The Oscillatory Discharge of a Large Accumulator. 259 



Supposing that the apparatus was arranged in a horizontal 

 plane, then the reflexions (under 45°) from the two plane 

 plates, one unsilvered and one sernisilvered with a transparent 

 film of silver, must weaken especially the horizontal vibrations 

 and hence almost annihilate their influence. For unsilvered 

 plates it is easily calculated that the intensity of horizontal 

 vibrations is under the circumstances stated weakened five or 

 six times more than that of the vertical ones. This ratio is 

 of course somewhat changed by the silver film. How much 

 cannot be said without knowledge of further details. 



If I have indicated the real cause Michelson has observed 

 a case analogous to my triplet a, when a nicol is interposed in 

 such a position as to get rid of the horizontal vibrations. Let 

 us hope that Prof. Michelson will soon give us his opinion. 



Amsterdam, July 10. 



[Note added August 10.] — Prof. Michelson kindly informs 

 me just now that he believes my explanation of the discre- 

 pancy in our results to be correct. 



XXXIV. The Oscillatory Discharge of a Large Accumulator. 

 By John Trowbridge*. 



ri'HE discharge from a large number of Plante cells is 

 *- characterized by a sibilant flame which, by quickly sepa- 

 rating the spark-terminals, can be drawn out to a length of 

 several feet. It closely resembles the light produced by 

 passing an electric spark through lycopodium powder. When 

 a photograph of this flaming discharge is examined it is seen 

 to have an intensely bright spark as a nucleus (fig. 1). On 

 account of the flaming discharge it is difficult to examine its 

 character by means of a revolving mirror. By employing, 

 however, two spark-gaps it seemed possible to ascertain 

 whether the discharge is oscillatory or not. 



In my experiments the circuit was made at the instant the 

 revolving mirror was in the position to reflect an image of 

 the discharge of the battery upon a sensitive plate. The 

 photographs obtained in this way showed disruptive dis- 

 charges superimposed upon continuous discharge. The latter, 

 however, masked any appearance of an oscillatory discharge. 

 It was evidently necessary to blow out the flaming discharge 

 in order to see if oscillations followed the pilot discharge. 

 The first experiment was made with 2500 cells arranged in 

 series ; and the flaming discharge was much lessened both by 

 the reduction in the number of cells and by a suitable arrange- 



* Communicated by tlie Author. 



