162 CATHODE RAYS. 



bearing' they have on the charges carried by the cathode rays, as well 

 as on the production of cathode rays outside the tube. The experi- 

 ments are of the following kind: In the tube (see fig.) A and B are 

 terminals. is a long side tube into which a closed metallic cylinder 

 fits lightly. This cylinder is made entirely of metal, except the end 

 farthest from the terminals, which is stopped by an ebonite plug, per- 

 forated by a small hole so as to make the pressure inside the cylinder 

 equal to that in the discharge tube. Inside the cylinder there is a 

 metal disk supported by a metal rod which passes through the ebonite 

 plug and is connected with an electrometer, the wires making this con- 

 nection being surrounded by tubes connected with the earth so as to 

 screen oft' electrostatic induction. If the end of the cylinder is made 

 of thin aluminum about one-twentieth of a millimeter thick, and a dis- 

 charge sent between the terminals, A being the cathode, then at pres- 

 sures far higher than those at which the cathode rays come off, the disk 

 inside the cylinder acquires a j)ositive charge ; and if it is charged up 

 independently the charge leaks away, and it leaks more rapidly when 

 the disk is charged negatively than when it is charged positively. There 

 is, however, a leak in both cases, showing that conduction has taken 

 place through the gas between the cylinder and the disk. As the pres- 

 sure in the tube is diminished the positive charge on the disk diminishes 

 until it becomes un appreciable. The leak from the disk, when it is 

 charged still continues, and is now equally rapid, whether the original 

 charge on the disk is positive or negative. When the pressure falls so 

 low that cathode rays begin to fall on the end of the cylinder, then the 

 disk acquires a negative charge, and the leak from the disk is more 

 rapid when it is charged positively than when it is charged negatively. 

 If the cathode rays are pulled off the end of the cylinder by a magnet, 

 then the negative charge on the disk and the rate of leak from the disk 

 when it is positively charged is very much diminished. A very inter- 

 esting point is that these effects, due to the cathode rays, are observed 

 behind comparatively thick walls. I have here a cylinder whose base 

 is brass about 1 millimeter thick, and yet when this is exposed to the 

 cathode rays the disk behind it gets a negative charge, and leaks if 

 charged positively. The effect is small, compared with that in the 

 cylinder with the thin aluminum base, but is quite appreciable. With 

 the cylinder with the thick end I have never been able to observe any 

 iBffect at the higher pressure when no cathode rays were coming -off. 

 The effect with the cylinder with the thin end was observed when the 

 discharge was produced by a large number of small storage cells, as 

 well as when it was produced by an induction coil. 



It would seem from this experiment that the incidence of the cathode 

 rays on a brass plate as much as 1 millimeter thick and connected with 

 the earth can put a rarefied gas shielded by the plate into a condition 

 in which it can conduct electricity, and that a body placed behind 

 this screen gets a negative charge^ so that the side of the brass away 



