of Magnesium , Copper, and Manganese. 303 



direction normal to the plane of: this oblong, its path is not a 

 closed one, hut forms a flattened helix between the negative 

 and positive ends of the graphite plate. If, instead of a 

 single particle, there be an uninterrupted and abundant 

 supply of them, such as would most likely be furnished by 

 the thermelectronic current from the small heap of carbor- 

 undum, the stream of particles would probably form a layer 

 of appreciable depth; and their combined system of paths, 

 each of which forms a helix, would wind itself round the 

 graphite plate like a screw-thread. This path system would 

 probably be confined to the region of the red fringe, for as 

 soon as a particle has passed out of the latter, it will 

 recombine to form a neutral molecule of luminous vapour. 

 In those places where the helix passes near the border of the 

 protected space or actually out of it, the electrified particles 

 travelling along it will come into contact with relatively 

 cool air and form a stria of bluish vapour. Evidently there 

 will be as mm y strise as there are turns in the helix. Hence, 

 on this view, the ions as they move in the direction of the 

 electric field are whirled round the plate of graphite by 

 the magnetic field due to the heating current, and the bluish 

 striae represent those portions of a helical path which cross 

 the boundary surface of the protected space. 



§ 4. Possible causes of the formation and downward drift 

 of the luminous vapour in the protected space. 



It has been shown that the emission centres of the 

 luminous vapour travel down to a considerable distance from 

 the electrically heated plate of graphite, even when a strong 

 magnetic force is acting upwards ; in fact, they behave in 

 this respect like neutral molecules. Their presence far 

 down in the protected space cannot, therefore, be attributed 

 solely to the action of the magnetic field established round 

 the plate by the heating current. Without for the present 

 attempting to find a definite solution of the problem, it may 

 be useful, however, to call attention to one or two possible 

 explanations. Thus the luminous vapour might originate 

 from the recombination of metal and carbon atoms near the 

 outer edges of the helical path, especially in those places 

 where the force of affinity between the atoms is great 

 enough to counteract the decomposing action of the therm- 

 electronic current. If large quantities of ions pass through 

 the fringe, it is conceivable that many of them will be pushed 

 out of the fringe owing to overcrowding. Once out of the 

 sphere of action of the thermelectronic current they will 

 immediately reassociate to form neutral molecules, and their 



