23(5 Dr. E. Goldstein on the Electric 



At the bend, on the side of the tube which forms the convexity 

 of the bend, there appears a bright phosphorescent surface. 

 The surface is a half ellipse, or, if no boundary on the one side 

 is to be detected, of parabolic form. The 

 axis of the parabola lies in a plane which 

 would cut the bent tube lengthwise into two 

 halves, meeting at the bend. The surface is 

 sharply bounded about the vertex, which is 

 turned towards the positive end of the tube. 

 At the opposite side turned towards the ka- 

 thode it gradually loses itself and presents an uncertain bound- 

 ary. If we call the greatest extension of the surface at right 

 angles to the axis measured on the circumference of the tube 

 its breadth, then its breadth is somewhat less than the half 

 circumference of the tube. The sharply-bounded end of the 

 surface reaches a little further towards the positive side than 

 the line in which the leading lines of the inner surface of that 

 branch of the bend which is on the negative side would cut 

 the other branch if produced. 



If, now, we employ a tube with several bends instead of one, 

 then we observe a phosphorescent surface of similar properties 

 on the convex side of each bend. From this it follows that 

 the phosphorescence is not caused by the rays of the kathode : 

 these could at most only produce phosphorescence at the 

 first bend, but in consequence of their rectilinear propagation 

 could not reach beyond it ; hence the phosphorescence is pro- 

 duced by the positive light itself. 



The phosphorescence of the surface at the bend is produced, 

 like that caused by the kathode-light, by a very thin layer 

 lying in close contact with the wall of the tube. This follows 

 from the sharp boundary which the surface shows at the end 

 turned towards the positive end of the tube ; and it follows 

 also from the fact that insulated wires suitably placed near the 

 bend cast sharp shadows on the phosphorescing part of the 

 wall of the tube. This last phenomenon shows, at the same 

 time, that the electric motion which manifests itself in the 

 light has a regular radiation. 



If instead of one wire two be placed to throw shadows, which 

 both lie in a plane coincident or parallel with the mesial plane 

 of the tube, their shadows are found to be superposed. Hence 

 it follows that this regular radiation of the positive light is 

 rectilinear. The position of the shadow shows that the phos- 

 phorescence is produced by rays very nearly parallel to the 

 axis of the tube, which radiate from the side of the kathode 

 towards the side of the anode. 



That the rectilinear direction of radiation does not com- 



