259 



assume, if attached with one of its points in the point of the negative 

 wire. It is well known that a magnetic curve is completely deter- 

 mined by a single one of its points. Therefore the whole light, 

 starting from all the different points of the negative wire, will be con- 

 centrated within a surface generated by a variable magnetic curve. 

 The form of this " magnetic surface " varies according to the form 

 of the negative wire, and its position with regard to the poles of the 

 electro-magnet. When the negative wire lies within the equatorial 

 plane, the magnetic surface assumes the shape of a vault ; when the 

 wire lies within the axial plane, the whole surface is contained within 

 the same plane, and generally bounded by very well defined magnetic 

 curves. 



The negative light partly depends upon the substance of the wire. 

 Particles of it, either pure or combined with the included gas, are 

 carried off to the interior surface of the tube, which, when platina 

 wires are used, consequently is blackened. If not acted upon by the 

 magnet, all the part of the surface surrounding the platina wire 

 becomes black ; if acted upon, only that line along which the surface 

 of the tube is intersected by the magnetic surface is blackened. In 

 this case, therefore, the particles separated from the wire move along 

 magnetic curves. 



I think it most probable that the luminous electric currents in 

 question are double currents, going from the wire to the glass, and 

 returning from the glass to the wire. 



The importance of the use of magnetic curves, or lines of magnetic 

 force, in experimental researches, has been shown by several philoso- 

 phers, especially by Mr. Faraday. Hitherto only filings of iron 

 enabled us to give in peculiar cases an imperfect image of these curves. 

 We may now trace through space such a curve in the most distinct 

 way and illuminate it with bright electric light. 



III. The light of the positive wire and its spirals under the mag- 

 netic action. 



The origin of the current takes place at the positive wire. If the 

 negative wire is not too far from the positive, most striking pheno- 

 mena are obtained when the magnet is acting on the formation of 

 the electric current. In these experiments I made use principally of 

 highly exhausted spheres, about two inches in diameter, through 



