122 Bujelow — The Inversion of Temperature Amplitudes 



per square centimeter per inimite, that is the actual energy of 

 the inconiinsi; solar radiation, and wo must for some years be 

 content with tlie avaihihlc (hita. The solar prominence num- 

 bers are from the Italian spectroscopic observations on the 

 edge of the solar disk, and these are only rough records of the 

 total action of the solar mass as an emitter of radiant energy. 

 Furthermore, the prominences break out earlier in the middle 

 latitudes than near the equator or the poles, so that the general 

 sumnuition of prominence numbers obscures the times of free 

 efficient emission to the earth. It is evident that systematic 

 spectroheliograph observations of the face of the sun should be 

 made in favorable climates, and the records should be care- 

 fully studied for the variations of the emission of solar energy 

 by months and by years. The European magnetic field is rep- 

 resented by the mean horizontal force variations at a few sta- 

 tions, three to five, whereas the mean magnetic variation should 

 be computed for all three components at about 10 stations for 

 the entire earth. It is proper to infer that the precise 

 synchronism shown between the prominences, sun-spots, and 

 magnetic field in the long period can be extended to the short 

 period by the expenditure of sufficient well-directed labor. 

 The temperature curves match the short-period variations of 

 the solar curves ajjproximately, that is enough to recognize the 

 same sequence of crests for each. We have shown that the 

 temperature data is in itself of the highest order. On the 

 other hand, it suffers from considerable iinsteadiness as regards 

 its synchronism with the solar energy on account of the inversion 

 effects produced by the atmospheric circulation. The varia- 

 tions of temperature are direct with the solar energy in the 

 Tropics, but inverse in the temperate zones. It follows that 

 the liigh pressure belt is the node of this temperature effect, 

 where the variation is indifferent. Since this belt is due to 

 the circulation, and since the location of the belt in the United 

 States is loosely defined, it is evident that some uncertainty in 

 the temperature amplitudes and excesses must be expected. 

 This high pressure belt usually runs from the Florida peninsula 

 to northern California, but it is much broken up by the anticy- 

 clonic and cjclonic circulation through which the air from the 

 Tropics mixes with that from the temperate zone. 



It is easy to see that if these several curves could be suitably 

 perfected, it would not be difficult to assert that certain com- 

 ing years would be prospectively warmer or colder than usual. 

 The years of maximum solar energy are years of minimum 

 temperature and minimum amplitude. An increase of energy 

 from the sun on the Tropics brings down colder air in the 

 temperate zone and holds it more fii-mly at the surface of the 

 earth, and the cold high areas are in excess, the winter and 



