DEGRADED CUMULUS.

A very similar line of argument applies to another well-known sign of rain-the appearance of cloud shown in Fig. 13, a, where a thin stripe of cloud seems to cross a well-formed cumulus.

This is a foreshortened view of a cumulus (b), and a degraded patch of cloud (c). Sometimes, but most unfortunately, this cloud is called cumulo-stratus, as by Howard and others. We shall presently see, however,

a. Cumulus

FIG. 13.-Cumulus, degraded cumulus, and line cumulus. crossed by another cloud. b, c. The same from another side. d. Line cumulus, or high cumulus. e. Degraded cumulus, lensshaped.

that it has nothing in common with true stratus, but is a mixture of pure cumulus with a degraded patch of cloud. The origin of this cloud is very simple. We have shown in the preceding diagram that a detached cumulus (Fig. 12, a 1) can become degraded into a flat, thin mass, with festooned base; but in certain conditions the failure of the rising current takes place more gradually, and then the base of the mass remains flat instead of becoming festooned.

This, too, is a sign of rain for the same reason as in the former instance. The existence of failing or abortive rising currents is of itself a sign of disturbed weather, and is really more of an accompaniment than a prognostic of rain. This cloud is common in the equatorial doldrums, and in any other part of the world where showers fall from cumulus. Thus we see that festooned, raggy, and this streaked cumulus are all associates of rain, and for similar reasons. A very similar degraded cumulus-patch is very common during the finest weather in the tradewind districts. The small isolated patches of cumulus, which are so common there, often seem to lose so much of their rising impulse that a rocky top cannot form, but at the same time the stoppage is not so sudden, or the cloud so heavy, as to develop festoons. Then we get a cloud nearly flat below, with a smooth round surface above, like a plano-convex lens, as in Fig. 13, e. But, as Ley finds an almost identical form as the embryo of a cumulus whose rising force is very weak, we must judge the import of this, as of every other cloud, by its surroundings.

MINOR VARIETIES.

Another form of cumulus is developed almost at the level of cirrus, in long thin lines made up of little heads of condensed vapour, sometimes called thunderheads. This is shown at the right-hand top corner of Fig. 13, d. It is only noticed here to guard against its being called cirro-cumulus. In practice this cloud is almost invariably produced in front of thunderstorms, and it is difficult to see how it can be formed otherwise than

G

by assuming the air to rise in a thin vertical curtain. In our chapter on line-thunderstorms, we shall find from other reasons that the air really does sometimes rise in long narrow sheets. This is the cumulus simplex of Weilbach and rain-cumulus of Howard; it has also been called high cumulus, line-cumulus, and turreted cumulus.

There is one other variety of cumulus, which need only be mentioned here.

Sometimes the top of the cumulus becomes hairy, as if it had been combed out; this cannot be explained, but is usually seen over heavy rain. But occasionally this peculiar process on the top of a rainy cumulus develops a sort of flat sheet of cloud, apparently touching the summit, and the cloud may conveniently be called cumulo-stratus.

STRATUS.

We now come to the second variety of clouds, to which the name of stratus is applied, because it always lies in a thin horizontal layer, like a stratum of rock or clay. Pure stratus has no sign of any hairy or threadlike structure except at the edges, for a stratum which shows much marking would be cirro-stratus, and has quite a different origin. Pure stratus is essentially a fine-weather cloud, and is especially characteristic of anticyclones. One very beautiful variety is often seen during a fine night, when the cloud forms thin broken flakes, something like mackerel sky, from which, however, it is really quite distinct.

In Howard's original work on clouds, "stratus" was

applied to ground-mist, but that idea is now entirely discarded by all meteorologists. What we call pure stratus is the "strato-pallium" of Weilbach, and the "stratus" of Hildebrandson. The origin of this cloud seems to be that when the air is tolerably still, and radiation is going on, the general mass of the air gets gradually cooler, till at last the temperature is reached at which some stratum touches the dew-point, and therefore condenses its moisture into cloud. Sometimes the cloud is formed by rising fog.

This at once explains both why the stratum of clouds should be flat and thin, and why this form of cloud should be characteristic of anticyclones. We can also understand why, under these conditions, the sky sometimes becomes overcast almost instantaneously. Very often a mass of fine-weather stratus is uniform in the centre, but hairy or striated at the edges, and we get a cloud indistinguishable by form alone from some kinds of strato-cirrus, though very different in origin and surroundings. Sometimes the lower surface of a sheet of cloud is festooned for a short time like the flat base of a cumulus. The cloud is probably not then pure radiation stratus, but a smooth form of strato-cumulus, which, by sudden failure of the generating current, begins to fall in lumps just like the festooned cumulus before described.

CIRRUS.

The third primary form of cloud is cirrus, a word taken from the Latin, and meaning literally "a curl of hair.” We have already explained the origin of pure cirrus and its

relation to pure cumulus, together with the rudimentary idea of the formation of a stripe of cloud from a current of vapour-laden air, which rises in currents of different velocities, but in the same direction.

CIRRUS-STRIPES.

When cirrus rises irregularly, and appears not to be all at the same level, we have seen that it is then pure cirrus; but there is a modified form, in which more or less of the sky is covered with long thin stripes of cirrus, all apparently at the same level. Technically this is known as cirro-filum (literally "hair-thread "), a name first suggested by Mr. Ley, and the term is suitable for international use; but we shall call them cirrus-stripes. As these are by far the most important form of cirrus for forecasting purposes, we shall devote several paragraphs to their consideration.

First, as to their origin. We have already explained how a stripe can be formed which moves end on to the wind that is propelling it, but most frequently we see the curious spectacle of a long stripe of cloud moving either broadside on or obliquely to its length. As we must suppose that a stripe always sails with the wind in which it floats, we have to find out how a stripe can be formed which moves across its length. At first sight this is one of the most puzzling phases of cloud-motion. These formations of cloud are, however, exactly analogous to the smoke left by a steamer running before the wind. If she runs faster than the wind, her smoke trails behind; but if the wind blows faster than she steams, then the smoke is