OF ARTS AND SCIENCES. 453 



that its indications may be read at any time or place, and witliout 

 delay. 



Self-recording Apparatus. 



We propose to place the following attachment on the multiple 

 quadrant electrometer in order to render it self-recording. A met:il 

 plate is placed just above, and a metal cylinder with its axis hori- 

 zontal just below, the metal index-pointer of the instrument. One ter- 

 minal of the secondary circuit of a small RuhmkorfF coil is connected 

 to the plate and the other to the cylinder. A strip of co-ordinate paper 

 passing through a solution of iodide of potassium to keep it moist is 

 drawn over the cylinder by clock-work. At regular intervals the 

 primaiy circuit of the Ruhmkorff coil is broken for an instant by an 

 automatic circuit-breaker, and a spark passes from the plate to the 

 cylinder through the pointer, and registers on the paper the position 

 of the pointer at the instant. Since the induced current will be of 

 short duration, the spark will register the position of the index before 

 the electrifications of the plate and cylinder can influence the needle. 

 In this manner a record of every five minutes, or of every single min- 

 ute, can be obtained without photography. The instrument need not 

 be placed in a dark room, but may be moved around at will. The cost 

 of the necessary apparatus for this registration will be more, but the 

 expense of running it will be less, than the photographic apparatus. 

 The great advantage of this apparatus will be that single observations 

 can be made at any time without disturbing the record. 



The preliminary experiments tried were successful ; but as we have 

 been unable to obtain a mechanician who could do the necessary 

 meclianical work properly, the attachment has not been placed on the 

 electrometer. 



The Galvanometer. 



A galvanometer was used to measure the potential of the atmos- 

 phere in the following manner. A condenser was charged by connect- 

 ing one jilate with the collector and the other plate with the ground. 

 The condenser was then discharged through a ballistic galvanometer. 

 By comparing the deflection of the needle with the deflection pro- 

 duced when the condenser is charged by a known electromotive force, 

 and then discharged through the galvanometer, the difference of 

 potential between the collector and the ground was obtained in 

 absolute measure. 



