upon the Electrical Discharge in Rarefied Gases. 135 



intervals are very often brought into existence by the magnet, 

 whose attractive or repulsive power concentrates the diffused 

 light. I imagine that this concentration of the light-stream 

 must be considered as the cause why the beautiful blue light in 

 a tube containing sulphurous acid, which had remained quite 

 unchanged during a long-continued succession of discharges 

 from Ruhmkorff's apparatus, on placing the tube upon the 

 magnet, was in a very short time converted into a washed-out 

 green, a colour corresponding to that produced by oxygen 

 gas. It seems to me as yet improbable that the magnet should 

 exert a direct influence upon the decomposition which takes 

 place. 



17. A tube containing bromine, 200 milhms. long and 10 

 millims. wide, teraiinated at both ends with bulbs, gave rise to 

 reddish and violet flashes, which, forming a pencil, continually 

 traversed the axis of the tube from one end to the other, being 

 surrounded by a difiiise green light. On placing the tube 

 equatorially upon the magnet, the flashes were, when the cur- 

 rent passed in one direction, drawn down so that they traversed 

 the lower part of the tube without rupture. They underwent, 

 however, thereby a change of colour, becoming bi'ight bluish 

 green, while a red mist appeared in the upper part of the tube. 



18. In another case, a short tube, terminating in a wider 

 bulb, was laid upon the approximated armatures. Before the 

 magnet was excited, the light on entei'ing the bulb, which rested 

 sideways on the two armatures, was dispersed in all directions : 

 after the magnet was excited the bulb was traversed by green 

 and reddish flashes. 



19. It being my intention to study the action of magnetism 

 upon the light which appears at the warmth-pole and which is 

 separated by the dark space from the true light-stream, I 

 brought appropriate tubes with their warmth-electrodes into 

 the neighbourhood of the pole; and I shall now consider the 

 new class of phtenomena which I thereby observed. In order 

 to see clearly the phaenomena to be described, it is necessary 

 that the electrode, where the warmth-pole has to be situated, 

 should penetrate far enough into the tube and that the tube be 

 sufficiently wide at this end. 



20. I chose, in the first place, a tube (fig. 5) 10 millims. 

 wide, upon whose two extremities glass bulbs of 35 millims. 

 diameter are fused. The length of the entire tube is 250 

 millims. The platinum electrodes extend into the bulbs as far 

 as their centres. Upon the electro-magnet two armatures were 

 placed, whose section was a square having a side of 70 millims. 

 The edges of the two armatures which were turned towards one 

 another were rounded off, and in the naiddle of each an obtuse 



