before the problem was solved by the intuition of Copernicus or Kepler and by the genius of Newton. When we reflect on the history of astronomy, we may well be hopeful as to the future of meteorology, though the full knowledge may not come in the time of us, or our sons, or our sons' sons, and all that we individually can do is to go on slowly and laboriously accumulating observations of every conceivable kind, not knowing when or how any one set of observations may contain the key to the problem; even as, in astronomy, the observation of certain small, apparently trivial, irregularities in the orbit of Uranus led to the discovery of Neptune.

Meantime, it cannot be too plainly understood that no student of meteorology on a scientific basis will venture to pronounce as to what the weather is likely to be three months, or three weeks-seldom even three days-in advance; that prognostications as to the coming seasons, even in a general way, such as a hot, a cold, or a wet summer, are, without exception, idle guesses, of no value to any one, except perhaps to the publishers of some weather-wise almanacks. No doubt such guesses often come true. A summer month must be hot or cold, wet or dry, or about the average; and a guess, one way or the other, has one good chance out of three. The odds are not more than two to one against it in any one place; and as prophecies, when made public, are supposed to apply generally throughout the United Kingdom, it must be hard indeed on the prophet if his accuracy does not shine forth somewhere. And each success gains many believers, people who will accept any prognostication for the future because they have known one appear to be true in the past. Where, however, prophecies are not mere guess-work, where they are based on some sort of calculation, they set odds at defiance. Every one perhaps knows that, in playing the three-card trick, the surest way to lose-when every way is pretty sure -is to watch the fall of the cards, and to stake on the evidence of the eye. Supposing the cards to be dealt honestly, a random guess has only two to one against it: a guess, sanctioned by judgement, is certainly wrong. So it is with the weather: reason and judgement almost certainly lead to erroneous conclusions. It is, for instance, still fresh in our memories how freely it was foretold last May that the summer was to be one of exceptional heat and drought, resembling the summer of 1893; or last December, that the winter was to be very wet and mild. Of the truth

of such prophecies, those who remembered them could judge in July or in February.

The fancies which have attached almost supernatural importance to the weather on certain Church festivalsnotably Candlemas and St. Swithun's-may, perhaps, have their origin in something more respectable. Observation shows that if clear, cold weather prevails in the early part of February, it is likely to last some time and with much rigour. Similarly, wet weather in the middle of July is likely to be continuous till it merges in the Lammas 'floods.' In former ages dates were more commonly defined by the name of the saint or festival than by the day of the month, and the significant weather would be noted as belonging to Candlemas or St. Swithun's, rather than to the beginning of February or the middle of July, just as we now more frequently speak of Michaelmas or Christmas than the end of September or December. But what meteorologists insist on is their absolute inability, at present, to foretell the coming weather for any longer time than is necessary for the changes reported by telegraph to reach any given place. This is merely a telegraphic extension of the range of vision. If we see signs of bad weather to windward a few miles off, we can judge that we are likely to have it within an hour or two; and similarly, if the telegraph sees them for us at a distance of 300 or 400 miles, we can judge of what may probably happen during the next twenty-four hours.

A few years ago a persistent attempt was made to extend the period by telegrams from New York. There is no doubt that these messages were carefully thought out; but their continued failure proved that it was, as yet, impossible to judge correctly from such distant indications. The mention of one rather peculiar series of errors will illustrate one difficulty in the way of such an extension. In January 1882, the barometer, here in England, stood for some weeks at about 31 inches, with the usual concomitant of remarkably fine, calm, cold weather. During this time repeated warnings of approaching storms were telegraphed from New York. Day after day these warnings were read with derision, for day after day, and week after week, the weather continued settled and very fine. It was not known till afterwards that these successive disturbances in the North Atlantic did actually come towards our shores, but were boomed off by the great pressure of the air over these islands, and were rolled down to the southward-as though along the cushion of a billiard

table-and vented their fury on the coast of Africa outside the Straits of Gibraltar. Similarly, during a great part of last January and February, a very high barometer prevailed over the Scandinavian peninsula, extending its influence over these islands, with, as before, fine, clear weather, whilst violent storms-such as made the passage of the Gascogne perilous-were raging in mid-Atlantic, and turning southwards into the Bay of Biscay, or to the coasts of Spain and Portugal.

It is probably now very generally understood that-in this country, at any rate-bad weather, storm, and rain, are associated with large whirls in the atmosphere-in technical language, cyclones-coming in from the Atlantic, following each other often in rapid succession. The rotatory motion of such whirls is invariably, in northern latitudes, anti-clockwise-contrary to the motion of the hands of a clock-round a centre of low pressure; the wind is generally violent; the rain is commonly heavy; though often, in summer, the wind is not very strong, and the rain may sometimes give place to a warm, noxious mugginess; but winter or summer, the approach of a cyclone means bad or disagreeable weather. Cyclones are seldom, if ever, stationary, but advance in every part of the world on settled routes, from which their divergence is slight, except under peculiar circumstances. Speaking with particular reference to the weather of this country, the track of a cyclone is most commonly from about west-south-west to east-north-east; so that first appearing on the west coast of Ireland, the centre of low barometric pressure will pass away over the north of England or the south of Scotland. It is not often that such a centre of low pressure passes south of the Humber, or north of Aberdeen. But the diameter of the whirl itself is commonly very large, so that the force of the cyclone is felt over the whole kingdom, and far to the southward. It will thus be seen that, as a general rule, in the English Channel and over the south of England, the wind during the passage of a cyclone blows successively from south, south-west, west, and north-west, the most violent squalls generally coming as the wind shifts from south-west to north-west. In the far north of Scotland, at the same time, the wind blows from south-east, changing to north-east.

Very different in every respect from a cyclone is its antitype, which has been appropriately named-in the first instance by Mr. Francis Galton-an anti-cyclone, a wind

system in which, in northern latitudes, the air rotates clockwise in the same direction as the hands of a clockround a centre of high pressure; but the wind is generally of no great force, the weather is fine and clear, and the anti-cyclone, once formed, is nearly stationary. If, as is frequently the case in this part of the world, the centre of high pressure is over the north of Sweden, we have in this country a persistence of easterly winds, ranging most commonly between north-east and south-east; but if the centre of high pressures moves toward the south-east and rests over south-western Russia, we get here a southerly or southwesterly wind, which yet has none of the characteristics ordinarily attributed to such winds: a wind, in fact, which has been neatly described as an easterly wind with a kink in it.

It will now be easily understood that the different parts of a cyclone or an anti-cyclone have their own peculiar weather; and that if the meteor is laid down on a map, a probable forecast of its motions and of the weather at the several places within its circumference may be made for several hours in advance. If the barometric readings reported from many stations are laid down on the map, and lines drawn joining those places where the reading is the same, it is at once seen whether these lines of equal barometer (isobars) tend to form a closed curve or not. Their shape shows whether they represent the distribution of air-pressure normal in the locality, or form part of a cyclonic disturbance or of a wedge interposed between two cyclones. The direction and force of the wind can be also drawn in to check the accuracy of the isobars; and from these delineations alone an opinion may be given. The direction of the wind is approximately along the isobar, inclining inwards; and its force is indicated by the distance from each other of successive isobars, supposed to be drawn for every tenth of an inch. By an obvious analogy, the ratio which the barometric difference bears to the distance is called the barometric gradient.' The steeper the gradient the stronger is the wind likely to be. This is a matter of observation, though as to the reason of it opinion is divided. But, so far as forecasting the weather is concerned, when the isobars are drawn on the chart, and corroborated by the direction and force of the wind, it is then possible to speak with some approach to certainty.

The normal weather system of this part of the world in winter is a barometric pressure of about 29.9 inches, slowly

becoming less towards the north, and greater towards the south, with a gentle west or south-west wind and soft blue sky, with scattered flocculent clouds: any variation from this at once speaks of disturbance and calls for examination. The first symptom of an approaching cyclone is a risepossibly a very considerable rise-of the barometer. This, if it comes from the westward-that is, is first observed on the west coast of Ireland-advancing eastward, is an almost certain prognostic, which is confirmed if the barometer, after its sudden rise, begins to fall as quickly as it rose. The attention of the observer is then given to determine the probable position of the centre of the cyclone, its probable track and the steepness of the gradients; from which, if they are known, the weather of the next twenty-four hours can be foretold.

As written down, this seems a very easy matter: in practice, it is far from being so; for, except in extreme cases, there is a great deal of uncertainty about the deductions. The gradients change; the whirl seems to lose or gain force; the barometric depression, to fill up or deepen; and the track, notwithstanding the great preponderance already mentioned, may and occasionally does differ widely from the normal. Sometimes the centre comes in by the south coast of Ireland and turns up St. George's Channel in a direction nearly due north; sometimes, after crossing the north of England in an east-north-east direction, it is borne back by a high pressure over the Baltic, and turns nearly due south down the North Sea; sometimes even it turns back, and crosses England once more from east to west. Such instances are exceptional, but they do occur; and as yet no one has been able to offer a perfectly satisfactory explanation of them. In this lies much of the difficulty, much of the uncertainty. As yet we know nothing of the cause or causes of a cyclone. We are ignorant of the cause or causes of the whirling motion, of the uniform direction of the whirl, of the low barometer in the centre. But until we have some knowledge of these, we cannot expect to have more than an empirical knowledge of the behaviour of the cyclone under different circumstances. So long as we are ignorant of the constitution of the meteor, we can have no real knowledge of the effect which changing conditions may produce. Various theories have, of course, been propounded, but they are only theories, or rather hypotheses, proof of which is altogether wanting. It may, however, be well to say a few words about them.