In the tropical regions of the South Pacific, from the barrier reefs of Australia through the numerous groups of islands to the Low Archipelago, and perhaps even to near the coast of South America, and from the equator to lat. 25° S., there is no doubt that true hurricane storms (cyclones) occur of as great violence at least as those in the North Pacific just alluded to; but from the scattered accounts of single ships, or missionary residents on the various islands, we cannot say anything positive as to their tracks, though they appear to come from the eastward amongst the islands, and sometimes to curve to the southward. The following are a few notes. The seasons at which they prevail seem also to be the same as those of the Mauritius and Bourbon.

At Viti-Levu, in the Feejee group, in February, 1841, a well-defined circular storm (cyclone), tolerably observed, seems to have moved to the southward, and, though it lasted four days, was not felt at Tonga, 8° or 10° to the S.E. of it.

At Apia Harbour, in the Samoan group (Navigator's Islands), lat. 14° S., on the 16th of December, 1840, a true hurricane storm (cyclone), of great violence, with a fall of 4 inches of the mercury (by a damaged barometer), was observed, moving from the North to the southward; and four years previous, another, also well defined, moving from the N.E. to the south-eastward, the change of wind being from S.E. to N.W. The space between the Samoan (Navigator's) Islands and Friendly Islands is said expressly to be subject to violent hurricanes, and that scarcely a year passes without some of the Friendly Islands suffering from them. Their violence is such that many of the American whalers have been made complete wrecks of by them; two were lost in 1842 (year uncertain) at the Navigator's Islands.

At the Kingsmill group, on the equator! violent storms, which appear to be typhoon-like, are experienced.

At Vavaoo, in the Friendly Islands, lat. 19° S., lon. 173° W., in 1837, the American whaler Independence was driven on shore by "a hurricane," and taken off by a shift of wind.

The account of the storm at Raratonga, in the Hervey Islands, in lat. 19° S., lon. 160° W., described by Mr. Williams, and quoted by Colonel Reid, gives us unfortunately nothing further than the certainty that hurricanes (cyclones) prevail there at times.

Mr. Thom (page 341) says:-"In December, 1842, H.M.S. Favourite, on her way from Tahiti to the Island of Mangeea, met with a storm of a rotary kind, and so severe, that the vessel was hove-to under a maintopsail. Capt. Williams was warned of a hurricane before his departure, which shows that storms of this kind are familiar to the natives."

At New Zealand there is no doubt that true rotary hurricanes (cyclones) sometimes occur, and these of considerable violence. In the U.S. Exploring Expedition, vol. ii. p. 381, is a very good account of one which occurred February 29, 1840, at the Bay of Islands, said to have been the severest which the missionaries had experienced there. It was felt at other stations, with all the veerings,

* United States' Exploring Expedition, vol. iii. p. 321.

calm centre, &c., of a true tropical hurricane (cyclone), its course being to the south-westward.*

On July 28, 1840, H.M.S. Buffalo was wrecked in a heavy gale, which lasted three days, at Mercury Bay, New Zealand. About that time also three American whalers were wrecked at Port Leschenhault, in one of the strongest hurricanes ever experienced by their commanders.

In the great space lying between Van Diemen's Land and Cape Horn, we have scarcely any observations of rotary storms, but in a capitally well-kept log of the ship Lord Lyndoch, Capt. Clapperton, now master- attendant at Calcutta, we find that in the month of December, 1820, in lat. 45° S., lon. 117° W., a gale was experienced which veered from N.W. to S.W. in fourteen hours, or about half a point an hour, in which time the ship, standing to the N. 60° E., made 83 miles on that course." If this was a rotary storm (and the barometer fell from 29.70 to 29.07), it passed here to the southward, on a track a little to the northward of West, and travelled at the rate of about 15 miles an hour. Judging from the fall and subsequent rise of the barometer, as well as the veering of the wind, there seems no reason to question that it was so.

In a newspaper article, copied from the Sydney Herald, are the following (imperfect) extracts and notes :-" So completely does the law of rotation appear to be from left to right, in gales of wind off the coast of Australia and on the neighbouring ocean, that it is scarcely possible to escape the observation, in perusing the log-books of any extended cruise. One further example to show this shall now be quoted.

"The whaler Merope left Sydney, March 22, 1840, with the wind at South, steering for Lord Howe's Island. On the 27th she was in lat. 34° 4′ S., lon 158° 35' E. The order of the wind's changes was as follows:-23rd, at S.E., veering N.E.; 24th, N.N.E. and N.E.; 25th, increasing from N.E., N.N.E., E.N.E., with a tempest; 26th, N.E. to North, with a confused sea, N. W., and drawing to West; 27th, S.W., South, S.S.E., and back to South. The wind thus completed a revolution in five days, on a direct course from left to right.

"Between Australia and America a similar course is pursued by the winds to that which is followed between the Cape of Good Hope and Cape Leemoin, and more than one instance has come before us of vessels having been driven all round the compass during a gale, not far from Cape Horn."

The following examples from Capt. Stokes' journal (Voyage of the Beagle, vol. iii.) show the general character of the gales on the West coast of South America. About the latitude of 50° S., April 5, 1828, a gale came ou from North off Cape Tres Puntas, blowing on the 6th, 7th, and 8th, from North, N.W., and S.W., with squalls, thick weather, and rain. It abated on the 9th, veered to the southward, and then to S.E., when it ceased. This was from left to right. On the 10th of April another gale came on from N.W., which as suddenly subsided in the western quarter. This, says Capt. Stokes, "was singular, for those we have experienced generally commenced at North, thence drew round to

It is said to have passed between the Bay of Islands and the River Thames at the rate of about 340 miles in thirty-six hours, or say, 10 miles an hour. Commodore Wilkes suggests that this may have been the same as that which occurred at the Feejee group, which is very probable.

westward, from which point to S.W. they blew with the greatest fury, and hauling to the southward, usually abated to the eastward of South" (page 192). These gales, therefore, rotate from left to right.

It would seem that as to violence, at least, hurricanes are felt as far South as Patagonia, where (Nautical Magazine, May, 1846) "a severe S.S.E. hurricane is said to have swept the coast from the Bay of Camaros to the Island of Desejada, occasioning the loss of twelve English and American vessels." We do not know if it was rotary, or what was its track.

In the Voyage of Don Juan de Ulloa, in 1743, speaking of the weather on the coast of South America, we find some account of storms which resemble rotary ones. In fact, he describes one which he encountered, in April, 1743, in lat. 40°S., which lasted in its full violence from March 29th till April 4th. It began at North, and twice shifted to South, returning in a few hours to North.

Tempestuous weather is equally common in the latitudes of 20° and 23° in the South Sea, as in the oceans of Europe. Along the coasts and adjacent seas the winter begins in the month of June and lasts till October or November, its greatest violence being past in August or September. Storms, which arise with great rapidity, are very frequent during the whole winter; northerly winds are very prevalent, and often of extreme violence, raising a tremendous sea. It often happens that these violent North winds, without the least sign of an approaching change, shift round instantly to the West, which change is called the Travesca, but continue to blow with the same force. Judging from this and the rest of the paragraphs, Mr. Piddington says it may be concluded that the storm tracks between 400 and 20° S., between the meridian of Juan Fernandez and the coast of South America, appear to be from westward (and probably from north-westward), toward the coasts.

From 20° S. to the equator, and thence to the Gulf of California, our information is still very deficient. Mr. Redfield and Colonel Reid incline to think that the storms on the coasts of Nicaragua, Guatemala, and Mexico,* are connected with those of the Gulf of Mexico, or perhaps originate there. In a recent memoir (American Journal of Arts and Sciences, March, 1846, No. 1, p. 164), Mr. Redfield says:

"According to Humboldt, both the eastern and Pacific coasts of Mexico are rendered inaccessible for several months by severe tempests, the northers prevailing in the Gulf of Mexico, while the navigation of the western (Pacific) coasts is very dangerous in July and August, when terrible hurricanes blow from S.W. At that time, and even in September and October, the ports of San Blas and Acapulco are of very difficult access. Even in the fine season, from October to

In a former page (part i. pp. 256-7), we have given some remarks on the winds on the Mexican coast, by Commander C. B. Hamilton, R.N. It will be seen, by referring to them, that the coast is highly dangerous in the bad season, between June and the 5th of November. The hurricanes of these months are so much dreaded that trade ceases on all the coast. That peculiar to the region is called the cordonazo, or cordonazo de San Francisco (the lash with St. Francisco's girdle), because it is thought to occur about St. Francisco's day (October 4th); but it appears that they may be expected at any time between the middle of June and the beginning of November. The squalls and gales usually commence about S.E., and fly quickly round to southward and S.W., sending in a very heavy sea, so that a ship caught at anchor off San Blas or Mazatlan, would have small chance of escape.

May, this coast is visited by impetuous winds from N.E. and N.N.E., known by the names of Papagallo (Papagayo) and Tehuantepec.

"It appears, in like manner, that the coasts of Nicaragua and Guatemala in the Pacific are visited by violent S.W. gales in August and September, known by the name of Tapayaguas, which are accompanied with thunder and excessive rain; while the Tehuantepec and Papagallo exert their violence during a clear sky. "This seems to show that the so-called Papagallo, Tehuantepec, and norther of Vera Cruz, are but the clear weather side of a revolving gale, like the north-wester of the coast of the United States; each in its turn being but part of a great vertical storm, which, in certain other portions of its area, or route, often exhibits abundance of rain.

"Humboldt suggested that these northerly winds may blow from the Atlantic and Gulf of Mexico to the Pacific, and that the Tehuantepec and Papagallo may be merely the effect, or rather continuation, of the North wind of the Mexican Gulf, and the brizottes of Santa Maria. But the vertical character and determinate progression of violent gales were then unknown, and I cannot doubt that the northers which visit the Pacific coast and the Gulf of Tehuantepec precede, in point of time, the same storms in the Gulf of Mexico, and are identical with them, having commonly in this region a northerly progression."

With this extract we will close this important subject. At some future period the navigator may be presented with something more determinate on the character of the Pacific hurricanes. The present notices will serve to draw his attention to these phenomena.

CHAPTER XXXVII.

THE CURRENTS OF THE PACIFIC OCEAN.

NEXT to a knowledge of the prevailing winds, that of the currents is the most important to the navigator. By the combined action of these two phenomena, and their effects are frequently coincident, a passage may be made in a much shorter period, though apparently by a circuitous route: and it will be the object. of the following remarks to point out what is known, and what may be anticipated to be found in the various portions of the Pacific.

Like many other branches of physics, the origin and exciting cause of currents is still involved in some obscurity, though all analogy points to the same source as that to which the regularity of the system of winds is owing-the revolution of the earth, and the consequent unequal distribution of heat. The action of the wind itself certainly has no small share in effecting the circulation, if it be not the sole exciting cause, as there is some reason to suppose. The effect of tides is

also supposed to be another element in their action, but as yet so little is absolutely known of the propagation or progress of the tidal wave, that no deductions can be confidently made from them.

Currents have been distinguished as of two classes-the drift current and the

stream current.

The drift or drift current is the mere effect of a constant or very prevalent wind on the surface water, impelling it to leeward until it meets with some obstacle which stops it, and occasions an accumulation, and consequently stream of current. The stream current is thus formed by the accumulated waters of a drift current. It is more limited, but it may be of any bulk, depth, or velocity.-(Purdy.)

That ocean waves have a progressive motion is evidenced by their forming the drift current, though it is denied by many that a wave has anything more than an undulatory movement.

There is much greater uncertainty in the ascertaining the rate of currents by the means ordinarily employed than is usually considered. It is pretty certain that much error of dead reckoning, now attributed to the effect of current, would be properly placed to other accounts. Without very great care in the navigating a ship, an exact estimate of their rate and direction cannot be made. Bad steerage, the heave of the sea, imperfection in the log or glass, uncertainty of the leeway, all tend to invalidate the estimate formed without certainty as to the extent of the allowances that must be made on these accounts.

Another point more particularly important in the Pacific is, the wide distribution of fixed points by which the dead reckoning may be positively corrected. The imperfection of observations for longitude, the errors in the rates of chronometers not ascertained but at long intervals, all tend to throw discredit on allowances for current-the usual scapegoat for all such errors.

There is one source of error which, until recent times, has passed unregarded by most, and that is, the local attraction of a ship on her compass. Invariably a compass will be directed to different points according as the direction of her head is varied. In many cases this will amount to one point or more, varying at different times and places according to the trim of the vessel or the magnetic position. This local deviation is usually greatest when the ship's head is East or West, or rather when at right angles to the magnetic meridian. With such an unsuspected source of error in laying down a ship's course, it cannot be a matter of surprise that great discrepancies will be found between her actual and calculated positions. And should all the causes of error that we have enumerated combine to act in one direction, and that direction be attributed to the effect of current, much confusion and doubt must naturally follow.

The current streams of the Pacific are not generally so strongly marked as they are in the Atlantic; and, in many cases, in the open ocean their entire effects would be comprised within the limits of the possible errors from the above sources. But we are not entirely dependent on desultory observations for our knowledge of the Pacific currents, imperfect though it be. Several navigators of high scientific character, with every appliance, have made the currents the subject of long-continued observation, and it is to their labours that we owe the positive groundwork on which the system rests.