and Departures in the United States. 



121 



The synchronism in the long period between the ttco solar 

 curves and three terrestrial curves cannot he questioned, that 

 for the magnetic field heing direct, and those for the temper- 

 ature system of the United States heing inverse. This result 

 confirms the data and conclusions of my papers of 1894, 1898, 

 1903, mentioned above. The synchronism between the curves 



Fig. 3. 



Fig. 2. Prominences, magnetic field, sun-spots and the temperature 

 amplitudes and excesses in the United States. 



of short period is not so well proncmnced as for the long- period, 

 and that for several reasons. The prominences and the mag- 

 netic curve are only approximate measures of the true solar 

 radiation output, for which they stand as substitutes. We have 

 no long-continued series of homogeneous radiation observations 

 from which to construct a curve representing the gram calorics 



