46 Prof, von Bezold's Researches on the Electrical Discharge. 



it takes, at any rate apparently, the path, of many hundred times 

 that length, through the wire D. 



If the wire D be still more lengthened, the phenomenon at first 

 remains, within tolerably wide limits, unchanged ; and not until 

 the length has nearly doubled does the figure at A become again 

 larger, until with still more considerable lengths the difference 

 in magnitude of the two figures again quite disappears. In this 

 case it was immaterial whether a thick or a thin, a good or a bad 

 conducting-wire was used, or whether it was tied in a tight knot 

 or made to describe an arc of a circle. I have not yet worked 

 with spirals. 



The phenomenon being so entirely novel, I thought it would be 

 interesting to investigate the deportment of the wire D in various 

 positions. Hence an alteration was made which is represented in 

 fig. 4. The conductors A, B, C are placed upon the plate and 

 are connected with each 

 other by two wires, D and 

 J)'. If, now, the lengths 

 are so chosen that at C as 

 large, and at A, on the 

 contrary, as small a figure 

 as possible is formed, the 

 figure at B is larger than 

 that at A and smaller 

 than that at C. But if 

 the length of the wire is 

 more considerable, the 

 sizes of thefigures AandC 

 become more nearly equal, 

 while B, with a suitable choice of the ratio D : D', becomes very 

 small or even quite disappears. With a striking-distance of 4*3 

 millims., and the lengths AF = 50 centims., D = 6*2 metres, 

 D'=8*l metres, the figures at A and C were large, while at B 

 very small stars only appeared. 



If any one of the conductors be raised from the plate, the 

 figures at the other conductors will not be in the least altered. 



This experiment teaches the new fact that the connexion of the 

 conductor with a blind-ending wire is sufficient materially to 

 alter the figure formed at the conductor, or to make it disap- 

 pear. The experiment becomes most instructive when close to 

 the conductor A a second spark-micrometer (/, fig. 5) is in- 

 troduced, one of whose knobs is connected with A, while the 

 other leads to the wire D (fig. 5) . If, then, the spark-micro- 

 meter / be first of all adjusted for a great distance and this 

 distance be gradually diminished, it will be seen how from the 

 moment the spark passes at / the figure at A becomes different 



