and between the 24th and 29th the Garonne rose above its banks at Bordeaux, as the Lot had been swelled by constant rain. In contrast to this, only 0.51 inches of rain fell in Algiers in January, so that the dryness of that month was remarkable.

The table of the barometrical maximum on the 13th and 14th of January, at the Prussian and Austrian stations, is here subjoined. It is calculated in English inches, like that for the previous December, which has been given above. We see by it that the effect of the meeting of the two currents was more decidedly marked to the SW. than to the NE.:

The sequence of the stations in the following table is not strictly from south to north, as the extent of the area over which the observations were extended would not permit of such an arrangement :

The last station, which is at a great height, seems to show that the equatorial current was never interrupted at this level, and that the polar current, blowing in the opposite direction, was only able to effect an entrance into the lower strata of the atmosphere. Subsequent to the maximum, southerly winds replaced those from the north over the whole district.

At the first appearance of the equatorial current, in the beginning of January, the polar current, finding its way blocked up in Europe, seems to have forced a path for itself in America. This is shown by the fact that on the 5th a storm raged, for from fifteen to eighteen hours, across the whole continent, from Virginia to California, with such violence as to stop even the railway traffic. This was followed by a very heavy fall of snow, while the thermometer fell to -15°.25.

A very characteristic and distinctive mark of the storms which arise from the interference of two opposite currents is, that the fluctuations of the barometer, which are observed when the wind changes its direction, are accompanied by great variations of temperature. On the path of an advancing cyclone the direction of the wind shifts from one point of the compass to that which is opposite to it and after the lull, the barometer begins to rise

again as quickly as it had fallen before it; but the temperature remains unchanged, inasmuch as, in its advancing course, the air, moving in spirals, may pass, perhaps, twice over each station, over which the cyclone travels. Consequently, in this instance, there is no connection between. the temperature and the direction of the wind. Another important difference is, that every storm in the torrid zone travels along a clearly defined track, along which no others ever return in the opposite direction; whereas the storms of the temperate zone are often characterised by such an oscillation, backwards and forwards, as that described. Hence it is at the very outset unjustifiable, when (in cases similar to that just considered) the currents which are striving to drive each other back produce a local whirlwind, to identify this with a cyclone. In this case the expression, 'clear-weather side of a storm,' which is used by Redfield, may possibly indicate winds which, even if they have a rotatory motion, are yet not necessarily cyclones. In such an instance as this the thermometer would serve as a guide, when the vane and the barometer leave us in uncertainty as to the character of the storm with which we have to deal.

THE STORM OF JANUARY 17, 1818.

I am disposed to attribute the great storm in Eastern Prussia, to which allusion has already been made (p. 157), to a strife between the polar and equatorial currents, in consequence of the following facts:

On December 10, 1817, a very severe frost set in at Archangel, with a NE. wind, so that the thermometer fell to -19°.75 on the 11th, and -33°25 on the 14th, and the barometer reached its highest level - 30.683 inches (reduced to 32°)- on the evening of the 13th, the wind

still continuing in the E. and SE. points. At Dantzic the barometrical maximum 30.214 inches was not observed until the 17th, the wind having been at first from the N., and then from the E. The cold at Archangel was mitigated by westerly winds, which prevailed for a short period, and the temperature rose to +11°.75 by the 20th; but as the wind returned to the NE. and SE., the temperature fell to -44°, and was -38°.9 on January 1, 1818. The barometer now began to indicate the approach of a south wind by its fall; and this wind set in on the 3rd, raising the thermometer to the freezing point on the 4th, so that the change of temperature in three days was 70°-4 (from -38°.9 to +31°5). This mild temperature continued, with slight intermissions, until the barometer reached its lowest level-28.252 inches on the evening of the 16th, 2.431 inches below the maximum of the preceding month. The shift of the wind during the hurricane of the 17th at Königsberg was S., SW., W., NW.; and fourteen miles from that city there was a thunderstorm. The barometer fell from 30-788 inches on the 3rd to 28.879 inches on the 18th1.909 inches in fifteen days; while the thermometer rose from -18°8 on the 3rd to 34°7 on the 4th (53°.5 in four days). The wind shifted from E. to SW.; and the warm weather lasted so long, that we soon saw the green fields without snow. At Dantzic the barometrical maximum, on the 3rd, was 30.586 inches, and the temperature on the next morning was -4°, with a SSE. wind. In the course of the succeeding night, sheet lightning heralded the approach of the equatorial current, which came on during the night of the 6th as a S. wind, and then shifted through SSW. to W., bringing on a thaw. The barometer fell to 28.890 inches on the 17th, the wind shifting from S. to W.; but rose to 30-152 inches by the 20th, when the

wind had become WNW. Howard says that during the violent storm of the 15th, the Elbe at Hamburg rose to such a height that the lower parts of the town were flooded, and the people had to go about the streets in boats. The storm did great damage at Stettin and Königsberg. At Edinburgh there was a perfect hurricane' at 5 P.M. on the 15th, the barometer having fallen 0.8 inches in the morning. At the west end of Princes' Street, the minarets of St. John's Chapel were all blown down, great and small, so that the top of the tower was a complete wreck. Howard gives the character of the month for London as a succession of gales.' This wild weather did not cease at once; for Buek (Climate and Weather of Hamburg '), who enumerates only four hurricanes at that place in the course of thirty years, mentions one on February 27, 1818, accompanied by a high tide. Howard says of February, Winds changeable; at the end of the month stormy from the W.' The Journal of Observations at Dantzic gives westerly winds during the whole month, with a barometrical minimum of 29.009 inches on the 26th, with a temperature of about 36° or 37°, and a high W. wind, succeeding a barometrical maximum of 30-286 inches on the 17th, showing that the struggle was not even then concluded.

6

It is evident that the original disturbance of the equilibrium, which we have been unable to examine more accurately, in consequence of the want of a sufficient number of stations of observation, was due to the intrusion of the equatorial current into a district in which the polar current had prevailed for a long time, and had produced a very intense degree of cold. The whole series of storms followed from this interference. We hear from Edinburgh, on January 3, Notwithstanding the severe and uninterrupted frost, there is very little snow lying;' and