186 Dr. E. Goldstein on the Electric 



We have seen that a kathode may be regarded as a system 

 of fine conducting pores in a non-conducting surface. The 

 kathode light must then consist of rays which possess the pro- 

 perties of the light of very narrow tubes ; and in fact not only 

 does the colour of the kathode-rays agree with the blue of 

 narrow tubes, but the spectrum of the kathode-light consists 

 precisely of the same four bands, with the same maxima of 

 light, as that of the blue of the narrow tubes. 



The suspicion expressed on p. 183, as to the true character 

 of an apparently simple discharge between two metallic 

 electrodes appears to be supported by the mode of action of 

 the magnet upon the discharge. It follows, in fact, from it that 

 each negative bundle, as well as each positive layer, constitutes 

 a whole in itself. 



Each negative bundle, in fact, kathode-light as well as 

 secondary negative light, as well as each separate layer of 

 positive light, upon being magnetized rolls itself together 

 to a single magnetic curve by itself, altogether independently 

 of the extent occupied in the unmagnetized condition by the 

 bundles and layers. A negative bundle 30 centims. long 

 forms only a single magnetic curve, just the same as a layer 

 of 2 millims. length.* 



In the same way the bundle issuing from a given point, 

 which when of a certain length forms a single curve, when by 

 exhaustion it is brought to a length three times, five times, or 

 ten times as great, gives always only a single curve, inasmuch 

 as the bundle, for example, in the equatorial position with 

 reference to the magnet rolls itself together from the ends of 



* The magnetic curves formed from the positive light may be clearly 

 distinguished for a considerable distance in the neighbourhood of the 

 kathode, and in the neighbourhood of secondary negative poles, as in 

 fig. 9. 



That they are not to be perceived in like manner in the remainder of 

 the deflected positive light is to be explained, as I have remarked in the 

 Monatsber d. Akad. 1876, p. 282, by the curvature of the wall of the 

 vessels commonly employed. The disturbing forces exerted by the 

 magnet drive the discharge, and consequently the magnetic curves com- 

 posed of its layers, towards the wall of the vessel. If this is in the plane 

 of the magnetic curve, more strongly curved than the curve itself, and in 

 the same direction, so that the wall is cut by the curve, only so much of 

 the magnetic curve can be visible as lies in free gas-space between the two 

 points of section. In consequence of this limitation by the wall, each 

 magnetic curve driven to the wall is more or less nearly reduced to a 

 point. 



The sum of the light-points belonging to successive curves gives that 

 narrow line in which the magnetized positive light appears for the greater 

 portion of its course. This line, hitherto regarded as a single portion of 

 the current deflected from its course while both ends are fixed, is rather 

 to be regarded as a succession of short magnetic curves. 



