304 CARNEGIE INSTITUTION OF WASHINGTON. 



From a preliminary investigation in 1909, it was found that an 

 increase in solar activity, as judged by sun-spot frequency, was in 

 general accompanied by a decrease in the so-called local magnetic 

 constant which, under certain assumptions, is proportional to the 

 intensity of the magnetization of the Earth. The present investigation 

 examines into the possibility of a relationship between fluctuations in 

 the solar radiation and certain changes of the Earth's magnetism. Use 

 was made for this purpose of Abbot's extensive series of solar-constant 

 values. 



As a provisional result, it was found that for a change of 10 per cent, 

 such as may be observed in the value of the solar constant, there is 

 apparently a change in the magnetic constant, above referred to, of 

 about 0.03 per cent of its value, decreased magnetic constant corre- 

 sponding to increased solar constant. The effect is most pronounced 

 in the sun-lit portion of the globe, and hence may var}^ according to 

 local apparent time, just as a certain class of magnetic disturbances 

 is known to do. It is furthermore of interest, in this connection, to 

 recall the class of magnetic disturbances which it v/as found did not 

 occur simultaneously over the globe, but, instead, each time were 

 recorded only at the observatories in the daylight zone. 



While the discovery that the Earth is not the only rotating celestial 

 body enveloped by a magnetic field may assist in solving the vexed 

 problem as to the origin of planetary magnetism in general, possibly 

 especial progress may be expected from the discovery of the agencies 

 responsible for the space and time variations in the particular magnetic 

 field under consideration. Each body may present characteristics of 

 its magnetic field having no counterpart elsewhere. Accordingly the 

 Department studies are at present directed mainly to the study of the 

 variations of the Earth's magnetic field and of the causes which make 

 it depart from the simple type generally postulated by theories. It 

 is believed that a long step forward has been taken toward the dis- 

 covery of the origin of the Earth's magnetism when once we thoroughly 

 understand what causes it to vary in the surprising manner shown, for 

 example, by the secular or long-period changes, by magnetic storms, by 

 the diurnal variation, etc. However, before long these studies may 

 be supplemented by certain laboratory investigations. 



An investigation pertaining to the variations of the Earth's mag- 

 netism was made during the past summer by Messrs. G. N. Armstrong^ 

 and C. R. Duvall. The title of this research is "On the study of 

 methods of seeking hidden periodicities, with some applications to 

 magnetic and sun-spot data." The main part of the work was devoted 

 to recurring-series methods, which were first investigated b^^ Lagrange 

 in 1772; also a brief study of some other methods was made. In his 



^Associated with the Department as mathematician, June 26-September 10, 1914. 



