240 Dr. E. Goldstein on the Electric 



3. The rays of the kathode-light are rectilinear both within 

 the third layer and within the second of a straight kathode- 

 wire ; but Crookes's theory implies that the discharge, and 

 consequently the repulsive electrification, at the kathode must 

 continue at least until the first repelled molecules have tra- 

 versed the diameter of the second layer ; whence it would 

 follow that the rays, at least within the second layer, must be 

 hyperbolic, and not straight. For so long as the repulsive 

 electrification continues, the repelled molecules must follow 

 lines of force ; and their form is determined by the condition 

 that about a straight thin wire equipotential surfaces form 

 confocal ellipsoids. 



Although the above has long been known to me, yet, in 

 order to further test Crookes's theory, I have made additional 

 experiments, which I will briefly describe. 



In a cylinder the kathode was formed by a plane surface 

 of metal, obliquely to which was placed a phosphorescent 

 plate, so that at the densities at which the discharge excites 

 phosphorescence a part of this plate lay within and a part 

 without the second layer of the kathode-light. But as, ac- 

 cording to Crookes, the second layer (in opposition to the rest 

 of the kathode-light) consists only of molecules not in collision, 

 and the phosphorescence of the wall is caused only by the 

 impact of such not yet intercepted molecules, the phospho- 

 rescent plate ought to show a sharply-marked contour-line 

 where the plate is intersected by the outer surface of the 

 second layer, on one side of which it should be brightly 

 phosphorescent, and on the other dull or altogether non- 

 luminous. Observation shows, however, that no such dividing 

 line exists ; the brightness of the phosphorescent plate alters 

 continuously from point to point, and is still of considerable 

 intensity in the neighbourhood of the outer kathode-light, 

 which, according to Crookes, consists almost entirely of 

 colliding particles incapable of exciting phosphorescence. 



Mr. Crookes must have observed that in particular cases 

 surfaces lying outside of the second layer still phosphoresce. 

 To explain this he assumes that a few molecules have much 

 overpassed the mean free path of the molecules thrown off 

 from the negative pole, and so reach and excite phosphorescent 

 screens lying outside the second layer. 



The mean free path of the molecules thrown off from the 

 negative pole agrees, according to Crookes, with that calcu- 

 lated from the kinetic theory of gases for the mean free path 

 in the gas contained in the discharge-tube. I had a cylin- 

 drical tube constructed, 90 centims. long, in which the kathode 

 was a plane surface of metal at one end of the tube and at 



