Page:Physical Geography of the Sea and its Meteorology.djvu/422

being assumed that the drift or flow is from the poles when the temperature of the surface water is below, and from the equatorial regions when it is above that due the latitude. Therefore, in a mere diagram, as this plate is, the numerous eddies and local currents which are found at sea are disregarded. Of all the currents in the sea, the Gulf Stream is the best defined; its limits, especially those of the left bank, are always well marked, and as a rule, those of the right bank, as high as the parallel of the thirty-fifth degree of latitude, are quite distinct, being often visible to the eye. The Gulf Stream shifts its channel (§ 124), but nevertheless its banks are often very distinct. Ships, in crossing! the edges of it, can sometimes know it by the colour of the water; at other times they find, as they pass along, the temperature of the water to change 8° or 10° in the course of as many minutes; as an example of this, I quote from the abstract log of the "Herculean," in which Captain William M. Chamberlain, being in latitude 33° 39' north, longitude 74° 56' west (about one hundred and thirty miles east of Cape Fear), remarks: "Moderate breezes, smooth sea, and fine weather. At ten o'clock fifty minutes, entered into the southern (right) edge of the Stream, and in eight minutes the water rose six degrees; tho edge of the stream was visible, as far as the eye could see, by the great rippling and large quantities of Gulf weed—more 'weed' than I ever saw before, and I have been many times along this route in the last twenty years." In this diagram, therefore, I have thought it useless to attempt a delineation of any of those currents, as the Kennell Current of the North Atlantic, the "connecting current" of the South, "Mentor's Counter Drift," "Rossel's Drift of the South Pacific," etc., which run now this way, now that, and which are frequently not felt by navigators at all. In overhauling the log-books for data for this chart, I have followed vessels with the water thermometer to and fro across the seas, and taken the registrations of it exclusively for my guide, without regard to the reported set of the currents. When, in any latitude, the temperature of the water has appeared too high or too low for the latitude, the inference has been that such water was warmed or cooled, as the case may be, in other latitudes, and that it has been conveyed to the place where found through the great channels of oceanic circulation. If too warm, it is supposed that it had its temperature raised in warmer latitudes, and therefore the channel in which it is found leads