30 J. E\ Moore — Electrical Discharge from the point 



being uniformly at a higher pressure (or molecular density) 

 than region (A), from which the discharge stream comes (on 

 account of the law of the conservation of matter), one 

 would apparently he making no mistake in concluding that the 

 discharge stream must convey gaseous matter from region (A) 

 to region (B). 



§ 5. In the foregoing considerations and experiments on dis- 

 charge streams, attention has been almost wholly directed to 

 the discharge stream from the cathode. But if discharge 

 streams are streams of gaseous matter (unless there be a con- 

 tinuous accumulation and rarefaction of the gaseous medium as 

 long as the discharge stream flows), then, according to the law of 

 the conservation of matter, there must be found to be second- 

 ary streams, which serve as return streams in the circulation of 

 the gaseous medium. These secondary streams have, in real- 

 ity, been found to exist. 



It is a fact generally observed, in tubes intended for the 

 generation of Kdntgen rays, that the glass walls of the tube in 

 front of the electrode struck by the discharge stream from the 

 cathode, fluoresce vigorously. This fluorescence of the walls 

 of the tube is sharply limited (in case of a plane electrode) by 

 the line of intersection of the plane of the electrode with 

 the walls of the tube. If we make the electrode cylindrical 

 in form, the fluorescence of the glass walls is no longer of uni- 

 form brilliancy, but appears as a bright line or band on 

 those portions of the tube lying in the direction of the center 

 of curvature of the cylindrical electrode. If, instead of making 

 the electrode cylindrical, we make it of spherical form, the 

 fluorescence of the glass walls no longer shows as a line or 

 band, but as a bright, sharply defined circular or elliptical spot 

 of light, according to the angle of inclination of the axis of 

 the spherical electrode to the glass wall where the fluorescent 

 spot occurs. 



Of the numerous tubes that have been made for the purpose 

 of studying these secondary discharge streams only two will be 

 described at this point. Fig. 5 represents a simple cylindrical 

 tube, provided with two precisely similar spherical aluminum 

 electrodes, of 2*5 cm radius of curvature and 2 # 7 cm in diameter. 



