launched into the air at any time, strong currents of electricity (usually positive) are obtained. This experiment, which I performed for the first time in company with William Thomson (now Lord Kelvin), many years ago, at the Philosophical Institution in Glasgow, was merely an extension of the ingenious experiments of Bénédict de Saussure (1767), made in the Alps with his electroscope: 1 whenever he raised the little instrument above his head, the pith-balls separated, showing that the electric tension of the atmosphere increases with the height above the ground, and is generally positive.

M. Ch. André 2 concludes from observations, made in a balloon, on the variation of the electric state of the higher regions of the air, that in fine weather the electric tension certainly does not increase with the altitude; but he does not appear to be entirely satisfied, at present, with these observations, and hopes to repeat them.

During a thunderstorm, this tension will change its sign many times, often, in fact, with every flash of lightning. In the experiment just alluded to, the isolated condenser, communicating by a long wire with Thomson's galvanometer-electroscope upon the table, was attached to a long pole: when this was thrust suddenly out of a window to a great height in the air, the needle of the galvanometer was immediately set in violent motion.

1 Bénédict de Saussure, Voyages dans les Alpes, 4 vols., 1779-1796. 2 Ch. André in the Comptes-rendus of the Paris Academy, 27th November 1893.

D

In recent times Mr. M'Adie has succeeded in getting electric sparks from a kite raised in the air when there were no clouds in the sky.

In August 1894, four soldiers were struck down, and severely injured, in presence of the Duke and Duchess of Connaught, at Aldershot, by a flash of lightning that destroyed a captive balloon to which they were attending. The day was characterized by a series of small thundershowers of short duration.

In dry summer weather, the dust raised by the French postilions in Languedoc, during the afternoon, has been known to remain suspended in the air the whole of the night, on account of its electric state. That this electricity is due to the friction of the horses' feet, may be demonstrated by throwing a little fine dry sand upon the condenser of a galvanometer-electroscope, and blowing it off again; in each case the needle of the instrument is vividly deflected.

I have described, in my work above mentioned, a number of wonderful phenomena due to atmospheric electricity, such as the emission of light from the hats, and raised hands, of travellers on the summit of a mountain, the buzzing noise emitted by their Alpenstocks when thrust into the soil, which has been compared to that of a boiling kettle; the marvellous fieldlightning of the Jura which, during electric disturbance of the air, plays over the pastures on the slopes of the mountains, and has also been witnessed in Mongolia. There, also, I have recorded the history of the discovery of lightning conductors by Benjamin Franklin (1749–

1752), and the discovery, a few years afterwards, by Thomas Ronayne, of London, concerning the electric nature of fogs (1761), a subject since investigated by Peltier and others.

An electrometer thrust into a dense London fog often shows enough electricity to send a telegram round the globe; and, in 1876, I published the following lines:-"If after ascertaining the nature (or sign) of this electricity, the fog could be supplied with a plentiful amount of opposite electricity, I have no doubt that it would be entirely dispersed in the course of a short time." 1

Independently of the beautiful phenomena produced by the Aurora borealis in the highly rarefied regions of the atmosphere, I should notice here those gleams and flashes of light, probably of an electric nature, which I witnessed accompanying the remarkable swarm of shooting stars in November 13-14, 1866, and which were noticed by others in the star-shower seen the following year in the West Indies. account of this phenomenon was given in my note to the Academy of Sciences at Paris in 1868.2

An

The phosphorescence of the air, which allows a traveller on the darkest nights, when there is no moon above the horizon, to find his way in the open country, and the phosphoric shining of isolated masses of cloud, often noticed, and described in my little

1 Phipson, Familiar Letters, etc., p. 37.

2 "Sur les phenomènes lumineux qui accompagnent les éssaims d'étoiles filantes" (in the Comptes-rendus, Paris, 1868).

52

THE EARTH'S ATMOSPHERE.

volume on Phosphorescence published in 1862 (my first work in the English language), must doubtless, as Humboldt suggested, be connected with the constantly varying electric state of the higher regions of the Earth's atmosphere. These phenomena, which were known to the Italian philosopher, Beccaria, in the middle of the eighteenth century, have been often witnessed by me.

I should also mention the phosphorescence of the snow and ice on the glaciers of the Alps: the darker the night the more brilliant is this effect, so that it appears sometimes like a second sunset. The snow which lies in the valleys of Piedmont, Switzerland, Valais, etc., is likewise affected in the same way; the bluish light emitted in these cases is due to phosphorescence by insolation. It is not remarked on snow which has fallen shortly before night, and which, consequently, has not been long exposed to the sun.

Although Franklin and his successors have perfectly demonstrated the identity of lightning and the electric spark-suspected years before by Otto von Guericke and Dr. Watt-yet the exact cause of the light that flashes through the atmosphere, and of its tiny representative which rushes between the poles of an electric battery, is quite unknown. By modifying the medium, we modify the spark; so that in highly rarefied air, for instance, it takes the appearances of the Aurora borealis. What is the cause of the sudden light which issues from the storm-cloud, or through the air-how is that light produced?

Some writers have imagined that lightning was due to the combustion of hydrogen gas, supposed to exist in the higher regions of our atmosphere; or to that of carburetted hydrogen (marsh gas) which has really been detected on several occasions in the air; or, again, to that of nitrogen, as in the celebrated experiments of Cavendish, who obtained nitric acid by passing electric sparks through atmospheric air. Nitric acid is, indeed, present, in minute quantities, in storm rain, and I am of opinion that all rain likewise contains it.

My own idea is that lightning is due to a vibration of the air similar to what occurs in the little instrument known as the briquet à air, where sudden compression of the air in a cylinder by a blow upon the piston will cause it to ignite a piece of tinder. A similar vibration of the air is caused by the rupture of the extremity of a Prince Rupert drop, which causes a sudden light, and by the discharge of an air-gun, also by the rupture of a glass vessel exhausted of air.

Similar vibrations, producing light, occur when crystals form in liquids, or by cooling after fusion, or by sublimation, or when divided suddenly by cleavagephenomena which I have witnessed scores of times.1 So that the minutest molecular vibration is capable of yielding a sudden flash of light.

The close calm weather which often precedes thunder

1 Phipson, Phosphorescence, or the Emission of Light by Minerals, etc., 1 vol., London, 1862.