May 19, 1 9 10] 



NA TURE 



i57 



he says, normally largest in the earth's equatorial regions, 

 where they consist mainly of a change in the horizontal 

 force, but they are also well marked in temperate latitudes. 

 The cause postulated by Birkeland is a circular electric 

 current in the plane of the earth's magnetic equator, at a 



/30- 



ir^/r^ 



KjUiT 



C-liS70 



Fig. 2. 



height of several thousand miles. An objection to this 

 explanation is that, according to Prof. Carl Stormer's ' 

 analysis, it is impossible for kathode rays emanating from 

 the sun to reach the earth's atmosphere at all, except in 

 a narrow band round each magnetic pole. The larger the 

 mass and the greater the velocity of the 

 particle for a given electrical charge, the 

 nearer can it approach the earth in the 

 equatorial plane, and the larger is the 

 radius of the zone surrounding each mag- 

 netic pole within which the particle can 

 actually reach the earth. The /3 particles 

 of radium, from their higher velocity, have 

 more penetrating power than ordinary 

 kathode rays, and are, in their turn, 

 eclipsed by the a raj's, the lesser velocity 

 of which is more than compensated by 

 their larger mass. According to Stormer, 

 the greatest angular distance from a mag- 

 netic pole at which average kathode rays 

 emanating from the sun can reach the 

 earth is only 24°, while the corresponding 

 angular distances for /3 and a rays are 

 respectively 4-1° and 12-7°. 



Undeterred by these mathematical re- 

 sults, Birkeland assumes that a type of 

 magnetic disturbance, which he calls the 

 " polar elementary " storm, is due to 

 kathode rays from the sun which get 



within a few hundred kilometres of the . ^ 



earth's surface at considerable distances 

 from a magnetic pole. The paths of 

 approach and retreat are supposed to be 

 radial, and the connecting part horizontal. These 

 " polar elementary " storms were observed on a good 

 many occasions at four temporary- Arctic observatories 

 provided with magnetographs, which Birkeland was 

 1 Archiz'ts <{csScien<es fhysiques et KatureUes., Gentva, 1907. 



NO. 2 1 16, VOL. 8^1 



a'ole to keep in action during the winter 1902-3. The 

 characteristics of " polar elementary " storms are their 

 comparatively simple character and short duration, and 

 the fact that their amplitude — unlike that of Birkeland 's 

 " equatorial " storms — is much larger in the Arctic than 

 elsewhere. These storms have at least 

 /3o' r- a general resemblance to a special tj'pe 



of disturbance ' of which I found 

 numerous examples in the records of 

 the National Ajitarctic Expedition of 

 1901— 4. 



Birkeland found that frequently, after 

 an " equatorial " storm had been in 

 progress for some hours, one or a series 

 of " polar elementary " storms inter- 

 vened. He obtained copies of the 

 curves taken at a number of observa- 

 tories on the days of the disturbances 

 recorded by his Arctic stations, and he 

 has reproduced these with his own 

 records in a most valuable series of 

 plates published in his recent monu- 

 mental work, " The Norwegian Aurora 

 Polaris Expedition, 1902-3," vol. i. 



Whilst recognising to the full the 

 devotion with which Prof. Birkeland 

 has prosecuted his investigations into 

 magnetic storms for more than a decade 

 of years, and while admiring the beauty 

 of his experiments, I have to admit that 

 I do not find his explanations con- 

 vincing. If " equatorial " storms are 

 due, as he believes, to electric currents 

 at great heights above the earth in the 

 magnetic equator, the disturbing force, 

 while approximately horizontal and in 

 the magnetic meridian at places near 

 the magnetic equator, should, even in 

 temperate latitudes, have a considerable 

 vertical component, and near the mag- 

 netic poles the vertical component 

 should largely predominate. I am un- 

 able to see these phenomena in the 

 curves of Birkeland's own plates. 

 Further, during the time of Birkeland's Arctic expedition 

 the Discovery was at work in the Antarctic, and the 

 simultaneous results obtained there do not seem capable 

 of explanation on his hypothesis. 



Fig. 3 affords one out of a number of examples of this. 



fyxa/rtA/ zt^ ifoi 



Fig. :?. — Magnetic Storm in Antarctic. 



It shows the declination (D), vertical force (V), and hori- 

 zontal force (H) traces at the Discovery* s winter quarters 

 on March 22, 1903, during a magnetic storm which forms 



1 "National Antarctic t- xpedition, 1901- 

 p. 186. 



Magnetx Observations, 



