260 



which two platina wires, either parallel or perpendicular to each 

 other, were conducted, whose shortest distance was about 0'8 of an 

 inch. 



When, for instance, the sphere with parallel wires is put on the 

 iron pieces of the electro-magnet so that both wires fall within the 

 equatorial plane and are vertical, the whole circular section of the 

 sphere passing through the negative wire is almost uniformly illu- 

 minated by violet light, while the light of the positive electrode 

 appears at one of its extremities, whence it moves, along an epibolic 

 curve, to the corresponding extremity of the negative wire. On re- 

 versing the polarity of the magnet, the illuminated epibolic curve 

 passes from one extremity of the positive wire to the other, while the 

 appearance of the negative light, after the reversion, is not at all 

 altered. 



When the positive wire terminates within the sphere, we get, 

 according to the magnetic polarity, either an epibolic curve, or dif- 

 fused light, starting from the free extremity of the positive wire 

 towards the surface of the circle illuminated by the negative light. 



When the sphere with crossed wires is put on the iron pieces so 

 that the negative wire becomes vertical, the positive one horizontal 

 and equatorial, the whole surface of the axial circle passing through 

 the vertical wire is illuminated, except the lower part, which is 

 bounded by the magnetic curve starting from the lower extremity of 

 this wire. There is no light seen on the positive wire, which is in- 

 tersected by the magnetic circular surface. On this surface the 

 shadow of the positive wire is most distinctly traced. [Shadows of 

 this description are, in the general case, produced by beams of light 

 starting from all points of the surface of the negative electrode, and 

 moving along the corresponding magnetic curves. Not even the po- 

 sitive wire deviates such beams of light from their curved paths.] 

 Nothing at all is changed by a change of magnetic polarity. 



When the sphere is turned round its vertical diameter till the 

 horizontal and positive wire passes from the equatorial direction into 

 the axial, the whole surface of the axial circle passing through the 

 negative wire is filled with illuminated magnetic curves. The positive 

 light starts from the middle of the horizontal wire, and moves round 

 it, within the equatorial plane, towards the negative wire without 

 interfering with the light emanating from this wire. It constitutes 



