fair weather being 4, it was found rarely to fall as low as 1; often during sudden showers it equals 20 or 30; during snowstorms with high wind it sometimes reaches 100; and during thunder-showers, 100 and even 200. A predominance of negative electricity is characteristic of thunderstorms. Hence some origin of atmospheric electricity, always operative, but in different degrees of intensity, must be sought. Clearly, any form of energy connected with mere cloud-formation will not answer the requirements.
2. The observations show also that the potential of the air is exceedingly fluctuating, no natural phenomenon being comparable with it in changeableness except wind-pressure. A change in the electrical state of the air indicates a corresponding change in the earth's surface. It is more reasonable to suppose that the earth, the great reservoir of electricity, should control the air and clouds electrically than that the clouds should control the earth.
3. The surface of the earth is perhaps never in electrical equilibrium; in other words, it is not an equipotential surface. Some time about 1865 Matteucci made prolonged experiments on earth-currents, and reached several interesting results. He found, for instance, that a tolerably steady current of electricity flowed through a line established along a meridian, uniformly from south to north; that fluctuating currents of low electromotive force flowed through an east and west line, sometimes in one direction, sometimes in the other; that when one terminal of a long line was in a valley and the other at a considerably higher elevation, a current flowed uniformly up the wire toward the more elevated end. A flash of lightning, in this case, was always accompanied by a sudden increase in the deflection of the galvanometer needle. ("Smithsonian Reports," 1867.)
4. Irregular and spontaneous earth-currents are the usual accompaniment of great terrestrial disturbances. James Graves showed in 1871 that spontaneous currents in the Atlantic cables frequently occur during earthquake-shocks. ("Journal Soc. Tel. Eng.," ii, pp. 80-120.) That spontaneous currents flow through land lines during auroral displays is a well-known fact. It is also asserted that any great meteorological change, as the motion of a heavy storm with considerable barometric fluctuation, is announced at a distance by irregular galvanic shocks through submarine cables.
5. Marked electrical disturbances in the atmosphere not infrequently accompany earthquakes and volcanic eruptions. A vivid flash of lightning was seen during an earthquake-shock in West Cumberland, England, on October 25, 1879. So frequently are these two phenomena conjoined that some writers have attributed South American earthquakes to electrical action. Lightning is often seen playing about the boundary between the condensing vapor from a volcano and the adjacent cool air.
6. Measurements of the potential of the air show that, as we pro-
