50 Messrs. Boys, Briscoe, and Watson on the 



pended so that the straight part was in the axis of rotation. 

 The other was then held with the curl close and parallel to 

 the curl of the other, and so that at first the current in each 

 should go round the same way, and then in opposite ways. 

 No action was observed in either case. When also cranked 

 or straight resonators were entirely immersed in water, no 

 effect at all was observed. 



Wo have given a short account of the early experiments 

 because at first they were so completely in accord with the 

 result of the theory put forward, though the action was more 

 powerful than had been expected, that for a time we were 

 deceived as to the real cause, and we wish to guard others 

 from falling into what is after all a very obvious error. The 

 later experiments showed either that the theory was only 

 true as regards electrostatical repulsion, the electrodynamic 

 attraction of the momentary currents being nil, or else that 

 the results already obtained were spurious. 



We have stated that the two resonators were suspended in 

 a glass jar, outside which the primary oscillator was placed, 

 and that while no repulsion was observed during the gradual 

 charging of the oscillator-ends to a higher potential than that 

 employed with the sparks, yet the moment the sparks passed 

 when the ends were nearer together, the suspended resonator 

 was observed to move in the expected direction. Now if the 

 glass jar were a conductor, the oscillator-ends might be 

 charged to any extent and no action would be felt inside ; if, 

 on the other hand, it were a perfect insulator, then during the 

 charging of the oscillator-ends before the spark the resonators 

 would have induced charges in their ends which would make 

 them repel one another. The glass jar exposed to the mois- 

 ture of the air was a sufficiently good conductor, on its outer 

 surface at any rate, to be able to screen the statical action of 

 the charges on the oscillator-ends which were by the arrange- 

 ment only slowly made to vary, and so the repulsion which 

 might have been looked for was not observed ; but the 

 moment a spark passed in the primary, the conditions were 

 changed. The fall of potential was instantaneous, but the 

 screening charge on the glass surface, which the previous 

 moment produced within the jar an effect equal and opposite 

 to that of the charges on the oscillator-ends, was unable 

 instantly to disappear owing to the imperfectly conducting 

 surface of the damp glass, and thus at this moment the reso- 

 nators were in the condition that they would have been in 

 had there been suddenly developed on the oscillator-ends 

 equal and opposite charges, but with no glass jar between. 

 This then was the cause of the repulsive impulse which was 

 observed at each spark when they followed one another at 



