the requisite inclination or gradient for the flow of the current, and the rising motion feeding the stream. In section (a) the falling motion is nearly coincident with the rising, as shown by the small arrows, the broken one representing direction of falling motion and the other the rising motion of the gradients. In section (b) these motions are at right angles, in section (c) nearly opposite and in (d) exactly so. Many of the features of cyclones are visible here. There is this difference, however, that in the whirlpool these rotating waves are stationary, as may be readily seen by any observer, whilst those in the cyclones have a progressive movement. I must not be understood as affirming that the cyclones are formed by the rapidity of the rotation alone. They are the result of the rotatory and progressive movements combined the former adding to their depth, the latter extending their area. Near the centre there is a calm area, and the steepest gradients are to be found in its vicinity.

But the most important inference to be drawn from a study of these motions is that every current presupposes both a rising and a falling gradient. It does not require much attention to perceive that to produce a current we must not only have a gradient, but a fountain or head to supply the stream. Of course in the atmosphere it is hardly possible to form a gradient without producing conditions favourable to the development of a current; at the same time it occasionally happens that extensive gradients-in some cases very steep-are formed without winds of very great strength being experienced. In every strongly marked current, therefore, we may confidently look for both a rising and a falling gradient. The south-westerly current provides a familiar example. When strong we have a rising barometer over France, and a falling motion from the north-west, the former supplying the wind and the latter forming the gradient along which the southwesterly current flows. A south-westerly current is, therefore, the product of a south-westerly rising and a north-westerly falling gradient. The necessity and utility of the rising and falling motion of the great gradients must now be apparent. I arrived at this conclusion, before I thought of the whirlpool, from an examination of the "Daily Weather Charts," and its truth will, I

think, be acknowledged on a very slight inspection of those charts which show well-developed currents.

As all our waves have a progressive movement, it is plain that the two sides of the wave at right angles to the line of progression are the most important, and we will confine our attention to these. No one who has watched the progress of the waves from the deck of a vessel could have failed to notice that the water was rising in front of the advancing wave and falling behind it, and when two waves met they caused a much higher compound wave, and after as it were a moment's tussle for the victory they passed through each other and left behind them a much deeper compound hollow. Similar effects undoubtedly take place in the atmosphere.

In treating further of these waves or gradients it will be convenient to adopt the nomenclature employed above in speaking of those producing the south-westerly current. When we have a motion from the south-east we will call it a south-easterly rising or falling gradient, according as the barometer rises or falls. A wave from the south-west will be called a south-westerly rising or falling wave, and so on. Now, as the wind, generally speaking, blows at right angles to the wave, by simply observing the direction of the wind, and the rise or fall of the barometer, we can tell how the wave motion is progressing, and also form a fair estimate of the movements of the respective gradients by considering the strength of the current and the rate and amount of the rise or fall of the barometer. Thus, a strong wind from the south-west, with a steady barometer, would indicate that the south-easterly gradient is rising as much as the north-westerly is falling; with a rising barometer that the south-easterly gradient predominates ; with a falling barometer that the north-westerly wave is the stronger.

To produce a south-westerly current we thus see that a northwesterly falling wave is required; a north-westerly current requires a north-easterly falling wave; a north-easterly current a south-easterly falling wave; and a south-easterly current a southwesterly falling wave. These currents will not be strong unless there are strong rising gradients from the opposite point of the

compass. No doubt the reaction of the air would of itself, for a short time, supply the current.

From a consideration of the whirlpool, we learn that the air of a south-westerly current must come round from the south-east; of a north-westerly from the south-west; of a north-easterly from the north-west; and of a south-easterly from the north-east. The effects of these currents must therefore be viewed in the light of the region whence they come. A south wind, coming as it does from the east, will, therefore, in winter bring a low, and in summer a high temperature, as in winter the continent has usually a lower and in summer a higher temperature than these islands.

Another principle deserves attention. It is quite evident that a current may become a wave and a wave a current. When the south-westerly current stops for some reason or other the wind usually veers to north-west, and the barometer rises; these changes indicating that the south-westerly current has become a south-westerly rising wave. A current may change into a wave either on account of the friction of the earth-perhaps more correctly the friction of the lowest stratum of air,-or on account of a stoppage in front, something there checking its course. the check is only due to friction the current will probably, after a short interval, resume its course. In a south-westerly current the south-westerly wind is seldom of long continuance, and the reason of this is not far to seek. We see the same principle exemplified in the bounding leaps which water takes in coming down a steep hill, also when a vehicle is driven down an incline, the hinder part has a tendency to rise up, and the driver requires to keep the horse's head well up as a kind of counterpoise. top of the current goes faster than the bottom, and a rising wave is the result. The wind, therefore, cannot continue long in a right line with the force of the current or tendency, but constantly veers or backs according as the wave approaches or recedes, at the same time its natural tendency is to place itself in a line with the current.

The

If the south-westerly current is thus only stopped by friction the barometer rises, and the wind veers to the north-west, the

north-west wind will rapidly die down; the wind will then back to south-west, and as the wave recedes the wind will back to south-east when the last traces of the wave have disappeared the wind veers rapidly to the south-west, blows for a short time in the line of the current, and so on. These may be called the normal changes which take place in currents, and I have already tabulated them in the December number.

In the wind itself, as distinguished from the general current, similar changes will, no doubt, occur. When its velocity is very great a wave will be formed, and whilst it is passing a lull takes place, and on its receding the full force of the storm is again felt. In the case of a very violent wind, as a hurricane, the rotating wave thus formed will be of considerable extent. So long as all the particles of a current move with uniform velocity a smooth flowing stream is the result, but when they move at different speeds waves are the inevitable consequence. The air, therefore, proceeds either by waves or currents, according to the existing conditions.

It is also evident that the current, besides being stopped by friction, may receive a check in front or at its source. In the former case, another current, proceeding in the opposite direction, is the probable cause; as, for instance, a north-easterly current meeting a south-westerly, when an area of high pressure would be formed. The two opposite waves would probably pass through each other, and leave behind an area of low pressure; and if at the same time, or rather during the formation of these falling gradients, there should be also north-easterly and south-westerly rising gradients, then gales from the south-east to the north-west would be produced.

Our most destructive storms are, I believe, brought about in this way. Two strong opposite waves meet, causing a rapid rise of the barometer; after a short time the barometer begins to fall, the waves recede, and a cyclone ensues.

The times of the oscillation of these waves have an important bearing on the results which they produce, but as the combinations which they are capable of forming are endless, I cannot enter further into the study of them here, but must now pass on to

consider the effect of the tendencies exhibited in the Forecast upon these currents and waves. According to the principles already

stated these results should be very similar to those shown in the following Table :

:

TABLE SHOWING EFFECT OF TENDENCIES UNDER VARYING CONDITIONS OF WIND AND BAROMETER.

During a strong south-westerly tendency, with the wind blowing strongly and steadily from the north-west and a rapidly falling barometer, a heavy north-westerly gale may be expected, with the wind in the south-east, a south-easterly gale. During northwesterly tendency, gales from the north-east or south-west may be experienced. During a north-easterly, gales from the southeast and north-west may be looked for and so on.

The currents or tendency due to the sun must be carefully considered as, generally speaking, the strongest gales will probably blow along its rising gradients, as shown in Figures 3 and 4, Part I., and the stormiest weather will likely be experienced when the gradients due to the sun and moon are opposite; so that in winter when the sun's gradient is for easterly winds over the