390 Royal Society :— 



sible to dimmish the strength of the magnetic poles. The discharge 

 was then effected either in the open air or in a closed chamber. 

 The laller Mas constructed of a short cylinder of glass, say 3 

 inches in length and 2 in diameter, having conical ends pointed 

 inwards, so as to receive the poles of the magnet. The chamber 

 was also furnished with a pipe and stopcock for the purpose of 

 exhaustion. 



The discharge from an induction-coil taken in air or other 

 gas at atmospheric pressure consists, as is well known, primarily of 

 the spark proper or bright line, irregular in form and instantaneous 

 in duration. But beside this, when the primary wire is thick 

 and the battery-current strong, the spark is enveloped in a bright 

 cloud, or rather flame, which is capable of being thrown on one 

 side, although not entirely detached from the spark, by a current 

 of air. This, when examined in a revolving mirror, is found to 

 be subsequent in time to the spark proper, and may be considered 

 to be due to the gas in the neighbourhood of the spark becoming 

 sufficiently heated to conduct part of the discharge, and to the 

 consequent combustion of any extraneous matter floating in the 

 medium. Such a view is supported by the fact that the colour of 

 this flame depends partly upon the nature of the gas in which the 

 discharge takes place, and partly upon that of any volatilizable 

 matter which may be introduced near the poles. 



The exciting of the magnet produces upon the spark proper no 

 appreciable effect ; but as soon as the flame is submitted to its 

 action it is spread out into a sheet, which arranges itself in a heli- 

 coid right-handed or left-handed according to the direction of the 

 current and of the magnetic polarity in obedience to Ampere's law. 



Effects substantially the same are produced whether the dis- 

 charge be taken in gas at atmospheric or at a less pressure. But 

 in the former case the helix has a lower, in the latter a steeper, 

 gradient ; that is to say, in the former case it presents a greater, 

 in the latter a less number of turns for a given interval between 

 the poles. 



But for producing the best effects, both of the rotating spark 

 and also of the spirals, there is a limit beyond which the exhaustion 

 should not be carried. At a pressure low enough to produce strati- 

 fication, or even short of it, the whole chamber is filled with the 

 discharge, and all traces of rotation and of spirals are obliterated. 

 The stage best suited to the purpose is that in which the dis- 

 charge has thickened in diameter, and where the spark proper has 

 been replaced by a suffused light of the thickness, say, of a quill. If 

 the negative terminal be a surface (say the naked surface of the 

 soft iron pole of the electromagnet) instead of a point, the sheet 

 does not become contracted at the negative end, but remains 

 spread out and cuts the iron pole in a line radiating from the 

 point. 



Various gases were tried — atmospheric air, carbonic acid, ether, 

 chloroform, coal-gas, hydrogen. Of these the first two succeeded 

 best. With air the illumination of the flame-sheet was rather 



