Bigelow. — The Inversion of Temperature Amplitudes. 115 



Art. VIII. — The Inversion of Temperature Amplitudes 

 and Departures in the United^ States; b}' Frank H. 

 Bigelow. 



The Phenomenon of Temperature Inversion. 



In several of mj papers on the temperatures of the United 

 States, I have given prehminary data showing generally that 

 there is a distinct tendency to invert the temperature varia- 

 tions in the United States relatively to the solar conditions as 

 given in the annual numbers for the frequency of the sunspots 

 and solar prominences. These are, this Journal, December 

 1894, inversion of temperatures in the 26"68 day solar magnetic 

 period ; AVeather Bureau Bulletin JSIo. 21, 1898, Solar and 

 Terrestrial Magnetism, pp. 121 and foil.; Monthly Weather 

 Review, November 1903, Synchronism of the solar promi- 

 nences with the terrestrial barometric pressures and the tem- 

 peratures. These papers indicated that the response to a 

 change in the solar energy when received at the earth is by no 

 means simple, but that the terrestrial effect is much com- 

 plicated by the circulation of the atmosphere, whereby 

 opposite temperature effects are produced in different latitudes. 

 Thus, an increase of solar action produces an increase of tem- 

 perature in the Tropics, but a relative decrease in the 

 temperate zones. This view was explained more fully in two 

 papers, this Journal, May, 1908, and April, 1910. Bulletin S, 

 U. S. Weather Bureau, 1909, contains the temperature data 

 of the United States, 1873-1905, reduced to a homogeneous 

 system, which is continued in the Annual Reports of the Chief 

 of the Weather Bureau, and these temperature data have been 

 used in preparing. Table I of this paper. We have the 

 independent records of upwards of 100 stations from which 

 the annual departures on the 33-year normal are computed. 

 These annual departures were entered on base maps of the 

 United States, and lines of equal departures were drawn. 

 This system of departures for the United States is so smoothly 

 developed that it is very easy to draw the lines of equal tem- 

 perature variation. There is one map for each month, and 

 one for the year, from January, 1873, to December, 1909, mak- 

 ing 37x13^481 separate charts, upon each of which there are 

 something like 100 entries of the temperature departures. 

 An example is given for May, 1896. Each entry on fig. 1 is 

 found by subtracting the 33-year normal of the station from 

 the monthly mean given in the Bulletin S table for the 

 several stations. An inspection of these 100 independent 

 values of the departure shows that there is a remarkably 

 harmonious agreement among them, which implies that the 



