remain colder than they are, were it not for the conveyance (convection is the technical word) of the heat from the lower strata to the upper regions in other words, from the ground floor to the attics. This convection plays such an important part in the atmospheric economy, and in the formation of clouds, rain, and storms of every kind, that it demands a brief consideration.

We have seen that in general, ascending air cools at the rate of 1° F. for every 183 feet. Consequently, so long as the rate of decrease of temperature with the height is equal to or slower than this, the atmosphere tends to remain in vertical equilibrium—that is to say, vertical motions will not arise spontaneously. The only interchange under such circumstances would be due to expansion and overflow, such as that described in Chap. IV., and which gives rise to the general circulation between the equator and the poles described in Chap. V. When, however, the atmosphere near the surface is not only heated by radiation of the changed solar rays by the earth, but is surcharged with vapour, it cools even at 62° F., our summer temperature in England, only 1° F. in every 400 feet of ascent, while at 92° F., as often occurs within the tropics, it cools only 1° F. in 500 feet. An upward movement once started, therefore, is able to continue, since the air is always warmer, and therefore lighter than that which it reaches above. Similar downward motions of the cool air above take place until a large proportion of the heat received near the surface is carried up aloft. The process is precisely analogous to that by which the hot water from the kitchen boiler is conveyed through the pipes to the cisterns at the top of a house.

These upward convection currents carry the life-giving heat to the cold regions above us, just as the arterial blood conveys warmth to our extremities, and are quite as necessary to the life of the atmosphere as the circulation of blood heat is to that of the animal. Were it not for this safe conveyance, moreover, of the suplus heat away from our midst, there would often be a dangerous accumulation which would render life more insupportable than it is, especially in the tropics. When the existing rate of decrease becomes anything like 1° F. in 100 feet, or greater, rapid convection sets in, even if the air is comparatively dry, and clouds are formed with rain, and often lightning. If the air over a large district is affected, as sometimes occurs in the tropics, a cyclone is formed by the inrush of surrounding air, or, if the action is very intense and quite local, a tornado or whirlwind may result. This may be termed convection run riot.

Prof. Abbé, of the U. S. Weather Bureau, considers the limit of convection currents to be about 30,000 feet, or about the height of Mount Everest. Above this the temperature diminishes very rapidly, as indeed we find from the observa. tions recorded on the free balloons, L'Aerophile in France and the Cirrus in Germany, which on March 21st, 1892, and July 7th, 1894, reached the same height of 10 miles. In the case of the French balloon, the temperature descended to 104° F. below zero, and at the same rate the cold of space-viz., minus 461°, would be reached at a height of about 30 miles.

The ordinary rate of decrease is in general about 1 in 320 feet after we rise above the first 100 feet.

From what has been said about the slower rate at which air saturated with vapour cools by ascent when the temperature is high near the surface than when it is low, we can readily understand why cumulus clouds and rain showers occur in the daytime and in warm latitudes more readily than at night and near the poles. In fact, since at freezing-point and at sea-level even saturated air would cool as rapidly as 1° F. in every 277 feet; if it were not for imported convection systems and clouds there would be very little ascent and precipitation of condensed vapour at all in the arctic zones.

CHAPTER VII.

THE DEW, FOG, AND CLOUDS OF THE ATMOSPHERE.

WHEN We gaze skywards and see the filmy wisps of high-cirrus cloud, touching as it were, the very vault of heaven, or when we notice the ragged scud of the approaching storm, half covering the low hills, we are witnessing one of the first stages by which the water of our atmosphere becomes visibly separated from its gaseous companions.

Another stage is manifested when the pearly drops of dew gather on the blushing petals of our roses, or the rain drops from the frowning storm clouds. Still another transformation scene, and the beautiful six-rayed flakes of snow fall like flowers, scattered by an angel hand, and cover up the gloomy earth with a mantle of dazzling white. Yet one more strange scene, and from the

fiery thunder-cloud white balls of ice rattle down as though from some aerial glacier.

The chameleon character of this same water element is indeed a most fortunate circumstance. Imagine what a dull world it would be without our gaudy sunset cloud tints. What a desert if it never rained.

It happens, however, that, unlike the other gases, water-vapour can undergo all its changes within the gamut of the temperatures we experience on this planet.

Solid at 32° F., liquid thence to 212° F., after which it becomes gas.

Moreover, the air is thirsty as it were, and so, from the liquid water, at all temperatures, and even the solid ice, vapour is ever ascending by evaporation and rendered invisible as it passes through the other gases.

There is a limit, however, to the capacity of air for such a temperance beverage, which, like the thirst of men, depends on the temperature. Thus, while a cubic foot of air at zero F. can hold but a grain of vapour, at 60° F. it can soak up 5 grains. At 80° F. as much as 11 grains can remain invisible in the same space.

To give a larger example. Suppose a room, 20 feet square by 10 feet high at 60° F., to be supplied with vapour until it could hold no more, then the air in such a room would weigh 304 lbs., while the vapour, if it were condensed to water, would weigh but 3 lbs., and fill three pint

measures.

When air can hold no more vapour it is said to be saturated, and since, when it is cooled, it is able to hold less and less water, it can, even when unsaturated, be made saturated by being cooled

down to a point of temperature some few degrees below. This point is called its dew point, and depends partly on how damp, partly on how warm, it was at first.

When very warm and moist, a very slight lowering of temperature produces condensation into cloud and finally rain. Hence clouds and rain will form easier in warm countries, though other conditions may make them more constant in cold countries.

What we ordinarily term the dampness of the air, is not simply a question of how much vapour is present, since warm air may hold more than cold air and yet feel drier.

It is determined by the nearness of the dewpoint to the existing temperature, and this depends on both the amount of vapour present and the temperature of the air in which it is dissolved.

Ordinarily the dampness in England is about 60 per cent. of what could be, but in very wet weather it rises to 90 per cent.

Over the ocean it is generally high both in warm and cold latitudes, while in the interior of continents and deserts it is occasionally as low as 15 per cent. As we rise above the earth towards the level of the lower clouds the dampness increases, until, at the cloud level, we reach dew- or cloud-point, where the air is saturated.

Dew itself is the moisture deposited on the surface of bodies near the earth's surface, which have cooled down by radiation below the dewpoint of the surrounding air.

Dr. Wells, in 1783, was the first to offer this explanation, and thought the moisture came entirely from the air around. Of late, however, Mr.