158 Trans. Acad. Sci. of St. Louis. 



It was found in this experiment, as in all others, that the 

 track of a ball discharge is a good conductor, so long as the 

 ball discharge is in existence at the end of the path. It affects 

 the operation of the machine in a very remarkable way, as has 

 been explained, when strong disruptive discharges are taking 

 place. If a ball discharge intersects the track of another ball 

 discharge, it will sometimes move along this track with great 

 speed, but sometimes it disappears and at the same time it 

 reappears on the same track further away from the cathode. 

 But as a rule these balls cross and recross old tracks while 

 running closely parallel to them, without being in the least 

 affected by their presence. 



It is not probable that these discharges are really spherical 

 in form. Sometimes they do not even appear spherical. The 

 phenomenon is apparently the result of a breaking down 

 under electric stress of the medium composing the sensitive 

 part of the film. The chemical action results in the formation 

 of a track along which a discharge passes to feed and main- 

 tain these luminous nuclei. This discharge along the track 

 between the cathode and ball is usually invisible, even in a 

 dark room, but its existence can always be shown by passing 

 the point of the teaser along the track. By separating the 

 knob of the machine from the disk bearing the discharge 

 point this silent brush discharge also becomes apparent. 



If a drop of water be put on a not too clean plate of glass, and 

 at the end of the discharge point a of diagram, the water is 

 drawn out into long narrow tracks. They originate at points 

 of maximum curvature, where a luminous point discharge 

 appears. This action results in the formation of a conducting 

 track which feeds the point discharge. But a similar action 

 will take place at the positive knob of the machine. Only 

 part of the conditions existing on the photographic plate are 

 found here. The medium is a conductor, which is distorted 

 by the acting forces. There is no chemical breaking down of 

 the medium, resulting in the formation of a conducting 

 channel. 



After a plate containing a ball discharge has been fixed and 

 dried, if it be replaced at either discharge knob of the machine, 

 luminous ball discharges form along the tracks, but the con- 



