ATMOSPHERIC ELECTRICITY. 93 



answer is given to them. The last mentioned view, propounded origi- 

 nally by Peltier, and latterly supported by Exner, is the simplest. If 

 we could allow that the earth, once electrified negatively, could remain 

 electrified forever, 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 nonconductor 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 neighbor- 

 hood, 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 not only the flame itself which con- 

 ducts, but also the gases rising from the flame. 1 The following experi- 

 ment will prove this. In Fig. 1, Plate I, 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 connected 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 neces- 

 sarily in a line with it. The 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 fibers of dust 

 attach themselves either to the jar or chimney. I have found it con- 

 venient to join a piece of amalgamated 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, dis- 

 charge by forming a contact with earth-connected plates C C, the 

 charging and discharging can be watched for a long time. It will be 

 noticed that the flame, being altogether surrounded by a tube of the 

 same potential, cannot be active in this case, but the conductivity must 

 be due to the gas as it escapes from the chimney. 



It follows from these experiments that every fire burnt on the surface 

 of the earth and every chimney through which products of combustion 

 pass act like very effective lightning conductors, and would conse- 

 quently discharge, slowly but surely, any electrification of the surface 

 of the earth. The peculiar immunity of factory chimneys against dam- 

 age by lightning appears from statistics collected by Hellmann in 



1 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). 



