time on the mountain heights. This most remarkable discovery is due to Dr. Viault, of Bordeaux (France), who made it originally on the slopes of the Andes, in South America, and has since confirmed it in his travels on the mountains of Europe.

With regard to the action of mountain air upon the mental powers, Professor Janssen, during his recent ascent of Mont Blanc, found that when he was not exhausted by physical exertion, his mental functions remained perfectly clear. When he took physical exercise he was rapidly fatigued, and quite unfit to perform any calculations, or other mental labour.

A number of interesting questions and some remarkable phenomena are connected with the height to which our atmosphere extends above the level of the sea. With this subject is connected the curious problem, whether gases, deprived of all pressure, can expand indefinitely, which formerly attracted the attention of my late friend M. Babinet, the eminent successor to François Arago, at the French Institute, and Bureau des Longitudes, and to which I hope to refer again.

Some have thought, indeed, that the Earth's atmosphere has no limit. Others suppose that at an elevation of forty to fifty miles the elasticity (tension) of the air, and the attraction of gravitation, must balance each other.

Applying to the total height of the Atmosphere the law of the diminution of density as observed in the lower strata of air in which we live, we arrive at the conclusion that at a height of about forty-five miles,

the atmospheric air must be as rarefied as in the exhausted recipient of the best-constructed air-pumps.

From a practical point of view, it may be stated, at once, that air in which man may live does not extend to ten miles above sea-level, probably not to eight miles, as proved by recent balloon ascents.

It has been attempted to ascertain the absolute height of the Earth's atmosphere (which the eminent meteorologist, Dr. Buist, declared to be unknown), by considering the height of the Aurora borealis, and that at which meteors are seen to shine. A distinguished American observer, Mr. Newton, once expressed the opinion that there must be some kind of an atmosphere at 500 miles above the surface of the Earth.

The height of luminous meteors, determined parallactically by various observers, gives results showing that they shine at very different heights, namely, 48 miles, 72 miles, and 132 miles.

The ingenious polarization experiments made by the French Astronomer, M. Liais, in his voyage to Rio de Janeiro, alluded to in my work on Meteors (chap. xix.), point to 200 miles as the height to which the Earth's atmosphere actually extends.

It seems to be beyond doubt that the Aurora borealis glows at a greater elevation than that at which shootingstars become visible.

The celebrated Laplace, in his Exposition du Système du Monde, says the solar rays reflected from the molecules of air before the rising, and after the setting of the sun, producing what is termed dawn and twilight, which

spread to more than 20 degrees from the sun, prove that the most distant molecules of the atmosphere are at least 30 miles above the Earth's surface. Now, if we suppose that a given volume of air, considered at the surface of the Earth, increases, as we rise, according to the square of the distance, the volume which corresponds to 45 or 50 miles, where the barometer would be at 0 inches (that is, where all pressure ceases), would be about 34,646 cubic miles, the cube-root of which gives about 32 miles, a figure approaching very nearly to that of Laplace just mentioned.

As we rise in the atmosphere it becomes more rarefied; the diminished pressure affects both the barometer and the boiling-points of liquids.

Practically speaking, the barometer falls about th of an inch for every 100 feet of elevation. Whilst at sea-level the atmospheric pressure of 15 lbs. to the square inch keeps the barometer at 30 inches, and the boiling-point of water at 212° Fahr., these figures both diminish in proportion to the altitude.1

Travellers sometimes make a rough guess at the altitude to which they have ascended on a mountain slope, by boiling the thermometer: there is a difference of some 4° to 5° Fahr. for every 1000 feet of elevation. Actual experiment has given the following results :—

1 Atmospheric pressure varies slightly at sea-level in various parts of the world; it is slightly less in the tropics, and increases towards the poles. The same has lately been found to be the case for gravitation (Defforges, 1893). This is as it should be, for after all, atmospheric pressure is merely the effect of gravitation upon the material of the Earth's atmosphere.

At sea-level in England, water boils, according to my thermometer, exactly at 212° Fahr., when the barometer is at 30 inches. On the summit of the Waldeck and Hartz mountains in Germany, at 2000 feet of altitude, I found, with the same thermometer, 208° Fahr. as the boiling-point of water. At a height of 6800 feet, on the summit of the St. Gothard (Alps), 200° Fahr. has been noted. At Quito (South America), at 9340 feet, water boils at 195° Fahr.; on the summit of Etna (10,900 feet), at 192° Fahr.; and on the highest summit of Mont Blanc, at an elevation of 15,600 feet, the temperature of boiling water was found to be 182° Fahr.

The refrigeration, or coldness of the atmosphere, increasing as we ascend, the thermometer is usually about 1° Fahr. lower for every 300 feet of altitude; but this varies somewhat with the locality, and with the direction of the wind at the time of observation.

The cold air of mountain heights throws out moisture in the shape of cloud; but as we ascend beyond this, the air becomes drier, and only on rare occasions, according to Fitz-Roy, have clouds been distinguished actually above the summits of the highest mountains.

Hitherto, no one has ascended in a balloon much above the summits of the highest mountains, says the eminent author just named; and it is not likely that it ever will be achieved, since man's existence at such altitudes is impossible. Small captive balloons provided with self-registering thermometers, and very light,

2

self-registering barometers made of aluminium, have quite recently been used in France, by Gustave Hermite, to ascertain the temperature of the atmosphere at great altitudes. At a height of 12,000 metres a temperature of 51° C., and at 10,000 metres - 41° C. were noted.1 Mendeleeff's calculations show -42° C.; Vallot's observation on Mont Blanc -45° to -47° C., as the temperature at the extreme limits of the Earth's atmosphere, and Glaisher from his dangerous balloon ascent in 1862, estimated -40° C. Forty degrees below freezing-point Centigrade is the temperature at which quicksilver becomes solid; it corresponds also to -40° Fahr. (72° below freezing-point Fahrenheit). Layers of fog, or cloud, lie at various heights, generally not exceeding two miles, and currents of wind set in different directions simultaneously; sometimes, clouds float in one current, sometimes in another, and occasionally between two currents (Fitz-Roy). On the slopes of mountains, and by balloon ascents, depths of more than 2000 feet of cloud have been measured.

Fogs, or clouds, are sometimes dry, and sometimes wet; in other words, they wet the Earth and objects on or above its surface, or they do not. This appears to be entirely dependent upon their electrical state (positive or negative).

A fog is exactly the same as a cloud; a traveller on a mountain is sensible of no difference; but the very

1 Hermite, Comptes-rendus of the Paris Academy, 22nd January

1894.

2 Vallot, Comptes-rendus, 5th February 1894.