GENERAL VIEW OF THE SUBJECT.

It is most interesting to note the unity which runs through the whole class of diurnal variations. The principal maxima and minima of temperature, wind, and partially weather at 2 p.m. and 4 a.m. respectively, are strictly analogous to each other, while the semi-diurnal features of wind and weather are analogous to the diurnal variation of pressure. The latter, to which we have scarcely alluded, has two minima at 4 a.m. and 4 p.m., and two maxima at 10 a.m. and 10 p.m. respectively. The single diurnal variations are undoubtedly due to the direct influence of the sun's heat; but the question how an influence such as that which runs its course only once in the twenty-four hours can induce a variation which has a semi-diurnal period, has, up to the present time, baffled the skill of meteorologists. It is, however, perfectly certain that no one is the cause of the others; all are equally the products of the same influence, and no comprehensive theory of diurnal variations will ever be complete which does not explain and co-relate all together. When we look at a series of synoptic charts for several consecutive days, we see that many cyclones go on their course often for two or three weeks, quite independent of diurnal changes. We may, therefore, perhaps suggest the following broad view of the relation of diurnal variation to the general character of weather. The whole atmosphere is circulating between the equator and the poles. Sometimes this flow of air takes the form of eddy known as a cyclone, sometimes that known as an anticyclone, and almost always one of the seven fundamental forms of

circulation. Every day, as the sun rises and sets on this system, he impresses either directly or indirectly a series of complex variations on every meteorological element, but does not change the intrinsic nature of any form of circulation.

The results of this chapter may therefore be summed up as follows. In every part of the world the diurnal variation is superimposed on the general character of the weather, which is due to the distribution of surrounding pressure. The resulting weather is the balance of the general character and diurnal variation; the prominence of the diurnal is a measure of the settled nature of the climate of any place.

The

All over the world there is a tendency to form both a single diurnal variation, which varies only once in the twenty-four hours, and a semi-diurnal variation, which has two maxima and two minima in the same time. origin of the first is undoubtedly the direct action of the sun; that of the latter cannot be at present explained. No diurnal variation has any effect on general weather, and can be neglected in all questions which relate to forecasting general changes. This independence is one of the most important principles in meteorology.

CHAPTER XII.

ANNUAL AND SECULAR VARIATIONS.

SEASONAL APPEARANCE OF THE SKY.

THE term "seasonal variation" is used in a twofold sense. In the simpler case, it refers to the minute differences in the appearance of the sky which are found at various seasons in cyclones, etc. For instance, the rear of a cyclone does not form cumulus cloud in the dry winter months of Continental Europe; only blue sky is seen. In damp England, cumulus is formed at all seasons; but is much denser and more strongly marked in summer than in winter. In like manner, a secondary which would develop thunder in summer in Great Britain would only produce heavy rain in winter. In this way seasonal is exactly analogous to diurnal variation, for it modifies but never changes the general character of the weather. The intensity alone is ever altered.

RECURRENT TYPES OF WEATHER.

But of far more importance is that form of seasonal variation which applies to the occurrence or recurrence of

similar weather about the same date every year. The nature of recurrent weather in the temperate region of variable pressure may be best illustrated by looking at the connection between the variable European types and the regular annual changes which take place in the tropics. In most equatorial and tropical climates there are only two or three seasons, which correspond to two or three positions of the equatorial low pressure and tropical belt of anticyclones. The monsoons of the Indian Ocean are the most striking and best known instances of weather that recurs at the same season of every year.

The English word "monsoon" is, in fact, derived from an Arabic word meaning "season." But in the regions which lie between the tropical and temperate zones the author has found that there are recurrent periods, intermediate both in their duration and the certainty of their return to the monsoons of India and the recurrent spells of European weather. One of the best known of these is the "Khamsin" (the fifty days), a hot, sandy south-east wind which blows regularly in Egypt from the end of March and through April for about fifty days. Klunzinger has given a whole series of persistent and recurrent types for the whole year in Kosseir, on the Red Sea, about one hundred miles south of Suez on the Egyptian side. The following may be considered a list of the chief recurrent periods of weather in Great Britain and North-Western Europe generally.

February 7 to 10.-A spell of cold weather, associated with the northerly type. This is the first of a series of six cold and three hot periods discovered by A. Buchan. He also noticed that during the cold periods the pressure

was higher to the north of Scotland, and lower to the south, and that during the warm periods pressure was higher over Scotland than in places to the north. This means that the cold periods were the result of the occurrence and persistence of either the northerly or easterly types of weather. We have been unable to find any allusion to this spell in European weather lore.

March. The proverbial east winds of this month are mostly due to the northerly type of weather. The occurrence of equinoctial gales about the 21st of the month is almost universally believed. It is, however, a curious fact, as has been pointed out by R. H. Scott, that the records of the British observatories give no decided indications of exceptionally strong winds at either equinox. Whether equinoctial gales really occur in the Mediterranean, and the idea has been carried from thence to England by the monks, or whether the weather in Great Britain might not be more properly called broken than stormy, we cannot say. The author, however, rather inclines to the latter view; for it is almost impossible to believe that an idea which has obtained such universal credence can be altogether destitute of some real foundation. The difficulties of settling a question like this bring forcibly before us the uncertainty of any numerical estimate of climate or weather.

April 11 to 14.-A cold spell; Buchan's second period, which he has identified with the popular "weather saw of the "borrowing days."

May 9 to 14.-A cold spell; Buchan's third period. This is the most celebrated of the cold periods, as it occurs over the greater part of Europe. A good many