6 74 



BIRKELAND. THE NORWEGIAN AURORA POLARIS EXPEDITION, igO2 1903. 



spots. This possesses considerable interest, inasmuch as the sun-spots in reality represent phenomen 

 of long duration, and not brief discharges. 



With the large, silver-coated globe, it was thus very difficult to obtain point-discharges when the 

 globe was cathode; they were pre-eminently continuous discharges from the entire surface or large por- 

 tions of it. 



When, however, the globe is made the anode, and the metal walls of the box, which are compara- 

 tively rough, unpolished, cast or rolled plates, are the cathode, a perfect firework-display of point- 

 discharges takes place, in rapid succession, from the inner walls of the box. Not only were the p., 



Fig. 259. 



luminous, but long pencils of rays passed from the points (almost like a kind of lightning) in to the 

 globe. Glowing metal particles were often torn from the points, especially from the steel plates, whence 

 particles shot inwards along the path of the current. 



In fig. 258, only the foot-points are visible, for when the anode-globe was non-magnetic, the flashes 

 in towards the globe, though fairly powerful, were too brief and of too little intensity too be fixed upon 

 the photographic plate with the camera used. When, on the other hand, the anode-globe was magnetised, 

 the flashes became more intense (see fig. 259), and the points of discharge were congregated in the vicinity 

 of the magnetic poles of the globe. The discharge-rays gathered in two zones about the poles of t 

 anode-globe, as might be expected; but there also appeared a faint band of light, of which an indicatioi 

 may be seen, round the magnetic equator of the anode-globe. 



In order to obtain point-discharges with my globe-cathode of 24 centimetres' diameter, 1 took 1 

 hemispherical shells of aluminium, and had them "sand-blasted" outside at a glass factory in the mann< 

 employed in the production of ground glass. 





