upon the Electric Current. 5 



magnetic polarity be inverted, it follows that after such an in- 

 version, exactly the same appearance is shown. 



The beautiful phsenomena first observed by me (see Phil. Mag. 

 for August and December 1858), as well as similar ones, after- 

 wards to be described, formed at the positive electrode, are 

 therefoi'e fully explained. 



82. (3) The discharge occurs between two given surfaces, or 

 between different portions of the same given surface ; so that the 

 elementary discharge-currents may, within certain limits, seek 

 their terminal points upon these surfaces. In this case also 

 experiment had preceded theory. The first observation of this 

 kind was made upon an evacuated ellipsoid, which formed the 

 central portion of a Geissler tube (63). If it was placed equa- 

 torially upon the approximated armatures, a vaulted arch con- 

 sisting of luminous magnetic curves was formed in its interior, 

 far removed from the two electrodes. After the eye had once 

 had its attention directed to these phsenomena, it recognized 

 them also under other and very different circumstances. 



83. It now only remains to make some preliminary remarks 

 concerning the application of Law II., which determines the only 

 possible path which the electric discharge, under the influence 

 of the magnet, may follow on the internal surface of the glass 

 tube ; that is, which determines the epipolar-magnetic curve. 

 Let us, for example, suppose that the internal surface has the 

 form of a triaxial ellipsoid, and that this lies with its middle 

 point above the middle, between the two horizontal armatures, 

 so that one of its tbin axes is vertical : it is clear that the section- 

 ellipse, in that portion of it which belongs to the principal 

 section which falls in the equatorial plane, belongs also evidently 

 to the curve in question ; for every magnetic curve which passes 

 from one of the two armatures to the other, cuts the equatorial 

 plane at right angles, so that the elements of that one of these 

 curves which cuts the ellipse, lie within the surface of the 

 ellipsoid. A regular current is only possible when it enters the 

 ellipsoid in one point of the ellipse and leaves it in another one. 

 This stream describes, then, on the ellipse between the two 

 points, one or the other path according to its direction (compare 

 the experiment of § II.). It is in consequence of a secondary 

 action (which always increases with the magnetic influence, and 

 is of very different intensity according to the different contents 

 of the tube) that the current, when drawn towards the magnet, 

 becomes disintegrated into an undulating light. This phseno- 

 inenon is conditioned by the occurrence of the action corre- 

 sponding to the third case (82) of Law I. ; and consequently the 

 iHuxplained (15) is rendered clear. If the current enters and 

 dej)arts tlirongh any two points which are not both points of 



