94 Prof. E . Wiedemann on the 



electrode. Its form is somewhat that of fig. 11. It is sharply 

 denned from the neighbourhood of the magnet to S, where its 

 two bounding lines meet; from there on, it is accompanied by 

 somewhat diffused light. 



The phenomenon also appears most distinctly when the 

 magnet-poles are at some distance from the kathode. Peculiar 

 phenomena are observed when, having produced the sickle- 

 shaped figure by means of a magnet placed under the tube, we 

 bring the one pole of a bar-magnet near to the tube from above. 

 The inner, weaker border undergoes only slight displacement, 

 but the outer, stronger border is greatly deformed. The con- 

 tour is shown in fig. 12: shows the weaker boundary nearest 



the kathode, — - the stronger boundary seen upon the wall of 

 the tube. The portion n m was, to begin with, upon the same 

 side of the tube upon which it still is; but the portion m I is 

 bent over from the other side of the tube, and has thus experi- 

 enced a rotation of nearly 1 .80°. 



The weaker boundary is therefore produced by rays which 

 resist bending and deflection much more than those which 

 produce the stronger boundary. 



In the above description no notice has been taken of the 

 behaviour of the positive column of light. We turn now to 

 consider it. 



If we move the Jamin magnet, in its equatorial position, 

 along the tube from the positive pole, the stratifications preserve 

 their position almost unaltered, although the green ring is 

 already visible, until the centre of the plane passing through 

 the ring cuts the first layer. At this moment it disappears, 

 uniting with the second; so that this becomes the first. This 

 itself then gradually advances with the magnet, and indeed 

 in proportion as the glow-rays are bent to the side, while new 

 layers continually appear from the anode. If we move the 

 magnet backwards, we find that the first layer divides as soon 

 as it has passed the ring. 



In most cases the persistent first anode-layer, as described 

 above, appears under the influence of the magnet. In some 

 cases, under conditions which I have not been able completely 

 to determine, we observe a simple displacement of the first 

 layer, whilst, however, there was always noticed a consider- 

 able expansion, e. g. with retrograde motion of the magnet, 

 corresponding to the previously mentioned division, and sub- 

 quent shortening of the layer. 



If the magnet in its motion reaches the limit of the dark 

 space, the first positive layer remains stationary for a moment; 

 and then, upon a further approach of the magnet to the kathode, 

 retreats, whilst at the same time the kathode-rays and glow- 



