254 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1925 



lar<;^e magnetic storms frequently follow each other at an interval 

 approximating the time of revolution of the sun and that such re- 

 currence has been traced for several rotation periods, not every 

 recurrence being accompanied by a visible sun spot, requires some 

 such explanation as that suggested by Maunder. 



Chapman, after a detailed analysis of a number of magnetic 

 storms, attempted to show how some of their characteristics may be 

 produced by streams of electric particles entering the atmosphere 

 from the sun, but this theory, like so many others, failed to fit all 

 the facts. 



The researches of Hale at the Mount Wilson Solar Observatory on 

 the evidence of magnetic polarity in the sun spots occurring in pairs 

 and on the general magnetic field of the sun have further stimulated 

 the eli'orts to trace a connection between solar and terrestrial mag- 

 netism and opened the way for a study of the details of the sun's 

 magnetic field. According to the most recent analysis by Scares, 

 the sun's magnetic axis makes an angle of about 6° with the axis 

 of rotation. 



When we consider the diurnal variation of the earth's magnetism 

 a different problem is presented. Here we have to deal with a 

 phenomenon which is a function of local mean time, as contrasted 

 with magnetic storms, which, as we have seen, occur everywhere 

 at practicall}'^ the same absolute time. Attempts have been made 

 to correlate the diurnal variation with changes of pressure, tempera- 

 ture, humidity, and other terrestrial phenomena which have a period 

 the same as the earth's rotation, but without success. Broadly 

 speaking the diurnal variation is a function of the position of the 

 sun above the horizon, distinctly a local phenomenon. The ex- 

 tremes and the principal portion of the variation occur during the 

 daytime. During the night hours there is comparatively little varia- 

 tion. In view of this fact, it occurred to Bauer that the inter- 

 position of the moon between the earth and the sun at the time of 

 a solar eclipse might have an appreciable effect on the earth's mag- 

 netism. Accordingly he arranged for special observations by ob- 

 servers of the Coast and (ieodetic Survey at the time of the total 

 eclipse of May 28, 1900, at stations extending from Alabama to 

 Maryland. The results indicated a small but definite disturbance 

 associated witli the passage of the moon's shadow across the place 

 of observation and of the character to be expected. Similar observa- 

 tions have been made at all accessible solar eclipses since that time, 

 principally on the initiative of the Department of Terrestrial Mag- 

 netism of the Carnegie Institution of Washington, but with the 

 cooperation of other observers in the countries crossed by or con- 

 tiguous to the belt of totality. These have in general confirmed the 

 results of the first series, though the effect produced by an eclipse 



