We shall take the equatorial region first, because we want to show the nature of weather-changes in that part of the world, but not to have to recur to the subject again till the end of this chapter. On all four days the

influence on Europe. On the first day, February 26 (Fig. 68), the isobar of 29.9 ins. (760 mm.) only shows one bend northwards, while the north-east and south-east trades are separated by a calm near the equator. By

next day this bend had become more pronounced, and moved a little more to the north-east. This latter motion is very interesting, for the prevailing wind is northeasterly, but the direction of the wind has not conformed to the bend of the isobars in the manner which might have been expected.

On the third day, February 28 (Fig. 70), very great changes have occurred. Under the col which lies near Bermuda, a second bend has made its appearance so as to greatly modify the trade-wind region in the West Indies; by next day (Fig. 71) this bend has developed into a well-defined cyclone of very moderate intensity,

which moved towards the north-east, and eventually affected the coasts of Great Britain. We thus see that the details of pressure-distribution are perpetually changing in this region, but never more than a certain amount. We can therefore easily understand why the weather which is always experienced in these latitudes is described

as generally easterly, variable in strength, with the weather fine or showery according to circumstances, but never following the cyclonic sequence of the temperate zone. This modified alternation of weather is called the fluctuation of its type, as opposed to a change of type, which would involve a totally different distribution of pressure.

From this digression on the trade-winds we must now return to the cyclone-traversed region of the temperate zone and the cols of the more tropical parts of the world. On February 26 (Fig. 68), we find a fragment of a large cyclone over Norway, a V over Great Britain, some complex secondaries over the Mediterranean, and an anticyclone over the United States. Note, however,

three innocent-looking bends on the north-west edge of the Atlantic anticyclone.

By next day the Norwegian cyclone and the British V have fused or merged into an irregular cyclone which covers Scandinavia; while the Mediterranean secondaries have also formed a new cyclone, and a corner of the Asiatic anticyclone just appears near the Black Sea. Further west, the three bends in the isobars which looked so harmless the preceding day are now reduced to two, but have gained intensity. One lies to the south of Iceland, and the wedge which precedes it determines the weather for this day in Great Britain. The other, which is less intense, lies south of Newfoundland; but the American anticyclone has somewhat retreated.

By next day, February 28 (Fig. 70), the Norwegian cyclone had nearly died out, while the Atlantic cyclone, with its associated wedge, had travelled eastwards and

much increased in intensity. In connection with this, a mass of secondaries had developed over Germany and Central Europe in the col which lay between the Atlantic and Asiatic anticyclones. Similar changes are most characteristic of European weather during the persistence of this type, and a knowledge of them is of the utmost importance in forecasting. The cyclone which has come in from the Atlantic is moving and will continue to move towards the north-east, and so far it might be said that it did not affect the forecasters in Central Europe; but when we know that the passage of the depression will develop secondaries and bad weather, it is evident that the indirect influence of the Atlantic cyclone is very great. In every part of the world we may say that the passage of a cyclone in the temperate zone will develop secondaries in the tropical col over which it passes. We may also use this as an illustration of the fact that the tracking of existing cyclones plays but a small part in forecasting, as compared with the larger question of detecting influences which will make new cyclones or destroy old ones.

The col nearly over Bermuda had developed a wellmarked inflection near the West Indies.

Lastly, by the morning of March 1 (Fig. 71) all these changes had somewhat developed. The British cyclone had begun to fill up, and the European secondaries had much diminished in intensity. This is an example of what we have already mentioned in the abstract—that a cyclone which is filling up is decreasing in intensity, and vice versa. In mid-Atlantic, the bend in the isobars near Bermuda, as before mentioned, has developed into a small