PROFESSOR IN BERLIN 345 



view, they must necessarily have another invisible returning 

 portion, somewhere in the region of the wall of the tube. This 

 is a point which I have often discussed with Dr. Goldstein. 

 In that case they could no more have external magnetic 

 action than closed currents proper, within the tube, since 

 they would form ring magnets. In the rectangular vessel 

 there would still be the possibility of giving such a form 

 to the invisible returning currents that the observed effect 

 should occur. Such an interpretation to me appears hardly 

 probable, since the cathode rays form a concentrated beam, 

 and personally I do not believe in its probability, but I fear 

 it is an objection that will occur to many readers.' 



To this letter Hertz made a full and most interesting 

 reply, in which he discusses the views and criticisms of 

 Helmholtz: 



' My warmest thanks for your kind letter. Your words are 

 the strongest and most agreeable spur to activity that could 

 be given me. May I make a few observations in reply ? I do 

 not want to inflict myself on you, but write in case you care 

 to read them. I had, as a matter of fact, considered the ideas 

 you express, but was inclined to think that the cathode rays 

 are produced by the longitudinal waves, which correspond to 

 the transverse vibrations of light. For it seems to me as if the 

 longitudinal waves, in a medium in which the plane of polariza- 

 tion of the transversal waves rotates, must be propagated along 

 curved lines, and thus the direction of rotation for light and 

 for the cathode rays would be identical. Then, if the arrow 

 xy gives the direction of the positive current, 

 produced by a magnetic field, the plane of 

 polarization for all gases hitherto investigated 

 will be turned in the direction of this arrow, 

 that is, a force is produced which acts along 

 AB, and produces a displacement at an angle to this, as CD. 

 There must also be longitudinal impulses propagated in a curve 

 that is deflected to the right hand. But an elastic wire in which 

 a positive current was flowing to the cathode would also be 

 deflected to the right hand, and so a confusion between the 

 two phenomena would be possible. The question no doubt is 

 whether these simple considerations will hold good for the 

 more exact application of the theory. I have not attempted 



