmosphere. 



4. The periodic temperature variations depend upon changes of 

 the ozone formation in the atmosphere. 



5. Finally, the hypothesis that the variations, for example in the 

 temperature and precipitation, depend upon variations of air pres- 

 sure and circulation of the atmosphere, which again depend upon 

 variations of the solar activity. 



The first named theory, namely, that variations in terrestrial tem- 

 perature are caused directly by variations in the same sense of the 

 solar radiation, has been put forth by a number of investigators ; 

 for example. Chambers, Newcomb, Abbot and Fowle (in the year 

 1908), and we find it still in the more recent investigations of 

 Arctowski and in part Huntington and many others.^ Remember- 

 ing Lockyer's spectroscopic investigations of the sun — in which he 

 showed with some certainty that the solar surface is warmest at 

 maximum of sun spots — it is surprising that it sho'uld still have been 

 thought that increased temperature of the earth at sun spot mini- 

 mum could be attributed to an increase in the output of solar radia- 

 tion. However, it was still conceivable that even if the real tempera- 

 ture of the sun increased it might be that the output of solar 

 radiation did not correspondingly increase. It might be considered 

 perhaps that formation of clouds or dust in the solar corona hindered 

 the outgoing radiation of the sun. But in the pyrheliometric and 

 bolometric measurements which were made by Langley and by 

 Abbot and Fowle after 1902, first in Washington and after 1905 

 on Mt. Wilson in America, and after their investigations on Mt. 

 Whitney in America, and in Algeria, it must be considered as having 

 been shown that the solar radiation which reaches the outside of 

 the earth's atmosphere experiences no variations which correspond 

 directly with the observed variations in the atmosphere at the sur- 

 face of the earth. These measurements indicate plainly that the 

 " solar constant " (that is, the solar radiation outside of our atmos- 

 phere) is considerably greater near the sun spot maximum than 

 near the sun spot minimum. Although the measurements do not 



^Huntington, however, later (1914-a) came to the view that the variations 

 in the solar activity cause first variations in the storminess, as we shall later 

 refer to more at length. 



