406 H. V. Gill — Electric Discharge in Geissler Tubes. 



been already indicated, this movement is dne to the fact that 

 after each extinction, the stratum is renewed at a slight distance 

 from its former position ; if the stratum always reappears at 

 the same place, there is no general movement. In Kundt's 

 experiment, if the length of the tube is an exact multiple of the 

 half-wave length of the note, stationary nodes are formed. But 

 if the length is not an exact multiple of the half-wave length 

 the nodes are not stationary, and at each vibration the node is 

 formed a little in advance of its former position; in this case 

 the heaps corresponding to the true nodes, when the length of 

 the tube is a multiple of the half-wave length, are not formed 

 with exactness, but between them are secondary smaller heaps 

 due to this progressive form of the wave. In order to obtain 

 the nodes well defined and without secondary heaps two means 

 may be employed. We may either alter the length of the 

 tube, which is the means employed by Kundt, or we might 

 alter the note so that its half-wave length became an exact 

 fraction of the length of the tube. If, however, we go beyond 

 the necessary alteration a new phase of indistinctness occurs, 

 which could be again remedied by a still further change in the 

 length of the tube or the period of the note. Compare the 

 following facts, bearing in mind that the oscillations are altered 

 by the resistance of the circuit. " When the strata are in 

 motion they preserve their respective distance apart, and their 

 number presents neither diminution or augmentation."* " This 

 flux can be regulated by introducing resistance into the circuit."* 

 " Oftener we can obtain a phase of stability by introducing 

 with care exterior resistances; if we continue to add resistance, 

 the introduction will produce a new phase of instability, and 

 on adding still greater resistance a phase of steadiness. "f Here 

 we see a remarkable confirmation of this theory. We must 

 remember that in the case of strata the u note" is very high \ 

 and consequently a slight change in the note which would pro- 

 duce steadiness would not produce a change in the wave-length 

 sufficiently great to be observed, hence the distance between 

 the nodes will apparently remain the same in both cases. This 

 might also be due to the "damping" effect of a wet thread, 

 etc., especially in the case of an induction coil, for in this case 

 the intensity, etc., of the spark is not constant during its dis- 

 charge. The damping would get rid of some of the final oscil- 

 lations which would render the period unsteady. 



* Spottiswoode, loc cit. 



f de la Rue and Muller, Phil. Trans., Pt. I, vol. lxix, fin. 



\ Although the period of a note must be within certain limits in order that it 

 may be heard as a musical note, jet vibrations having a frequency far beyond this 

 limit can be propagated in air. '' Vibrations transmitted as waves through steel 

 or air or water cannot be more frequent than a certain number, which 1 cannot 

 now put a figure to, but which I say, may be reckoned in hundreds of thousands 

 or a few million per second." (Lord Kelvin, Nature, March 6th, 1884.) 



