1882.] on Matter and Magneto-Electric Action. 79 



terminal, minus the length of the intervening flame. The per- 

 manently heated part of the flame will act in the same manner in 

 extending the effective length of the discharge. 



The discharge which we are now examining is not homogeneous 

 throughout, but consists of more than one layer. The flame, which, 

 from the fact of its forming the outer sheath of the discharge, is the 

 most prominent feature, consists mainly of heated but solid particles 

 emanating from the terminals. That this is the case may be inferred 

 in a general way from the colours which the flame assumes when 

 different substances are placed upon the terminals; for example, 

 lithium or sodium. The spectrum of the flame appears to be always 

 continuous. A convenient substance to affix to the terminals is boron 

 glass, on account of the brilliancy to which it gives rise in the dis- 

 charge ; this will enable us to project the phenomenon. Within this 

 sheath of flame, the discharge consists of the pink light characteristic 

 of air, and in the centre of all the true bright spark. There is reason 

 to think that, under certain circumstances, there are more layers to 

 be seen ; but the above division is sufficient for our present purpose. 

 In this somewhat complicated structure, the pink light corresponds 

 to the arc, and the flame to a similar accompaniment which is seen 

 playing about the upper carbon in electric lamps when a current of 

 great strength is used. 



From this account of the methods here employed I now turn to 

 the main question. In the investigation, to which allusion was made 

 at the beginning of this lecture, it occurred to us that an examination 

 of the effects of a magnetic field on discharges of this character 

 through air or other gases at atmospheric pressure, and a comparison 

 with those obtained at lower pressures, might throw some fresh 

 light on the nature of electrical discharges in general. It is these 

 phenomena to which 1 now propose to ask your attention. 



When the discharge, originally in the form of a vertical spindle, 

 is submitted to the action of a magnet whose poles are horizontal, it 

 spreads out into two nearly semicircular disks, one due to the dis- 

 charges in one direction, and the other to those in the opposite 

 direction. As the magnetism is strengthened, the flame retreats 

 towards the edge of the disks, and ultimately disappears. The disk 

 then consists mainly of the pink discharge ; but with a still stronger 

 magnetic field, it is traversed at intervals by bright semicircular 

 sparks at various distances from the centre. In every case, bright 

 sparks pass directly between the terminals at the opening of each 

 separate discharge. 



In order further to disentangle the parts of this phenomenon, 

 recourse was had in the original experiments to a revolving mirror. The 

 light in the disks is insufficient to allow of a projection of the effects, 

 but the accompanying diagrams represent the appearances seen in the 

 mirror. Fig. 1 shows the arrangement of the terminals and the 

 magnetic poles ; Fig. 2 the appearance of the discharges in a plane 

 at right angles to that of Fig. 1 ; Fig. 3 the appearance of three 



