316 PETER GUTHRIE TAIT 



potential in a large sphere, so built up, is in fact directly proportional to its surface 

 as compared with that of any one of the smaller equal spheres of which it is built. 



Now, the number of particles of vapour which go to the formation of a single 

 average rain-drop is expressed in billions of billions; so that the potential of the 

 drop would be many thousands of billion times as great as that of a particle of 

 vapour. On the very lowest estimate this would be incomparably greater than any 

 potential we can hope to produce by means of electrical machines. 



But this attempt at explanation of atmospheric electricity presents two for- 

 midable difficulties at the very outset. 



1. How should the smaller cloud-particles ever unite if they be charged to 

 such high potentials, which of course must produce intense repulsions between them ? 



2. Granting that, in spite of this, they do so unite, how are they separated 

 from the mass of negatively electrified air in which they took their origin? 



I think it is probable that the second objection is more imaginary than real, 

 since there is no doubt that the diffusion of gases would speedily lead to a great 

 spreading about of the negatively electrified particles of air from among the precipi- 

 tated cloud-particles into the less highly electrified air surrounding the cloud. And 

 if the surrounding air were equally electrified with that mixed with the cloud, there 

 would be no electric force preventing gravity from doing its usual work. This 

 objection, in fact, holds only for the final separation of the whole moisture from the 

 oppositely electrified air ; and gravity may be trusted to accomplish this. That 

 gravity is an efficient agent in this separation is the opinion of Prof. Stokes. It 

 must be observed that as soon as the charge on each of the drops in a cloud rises 

 sufficiently, the electricity will pass by discharge to those which form the bounding 

 layer of the cloud. 



The first objection is at least partially met by the remark that in a cloud-mass 

 when just formed, if it be at all uniform, the electric attractions and repulsions 

 would approximately balance one another at every point, so that the mutual repul- 

 sion of any two water-drops would be almost compensated except when they came 

 very close to one another. 



But there is nothing in this explanation inconsistent with the possibility that 

 the particles of water may be caused to fly about repeatedly from cloud to cloud, 

 or from cloud to an electrified mass of air ; and in many of these regions the air, 

 already in great part deprived of its moisture, may have become much cooled by 

 expansion as it ascends, so that the usual explanation of the production of hail is 

 not, at least to any great extent, interfered with. 



I may here refer to some phenomena which seem to offer, if closely investigated, 

 the opportunity for the large scale investigations which, as I shall presently show, 

 will probably be required to settle the source or sources of atmospheric electricity. 



First, the important fact, well known nearly 2,000 years ago, that the column 

 of smoke and vapour discharged by an active volcano gives out flashes of veritable 

 lightning. In more modern times this has been repeatedly observed in the eruptions 

 of Vesuvius and other volcanos such as, for instance, the island of Sabrina. It is 

 recorded by Sir W. Hamilton (British ambassador at Naples at the end of last 

 century) that in the eruption of 1794 these flashes were accompanied by violent 



