44 Prof. E. Wiedemann on the 



the end of the wire a; this also becomes smaller and smaller, 

 without however entirely vanishing ; the positive light radiates 

 upwards until the end of the positive electrode has broken 

 through the boundary of the dark negative space about the 

 kathode ; then it at once flows downwards like a small lumi- 

 nous waterfall. If the anode rises still higher and approaches 

 the negative plate, the discharge ceases to issue from it and takes 

 place, in highly inconstant fashion, between the upper surface 

 of the mercury-column in the U-tube and k. I found thus, 

 as Herr Goldstein has found, that if a positive electrode 

 approaches a negative pole, the end of the positive column 

 remains unchanged. But whilst he asserts that when the 

 positive pole approaches the negative within a distance equal 

 to or less than the interval between the negative pole and the 

 first positive layer of the positive light, at the density employed 

 the positive light disappears altogether, my often repeated 

 experiments showed that there always remains on the positive 

 electrode a little cap of reddish-yellow positive light even when 

 the electrode has much over-passed the end of the positive 

 light ; it becomes smaller only gradually. It seems as if this 

 cap, adhering to the positive electrode, were firmly united to it. 



Whilst the mode of combination of the positive and nega- 

 tive light can best be followed with electrodes of the form 

 described, certain other phenomena appear much more dis- 

 tinctly if we employ a second plate for the movable positive 

 electrode. 



If we allow the anode to approach the kathode, all the 

 phenomena are regular so long as the anode does not penetrate 

 into the dark space about the kathode, i. e. layer after layer 

 disappears, and when the last has been traversed there remains 

 a small cap of positive light upon the plate, which ascends 

 with it but becomes gradually smaller. If, however, the plate 

 penetrates into the dark space, we always observe a deforma- 

 tion of the negative stratifications, the blue light is driven 

 behind the kathode, whereas it was before only in front, and 

 there forms on the wall a narrow strip of positive light, as also 

 Hittorf * observed upon close approach of anode and kathode. 

 The kathode-rays themselves, which usually proceed from the 

 centre of the kathode, are compressed altogether to the edge of 

 the kathode, from which they issue like a star, whilst the 

 middle of the kathode remains quite dark | ; moreover kathode 



* Pogg\ Ann. vol. cxxxvi. pp. 1, 197 (1869). 



t This experiment shows that the kathode-rays do not always run in 

 the direction of the lines of force. In the case of two parallel plates the 

 lines of force are at right angles to the plates, and in the space between 

 them the potential falls most rapidly, nevertheless the kathode-rays issue 

 from the edge of the plate. 



