263 



opinion on the nature of those currents ; for my part, I think they 

 are induced currents returning on their own path. This opinion was 

 supported by the subsequent experiments. 



I got a hollow and highly exhausted sphere of glass about 2 inches 

 in diameter, into which no wire entered, and placed it on the iron 

 pieces of the electro-magnet. On touching the sphere with one 

 of the wires of RuhmkorfFs apparatus, diffused light was spread 

 through it. By the excited electro-magnet this diffused light was con- 

 centrated into a luminous stream, which moved along the magnetic 

 curve determined by the point touched, from this point to another one 

 in which the interior surface of the sphere is intersected a second time 

 by that curve, and evidently returned in the same curve from the second 

 point to the first. When the point where the sphere is touched by 

 the wire falls within the equatorial plane, no part of the magnetic 

 curve in question, which in this case is a tangent to the sphere, enters 

 it. In this peculiar case, the phenomenon is entirely changed, the 

 light of the induced current constitutes an ascending or descending 

 stream proceeding along the epibolic curve within the equatorial 

 plane. In reversing either the polarity of the magnet or the 

 direction of the discharge, the ascending stream is transformed into 

 a descending one, or vice versd. But there is no change of the 

 appearance in the general case. 



When the sphere is simultaneously touched by the two wires of 

 Ruhmkorff s apparatus in two different points, we get within it two 

 independent luminous currents showing no reciprocal action on each 

 other ; and it is only when both points belong either to the same 

 magnetic curve or to the epibolic one, that there is but one luminous 

 arch along that magnetic or the epibolic curve. 



All this shows that Gassiot's induced streams are subject to the 

 same laws as the direct currents are, formerly observed by myself. 



In all experiments mentioned in this section, the discharge of 

 Ruhmkorff 's apparatus may be replaced by a spark taken from the 

 conductor of an electrical machine. 



V. Fluorescence produced by the electric discharge. 

 Glass shows in sunlight scarcely any kind of fluorescence. A glass 

 tube, including residual traces of a proper gas, becomes highly 

 fluorescent when the electric discharge is sent through it : a fine green 



