838 Dr. E. Goldstein on the Elect 



'IC 



wall of the first bond, and excite a green phosphorescence at 

 the point b, of the form already described. Close behind the 

 bend the positive light shows an obliquely-placed maximum of 

 light, which is sharply bounded on the negative side and falls 

 On very rapidly, whilst on the positive side the brightness de- 

 creases very gradually. 



The positive layers traversing the branch 2 reach the 

 branch 3 at the point y, but do not at once bend round in the 

 direction of 3, but continue as far as the wall opposite to the 

 opening 7 as a well-defined conical bundle of small angle 

 parallel to the axis of 2 ; the section of this part of the tube by 

 the bundle issuing from 7 appears as a bright sharply-defined 

 green phosphorescent surface. 



Near the bundle there is in 3 a new maximum of positive 

 light, separated from it by a feebly-illuminated interval. This 

 maximum, like the first, is convex towards the negative side, 

 its brightness falling off rapidly on that side, gradually on the 

 positive side, thus possessing exactly the cup-like appearance 

 of the boundary of one of the stratifications of positive 

 light turned towards the kathode : its axis also coincides with 

 the axis of the branch 3. The blind prolongation X near 7 

 remains completely empty and shoics no light. Precisely the 

 same appearances are seen at /3 and a : here also well-defined 

 bundles issue, which continue as far as the wall opposite 

 -the opening, and produce there, where they cut it, sharply- 

 bounded areas of phosphorescence. After each bundle at /3 

 and a there follows a positive maximum symmetrically placed 

 with reference to the axis of the tube. After the bundle which 

 produces a green surface at the bend a, there follows again an 

 obliquely-placed maximum of positive light. The blind pro- 

 longations X all remain empty and dark. If now the direc- 

 tion of the current be reversed, the discharge presents the 

 following form (fig. 12): — The positive light which begins 

 at the opening of 6 produces a green surface at the bend a, 

 the sharply -bounded edge of which is now turned in the op- 

 posite direction (towards 5, instead of towards 6 as in the 

 previous case). Behind the bend, again, there is an obliquely- 

 placed maximum of positive light, which falls away sharply 

 on the negative side ; the positive rays in 5 do not bend 

 round at the point a into the tube 4, but continue with de- 

 creasing intensity into a prolongation X, which they completely 

 fill. At the closed end of X is seen green phosphorescence; but 

 there is now no phosphorescence at the part of the tube 5 

 lying opposite the opening a. At the entrance of 4 there is, 

 again, a cup-shapecl positive maximum, above which and di- 

 stinctly separate passes the bundle in 5. The same appear- 



