RECURRENCE.

We have already explained the tendency of certain kinds of weather to recur about the same date every year so fully in our chapter on Seasonal Variations, that it is unnecessary here to do more than allude to that great principle of meteorology. We shall, however, better understand now how the recurrence of weather is the secondary product of the recurrence of a certain type of pressure-distribution; and that to be a true periodicity of cold, for instance, it is not only cold in the abstract, but cold of the same type which must recur about the same date in most years.

DEPENDENCE.

By "dependence" of type or weather is meant the supposed connection between the occurrence of any particular type at one season of the year, and the consequent occurrence of it or of another type at another season.

For instance, there is a common saying in Great Britain, that if easterly winds prevail about the time of the spring equinox, then a great preponderance of easterly winds may be expected during the summer. Put into the language of synoptic charts and types, this means if the easterly type happens to prevail about the 23rd of March, then there will be a tendency of that type to occur more often than usual in the course of the

summer.

Again, in most temperate countries, hot summers are popularly supposed to be followed by cold winters, and

the latter are thought to depend in some way on the former. This is much more difficult to express in synoptic language, for heat and cold are not always produced by the same causes, and, unless the same type of summer is followed by the same type of winter, the apparent relation of the two seasons is illusory.

The same conception of the dependence of one season on the other is found in the tropics. H. F. Blanford has found that in India there is an apparent dependence or sequence of the summer wet season on the preceding winter rains.

At present we can do little more than note such a relationship of seasons, and cannot say whether there is even such a dependence at all. The older weather-lore seems to have been founded partly on observation, partly on an intuitive belief in the general balance of nature. In the main, the course of nature is constant; if the summer is hotter than usual, a cold winter is required to restore equilibrium, and so on for any other phenomenon of weather.

The middle stage of meteorological investigation seeks to find proof of such relation by comparing statistics of rain at different seasons. Here, of course, the great difficulty is to be certain whether all the rains which we compare are really of the same type.

The latest phase of thought would look for some connection or sequence between the forms and intensities of atmospheric eddies. All we can do is to note the facts for future research; and to remark that at the present time no use can be made of dependence of type in practical weather-forecasting.

CHANGE OF TYPE.

So far we have supposed well-defined specimens of each type, but in practice we meet with many transitional forms. Thus the southerly type may merge by insensible gradations into either the easterly or westerly, but in no case can it grow into the northerly. Similarly the westerly type may approximate on either side towards the southerly or northerly, but never jump suddenly into the easterly. In like manner the northerly and easterly types can only merge into those next to themselves on either side, but never into their opposites. This is obvious when we reflect that the types are determined by the surrounding anticyclones, and that a slight shift of one of these latter may modify the type very materially on either side, while a change to an opposite type would involve a total rearrangement of pressure over the whole northern hemisphere.

In a few cases we have been able to point out signs of an impending change of type, but unfortunately the forecaster is often confronted by very sudden alterations in the whole distribution of pressure over the northern hemisphere. Future research may perhaps some day lead to the detection of more certain symptoms of change, though at present we can say but little.

NORTH-EAST MONSOON.

But perhaps the nature of European types will be more readily comprehended if we give some illustrations of the Indian monsoons. These will be very valuable,

both as showing weather-features of a totally different character from any which we have hitherto examined, and as explaining the connection between the fluctuations of weather in the tropics and the more variable changes of the temperate regions. In Figs. 80 and 81 we therefore give isobaric and isothermal charts for January 4 and 5,

FIG. 80.-North-east monsoon; great cold.

1878, at about 6.30 p.m., Calcutta time-that is, during the season of the north-east monsoon in the Indian Ocean. These charts commence in longitude 40° east of Greenwich, where our Atlantic maps left off, and so continue on the same projection our survey of the world 60° further

east.

On both days we find an anticyclone, exceeding 31.0 ins. in height, resting over Tartary, to the east of Lake Aral. North of this, pressure slopes away towards the Arctic Ocean; southwards the pressure falls away to the equator. In fact, this anticyclone is probably the counterpart of the Atlantic anticyclone; while the low

FIG. 81.-North-east monsoon; great cold.

pressure over Southern India corresponds to the tradewind slope, which we also saw in the Atlantic. The most noticeable feature in both these charts is the persistence of the central Asiatic anticyclone and of the southern slope of low pressure, while the northern slope is more variable; just as we saw in the Atlantic.