462 



Professor Arthur Schuster 



[Feb. 22, 



some definite answer is given to them. The last-mentioned view, 

 propounded originally by Peltier, and latterly supported by Exner, 

 is the simplest. If we could allow that the earth, once electrified 

 negatively, could remain electrified for ever, the corresponding 

 positive electrification being outside our atmosphere altogether, the 

 chief difficulty of atmospheric electricity would be removed, and the 

 normal fall of potential at the surface would be explained by the 

 permanent negative electrification of the surface. 



Unfortunately, this view, to be tenable, has to assume that the 

 atmosphere is a complete non-conductor to the normal electric stress, 

 and this is known not to be the case. We know of several causes 

 which break down the insulating properties 

 of air. If two pith balls are electrified and 

 repel each other, and a match be lit in their 

 neighbourhood, the pith balls come together, 

 showing that they have lost their charge, and 

 consequently that the flame of the match has 

 destroyed the insulating power of air. It is 

 3 not only the flame itself which conducts, but 

 also the gases rising from the flame.* The 

 C c' following experiment will prove this. In 

 Fig. 1, A represents a metallic tube bent 

 round at the upper end, and containing at 

 its lower end a Bunsen burner in metallic 

 contact with the tube, which is also con- 

 nected to an electroscope. The tripod T 

 which supports the tube is insulated by 

 blocks of paraffin. A Leyden jar, L, on a 

 separate support, is placed so that the knob 

 stands at about the level of the upper part 



' of the tube, which acts as chimney to the 



flame. The knob of the jar may be a few 

 inches away from the opening of the chimney, 

 and not necessarily in a line with it. The 

 F IG> i. experiment succeeds, although the gases 



rising from the burner may not come into 

 contact with any part of the jar. The jar is charged, and care 

 must be taken that no fibres of dust attach themselves either to the 

 jar or chimney. I have found it convenient to join a piece of amal- 

 gamated zinc to the end of the chimney. Under these circumstances 

 the charge of the jar will be found to leak across to the tube, and the 

 leaves of it will diverge. If, as in Exner's form of electroscope, the 

 leaves, on reaching a certain divergence, discharge by forming a con- 

 tact with earth-connected plates, C C, the charging and discharging 



* The most complete investigation of the conduction of gases rising from 

 flames is contained in a series of papers by Giese (' Wiedemann's Annalen,' 

 vol. xvii.). 



