through Rarefied Gases. 355 



G. Since Pliicker's time it was believed that, while the ne- 

 gative light sets in magnetic curves in the magnetic field, the 

 stratified positive light behaves like a solid conductor fixed at 

 both ends, and is accordingly deflected according to Ampere's 

 Lnv (Hittorf). I have found that the positive light behaves 

 in the magnetic field exactly like the negative light ; a smaller 

 force even is necessary to set the positive light into the curves 

 than that required to do the same for the negative light. 



In order to show this behaviour of the positive light, it is 

 only necessary that the vessel containing the gas should have 

 such a shape and position that its curvature at the points of 

 contact with the lines of magnetic force would be greater than 

 the curvature of the line of magnetic force at the same point. 

 Only when this condition is fulfilled can the discharge set into 

 the magnetic curve without touching the vessel. 



7. Only the immediate neighbourhood of the positive pole 

 seems to be distinguished by a peculiar behaviour in the mag- 

 netic field. At the negative pole the magnetic force, as shown 

 by Pliicker, gives rise to a sheet which is situated axially. 

 At the positive pole, under great exhaustion, a similar sheet 

 appears surrounding an equatorial electrode ; but its position 

 is equatorial. If the electrode has the usual form, the sheet is 

 of an oval form and consists of three parts. The first part, 

 which is the one next to the electrode, is nearly elliptical, and 

 does not give out any light ; the two remaining parts surround 

 the first one in two spiral curves. If the position of the posi- 

 tive pole is axial, a cylinder of light is seen which is also sepa- 

 rated from the metallic surface by a dark space. 



8. The form which the electric light assumes in the mag- 

 netic field depends on the quantit}^ of electricity which passes 

 in each discharge. Intense discharges, such as we obtain by 

 introducing air-spaces in the outer circuit or by means of 

 condensers, show phenomena which have not hitherto been 

 observed. Whatever the form of the tube is in which these 

 discharges take place, the phenomena are essentially the same 

 as in simple cylindrical tubes the electrodes of which coincide 

 with the axis of the cylinder. 



For simplicity's sake I refer, therefore, all phenomena to 

 such cylindrical tubes. I call an axial position any position 

 in which the axis of the cylindrical tube is parallel to the line 

 joining the magnetic poles. The tube may either be situated 

 above or by the side of the magnetic poles. The tube is sup- 

 posed to be longer than the distance separating the two poles 

 of the magnet. 



The discharges, under these circumstances, always take the 

 form of a Z, the plane of which is equatorial or axial, accox'd- 



2A2 



