On the theory of the Solar diurnal variation of the Earth’s magnetism

Abstract

1. The most likely explanation of the solar diurnal magnetic variation (conveniently denoted by S) has for more than a generation seemed to be that afforded by the "dynamo ” theory, originally proposed by Balfour Stewart and further developed by Schuster and the writer. It attributes S to overhead electric currents, induced by convective motion of the air across the earth’s magnetic field. The form and intensity of the current-system can be inferred from the spherical harmonic analysis of S; J. Bartels, on the basis of the writer’s analysis, has drawn the current-diagram reproduced in the figure (p. 383). It will be seen that the current-system consists of four circuits, two north and two south of the equator, each flowing round a centre or focus. At the equinoxes, to which the figure refers, the system is approximately symmetrical with respect to the equator. The principal pair of circuits are situated in the sunlit hemisphere, and each circuit of this pair carried 62,000 amperes at sun­spot minimum; their foci lie approximately on the 11 hour meridian. The other pair of circuits, each carrying 32,000 amperes, are less intense but more extensive, covering about 15 hours in longitude. The equatorial current-intensity, per centimetre measured along the meridian, has an eastward maxi­mum of 2 .10 -5 e. m. u. at about 11h, and a westward maximum of 10 -5 e. m. u. at about 17h. 2. The dynamo theory can give a good account of the general form of this current-system, but meets serious difficulties when it attempts to explain the phase and intensity of the currents. This is because the convective motion which induces the currents, and the total conductivity of the layer in which they flow, are unknown. But if it is assumed that the convective motion is the same as at ground level, it appears that the day current-foci should be on the meridian 13h or 14h, distinctly later than the observed position. Further, the total electric conductivity (or ∫ σ dh , the integral of the specific conduc­tivity σ throughout the thickness of the layer) must be large, of the order 10 -5 e. m. u.