662 



BIKKKLAND. THK NORWEGIAN AURORA POLARIS EXPEDITION, IQO2 1903. 



b/J 

 U. 



sun, stood very nearly in the plane of the sun's equator. This would 

 have to be upon the assumption that in the spaces far from the sun 

 there is a gas that can become electrically luminescent, or, in an electric state 

 able to reflect sunlight. 



It is possible to believe, however, that the sun's chromosphere, which 

 is a sharply-defined envelope of hydrogen, is again surrounded by an 

 envelope of coronium, of almost limitless extent. 



Analogies from the earth's atmosphere, whose nature has been mail- 

 clearer through the latest researches of HANN, HUMPHREY (') and WEGKXKK, 

 seem at any rate to indicate that the above-mentioned assumpti< 

 probable. 



Wegener has recently ( 2 ) shown that there must be new fundamental 

 layer-limits in the earth's atmosphere. Above a covering of hydn 

 which prevails from a height of 75 to 200 kilometres, a new gas is tn 

 be found, which he calls geocoronium, extending up to such heights that 

 the steady auroral arcs, for instance, that are observed as much as 600 

 kilometres above the earth, would be due to electric luminescence in 

 this gas. 



129. We will now pass on to experiments that in my opinion have 

 brought about the most important discoveries in the long chain of exp 

 mental analogies to terrestrial and cosmic phenomena that I have produced. 

 In the experiments represented in figs. 248 a e, there are some small 

 patches on the globe, which are due to a kind of discharge that, under 

 ordinary circumstances, is disruptive, and which radiates from points on 

 the cathode. If the globe has a smooth surface and is not magm \ 

 the disruptive discharges come rapidly one after another, and are distril- 

 more or less uniformly all over the globe (see a). On the other hand, 

 if the globe is magnetised, even very slightly, the patches from which the 

 disruptive discharges issue, arrange themselves then in two zones parallel 

 with the magnetic equator of the globe; and the more powerfully the globe 

 is magnetised, the nearer do they come to the equator (see b, c, d|. With a 

 constant magnetisation, the zones of patches will be found near the equator 

 if the discharge-tension is low, but far from the equator if the tension is high. 

 Fig. 248 e shows the phenomenon seen from below. 

 If the pressure of the gas is very small during these discharges, 

 there issues (fig. 249, globe not magnetised) from each of the patches 

 narrow pencil of cathode-rays so intense that the gas is illuminated all 

 along the pencil up to the wall of the tube. This splendid phenomenon 

 recalls our hypothesis according to which sun-spots sometimes send out 

 into space long pencils of cathode-rays. 



SCHUSTER has recently ( 3 ) made some serious objections to the hypo- 

 thesis that sun-spots emit direct, rather well-defined pencils of cathode-rays, a hypothesis which was put 

 forward by me in 1899 and 1900, and by MOUNDER in 1904. 



(') HUMPHREY, Distribution of Gases in the Atmosphere. Bull, of the Mount Wealher Observatory, II, 2. 

 ( 3 ) WEGENER, Zeitschrift fiir anorganische Chemie, B. 75, p. 107. 1913. 

 ( 3 ) The Origin of Magnetic Storms. Proc. of Roy. Soc., 1911. 



