234 Mr. J. A. McClelland. 



A tube of the form fig. 3 was constructed, with the terminals A 

 and C about 40 mm. apart. This tube has a narrow side tube B, as 

 in fig*ure, the shortest distance between the terminals by this path 

 being about 160 mm. 



Fm. 3. 



Fig. 4. 



As the tube is being exhausted the discharge at first passes by the 

 shortest path from A to 0, but when the dark space round the 

 cathode C spreads out until it reaches the walls of the tube, a point 

 is reached at which the positive column starts behind the disk A, 

 and travels by the side tube B. 



The part between A and C which was formerly occupied by the 

 positive column is now entirely free from glow, bat there is still a 

 stream of negative rays from C towards A. Of course, there is also 

 a less intense stream of rays behind C, but still, if we measure the 

 current carried by the rays from C to A, we have an inferior limit. 



The reason why the much longer path by the side tube B is the 

 path of least resistance, seems to be, because by that path the posi- 

 tive column is enabled to approach closer to the negative electrode, 

 the intensity of the cathode rays being less behind C than from C 

 towards A. When the positive column is travelling by the side tube 

 B, if we bring a magnet up to the tube between A and C, so as to 

 deflect the cathode rays on to the side of the tube near C, the whole 



