66

This line is called the "trough" of the cyclone, because if we look at the barometer-trace at any one place, the ups" and "downs suggest the analogy of waves, so that the lowest part of a trace may be called a "trough." Or we may look at the cyclone as a circular eddy, moving in a given direction, and so far presenting some analogy to a wave. Here we are face to face with the primary difficulty of understanding synoptic charts. When we look at any chart of a cyclone which represents the state of things existing at some one moment, there is little to suggest the idea of a trough, because the latter depends on the motion of the cyclone, which cannot be shown on a chart. Perhaps the following illustration may help to explain the nature of the trough. Suppose the cyclone represented the inside of a conical crater, if we walked along the line that marks the path of the centre from the word "FRONT" to the word "REAR" on the diagram, we should pass over the centre of the crater, and be walking downhill all the time till we reached the bottom, and uphill afterwards. But now, if we walked across the crater on any other line parallel to this one, say from the word "pale" to the word "COOL," we shall equally walk downhill till we arrive at the point occupied by the letter q in the word "squalls." At q we should still be on the side of the crater, and some distance from the centre, but after passing q we should begin to walk uphill.

When we have once realized the meaning of the trough, we shall never fall into the very common error of thinking that because our barometer has begun to rise, the centre of the cyclone has necessarily passed over us. It is probably only the trough, but we will explain

afterwards how we can tell whether it is the centre or not.

So far for the shape and names of the different portions of the cyclone. Now for the wind. A glance at the arrows will show that, broadly speaking, the wind rotates round the centre in a direction opposite to the motion of the hands of a watch. That is to say, that in the extreme front, following the outer isobar, the wind is from the south-east; further round, it is from the east-north-east; still further, from the north-north-west; then from about west; and, finally, from the south-west. Then we note that in front the wind is slightly incurved towards the centre, and therefore blows somewhat across the isobars, while in rear it has little or no incurvature, and blows nearly parallel to the isobars. The velocity or force of the wind will depend on the closeness of the isobars. In the diagram they are much closer set in rear than in front of the cyclone, and therefore the wind is strongest behind the centre.

In our chapter on clouds we shall have to go much more minutely into the nature of wind, both on the surface and in the upper currents; but here we wish to confine our attention as far as possible to the weather and appearance of the sky.

For the same reason, the details of gradients will not be developed till we come to our chapter on wind and calms.

The weather in a cyclone is somewhat complicated. Some characteristic features depend on the position of the trough, and have nothing to do with the centre. For instance, the weather and sky over the whole front of the

cyclone—that is, all that lies in front of the trough—is characterized by a muggy, oppressive feel of the air, and a dirty, gloomy sky of a stratiform type, whether it is actually raining or only cloudy. On the other side, the whole of the rear is characterized by a sharp, brisk feel of the air, and a hard, firm sky of cumulus type.

But, on the contrary, other characteristic features are related to the centre, and have little to do with the trough. The rotation of the wind, though slightly modified near the trough, is in the main related to the centre, and the broad features of the weather in a cyclone are—a patch of rain near the centre, a ring of cloud surrounding the rain, and blue sky outside the whole system. The centre of the rain-area is rarely concentric with the isobars. It usually extends further in front than in rear, and more to the south than to the north, but is still primarily related to the centre.

This will be readily seen by reference to the diagram; there the drizzle and driving rain extend some distance to the right front, while almost directly behind the centre patches of blue sky become visible. Thus a cyclone has, as it were, a double symmetry: that is to say, one set of phenomena, such as warmth, cloud character, etc., which are symmetrically disposed in front and rear of the trough; and another set, such as wind and rain, which are symmetrically arranged round the centre. There is reason to believe that what we may call the circular symmetry of a cyclone is due to the rotation of the air, while the properties which are related to the trough are due to the forward motion of the whole system.

As this is a somewhat difficult conception, perhaps

the following analogy may not be out of place. Let us consider the twofold distribution of the population of London. As regards density, we find a comparatively thinly populated district in the centre of London—that is, in the City proper. Round this there is a tolerably symmetrical ring of very densely populated streets, outside of which the population thins away towards the suburbs. But at the same time London is divided into very well-defined halves of comparative poverty and wealth-the east and west ends respectively. This is a far more strongly marked distinction than any which is found between the north and south sides of London, in spite of a river that might have been supposed to make a natural boundary. This distinction into an east and west end is always attributed to the general march of the population westwards. Thus the front and rear of the moving population have a symmetry independent of the density of the population round a centre.

Returning now to the details of weather in a cyclone, we have marked on the diagram the kind of weather and cloud which would be found in different parts of a cyclone. The first thing which will strike us is that the descriptive epithets applied to the sky contain the phraseology of the most familiar prognostics. At the extreme front we see marked "pale moon," "watery sun," which means that in that portion of a cyclone the moon or sun will look pale or watery through a peculiar kind of sky. But all over the world a pale moon and watery sun are known as prognostics of rain. Why are they so? The reason we can now explain. Since a cyclone is usually moving, after the front part where the sky gives

D

a watery look to the sun has passed over the observer, the rainy portion will also have to come over him before he experiences the blue sky on the other side of the cyclone.

Suppose the cyclone stood still for a week, then the observer would see a watery sky for a week, without any rain following. Suppose the cyclone came on so far as to bring him under a watery sky, and then died out or moved in another direction, then, after seeing a watery sky, no rain would fall, but the sky would clear. The prognostic would then be said to fail, but the word is only partially applicable. The watery sky was formed and seen by the observer, because he was in the appropriate portion of the cyclone, and so far the prognostic told its story correctly-viz. that the observer was in the front of the rainy area of a cyclone. The prognostic failed in its ordinary indication because the cyclone did not move on as usual, but died out, and therefore never brought its rainy portion over the observer. This is the commonest source of the so-called failure of a rain-prognostic in Great Britain. The reason why all rain is not preceded by a watery sky is because there are other sources of rain besides a cyclone, which are preceded by a different set of weather-signs. Such is the whole theory of prognostics.

The same reasoning which applies to a watery sky holds good for every other cyclone-prognostic. We shall have explained why any prognostic portends rain when we have shown that the kind of sky or other appearance which forms the prognostic belongs to the front of the rainy portion of a cyclone. Conversely we shall have explained why any prognostic indicates finer weather when we have shown that the kind of sky belongs to the rear of a