242 Royal Society : — Messsrs. De La Rue, Miiller, and 



S Z represents the battery, V the vacuum-tube, C the coil-conden- 

 ser; one terminal is connected with the end A of the wire A A', and 

 the other terminal with the end B of the second wire B B'; connexions 

 are also led to the wires of the vacuum-tube. The ends A' and B' 

 are left free ; and it is clear that the coil forms a sort of Leyden 

 jar when thus used : an interval, however short it may be, must 

 elapse in accumulating a charge which at intervals discharges itself 

 and causes a greater flow in the vacuum-tube in addition to that 

 which passes continuously. It may be stated that the capacity of the 

 accumulator has to be carefully adjusted to prevent any cessation 

 of the current, to avoid, in fact, a snapping discharge at distant 

 intervals. The periodic overflows, so to speak, which increase the 

 current from time to time, would seem to have a tendency to cause 

 au interference of the current-waves, and to produce nodes of 

 greater resistance in the medium, as evinced by the stratification 

 which becomes apparent. To the eye no pulsation in the current 

 is apparent ; and in order to convince ourselves whether or not 

 there was really any fluctuation in the current when the apparatus 

 was thus coupled up with the battery, we made several experi- 

 ments, and ultimately hit upon the following arrangement (fig. 3): — 



The primary wire p p' of a small induction-coil, both with and 

 without the iron core, was introduced into the circuit as well as the 

 vacuum-tube V ; to the secondary wire s s' of the induction-coil was 

 connected a second vacuum-tube, V 2 . Under these circumstances 

 there was no change in the appearance of the discharge in V, 

 in consequence of the introduction of the induction-coil, the termi- 

 nals being still surrounded by the soft nebulous light before spoken 

 of : no luminosity appeared in the second vacuum-tube V* in con- 



