164 ANNUAL REPORTS OF DEPARTMENT OF AGRICULTURE. 



The data secured by means of kites and balloons point to another 

 equally important conclusion with respect to the time that tempera- 

 ture changes occur at the surface and at altitudes between 1 and 2 

 miles above, viz, that the changes occur practically simultaneously 

 at both levels. While this conclusion is probably not final, it is in 

 strong contradiction to statements Avhich have been made in this 

 country to the eli'ect that the temi)erature changes at relatively high 

 levels foreshadow those which take place in low levels twenty-four 

 hours afterward. 



The temperature gradient — that is, the rate of decrease of tempera- 

 ture for each 100 meters of ascent — immediately before the so-called 

 heat thunderstorms does not accord with that called for by theory. 

 Likewise in heated terms, or the so-called " heat waves," the un- 

 usually high temperatures appear to be confined to the air strata 

 next the earth's surface and up to less than half a mile above. The 

 heat wave does not advance abruptly as a wall of high temperature, 

 but rather builds gradually over the region affected. 



ATMOSPHERIC ELECTRICITY. 



During the year apparatus was installed by means of which the 

 potential gradient on the kite wire could be read off conveniently. 

 This work, however, should be considered as merely preliminary to a 

 more complete determination of the ionic content and movement in 

 the free air. 



No change has been made in the magnetic work during the year. 

 Photographic registers of the declination, the inclination, and the 

 horizontal intensity are made daily, and the photographic sheets are 

 properly checked so that hourly values may be drawn therefrom 

 hereafter. 



SOLAR RADIATION. 



Measurements of the intensity of solar radiation and the percentage 

 of polarization of sky light have been made at Washington and 

 Mount Weather, as in previous years. A five-year series of observa- 

 tions at Washington was completed on April 30, 1910, and the results 

 have been summarized for publication in the Mount Weather Ob- 

 servatory Bulletin. 



In addition to computations of the value of the solar constant, 

 which are necessarily confined to observations obtained on days when 

 the sky is exceptionally clear, all the pyrheliometric measurements 

 made during the five-3'ear period, 7,350 in number, have been utilized 

 in determining the mean rate at which solar radiation is received at 

 Washington, D. C, with a cloudless sky for different angles of the 

 sun and with average solar declination for each month. 



These monthly and annual mean rates show greater departures from 

 the normal than do the computed values of the solar constant. The pyr- 

 heliometric observations obtained at European observatories during 

 the past twenty-six years also show marked fluctuations. It therefore 

 appears to be desirable to undertake a systematic study of the rate at 

 which solar radiation is received at different points in the United 

 States, and of the variations that occur in this rate from year to year. 

 For this purpose, pyrheliometric records as nearly complete as pos- 

 sible will be obtained at four or five stations, so located as to be fairly 



