10 Trans. Acad. Sci. of St. Louis. 



cular impact is involved in the progressive elongation 

 of this drainage colunm, but it is the presence of the 

 drainage or conduction column which brings about the 

 discharge from the cathode knob. 



Plate VI, and many others which have been secured 

 show very clearly the spitting off of the discharge into 

 the air at sharp angles in the discharge channel. As 

 was stated in the first paper presented (No. 1 of Vol. 

 XIX), some results of this kind around a sharp angle 

 in a wire were secured. It was then thought that the 

 direction of flow of the discharge might thus be deter- 

 mined. Many plates like this here shown reveal this 

 effect. At the end of the channel nearest to the negative 

 terminal, the spitting off at an angle is such as would 

 be explained by a momentum effect of the negative par- 

 ticles. At points in the channel near to the copper plate, 

 effects are shown which could be explained as a spitting 

 off of a ^^ positive discharge^' in passing around the angle 

 in the opposite direction. 



In experiments with the angle in the wire, when the 

 plate was enclosed in a hard rubber holder, and placed 

 at the angle it required an exposure to thousands of 

 spark discharges when the angle was in the positive wire, 

 in order to obtain an appreciable effect where three or 

 four or even a single discharge would produce like 

 effects when the angle was in the negative line. In the 

 positive line when the plate was enclosed the effects were 

 , always on the side of the angle which indicated that a 

 negative discharge from the ground had produced the 

 effect. In Fig. A, Plate V, of the present paper, these 

 camera photographs of a disruptive spark in air, show 

 near the positive end an elongation of the drainage 

 channel at an angle, in a manner which may be explained 

 as a momentum effect of the air molecules, while they 

 are parting with negative corpuscles, and where the main 

 drainage column is suddenly changing direction. The 

 fogging effect upon a photographic plate placed on one 

 side or the other of an angle in a wire gave results which 



