There are other differences, but the one on which our author lays most stress, as the almost infallible criterion of 'true convergence,' is found in the direction in which the cirri move; if they come from the point of convergence, this 'convergence is real, because there are hardly any instances 'of parallel cirri moving in the direction of their orienta'tion.' It is this convergence that Fr. Algué considers the distinctive 'distant signal' of a cyclone, and he gives the following interesting experience of it as such:

'When I was in Havana in February 1893, I noticed with Fr. Viñes an arborescent ramification of cirrus and cirro-stratus, which was convergent toward the N.N.W. The great persistence of this convergence and a quick shift of the great tuft of cloud toward the north and north-east convinced us that we had to do with cirri radiating from a centre which was crossing the North American continent in higher latitudes. On the following day the telegraph informed us that a terrible snowstorm-a veritable cyclone-had passed at a considerable distance to the southward of New York; the centre was over 600 miles distance from Havana.'

In this deduction Fr. Algué differs-in our opinion correctly from Fr. Froc, who, in an essay on "The "Iltis" Typhoon,' * says:

The first sign is incontestably the swell, for the appearance of cirri, a valuable indication pointed out some years ago by the Rev. B. Viñes, S.J., for the Antilles, and the Rev. F. Faura, S.J., for the China Sea, cannot really be given seriously as a general forerunner for the seas of the Far East, at least to the north of the 28th parallel; these clouds may appear when no typhoon is in existence, and also typhoons make their appearance without being preceded by such cirri.'

On which Fr. Algué comments:

If we admit that in the central part of a cyclone ascending aërial currents exist in such a way as we have explained, we cannot for a moment deny the existence of cirrus rising from the centre; whether the cirrus is visible or not depends on the amount of clouds.... Neither Viñes nor Faura mentions that all cirri are indications of a cyclone; they are so only under certain circumstances-viz. in case of their being convergent and then showing a certain definite disposition.'

And on the principle that a handful of practice is worth a bushel of mere theorising, he refers to his own observations of cirrus at the very time of this particular typhoon,

* Zi-ka-wei, 1896.

which did not itself visit Manila. On July 17, 1896, he recorded

i

an arborescent ramification of cirrus convergent towards the east; and on the afternoon of the 18th a convergence of the cirrus clouds towards E.N.E. was noticeable. Observations which arrived subsequently confirmed the idea that there was a centre of a cyclone on those days and in the indicated direction. ... The cyclone took a northerly direction after the 17th; we were able to conjecture this by the movements of the radiating point of the cirrus clouds. It shifted from east to E.N.E. and N.E. These are movements which cirrus, under normal conditions, never shows, and hence they furnish another confirmation of the conjecture that these cirri were of cyclonic origin. The centre of the cyclone moving northward in all probability came closest to the north coast of Luzon on the afternoon of the 18th. There is no reason to suppose that a cyclone loses its apparatus of cirri whilst it moves in higher latitudes; on the contrary, they probably continue to issue from the vortex, though local circumstances, especially the heavy cloud covering of the sky, may often interfere with observation. This was the typhoon which travelled as far as the Yellow Sea, and which obtained a melancholy notoriety through the loss of the German gunboat "Iltis." "

When Fr. Froc put together the pamphlet from which we have quoted, he had not seen Fr. Algué's observations at Manila; and having misunderstood the suggestion of Viñes and Faura as to the interpretation of the cirrus, he did not push his inquiries in that direction. Now that we have this remarkable proof of the value and importance of these observations, there can be little doubt that it is to them navigators must look for early warnings of an approaching typhoon, or even-universally-of an approaching cyclone; and as cirrus can probably be more often seen, and can certainly be more accurately observed on shore than at sea, arrangements ought to be made for a speedy diffusion of the information.

There is, however, one point with regard to the observations of cirrus on which Fr. Algué is not quite clear. It almost seems as if he held that their indications are considerably in advance of those given by the barometer, counting them, apparently, from the beginning of its distinct fall. But it has long been known that within the tropicswhere, in normal circumstances, the fluctuations of the barometer are so regular that they may almost be used to set a watch by-the first indication which the barometer gives of an approaching cyclone is a certain small rise out of due course, a rise, in fact, which may sometimes appear as an irregular fall. Theoretically it would seem as if this

and the cirrus ought to accompany each other; the moist air thrown out into the upper regions of the atmosphere, and radiating from the centre, gives the peculiar formation of the cirrus; the same air piled up outside the whirlcorresponding to the overflow if the water in the experimental beaker is stirred too briskly-shows itself in the rise of the barometer; and thus M. Doberck says:

The earliest signs of a typhoon are clouds of the cirrus typelooking like fine hair, feathers, or pale white tufts of wool-travelling from the east or thereabout, their direction backing toward the north; a slight rise in the barometer; clear and dry but hot weather; calms or very light winds. If cirrus come from the west they prove that there is no typhoon.'

But in saying this he does not in any way depreciate the value of the swell of the sea as an indication, which, he says, may be noticed at a distance of fully 600 miles from the centre if no land intervenes. Evidently the Philippine Islands shelter the southern part of the China Sea, but further north M. Doberck indirectly confirms Fr. Froc's statement made, we may assume, with reference to his own station near Shanghai. He says:

'North of Formosa, i.e. between the east coast of China and southwest Japan, there is usually a heavy E. to S.E. swell when a typhoon centre is approaching from S.E. or E., and the sea there gets very high when the centre is yet at a great distance.'

It is, however, the extreme importance of the cirrus observations on which we wish to insist; the subject calls for exhaustive discussion, and may well be commended to the most careful attention of observers. There is no reason to suppose that, even in temperate latitudes, the cirrus, if it could be seen, would not give early indications of cyclonic weather. Observations of it would thus have a very real commercial value, but meteorologists are cautious and refuse to be hurried in their forecasts; shipowners are willing to take the risk and dislike new ways, and the Treasury readily acquiesces; so the one high-level observing station in Great Britain has, within these last few months, been closed for want of funds.

All this belongs to a timely knowledge of the approach of a cyclone, and has a two-fold value when rightly understood: it is not only a warning against coming bad weather, but is also the indication of a possible persistence of favourable weather. According to M. Doberck:

"The fine weather [which tells of a distant typhoon] lasts for days, and the existence of a typhoon at a great distance is the cause of fine weather all round, and contributes therefore to the safety of ships at sea, a fact that is not sufficiently appreciated by mariners.'

Naturally the mariner's first impulse is to make all snug in a safe harbour and stay there till the danger has passed; but he may thus lose a grand opportunity.

In all cases, cyclones follow the drift of the prevailing wind, and the typhoons of the Far East come in from the Pacific, following the course of the trade wind, and with that curving towards the north into the Yellow Sea, or frequently recurving towards the north-east, along the south coast of Japan; the path in such cases, when laid down on the chart, having a general resemblance to a parabola. The agreement is so close that it is impossible to avoid the suggestion that we have here a clear case of cause and effect, and that the whirling wind is borne along in the great body of the 'trade' in the same way that an eddy is borne along in a millrace. But as the whirl itself and its peculiar violence are due to several causes, some mechanical, some physical, so it may very probably be that there are various causes conducing to lead the cyclone on its path, and that, irrespective of the course of the trade wind, it is guided by the continual indraft of vapour and the consequent condensation in front, which, as it were, sucks it on towards the maximum condensation, which, in the Atlantic, lies along the course of the Gulf Stream, where the familiar name of the weather 'breeder' tells of its association with the bad weather; or in the Far East, along the course of the corresponding stream of the Black Water, the Kuro Siwo. The reflex pressure of the relatively stagnant air over the continent, whether of America or of Asia, must act in the same direction, and the path of the meteors, though primarily traced out by the course of the trade wind, is confirmed by several concurrent causes.

As the centre approaches any particular locality, the bad weather begins. When within 500 miles, the barometer falls, slowly at first; the air is close and very hot;* within 300 miles there is a very heavy cross sea and the temperature begins to fall; within 200 the temperature falls quickly. Within about 150 miles heavy rain begins, and at about 60 miles it falls in torrents; the temperature near the centre in Hong Kong is often about 78°, and over the China

*Cf. the 96° recorded on board the Glory' (ante, p. 220).

Sea 76°. As a general rule, the wind increases in force as the distance of the centre diminishes. M. Doberck gives the following figures as a sort of average, the strength of the wind being represented by the numbers of the Beaufort scale, from 0 to 12. At a distance of 250 miles, the wind has a force 5; at 220 it is 6; at 180, 7; at 145, 8; at 110, 9; at 75, 10; at 50, 11; and within 35, 12. But frequently the wind is very much stronger, and towards the centre has a violence far beyond anything denoted by the scale. This violence has long been proverbial, so that, by a customary exaggeration, any strong wind is often called a hurricane by people who have no clear notion of what a hurricane really is; but after all, the special danger of it lies not so much in the violence as in the accompanying unsteadiness both in force and direction. On this, M. Doberck gives some interesting notes, intended more immediately as a caution to mariners.

'When the wind rises in a typhoon,' he says, 'it blows in gusts, and the mercury heaves in the barometer ["pumps" is the more usual expression]. When the wind has reached force 11 it blows in fierce squalls of sometimes about ten minutes' duration, while the mercury heaves up and down as much as a tenth of an inch. The mercury often gives a jump upwards as the wind begins to veer in a squall; then it drops down and gives another upward jump as the wind comes back to nearly its old direction. . . . At the time when the centre is nearest, a fierce squall is usually felt, and in that squall the direction of the wind changes considerably and the barometer begins to rise. The squalls appear to be caused by an up-and-down movement of th: air.'

It is, indeed, difficult to explain the extraordinary violence of the squalls and the low temperature in any other way. When, however, M. Doberck goes on to speak of details outside his special knowledge, we have not that entire confidence in his statements which we are willing to give to his meteorological discussions. He says, for instance-'It is a fact that more damage to vessels is caused by the fearful seas than by the wind'; but we believe the fact to be exceedingly doubtful, and in whatever degree it is true, it can be so only in respect of steamers; it is certainly not true of sailing ships. And when he continues-Fortunately masters of ships are now making use of oil to calm the waves, a remedy which has been used with good effect for the last three thousand years,' he is surely drawing largely on his imagination, or at least representing what he thinks ought to be as what is; for it is not now nor has it been at any