ing current of air. This involves two other functions: 1. A rushing in of the air at the under part of the ascending current; 2. An outrushing at the upper. Upon the former of these functions, combined with modifying circumstances, depend the peculiar character and career of the under-currents and of the clouds they bear; upon the latter, combined with the same circumstances, the proportions and direction of motion of the upper currents and of the heavy masses of clouds they bear. There appears to be nothing in the nature of the cyclone itself which can determine the motion of either the upper or under current more toward any one point of the compass than toward the others. This direction of motion relative to the ascending column depends upon the direction and velocity of motion of the latter, and of the atmospheric strata in which the influx and efflux take place, modified to some extent by the differing velocities of revolution of the surface of the earth at different parallels of latitude, by the form of the earth's surface, and by the variation in the constitution of the atmosphere. If the cyclone column and the atmospheric strata which it penetrates move in the same direction, and with the same velocity, the influx and efflux will take place in nearly equal quantity, on all sides of the column. If they move with different velocities, the directions of exaggeration and diminution of the influx and efflux can be calculated in the same way as the direction of a vane on a ship's mast, given the directions and velocities of the motion of the wind and of the ship.
This gives the general idea of the cyclone for all space and time, but not the sources of its power. These are to be found in the less specific gravity, potential or actual, of the lower atmospheric strata as compared with that of the higher. Air, and gases generally, expand when heated, and become specifically lighter. It thus tends to rise above the superincumbent colder air. (We see this illustrated in our chimneys every day.) If the air over the surface of a plain becomes heated by contact with it, its specific levity is increased, and it tends to rise. But the density, and therefore weight of air, the temperature and humidity being the same, is inversely as the pressure upon it. Consequently, so long as the diminution of specific gravity caused by increased temperature is balanced in the strata above by the diminution of pressure due to elevation, the heated air cannot ascend very fast. If it had a chimney to rush up, the case would be different; but, not having one, it can only rise slowly by intermingling itself with the superincumbent air. As soon, however, as the diminution of specific gravity due to higher temperature is greater than that due to diminished pressure, the lower stratum will break itself a way through the air above it, and rush up through the opening. This result may be attained, and is generally reached, by favoring circumstances, long before equilibrium is totally destroyed. When once the' heated air has thus at a particular point formed itself a channel of escape, the
