126 Mr. E. L. Nichols on the Alternating 



were sufficient to throw the apparatus out of adjustment, and 

 the arc, once ruptured, would not reappear spontaneously. It 

 could be re-established, however, by the momentary introduc- 

 tion of a bit of metal between the ball and point, or even by 

 the interposition of a candle-flame. The complete stability of 

 the discharge was finally secured by driving the dynamo by 

 means of a motor, the latter being supplied from a storage 

 battery. 



The main portion of the investigation consisted in the deter- 

 mination of the periodic changes of electromotive force and 

 current during a complete cycle, when no arc existed, and of 

 the modifications introduced into the curves of potential and 

 current by the discharge between ball and point. Throughout 

 the entire series of measurements, the striking-distance was 

 greater than the critical values already defined, a condition the 

 maintenance of which was secured by watching the indications 

 of the galvanometer. 



The instrument used in the measurement of electromotive 

 force was a Thomson mirror-galvanometer of ten thousand 

 ohms resistance. The galvanometer line was carried to the 

 dynamo, where, by means of an instantaneous contact device, 

 the circuit was closed during an interval of exceedingly short 

 duration, once in every revolution. The device consisted of a 

 wooden disk, mounted upon the shaft of the machine. A 

 single bar of brass, on the periphery of the disk, passed under 

 a brush at every turn. This bar was connected metallically 

 with a brass collar on the shaft, and a second brush bearing 

 upon the collar completed the circuit. By thus closing the 

 line through the galvanometer, for an instant, once in a revolu- 

 tion, the electromotive force of the second circuit, at that 

 particular point of the cycle for which the contacts were made, 

 could be measured ; and since the brush was adjustable through 

 considerable range, the entire cycle could be explored. 



The arrangement of the entire apparatus is shown in fig. 1. 

 PC and SC are the primary and secondary coils of the induc- 

 torium, p and b are respectively the point and ball. R is a 

 non-inductive resistance, g the indicating galvanometer, in 

 parallel with the ball and point, r a non-inductive resistance, 

 s a switch by means of wdiich the Thomson galvanometer 

 could be shunted at will around R or g. K is the instanta- 

 neous contact-device, and G is the Thomson galvanometer. 

 When the Thomson galvanometer was shunted around R, 

 which was placed in the main circuit leading from the 

 induction-coil, it served to indicate the current flowing in 

 that circuit during that portion of the cycle for which contact 

 was being made ; when connected in shunt with the gal- 



