lence of the NE. wind below, between that current and the SW. current above, but was to be found nearly in the centre of the under current, at a height of about 4,000 or 5,000 feet above the sea-level.

The fact that Goodrich found, in April, a SW. wind at the top of Mouna Kaa, while the NE. Trade-wind was blowing at the lower levels of the island of Hawaii, proves that the phenomena observed at the Canary Islands are true for all places situated at the outer edge of the Tradewind zone.

If we were able to determine by actual experiments on a series of lofty peaks the direction of the wind at various heights, we should be in a position to gain a more accurate insight into the ratio between the volumes of the direct and counter currents. The barometer gives us the total pressure produced by the conjoint action of both currents, and by means of this instrument we can see very clearly that the atmospheric pressure at the interior edges of the Tradewind zones, where the air rises, is materially less than at their outer edges, where the upper Trade-wind descends. A. von Humboldt was the first to draw attention to the diminution of pressure in the vicinity of the equator, and L. von Buch to the relatively increased pressure in the vicinity of the Canary Islands; but A. Erman and Herschel were the first to prove that the passage of the one condition into the other was gradual. It is very clearly perceptible in the new tables of observations made by the Dutch marine. We may also observe in these tables the close accordance between the variable place of ascent and the district of maximum temperature, pointed out above (page 21). The following table gives, in decimals of an inch, the differences from the mean height of the barometer observed in the Atlantic Ocean, from lat. 35° N. to 36° S., and also the limits of the zone of Calms.

Were the district of maximum rarefaction to reach higher latitudes in summer, the SE. Trade would extend still further over the equator into the northern hemisphere, and its direction would be altered through south finally into south-west, in consequence of its now passing over a series of points, whose velocity of rotation is continually decreasing. Under these circumstances, all places situated in the northern portion of the torrid zone would lie in the region of this southern Trade-wind, whose direction would have been changed to south-west, as long as the sun was on the northern side of the equator. During the winter they would lie in the northern Trade-wind, and the region of Calms would pass over them twice a year. Thus their permanent Trade-wind would be changed into a wind which changed periodically (a Monsoon). This really happens in the Indian Ocean, and we must therefore submit the laws of its winds to a special investigation.

43

II. The Yearly Periodical Winds.

1. THE MONSOONS OF THE EAST INDIES.

Ir would be more correct to describe the Trade-wind as an imperfectly developed monsoon, than to consider the latter as a modification of the former. Monsun (Malay musim) is derived from mausim, the Arabic for season. The Greeks learnt the fact of its existence during the expeditions under Alexander. Arrian says that according to the observations of Hippalus, who gave it the name of Libonotus, this wind appears in the Indian Ocean at the same time as the Etesian winds prevail in the Mediterranean, and that navigation from the ports was not practicable until the appearance of this southerly wind, which blew from the sea towards the land. In this he agrees with Aristotle, who describes expressly the regular alternation of the winds in those seas. Marco Polo first heard of it at Mangi, as the inhabitants of that place sail in winter to the Spice Islands near Zipangri (Ceylon), and return in summer with a wind in the opposite direction. The knowledge which the Arabs had of the phenomenon went very much into detail; for Sidi Ali, in his work Mohit, on the navigation of the Indian Ocean, compiled out of ten Arabic works, and published in 1554, gives the time of commencement of the Monsoon at fifty distinct places.

Halley (Phil. Trans. 1686, p. 158), describes them in the following words:

Between the latitudes of 10° and 30° south, between Madagascar and Hollandia Nova, the general Trade-wind about the SE. by E. is found to blow all the year long, to all intents and purposes, after the same manner as in the same latitudes in the Ethiopic Ocean, as it is described in the fourth remark aforegoing.

The aforesaid SE. winds extend to within 2° of the equator, during the months of June, July, August, &c., to November, at which time between the south latitudes of 3° and 10°, being near the meridian of the north end of Madagascar, and between 2° and 12° south latitude, being near Sumatra and Java, the contrary winds from the NW. or between the north and west, set in and blow for half the year, viz., from the beginning of December till May; and this monsoon is observed as far as the Molucca Isles, of which more anon.

To the northward of 3° south latitude, over the whole Arabian or Indian Sea and Gulf of Bengal, from Sumatra to the coast of Africa, there is another monsoon, blowing from October to April upon the north-east points; but in the other half year, from April to October, upon the opposite points of SW. and WSW., and that with rather more force than the other, accompanied with dark rainy weather, whereas the NE. blows clear. 'Tis likewise to be noted that the winds are not so constant, either in strength or point, in the Gulf of Bengal, as they are in the Indian Sea, where a certain steady gale scarce ever fails. "Tis also remarkable that the SW. winds in these seas are generally more southerly on the African side, more westerly on the Indian.

To the eastward of Sumatra and Malacca, to the northward of the line, and along the coast of Camboia and China, the monsoons blow north and south,- that is to say, the NE. winds are much northerly, and the SW. much southerly. This constitution reaches to the eastwards of the Philippine Isles, and as far northerly as Japan. The northern monsoon setting in in these seas in October or November, and the southern in May, blowing all the summer months. Here it is to be noted that the points of the compass, from whence the wind comes in

these parts of the world, are not so fixed as in those lately described; for the southerly will frequently pass a point or two to the eastwards of the south, and the northerly as much to the westwards of the north, which seems occasioned by the great quantity of land which is interspersed in these seas.

In the same meridians, but to the southwards of the equator, being that tract lying between Sumatra and Java to the west, and New Guinea to the east, the same northerly and southerly monsoons are observed,—but with this difference, that the inclination of the northerly is towards the NW., and of the southerly towards the SE.; but the plaga venti are not more constant here than in the former, viz. variable 5 or 6 points. Besides, the times of the change of these winds are not the same as in the Chinese seas, but about a month or six weeks later.

These contrary winds do not shift all at once, but in some places the time of the change is attended with calms, in others with variable winds; and it is particularly remarkable that the end of the westerly monsoon on the coast of Coromandel, and the two last months of the southerly monsoon in the seas of China, are very subject to be tempestuous. The violence of these storms is such that they seem to be of the nature of the West India hurricanes, and render the navigation of these parts very unsafe about that time of the year. These tempests are by our seamen usually termed the breaking-up of the monsoons.

Capper (Observations on the Winds and Monsoons, London, 1801, p. 42) gives a more accurate account of these intermediate periods

On the Coromandel coast, between the end of one monsoon and the commencement of the other, the winds are variable, partaking of both directions. Calms often last through the whole of September, and into October. As soon as the sun commences to approach the zenith from the southward, the NE. monsoon loses its power, and there is a daily alternation of land and sea breezes,-a phenomenon not observed at its commencement. At this time the wind on the coast seems to