120 Bigelow — The Inversion of Temperature AmplUucles 



States is of one departure sign, while the Coast States, Atlantic, 

 Gulf of Mexico, Pacilic, in whole or in part, are of the oppo- 

 site sign. 



(2) The strength of the amplitudes in winter is about twice 

 as great, sometimes three times as great, as in the summer. In 

 the winter the sun and the warm zones withdraw to the south, 

 and the cold polar air moves southward. The strength of 

 amplitude means that alternately warm and cold areas of air, 

 warm cyclonic streams from the south, and cold anticyclonic 

 streams from the north, pass over the United States in a 

 turbulent circulation, the crests and hollows further apart in 

 some years than in others. If the mean amplitude is 029, then 

 the range in amplitude is from 426 in 1874 to 842 iu 1878, or 

 1907, or as per cent for the entire series. In some months the 

 maximum temperature departure is between 20° and 25° F., in 

 other months only 2° or 3° F. The fundamental question is 

 whether these variations are purely accidental or not, and this 

 will be taken up in the following section. 



(3) The excesses of temperature for the United States as a 

 whole are shown in the last column of Table I, the maximum 

 relative number being +470 in 1878 and —399 in 1875, indi- 

 cating that during some years the temperature is maintained 

 above the normal or average, and below the normal in others. 

 The important point is to correlate the amplitudes and the 

 excesses with solar phenomena in order to learn whether there 

 is synchronism annually or not, and if any to classify its laws. 



Synchronism between the variations in the numbers representing 

 the frequency of the solar prominences, the sun-spots, the Euro- 

 pean terrestrial magnetic field, and the amplitudes or excesses 

 of the temp>erature departures in the United States. 



The curves representing these fre(pcency numbers are 

 brought together in fig. 2. The curve of solar prominences 

 is plotted from Table 2, this Journal, May, 1908, p. 417 : that 

 for the European magnetic field is from Table 3, same paper ; 

 that for the sun-spot numbers is from Wolfer's frequency 

 numbers ; that for the temperature amplitudes is from the 

 15th column of Table 1 of this paper; and that for the tem- 

 perature excesses is from the last column of the same table. I 

 have not had an opportunity to extend the first two curves, of 

 the prominences and the magnetic field, beyond the year 1905, 

 as should be done. These curves are derived from relative 

 numbers, and they are plotted on independent scales. The 

 heavy mean line represents the consecutive means of the long- 

 period, while the variations between this line and the original 

 curve give the short-period residuals. It should be carefully 

 noted that the curves of the temperature amplitudes and ex- 

 cesses are plotted inversely to the three other curves. 



