Of the water carried westwards by the great equatorial movement, the most important portion after reaching Brazil is carried northwards towards the West Indies. The reason of this is obviously to be found in the fact that Cape San Roque, forming the jutting angle of Brazil, lies several degrees south of the equator. The portion carried southward forms the Brazil Current, and after travelling along the shores of South America almost as far as the mouth of the La Plata, acquires gradually an eastwardly motion which eventually carries it back across the Atlantic towards the Cape of Good Hope, there to pass northwards, and so again to traverse the Bight of Biafra. The surface-circulation in the South Atlantic is thus seen to be comparatively simple.

The larger portion of the equatorial current is carried less quickly northward, because the northern shore-line of Brazil and Guiana is inclined at a much smaller angle than the south-eastern to the westwardly course of the great equatorial currents. Thus the water which is carried towards the West Indies has time to acquire under the tropical sun a much higher temperature than it had possessed when traversing the Gulf of Guinea. It is divided into two parts by the quasi-barrier which the West Indian Islands (or rather the semi-submerged mountains of which they form the crests) oppose to its progress. A comparatively small portion finds its way into the Caribbean Sea, and making the circuit of the Gulf of Mexico, passes out eastwards round the peninsula of Florida. We may fairly assume that this portion is comparatively small; simply because this true gulf stream, passing between Cuba and Florida on an eastern course, would continue so to move for at least some considerable distance, were it not in some way deflected. But it actually turns almost due northwards after passing through the Bahama Sea, traversing the Bernini Narrows on this course, and so onwards towards Hatteras. This would seem to imply that the true Gulf Stream is pressed northwards by the arrival of a much larger body of water which has travelled outside the West Indies. It is true that the diversion of the Gulf Stream northwards may be really caused by the great Bahama Bank. But this would equally establish our position; for if the Bahama Bank is thus effective in diverting the whole of this now swiftly moving current, the Windward Isles may be assumed to be correspondingly effective in diverting the greater portion of the sluggish equatorial current. Moreover, if we remember how shoals commonly take their origin, we may consider that the very existence of the Bahama Bank is probably due to the former encounter of the two important branches of the equatorial current-the part which had circled the Gulf of Mexico and the part which had travelled outside the West Indies. Thus, the northerly course finally taken by the Gulf Stream implies that the latter portion had prevailed over the former, and therefore that it is the most considerable portion. I must mention, however, that the Edinburgh Reviewer holds the part which enters the Caribbean Sea to be the larger.

Oceanic Circulation 2

Be this as it may, the Gulf Stream proper has acquired, during its circuit, characteristics perfectly distinct from those which it had had when entering the Caribbean Sea, or from those possessed by the remaining portion when approaching the Bahamas. In the first place, having traversed a much longer course under the same intense tropical heat, the Gulf Stream has become much warmer than the outer stream. In the second place (probably from having traversed the outlets of the Mississippi, and so carrying with it the finely-divided matter brought down by that river), the Gulf Stream has acquired a peculiar blue colour, somewhat resembling that recognised in most of the Swiss lakes.1

1 This explanation of the colour of the Gulf Stream seems the best that has hitherto been offered. The Edinburgh Reviewer thus states the matter:' The remarkable blueness which distinguishes the water of the Gulf Stream from the oceanic water through which it flows may be due to its retaining in suspension the finest of the sedimentary particles tucket the breadth of the current is about 410 miles, its winter surface temperature only 10° below that which it had in the Florida Channel, and its rate of flow still nearly one mile per hour. It has at this part of its course acquired a good deal of easting, a circumstance which must (unquestionably, we conceive) be ascribed to the fact that it brings from low latitudes the more rapid easterly rotation movement of the earth. The same would, of course, apply to the less characteristic but larger current which has arrived at the same latitudes without circuiting the Gulf of Mexico.

Thirdly, its course having carried it into narrow channels, it has required a relatively rapid rate of outflow, insomuch that the surface flow of the current on its outward passage through the Narrows of Bernini, takes place at the rate of from 2 1/2 to 4 miles per hour. Its width here is at the surface not more than about 25 miles, its maximum depth rather more than a quarter of a mile (about two-fifths of the channel's maximum depth), and its mean rate of flow probably about 50 miles per day.

I shall not follow the Edinburgh Reviewer in considering the details of the progress of the Gulf Stream from the Narrows of Bernini to Cape Hatteras, because, though in themselves of the utmost interest and importance, these details throw no special light on the general subject of oceanic circulation. Suffice it that as far as Hatteras the Gulf Stream remains distinctly recognisable, and that even off Sandy Hook (New York) its surface temperature is little reduced, and its velocity still amounts to about one mile per hour. Off Nanbrought down by that river, the coarser having been deposited near its (the river's) mouth; just as the intense blueness of the waters of Lake Geneva depends on its retention of the finest sedimentary particles brought down by the Rhone in the upper part of its course, while that of the waters of the Mediterranean is due to its pervasion by the like particles brought down by the river Rhone and other rivers, which discharge themselves into its western basin, and by the Nile into its eastern.' It will be remembered that Prof. Tyndall, by researches carried on during the return of the Urgent from the eclipse expedition of 1870, was enabled to throw considerable light on the cause of the colour and shades of colour in water of greater or less depth. See also Dr. Carpenter's ' Report of Researches in the Mediterranean,' in the ' Proceedings of the Royal Society,' vol. xix. p. 200.