tude at noon 28° 7′ N. Next day the heat was only 71°; we were then in latitude 29° 12′; the heat of the water, therefore, was now leffening very faft in proportion to the change of latitude. The 25th our latitude was 31° 3'; but though we had thus gone almoft 2° farther to the northward, the heat of the fea was this day rather increafed, it being 72° in the morning, and 72° in the evening. Next day, the 26th of April, at half after eight in the morning, I again plunged the thermometer into fea-water, and was greatly furprised to fee the quickfilver rife to 78°, higher than I had ever obferved it, even within the tropic. As the difference was too great to be imputed to any accidental variation, I immediately conceived that we must have come into the Gulf-ftream, the water of which ftill retained great part of the heat that it had acquired in the torrid zone. This idea was confirmed by the fubfequent regular and quick diminution of the heat; the fhip's run for a quarter of an hour had leffened it 2°; the thermometer, at three quarters after eight, being raifed by feawater fresh drawn only to 76°; by nine the heat was reduced to 73°, and in a quarter of an hour more, to 71° nearly all this time the wind blew fresh, and we were going even knots an hour on a north-western courfe. The water now began to lose the fine tranfparent blue colour of the ocean, and to affume fomething of a greenish olive tinge, a well-known indication of foundings. Accordingly, between four and five in the afternoon, ground was ftruck with the lead at the depth of

eighty fathoms, the heat of the fea being then reduced to 69°. In the course of the following night and next day, as we came into fhallower water and nearer the land, the temperature of the fea gradually funk to 65°, which was nearly that of the air at the time.

Unfortunately, bad weather on the 26th prevented us from taking an obfervation of the fun; but on the 27th, though it was then clou dy at noon, we calculated the la titude from two altitudes, and found it to be 33° 26' N. The difference of this latitude from that which we had obferved on the 25th, being 2° 23', was fo much greater than could be deduced from the fhip's run marked in the log-book, as to convince the feamen that we had been fet many miles to the northward by the current.

On the 25th at noon, the longitude by our reckoning was 74° W. and I believe the computation to have been pretty juft; but the foundings, together with the lat tude, will determine the pot where, thefe obfervations were made, better than any reckoning from the eastward. The hip's run on the 26th, from nine in the forenoon to four in the afternoon, was about ten leagues on a northweft-by-north courfe; foon afterwards we hove-to in order to found, and finding bottom, we went very flowly all night, and till noon the next day.

From thefe obfervations, I think it may be concluded, that the Gulf-ftream, about the d degree of north latitude, and the 76th degree of longitude weft of Greenwich, is, in the month of

variations of temperature on the furface not being fufficient to affea materially that of the general mafs. At the tropic of Cancer I found the heat to be 77°; the ftream, therefore, in its whole courfe from the Gulf of Florida, may be fuppofed to have been conftantly running through water from 4 to 6 colder than itself, and yet it had loft only 4o of heat, though the furrounding water, where I obferved it, was 10° below the fuppofed original temperature of the water which forms the current. From this fmall di minution of the heat, in a dif tance probably of 300 miles, fome idea may be acquired of the vaft body of fluid which fets out of the Gulf of Mexico, and of the great velocity of its motion. Numerous obfervations on the temperature of this ftream, in every part of it, and at different feafons of the year, compared with the heat of the water in the furrounding feas, both within and without the tropic, would, I apprehend, be the beit means of afcertaining its nature, and determining every material circumftance of its movement, efpecially if the effect of the current in pushing fhips to the northward is carefully attended to, at the fame time with the obfervations upon its heat.

On the 25th of September, 1777, as the fhips which had transported Sir William Howe's' army up the Chefapeak Bay were returning towards the Delaware, with the fick and ftores, they were overtaken, between Cape Charles and Cape Hinlopen, by a violent gale of wind, which, after fome Variation, fixed ultimately at N. N. E. and continued five days

without intermiffion. It blew fa hard that we were conftantly lofing ground, and driving to the fouthward: we alto purposely made fome cafting to keep clear of the dangerous fhoals which lie off Cape Hatteras.

The 28th at noon our latitude was 36° 40' N. and the heat of the fea all day about 65°. On the 29th our latitude was 36° 2′; we had, therefore, in the courfe of thefe twenty-four hours, been driven by the wind 38 nautical miles to the fouthward: the temperature of the fea continued nearly at 65°. Next day, the 30th, our latitude at noon was 35° 44', only 18 miles farther to the fouthward, though in the opinion of the feamen aboard, as well as my own, it had blown at leaft as hard on this as any of the preceding days, and we had not been able to carry more fail; confequently, it may be concluded that fome current had fet the ship 20 miles to the northward. To know whether this was the Gulf-ftream, let us confult the thermometer. At half after nine in the forenoon of this day, the heat of this water was 76°: no less than eleven degrees above the temperature of the fea before we came into the current!

of October, our latitude was 369 22, 38 miles farther to the north than we had been the day before; the difference, 22 miles, must be attributed to the Gulf-ftream.This, however, is only part of the effect which the current would have produced upon the fhip if we had continued in it the whole four and twenty hours; for, though we were fill in the ftream at five in the afternoon of the 30th, as appeared by the heat of the water being then above 75°, and at eight in the evening the heat being till 74°, yet by feven next morning we were certainly got clear of it, the heat of the fea being then reduced to its former standard of 65°. On this occafion, therefore, we did not crofs the ftream, but having fallen in with it obliquely on the western fide, we pufhed out again on the fame fide as foon as the gale abated.

Thefe obfervations having been made three degrees to the northward of my former ones, it is curious to obferve, that the heat of the Gulf-ftream was about 2° lefs. The feasons of the year, indeed, were very different; but, perhaps, under fuch circumstances that their effects were nearly balanced. In the latter obfervations the meridian altitude of the fun was lefs; but then a hot fummer preceded them whereas in the former, though the fun's power was become very great, yet the winter had been past but a fhort time. Calculating upon this proportion, we may be led to fulpect that about the 27th degree of latitude, which is as foon as the ftream has got clear of the Gulf of Florida, it begins fenfibly to lofe its heat

from 82°, the fuppofed tempera ture of the Gulf of Mexico, and continues to lofe it at the rate of about 2 of Fahrenheit's fcale to every 3° of latitude, with fome variation, probably as the furrounding fea and the air are warmer or colder at different feafons of the year.

The preceding facts had made me very defirous of obferving the heat of the Gulf-ftream on my paffage homeward; but a violent gale of wind, which came on two days after we had failed from Sandy Hook, difabled every perfonaboard, who knew how to handle a thermometer, from keeping the deck. The mafter of the hip, however, an intelligent man, to whom I had communicated my views, affured me, that on the fecond day of the gale the water felt to him remarkably warm; we were then near the 70th degree of west longitude. This agrees very well with the common remark of feamen, who allege, that they are frequently fenfible of the Gulf-ftream off Nantucket fhoals, a distance of more than 1000 miles from the Gulf of Florida! According to the calculation I have before adopted, of a lofs of two degrees of heat for every 3° of latitude, the temperature of the Gulf-ftream here would be nearly 73°; the difference of which from 59°, the heat that I obferved in the fea-water both before and after the gale, might eafily be perceived by the mafter of the veffel. This was in the winter feafon, at the end of December.

An opinion prevails among feamen, that there is fomething peculiar in the weather about the Gulf-ftream. As far as I could

judge,

judge, the heat of the air was confiderably increafed by it, as might be expected; but whether to a degree or extent fufficient for producing any material changes in the atmosphere, must be determined by future obfervations,

Perhaps other currents may be found which, fluing from places warmer or colder than the furrounding fea, differ from it in their temperature fo much as to be difcovered by the thermometer. Should there be many fuch, this inftrument will come to be ranked among the most valuable at fea; as the difficulty of ascertaining currents is well known to be one of the greatest defects in the prefent art of navigation.

In the mean time, I hope the obfervations which have been here related are fufficient to prove, that in crolling the Gulf-ftream very effential advantages may be derived from the ufe of the thermometer: for, if the matter of a fhip bound to any of the fouthern provinces of North America, will be careful to try the heat of the fea frequently, he muft difcover very accurately his entrance into the Gulf-ftream, by the fudden increase of the heat; and a continuance of the fame experiments will fhew him, with equal exactnefs, how long he remains in it. Hence he will always be able to make a proper allowance for the number of miles that the fhip is fet to the northward, by multiplying the time into the velocity of the current. Though this velocity is hitherto very imperfectly known, for want of fome method of determining how long the current acted upon the fhips, yet all uncertainty arising from thence muft foon ceale, as a few experiments upon

the heat of the ftream, compared with the ship's run checked by obfervations of the latitude, will afcertain its motion with fufficient precifion. From differences in the wind, and perhaps other circum. ftances, it is probable that there may be fome variations in the ve locity of the current; and it will be curious to obferve, whether thefe variations may not frequently be pointed out by a difference in its temperature; as the quicker the current moves, the lefs heat is likely to be lost, and conlequently the hotter will the water be. In this obfervation, however, the feafon of the year must always be confidered; partly, because it may, perhaps, in fome degree affect the original temperature of the water in the Gulf of Mexico; but principally, because the actual heat of the stream must be greater or lefs in proportion as the track of the fea through which it has flown was warmer or colder. In winter, I fhould fuppofe that the heat of the ftream itself would be rather lefs than in fummer; but that the difference between it and the furrounding fea would be much greater; and I can conceive that, in the middle of fummer, though the fream had lost very little of its original heat, yet the fea might, in fome parts, acquire fo nearly the fame temperature, as to render it fcarcely poffible to distinguish by the thermometer when a hip en. tered into the current.

Befides the convenience of corresting a fhip's courfe, by knowing how to make a proper allowance for the diftance fhe is fet to the northward by the current, a method of determining with cer tainty when he enters into the Gulf-fream is attended with the

further