If we take a shot of small size and a feather of the same weight, we readily note that their rate of falling through the air may vary in the proportion of ten to one or more. It is easy to conceive that this difference is due to the very much less friction which the smaller body encounters in its motion by the particles of air. With this point in mind, the student should observe that the surface presented by solid bodies in relation to their solid contents is the greater the smaller the diameter. A rough, though not very satisfactory, instance of this principle may be had by comparing the surface and interior contents of two boxes, one ten feet square and the other one foot square. The larger has six hundred feet of surface to one thousand cubic feet of interior, or about half a square foot of outer surface to the cubic foot of contents; while the smaller box has six feet of surface for the single cubic foot of interior, or about ten times the proportion of exterior to contents. The result is that the smaller particles encounter more friction in moving toward the earth, until, in the case of finely divided matter, such as the particles of carbon in the smoke from an ordinary fire, the rate of down-falling may be so small as to have little effect in the turbulent conditions of atmospheric motion.

Pocket Creek, Cape Ann, Massachusetts. Note the relatively even size of the pebbles, and the splash wave which sets them in motion.

The little drops of water which gather round dust motes, falling but slowly toward the earth, are free to obey the attractions which they exercise upon each other—impulses which are partly gravitative and partly electrical. We have no precise knowledge concerning these movements, further than that they serve to aggregate the myriad little floats into cloud forms, in which the rafts are brought near together, but do not actually touch each other. They are possibly kept apart by electrical repulsion. In this state of association without union the divided water may undergo the curiously modified aggregations which give us the varied forms of clouds. As yet we know little as to the cause of cloud shapes. We remark the fact that in the higher of these agglomerations of condensed vapour, the clouds which float at an elevation of from twenty to thirty thousand feet or more, the masses are generally thin, and arranged more or less in a leaflike form, though even here a tendency to produce spherical clouds is apparent. In this high realm floating water is probably in the frozen state, answering to the form of dew, which we call hoar frost. The lower clouds, gathering in the still air, show very plainly the tendency to agglomerate into spheres, which appears to be characteristic of all vaporous material which is free to move by its own impulses. It is probable that the spherical shape of clouds is more or less due to the same conditions as gathered the stellar matter from the ancient nebular chaos into the celestial spheres. Upon these spherical aggregations of the clouds the winds act in extremely varied ways. The cloud may be rubbed between opposite currents, and so flattened out into a long streamer; it may take the same form by being carried off by a current in the manner of smoke from a fire; the spheres may be kept together, so as to form the patchwork which we call "mackerel" sky; or they may be actually confounded with each other in a vast common cloud-heap. In general, where the process of aggregation of two cloud bodies occurs, changes of temperature are induced in the masses which are mixed together. If the temperature resulting from this association of cloud masses is an average increase, the cloud may become lighter, and in the manner of a balloon move upward. Each of the motes in the cloud with its charge of vapour may be compared with the ballast of the balloon; if they are warmed, they send forth a part of their load of condensed water again to the state of invisible vapour. Rising to a point where it cools, the vapour gathers back on the rafts and tends again to weight the cloud downward. The ballast of an ordinary balloon has to be thrown away from its car; but if some arrangement for condensing the moisture from the air could be contrived, a balloon might be brought into the adjustable state of a cloud, going up or down according as it was heated or cooled.

When the formation of the drop of water or snowflake begins, the mass is very small. If in descending it encounters great thickness of cloud, the bit may grow by further condensation until it becomes relatively large. Generally in this way we may account for the diversities in the size of raindrops or snowflakes. It often happens that the particles after taking on the form of snowflakes encounter in their descent air so warm that they melt into raindrops, or, if only partly melted, reach the surface as sleet. Or, starting as raindrops, they may freeze, and in this simple state may reach the earth, or after freezing they may gather other frozen water about them, so that the hailstone has a complicated structure which, from the point of view of classification, is between a raindrop and a snowflake.

In the process of condensation—indeed, in the steps which precede the formation of rain and snow—there is often more or less trace of electrical action; in fact, a part of the energy which was involved in the vapourization of water, on its condensation, even on the dust motes appears to be converted into electrical action, which probably operates in part to keep the little aggregates of water asunder. When they coalesce in drops or flakes, this electricity often assumes the form of lightning, which represents the swift passage of the electric store from a region where it is most abundant to one where it is less so. The variations in this process of conveying the electricity are probably great. In general, it probably passes, much as an electric current is conveyed, through a wire from the battery which produces the force. In other cases, where the tension is high, or, in other words, where the discharge has to be hastened, we have the phenomena of lightning in which the current burns its way along its path, as it may traverse a slender wire, vapourizing it as it goes. In general, the lightning flash expends its force on the air conductors, or lines of the moist atmosphere along which it breaks its path, its energy returning into the vapour which it forms or the heat which it produces in the other parts of the air. In some cases, probably not one in the thousand of the flashes, the charge is so heavy that it is not used up in its descent toward the earth, and so electrifies, or, as we say, strikes, some object attached to the earth, through which it passes to the underlying moisture, where it finds a convenient place to take on a quiet form. Almost all these hurried movements of electrical energy which intensely heat and light the air which they traverse fly from one part of a cloud to another, or cross from cloud sphere to cloud sphere; of those which start toward the earth, many are exhausted before they reach its surface, and even those that strike convey but a portion of their original impulse to the ground.

The wearing-out effect of lightning in its journey along the air conductors in its flaming passages is well illustrated by what happens when the charge strikes a wire which is not large enough freely to convey it. The wire is heated, generally made white hot, often melted, and perhaps scattered in the form of vapour. In doing this work the electricity may, and often is, utterly dissipated—that is, changed into heat. It has been proposed to take advantage of this principle in protecting buildings from lightning by placing in them many thin wires, along which the current will try to make its way, being exhausted in melting or vaporizing the metal through which it passes.

There are certain other forms of lightning, or at least of electrical discharges, which produce light and which may best be described in this connection. It occasionally happens that the earth becomes so charged that the current proceeds from its surface to the clouds. More rarely, and under conditions which we do not understand, the electric energy is gathered into a ball-like form, which may move slowly along the surface until it suddenly explodes. It is a common feature of all these forms of lightning which we have noted that they ordinarily make in their movement considerable noise. This is due to the sudden displacement of the air which they traverse—displacement due to the action of heat in separating the particles. It is in all essential regards similar to the sounds made by projectiles, such as meteors or swift cannon shots, as they fly through the air. It is even more comparable to the sound produced by exploding gunpowder. The first sound effect from the lightning stroke is a single rending note, which endures no longer—indeed, not as long—as the explosion of a cannon. Heard near by, this note is very sharp, reminding one of the sound made by the breaking of glass. The rolling, continuous sound which we commonly hear in thunder is, as in the case of the noise produced by cannon, due to echo from the clouds and the earth. Thunder is ordinarily much more prolonged and impressive in a mountainous country than in a region of plains, because the steeps about the hearer reverberate the original single crash.

The distribution of thunderstorms is as yet not well understood, but it appears in many cases that they are attendants on the advancing face of cyclones and hurricanes, the area in front of these great whirlstorms being subjected to the condensation and irregular air movements which lead to the development of much electrical energy. There are, however, certain parts of the earth which are particularly subjected to lightning flashes. They are common in the region near the equator, where the ascending currents bring about heavy rains, which mean a rapid condensation and consequent liberation of electrical energy. They diminish in frequency toward the arctic regions. An observer at the pole would probably fail ever to perceive strong flashes. For the same reason thunderstorms are more frequent in summer, the time when the difference in temperature between the surface and the upper air is greatest, when, therefore, the uprushes of air are likely to be most violent. They appear to be more common in the night than in the daytime, for the reason that condensation is favoured by the cooling which occurs in the dark half of the day. It is rare, indeed, that a thunderstorm occurs near midday, a period when the air is in most cases taking up moisture on account of the swiftly increasing heat.