anticyclone reaches so far north as to suggest some features of the northerly type. The col of low pressure below these two anticyclones is the theatre of cyclone activity, and we will now describe how the weather in Western Europe was affected by these changes. On the morning of February 25 (Fig. 76), we find the Scandi

navian anticyclone almost meeting a wedge of high pressure stretching northwards from the Atlantic anticyclone to Greenland. The pressure for several days previous had belonged to the northerly type, with an anticyclone over Greenland, which had now drifted eastwards and joined the Scandinavian anticyclone. To the

south of this at least three cyclones are found: one over the Azores, another at the entrance to the English Channel, the third over Italy. These must all be treated as belonging to the same system, as they are all formed in the same pit of low pressure. The weather, of course, is bad all over France, Germany, and Italy. The American

reports are meagre, but point to the existence of a cyclone in Lower Canada.

By next day (Fig. 77), the Scandinavian anticyclone has increased in height, while the Atlantic one has retreated nearly to its usual position. The Italian cyclone has moved a little to the north-east, while that

in the Bay of Biscay has apparently moved a very little to the south-west, and so far absorbed the Azores depression that the latter has become degraded into a secondary. Here we have the same fusion of cyclones which we have seen in all the other types, combined with the stationary character which is so peculiar to this class of weather.

Across the Atlantic an intense secondary has formed over New Brunswick, while another shallow one has pushed itself into the col between the Scandinavian and Atlantic cyclones.

On February 27 (Fig. 78) these changes have made further progress. Though the general position of the

European area of low pressure has not materially altered, the cyclones which lie within it have decreased in complexity, though a new depression has formed in a col between the Azores and the Canaries. The two American secondaries have fused into one large primary, and a large col covers the Central Atlantic.

Lastly, on the 28th (Fig. 79), while the Scandinavian anticyclone has diminished in height and area, the Atlantic anticyclone, on the contrary, has increased no less than 0.4 inch (10 mm.) in height, and much increased in size. The size and intensity of the European low pressure has diminished, but its components are more complex; so

that while weather has improved over Great Britain, it is worse in many parts of France and Italy. Across the Atlantic, all we can say is that where one large cyclone was yesterday, there are now two secondaries: one intense over Nova Scotia, another slight in the Atlantic col. This is one of the numerous cases where it is impossible to trace the exact history of pressure-changes.

The general character of the weather in Great Britain during the persistence of this type is very well marked. The sky is usually black, and, even if there is a certain amount of blue overhead, the horizon has a peculiar black, misty look, popularly known as an "eastern haze." This is quite different from the misty horizon of a calm day in the westerly type, and is associated with the peculiar bitter feel of an east wind. A well-known

saying is

"When the wind is in the east,

It's good for neither man nor beast;

and this is certainly no exaggeration.

This is the most striking illustration we can have of the general principle that no instrumental records can take the place of verbal description. We might find two north-east winds recorded automatically, of exactly the same velocity and temperature-one on the northern side of a cyclone of the westerly type, the other at the edge of the Scandinavian anticyclone in the easterly type. Read mechanically, they might be taken to be identical, while practically they are very different. Unfortunately we can give no explanation of the malignant nature of true east winds.

The temperature is generally low, but more variable