when the hurricanes were at Bermudas, the Leeward or Caribee Islands had no hurricane: nor had those islands the hurricane when Barbadoes had it.

It may be also objected, why the hurricane was never known to go farther to the westward than Porto Rico, which lies in or near the latitude of those islands of St. Christopher? To this I answer, that from Porto Rico, downwards, both that island and Hispaniola, as well as other adjacent islands, are of vast magnitude, and very high lands, that of themselves most commonly give reversal or westerly winds at night, through the year; for there, for the reasons aforesaid, the easterly wind, towards night, calms, and those lands afford a land-wind, which the other islands cannot do, by reason of the smallness of those Caribee Islands; but very near the shore, the trade-wind having its full power till this general whirlwind comes, for the reasons aforesaid. I do imagine likewise, to the southwards of Barbadoes, where the tornadoes come fre quently, there are no hurricanes; nor was there at Barbadoes, when these tornadoes commonly came there, which made some small reversal, though it was but for two or three hours: yet the easterly wind, giving some way by the sun's declining from that zenith, prevents this furious reverse, where it has no vent till it is forced by the violence of the two winds.

[Phil. Trans. 1698.

2. Hurricanes of the Indian Coast.

Dr. Halley seems to consider the hurricanes which blow occa. sionally in the month of October in the Gulf of Bengal, as of a similar nature to those in the West Indies, in which probably he is right; but at the same time it is evident, that he has been mis. informed, respecting the time they generally happen in the East. He observes that our seamen suppose them to be the breaking up of the monsoon. In this circumstance the mariners have misled the philosopher; for the hurricanes seldom happen near the change or breaking up of the monsoons, but generally many days after their commencement, and sometimes about the middle of them. Both the N.E. and S.W. monsoons blow at first in fresh gales, but neither of them encrease to violent hurricanes. It is from very obvious causes already sufficiently explained, that the one dies gradually away before the other begins. But we will first

adduce unquestionable proofs of these facts, and then endeavour to ascertain the causes of them.

The first hurricane on the coast of Coromandel, mentioned by Mr. Orme, in his History of Hindustan, was that which destroyed Le Bourdonai's fleet, after he had taken Madras in the year 1746. He attacked this fort in September, which surrendered to him in less than a month, on condition that private property should be protected. But Dupluix, the governor of Pondicherry, disputed the right of the admiral to make such a capitulation, and insisted on his seizing all property, both public and private. The correspondence on this subject, in which the virtuous admiral strenu ously defended the rights of individuals and his own honour, detained him at Madras with his squadron much longer than otherwise he intended to have stayed; and on the 2d of October came on a hurricane, which in a few hours destroyed almost the whole French fleet, and in which twenty other ships of different nations were driven on shore. One of the ships, says Mr. Orme, foundered in an instant, and only six of the crew were saved. But it must be remembered that four vessels laden with effects sent from Madras, with three others lately arrived at Pondicherry from Europe, were not affected by this hurricane; the violence of which, therefore, did not extend more than sixty or eighty miles to the southward.

On the 31st of October, 1753, Mr. Orme mentions also a vio lent hurricane on land, which was felt mostly near Wandiwash; but as the same author, who is in general equally minute and correct, takes no notice of any bad consequences happening from it at sea, we may reasonably suppose that it did no mischief either at Madras or Pondicherry, although its principal violence was felt nearly half-way between both, and not more than sixty miles in a direct line from either.

The next, which occurred during the N. E. monsoon, was on the 30th of December, 1760, during the siege of Pondicherry. On the evening of that day the weather was fair, the rains had ceased, and there were regular land and sea breezes; but a heavy swell rolled in on the shore from the S.E. The next morning the sky was of a dusky hue, accompanied with a closeness in the air, but without that wild irregularity which prognosticates a hurricane. Towards the evening, however, the wind freshened from the N.W. and at eight at night encreased considerably. About midnight the wind

veered round to the N.E. fell calm with a thick haze, and in a few minutes after flew round to the S.E. whence it blew with great violence. Almost all the ships might have been saved, had they taken the advantage of the wind blowing off the land, but the roaring of the wind and sea prevented the captains from hearing the signals for standing out to sea, and many of the ships were wrecked. The Newcastle, Queenborough, and Protector, were driven on shore a few miles S. of Pondicherry, and the crews were saved. The Norfolk, Admiral Stevens, returned next day; and on the 7th came in the Salisbury from Trinco Trincomalay S. and the Tiger from Madras N.; so that in these opposite directions of E.N. and S. the violence of the storm had not been felt. It is observed by mariners in the East Indies, that these hurricanes usually happen once in five years; but for this opinion I can find no reason, either from what I have heard from others or have myself observed.

The next in succession to that of 1760-61, was in 1763. On the 20th of October in that year, many days after the N.E. monsoon had apparently commenced, the wind began to slacken, and the clouds in the evening appeared uncommonly red, particu. larly on the day preceding the hurricane. On the morning of the 21st a strong wind blew off the land, and in the course of a few hours flew all round the compass. At this time the Norfolk man of war, Admiral Cornish, with the America and Weymouth, and the Royal Charlotte country ship of four hundred tons, remained in Madras roads, with several other country vessels. The wind began to blow from the N.W. and continued from that quarter from three or four hours, of which time the men of war availed themselves to put to sea, but it then suddenly changed to the eastward, and prevented most of the country ships from following their example. After having blown with incessant violence for fourteen hours, and with almost equal strength from every point of the compass, it at length ceased, but literally left only wrecks behind. All the vessels at an anchor were lost, and almost every person on board perished; but the men of war and the Royal Charlotte returned into the roads on the 24th. The former had felt the gale very severely whilst near the coast, but without sustaining any material injury; the latter vessel likewise, from staying rather too long at anchor, had lost her fore and main.masts, and was otherwise much da maged.

The last of these hurricanes on the coast of Coromandel, which it seems necessary to mention, is that which happened on the 29th of October, 1768. Of this sufficient notice was given, but the officers of the Chatham Indiaman, then in the road, did not avail themselves of it; for on the preceding evening the sea was violently agitated, the sun set in a haze deeply tinged with red, with every other prognostic of a gale of wind. But unfortunately there had been a misunderstanding between the cap. tain and officers, and the former being on shore, the latter, probably waiting for orders, remained at anchor, notwithstanding they might have put to sea with the N.W. wind, which, as usual at the commencement of these hurricanes, blew off the land. The governor and council, who foresaw the danger even time enough to have prevented the loss of the ship, ordered signal guns to be fired with shot, by way of directing the officers to weigh anchor and stand out to sea; but either they did not hear the guns, or were too punctilious in waiting for orders, and in consequence of this inflexibility were lost, for the ship was never seen or heard of after the close of the evening of the 29th. It is possible they were not able to distinguish the signal guns, for many of the inha. bitants of the fort, during the violence of the hurricane, did not hear them, and the flashes of the guns might be mistaken by the officers of the ship for those of lightning. The vessels, lying at this time at a single anchor in the open road of Pondicherry, were not in the least disturbed by this hurricane; neither were the effects of it in the smallest degree felt at any of our settlements to the northward. Ships which put to sea in due time very soon get beyond their influence to the eastward, and it is very well known that they never extend far inland. All these circumstances properly considered, clearly manifest the nature of these winds, or rather positively prove them to be whirlwinds, whose diameter cannot be more than 120 miles, and the vortex seems generally near Madras or Pulicat, where a branch of the Ballagat Mountains extends towards the sea. Those which happen in the N.E. monsoon generally fall with the most violence within a few leagues of this place, and never, I believe, reach to the S. of Porto Novo.

But at the commencement of the S. W. monsoon, violent gales are sometimes felt on the east side of Ceylon and the southern extremity of the coast; these however should be considered rather as

the tail of that on the Malabar coast, which extends itself over Cape Comorin, near the southern extremity of the peninsula. In that quarter, however, such gales seldom occur, and are always of short duration.

One instance only is to be found in Mr. Orme's History of a violent hurricane to the southward. In this instance, which happened on the 13th of April, 1749, near Porto Novo, on the coast of Coromandel, two of the company's ships were stranded near Cudda. lore; and the Namur, one of Admiral Boscawen's squadron, with the Apollo hospital ship, foundered. This is the only instance known to me, in thirty-five years, of a hurricane on the Coroman. del coast during the S. W. monsoon, and the effects of this were not felt beyond 11 degrees N.

On the coast of Malabar, however, this monsoon frequently blows with considerable strength at the commencement; but it must be observed that it does not begin at the same time on all parts of the coast, nor does it proceed rapidly in its course towards the N. For although the change of the monsoon generally takes place at Anjengo about the time the sun becomes vertical at that place, it never reaches Bombay before the middle or rather the end of May; the latitude of the former is about 80° 30′ N. and of the latter 19 degrees. On the 12th of April the sun is vertical at Anjengo, and about the 15th of May at Bombay. If then the difference of latitude and declination be compared, it will be found that the sun and the monsoon move almost precisely together, at the rate of about twenty miles per day: a circumstance, which above all others tends to prove that the sun's motion in the ecliptic is the primary, if not the sole case of the motion of the air, or rather of the course of the wind, at least in this part of the world, I mean on the Mala bar coast.

Monsieur D'Apres however remarks, that the N.E. monsoon, in the Mosambique Channel, begins at the north end of Madagascar, and amongst the Comero islands, in the first week of November; and at St. Augustine's Bay to the southward, at the end of the same month. If the distance of those two places in like manner be divided by the number of days, it will be found to correspond nearly with the daily difference of the sun's declination; consequently this fact will further corroborate the truth of this hypothesis; for the correspondence between the motion of the monsoon and the