10 E. L. Nichols — Alternating Electric Arc. 



ally into its path ; also that the normal arc on approaching the 

 point, avoided the apex and entered the wire from the side, 

 never passing in at the point itself. 

 From these observations it appears : 



1. That the discharge from the ball (positive) leaves the 

 latter in a direction normal to the surface, but that it enters 

 the other terminal at some distance from the apex. 



2. That the discharge from the point (positive) leaves the 

 very apex of the latter, but is deflected into a course nearly 45° 

 from the axis and reaches the ball obliquely at some point on 

 its side. 



Taking these observations into consideration, the explana- 

 tion of the cessation of the discharge from the point at the 

 critical distance and of the establishment of the " one way " 

 arc, follows at once. The two paths of discharge differ in 

 length and for a given electromotive force, the maximum 

 striking distance is sooner reached in the case of the oblique 

 than of the normal arc, so that the latter continues to pass at 

 greater distances (of ball and point) than the former. 



Further inspection of the images of the two arcs in the 

 revolving mirror, revealed another curious fact The mirror 

 was set up with its axis of revolution parallel to the common 

 axis of the ball and point. The image of an instantaneous 

 spark following the line of the normal arc would therefore be 

 a line parallel to the axis of the mirror. Since the duration of 

 the discharge was nearly .001 seconds, this linear image was 

 expanded into a broad rectangular band. The image of any 

 oblique discharge would in general be an oblique parallelogram. 

 The image of the discharge from point to ball, however, was 

 not of that form. It appeared rather as a warped surface the 

 form of which could be explained only by supposing that the 

 discharge at first followed the normal path to the ball, and was 

 gradually displaced as the cycle progressed, until it reached its 

 extreme position at 45° to the axis, just before the rupture of 

 the arc. 



Definite results were obtained only while the platinum sur- 

 face was new and bright. The region where the normal arc 

 left the ball soon became tarnished and corroded and there was 

 an increasing tendency on the part of the oblique arc to leave 

 its own path and join the other. 



When the ball was supplanted by a platinum wire, l mm in 

 diameter, with rounded tip, the object being to force the two 

 discharges into a common path, it was found that the arc from 

 the point (positive) avoided the end of the wire altogether, 

 and struck in upon the cylindrical surface beyond. When the 

 end of this wire was surrounded by an insulated platinum ring 

 which was connected with the terminal of the induction coil 



