210 



Sir J. J. Thomson : Furth 



er 



space between the sides of the cathode and the neck of 

 the flask must be small, otherwise when monatomic gases 

 such as helium or argon are used there is a considerable 

 deposition of metal on the neck of the flask ; the discharge 

 then tends to start from this deposit instead of from the 

 cathode, and very few positive rays pass down through the 

 hole in the cathode. 



For the channel through the cathode we generally use fine 

 copper tubes drawn out until the opening is sufficiently 

 reduced ; the length of the tube is about 8 cm., and the 

 diameters of the tubes we have used vary from *1 to *01 mm. 



It is a matter of some difficulty to get the finest tubes 

 sufficiently straight to allow a particle moving in a straight 

 line to pass through the tube without striking against the 

 sides. The following method was used when it was necessary 

 to reduce the size of the tube as much as possible. Two flat 

 pieces of steel were made as accurately plane as possible by 

 the use of surface plates ; a shallow fine scratch was ruled on 



Fig. 1. 



one of the plates, and the two plates were then bolted 

 together, the scratch making an exceedingly fine channel 

 between the plates. With these very fine tubes it is necessary 

 when taking photographs to give a very long exposure, 

 2 or 3 hours ; and to adopt special arrangements for 

 keeping the magnetic and electric forces constant during 

 this interval, as fluctuations in the intensity of these forces 

 broaden the lines and neutralize the benefit derived from 

 the smallness of the tube. 



The photographic plates with which we have obtained the 

 best results are the Imperial Sovereign. Inasmuch as the 



