Now, the study of the submarine currents has of late years thrown considerable light on the whole question of oceanic circulation, and has supplied the solution of some problems which had formerly appeared altogether perplexing.

We owe to Drs. Carpenter and Wyville Thomson some of the most important facts recently ascertained. Others, however, have shared in the work. I would, indeed, particularly invite attention to the fact that I do not here pretend to give anything like a complete history of recent investigations into the subject. I select only those facts which bear most significantly on the wider relations - the more marked features-of oceanic circulation.

In the first place, a result which had long perplexed physical geographers has been shown to be erroneous. It had been supposed that the temperature of sea-water below a certain depth is in all latitudes constant, and about seven degrees above the temperature at which fresh water freezes. Sir John Herschel, in his 'Physical Geography,' adopted this supposed discovery as well established. Now, let one consequence of such a relation be carefully noted. The surface water in the tropics is warmer than this supposed constant bottom-temperature; the surface water in arctic regions is cooler; at some intermediate latitude the surface water has the same temperature as the water at the bottom. Hence in this intermediate latitude the water is uniformly warm (according to the supposed relation) from the surface to the bottom. We may therefore regard the water in this latitude as constituting, in effect, a constant barrier between the tropical waters and the arctic waters. Without regarding it as absolutely immovable we should yet be compelled to regard it as so far steadfast as to negative the theory of the existence of submarine currents of an importance corresponding to that of the surface currents. Accordingly, the theory put forward by Humboldt and Pouillet to the effect that there is an interchange of waters between polar and equatorial regions was discredited by this supposed discovery.

Drs. Carpenter and Wyville Thomson, however, have been able to show that no such relation exists. There are vast submarine regions of the Atlantic where the temperature of the water is far lower than the constant and uniform temperature believed in by Sir John Her-schel. The temperature is, indeed, in places, as low, or very nearly so, as the freezing-point of fresh water, under a surface-temperature 20 degrees or so higher. But in other regions having the same surface-temperature the depths are 10, 12, or 14 degrees higher than that of freezing fresh water. Moreover these relations are constant, so far as the deep water is concerned.

Now, there can be only one interpretation of the circumstances here mentioned. If the depths of the ocean were unmoved by any process of submarine circulation there can be no question that a general uniformity of deep sea temperature would prevail in given latitudes. We should not find the bottom water in one region 13 or 14 degrees warmer than the water in a closely adjacent region. We have only to inquire what is the case in inland seas where no great influx of water of alien temperature can take place, to see that this must be so. Take, for instance, the Mediterranean. Here we learn from Dr. Carpenter's recent researches that ' the temperature at every depth beneath 100 fathoms is found to be uniform, even down to a bottom of 1,900 fathoms; as had, indeed, been previously ascertained by Captain Spratt, although his observations, made with thermometers not protected against pressure, set this uniform temperature too high. In the western basin of the Mediterranean, as shown by the Porcupine observations of 1870, the uniform temperature is 54 or 55 degrees; being, in fact, the winter temperature of the entire contents of the basin, from the surface downwards; and being also, it would appear, the mean temperature of the crust of the earth in that region.' We learn, then, two things - viz., first, that where extensive submarine motions are impossible, a constant submarine temperature may be expected to prevail in the same latitudes; and, secondly, that in the latitude of the Mediterranean the submarine temperature is about 54 or 55 degrees Fahr. Thus, it is clear, in the first place, that the varieties of temperature observed in the depths of the Atlantic must be due to the continual arrival of water of the observed temperatures, at a rate great enough to prevent the deep water from acquiring a constant temperature; and in the second place it becomes possible to tell whence the submarine currents are flowing. If they are cooler than they should be supposing latitude alone in question, then they are travelling from arctic towards tropical regions, and vice versa. On this last point no doubt remains. In a latitude corresponding to that of the Mediterranean basin, the depths of the Atlantic are far colder, even in their warmest portions, than they would be if latitude alone were in question. 'In regard to surface-temperature,' says Dr. Carpenter, 'there is no indication of any essential difference between the Mediterranean and the

Eastern Atlantic' in the same latitudes; 'and the thickness of the stratum that undergoes superheating during the summer is about the same. ... At the depth of a hundred fathoms, in the Atlantic as in the Mediterranean, the effect of the superheating seems extinct, the thermometer standing at about 53 degrees; and beneath this' (in the Atlantic only), 'there is a slow and tolerably uniform reduction at the rate of about two-thirds of a degree for every fathom, down to 700, at which depth the thermometer registers 49 degrees. But the rate of reduction then suddenly changes in the most marked manner; the thermometer showing a fall of no less than nine degrees in the next 200 fathoms, so that at 900 fathoms it stands at 40 degrees. Beneath this depth the reduction again becomes very gradual; the temperatures shown at 1,500, 2,000, and 2,435 fathoms (the last being the deepest reliable temperature sounding yet obtained) being, respectively, 38, 37, and 36 1/2 degrees.'