[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

in a north and south line, and at a distance of about two hundred miles apart. The progressive eastward movement of the whole system is indicated by the compression of the lines on its eastern or advancing side. Each circle represents 2/10 in. of pressure, or from 30 to 30.4. The small arrows show wind-rotation and the large feathered arrow the forward movement.

Anticyclone (Southern Hemisphere). Fig 1

It will be seen that at a pressure increases from 30 in. to about 30.3 in., the wind changing from the southward toward east, and that when the maximum pressure has passed to the right of–i.e., to the eastward of–a, but also lies to the southward of it, pressure decreases to 30 in., with the wind changing to north-east; at b pressure increases to 30.4 in., the wind continuing in the south, because the position of b is such that it lies directly in the route of the anticyclone maximum pressure, and when this has passed to the eastward of b the wind comes from north and pressure decreases to 30 in., but at c pressure increases only about 2/10 in. or to 30.2 in., with south-west wind changing toward west; and as the maximum pressure passes to the eastward of c, but lies also to the northward of it, pressure decreases to 30 in., with wind changing to the north-west. It will be seen by this figure that, although the wind-rotation of the whole system is from right to left, the changes in the direction of the wind as experienced at a and c are in quite opposite directions.

Nearly all anticyclones have more than one maximum area. This is shown by the occurrence of partial decreases of pressure, accompanied by changes of wind to the northward, and similar to that which has been shown to occur when a high pressure passes away and is followed by partial increases, accompanied by southerly wind, corresponding to those shown as taking place on the first approach of high pressure. Diagram No. 2 shows an anticyclonic pressure curve of the Southern Hemisphere containing three maximum points: at d pressure is at 30 in., with southerly winds, and increases to e; then decreases with northerly winds to f, followed by a second increase and southerly winds to g; then decreases with northerly winds to k, and increases with southerly winds to m, finally decreasing with northerly winds to 30 in.

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

Fig. 3 shows a short series of these waves with preceding and following high pressures. Each vertical space represents one day; each pressure-line represents 2/10 in.; the small arrows show the wind-movement, and the long feathered arrow indicates the movement of all these pressures from west to east.

Wave-Pressures (Southern Hemisphere). Fig 3.

Using the same positions as in fig. 1, the rear or left-hand side of the anticyclone is shown as now being indented by the advancing depression D, and as this progresses eastward the wind at a changes to the north-west, pressure continuing to decrease slowly until a little below 29.80, when the wind comes from the west; but at b the decrease is to 29.60, also with north-west changing to west, while at c pressure falls to 29.35 with the same wind-movement. As D passes to the right or eastward pressure increases with wind from the south-west at all three stations–a to 30 in., b to 29.80, c to about 29.65. This change of wind–i.e., back to west from the south-west, or from left to right–is the reverse of that which precedes it, and, as E is now approaching, this backing movement is continued further towards the north at each position; it causes a second decrease of pressure which at a falls to about 29.70, at b to about 29.45, and at c to about 29.10. The wind-movement has now been from right to left at all three places, and south-west winds with rising pressure occur again as E continues its easterly movement, the change at a being an increase to about 30, showing that this place is now within anticyclonic pressure; but b and c remain within low - pressure limits, the increase being at b to 29.90, c to 29.80, with the wind back again to west. The approach of F causes pressure to decrease again with

movement of an anticyclone, followed by a cyclone from a north-westerly direction. Each vertical space represents one day. Each oval shows a decrease of two-tenths pressure toward the central one. The small arrows indicate the wind-rotation; the long feathered arrow on the left shows that the cyclone is travelling towards south-east, and the one on the right that the anticyclone is progressing eastwards. The left-hand side or rear part of the high pressure is shown slightly indented as the effect of the approaching cyclone, whose front curves are compacted in the line of its advance.

Cyclone and Anticyclone (Southern Hemisphere). Fig 5

Using three positions–viz., a, b, and c–situated upon a line from north-east to south-west, the diagram shows that the movement of the high pressure to the eastward has caused a decrease of pressure accompanied by northerly winds at each station, and that as the cyclone approaches pressure continues to decrease, and has at a fallen to 29.90, wind about north-north-east; at b to 29.85, wind north-north-east; and at c to 29.90, wind north-east. The continued advance of the cyclone causes further changes; at a the decrease is to 29.60, wind north-west; at c the wind has changed from north-east through east to south - east, and pressure has fallen to 29.40. The minimum pressure of the cyclone is now at its nearest to both of these positions, but at b, which lies directly in the track of the cyclone centre, pressure has fallen rapidly to 29.05, and the wind, which remained at about north-east, varies rapidly from all directions while the centre is passing over b, finally steadying in the south-

the coast, because they do not all originate in the same place, and also some travel into higher latitudes than others. There are really three routes: (1.) Approaching from near Norfolk Island towards but passing eastward of North Cape, and continuing thence toward south-east they pass to the northward of East Cape. (2.) From near Lord Howe Island travelling south-eastward towards Cook Strait, or to some point on the west coast of the South Island between Cape Farewell and Hokitika (but also occasionally further southward), and, crossing the South Island, continue towards south-east. (3.) Approach East Cape from about north or north-north-west, bur passing eastward of it, and when southward of it take a more south-easterly direction.

The lowest pressure of a cyclone travelling on about route (1) usually passes North Cape about two days and a half after passing Norfolk Island; those of route (2) appear on the west coast about four days after passing the meridian of Lord Howe Island; and those of route (3) have occupied about five days from the western part of the Fijian Islands. These periods are sometimes apparently accelerated by the expansion of the storm-area. Cyclones have sometimes two and occasionally three areas of minimum pressure. This is shown by there being only a partial restoration after the first decrease, and this movement is followed by the wind soon changing back to the direction it had at first, accompanied by another and immediate decrease of pressure, these changes being sometimes repeated a third time. The pressure curves registered under these conditions have a resemblance to those of the westerly wave-pressures, but with this difference: that of two successive minima in a cyclone, the second may equal but never exceeds the first, and the third is always the least. These disturbances have always increased considerably in area after leaving tropical latitudes, and have lost the extreme wind-velocity they had when nearer to their point of origin; but some that arrive here have winds of nearly hurricane force, and include the whole country within their area.

The upper and intermediate clouds which precede and accompany these low pressures are cirrus, cirro-stratus, and alto-stratus, and when the latter cloud is conspicuous, remarkably well-defined halos with decided prismatic colouring are seen. All these clouds come from the eastward, are generally moving fast, and the ends of the cirrus are curled back toward the direction they are coming from. The sunset colour with these disturbances is always of a decided yellow, with occasionally a tinge of orange towards the south. The yellow sometimes deepens very quickly, and in such instances all colour disappears almost immediately afterwards.

movement of the belt shows that the wind-currents of the Equatorial low pressure must mainly travel from the eastward. The mechanical movement is continued on the polar side of both the high-pressure systems by belts of westerly winds which extend round the earth; but, owing to the continual changes in the dimensions of the high pressures, these belts of westerly winds are of varying width and sinuous form, causing deflection of the westerly current to between south-west and north-west in the Northern Hemisphere, and to between north-west and south-west in the Southern Hemisphere, resulting in wavelike depressions whose route is towards east. Pressure within both of these belts decreases polewards, and their latitudinal limits extend from the edges of the anticyclonic belts to between the 60th and 70th degrees of both north and south latitudes.

Cyclones originate within the limits of the Equatorial belt, and in both hemispheres travel first towards the west, those of the Northern Hemisphere curving gradually toward north and east, those of the Southern Hemisphere toward south and east. Some of these disturbances have sufficient energy to force their way between the systems of the anticyclone belts and into the regions of the westerly winds, to which they convey both heat and moisture.

Continuing the consideration of mechanical effort, it is evident that the rotation of the two discs is such that the movement of their poleward edges is towards the right–i.e., to the eastward–therefore a belt resting on these edges must move parallel to them, and to this movement the west-wind belts correspond.

On the northern side of the west-wind belt of the Northern Hemisphere, and on the southern side of that of the Southern Hemisphere, pressure is lower than in all other regions, and therefore it should seem that there should be continuous westerly winds circulating round a minimum pressure at each pole. But it is well known that easterly winds are experienced in very high latitudes, and from this it is evident that some other arrangement of winds and pressure exists within these areas. Further consideration of the mechanical effort already mentioned will show that the movement of the west-wind belts must influence whatever pressures he poleward of them, and that their action on a disc lying between them and the pole would be to cause the disc to rotate from right to left–i.e., from west to east–and therefore the edge nearest the pole must move from east to west. This rotation corresponds to the wind-revolution of the anticyclone of the opposite hemisphere, and must be accompanied by an increase of pressure toward the centre of the disc, although pressure within that area is low.

The atmospheric circulation of both Poles should therefore be principally anticyclonic, but with lower pressure than in other parts of the world, and consists of a series of anticyclones having progressive movement towards east, their limits poleward being in very high latitudes, probably to between 85° and 87°.

But there are other easterly winds which, though they may not occur frequently, have been experienced in high latitudes within which westerly winds predominate–i.e., on the poleward edges of both west-wind belts–and they therefore do not belong to any system of anticyclonic winds. Now, precisely the same effect–viz., easterly winds in west-wind regions–is produced in temperate latitudes when a cyclone from equatorial latitudes has advanced sufficiently far to carry its easterly winds into the west-wind belt.

There thus seems probability that there are low pressures within both Arctic and Antarctic Circles, and if so their wind-rotation must be cyclonic, and corresponding to that of the cyclone of the opposite hemisphere, the area within which they originate being near the Poles–arctic cyclones travelling first from the Pole toward west, then curving towards south and south-east; antarctic cyclones travelling from the Pole toward west, curying toward north and north - east. These low pressures would convey cold and moisture, as snow, into the west-wind belts, but though cold they would be warmer than the anticyclonic systems of either Pole. Between the poleward edges of the anticyclones and the Poles there should be variable winds

If the mechanical principle upon which the horizontal circulation of the atmosphere is arranged be such as has now been described, then the corresponding vertical movement should be that all anticyclonic systems are descending currents, and all others are ascending currents.

The governing principle regulating Horizontal atmospheric circulation may be described as being that which controls the movements of three belts placed alternately between four circular discs, which press upon them sufficiently to insure friction, as in fig. 7, in which it is evident that if the central belt be pulled toward the left-hand side of the figure all the other belts and discs must move in accordance.

The governing principle of Vertical atmospheric movement should be that which controls the uptake and downcomer circulation of the multitubular boiler.

The influence of the sun upon the whole atmosphere is to cause all its systems to move southward from the time he has attained to his maximum north declination until he has reached his maximum south declination, from which