that observed in front of a cyclone which is advancing in a direction perpendicular to the coast line. As for the Typhoons which strike the coast of China, and approach it from the SW., the point where they change their course must be on the continent, and their subsequent progress in the temperate zone could be only traced in the Japanese Seas. These Typhoons are frequently accompanied by extensive inundations of the coasts by the sea, of which Macgowan's work* contains numerous instances taken from Chinese sources. In the year 1748, 20,000 persons were destroyed by an inundation of the sea during a storm of this character. The violence which these storms exhibit even on the eastern edge of the Monsoon region is apparent from the fact, that in one storm which passed over the island Guaru,one of the Ladrone Islands, in September 1855, more than 8,000 people were rendered houseless in the space of twenty minutes by the fall of their roofs. It is moreover evident that the connection between a storm in the temperate zone, and the original cyclone in the torrid zone, to which it owes its origin, need not necessarily be traceable, as a continuously advancing minimum, in the lower strata of the atmosphere. We have explained above (p. 36) the circumstance that the predominant winds in the upper strata, between the topics, are SW., so that the identical conditions are from the very first presented to the upper part of the cyclone, which cannot be experienced by the lower part before it leaves the zone of the Trade-winds. Hence the upper portion will dilate at once, and advance in a direction different to that in which the lower portion moves. A very good proof of this connection, which was pointed out by me * On the Cosmical Phenomena observed in the neighbourhood of Shanghai during the past Thirteen Centuries.- Journal of the North China Branch of the Royal Asiatic Society, 1860, ii. 55. in the year 1842, is given by the hurricane at the Havannah on October 12, 1846, in which case the effect of the diminution of pressure has been already alluded to (p. 202),* and the hurricane of October 17, in the same year in the south of France, which was accompanied by torrents of rain and the fall of dust. The description given by Lortet of this storm reminds us at once of the description of a tropical cyclone. The tempest at La Verpellière, between Lyons and Grenoble, came over the chain of hills in the Ardeche district with a NW. wind. At the same time, from seven o'clock on the morning of October 17, the sky became extraordinarily overcast over Grenoble. The stifling blasts of a south-easterly sirocco were felt there; blood-coloured rain fell there, together with a reddish dust, which covered the diligences near Lyons to the depth of one or two lines. This occurred, according to Dupasquier, during a calm, or while the wind was from the S., and very light. The rain was not excessive; but the appearance of the sky was terrific. Two banks of clouds one in the S., the other in the NW. were the foci of the storm. The wind shifted from minute to minute; flashes of lightning of extraor - : * The fall of the barometer, for each hour, if we count the time from the period at which the centre reached the town, is as follows (in inches): Between 12 and 6 hours before, 0·077 Between 4 and 2 hours before 0.290 Accordingly, in the four hours immediately preceding the passage of the vortex, it was 1.56 inches, and in the last two alone 0.980 inches. dinary brightness crossed the sky, not vertically, but horizontally, flying round more than one-third part of the horizon. At every flash birds of passage on the wing redoubled their shrieks of terror. In the streets, in open rooms, and in the chimneys, the inhabitants caught ducks, quails, fieldfares, nightingales, flycatchers, and other birds. The fall of the barometer at the different stations is given in the following table: Bordeaux, from 13th to 16th, 0.743 inches, and up to the 20th rose 0.711 inches. According to Ehrenberg's microscopical analysis, there is nothing either in the internal or the external character of the dust which fell on that occasion which would lead him to attribute its origin to Africa. On the contrary, many of the forms therein contained were, either mainly, or in a great degree, indigenous to South America. It could not have come from the interior of a continent, but must have been carried from the sea-coast (if we may assume that it all came from one district), inasmuch as it contained existing marine forms.* This case which we have just described forms a very natural introduction to the consideration of the storms which arise outside the zone of the Trade-winds. *Berichte der Berliner Akademie der Wissenschaften (Reports of the Berlin Academy of Science), 1846, 227. 221 II.-Storms which arise at the exterior edge of the Zone of the Trade-winds. In the preceding pages we have referred the West India hurricanes to the interference of lateral cross-currents with the upper Trade-wind on its return from the equator, portions of which, being forced to enter the lower strata of the atmosphere, meet with a constant wind, moving in a direction opposite to their own, and thus produce a cyclone. Outside the Trade-wind area the upper current descends to the surface of the earth, and is predominant there in different districts at different times, while the under-current in the opposite direction is not constant. Here, then, we shall find that the conditions of interference will constantly be presented, but the currents will be directly opposed to each other, so that they will only check each other's progress. In this case, a partial gyration indicates nothing more than that one of the currents has at last overcome the resistance of the other, a success which is just as liable to be reversed as not. In all such cases, the warm equatorial current will produce a considerable depression of the barometer, inasmuch as in its progress into colder regions it loses the moisture which it had brought with it, in repeated discharges of rain. This fall of the barometer will be observable along a line perpendicular to the direction of the current, rather than at the centre of a rotating mass of air. In the case of cyclones the height of the barometer increases in all directions from the centre; in the case now under consideration the barometer will rise to a great height over an extensive area, affected by the polar current, which has been checked in its progress. This contrast of two districts, with respectively high and low levels of the barometer, will also be marked by great differences of temperature between the regions in question. The investigation of such phenomena requires observations from a great number of stations scattered as thickly as possible over an extensive area. In the consideration of the preceding class of storms, we have derived our information principally from the logs of ships, but in this class we are referred, by the nature of the case, principally to observations made on land. In a larger work of mine,* I have described at length several instances of this interference of and strife between two opposite currents, of which I shall quote a few cases here. * THE CONFLICT OF OPPOSITE CURRENTS IN JANUARY 1850. The month of November 1849 had been very warm throughout Germany, but at the end of the month an intense degree of cold, accompanied by easterly winds, was felt in European Russia and the east of Germany. This was broken for a short time, in the middle of December, by southerly and westerly winds; but towards the end of the year it recommenced, and in January, with easterly winds, reached a degree of intensity in eastern Germany such as had not been recorded at many of the stations since the observations commenced. At Posen the thermometer fell to -37°7 Fahr. ; at Bromberg * Darstellung der Wärme-Erscheinungen durch fünftägige Mittel von 1782 bis 1855, mit besonderer Berücksichtigung strenger Winter (Representation of the Phenomena of Temperature by Five-Day means, from 1782 to 1855, with a Special Reference to Severe Winters). Berlin, 1856, folio, 113 pp. |