Having overthrown this old theory of the Gulf Stream, Captain Livingston attempted to set up one which is equally unfounded. He ascribed the current to the sun’s apparent yearly motion and the influence thus exerted on the waters of the Atlantic. A sort of yearly tide is conceived, according to this theory, to be the true parent of the Gulf current. It need hardly be said, however, that a phenomenon which remains without change through the winter and summer seasons cannot possibly be referred to the operation of such a cause as a yearly tide. It is to Dr. Franklin that we owe the first theory of the Gulf Stream which has met with general acceptance. He held that the Gulf Stream is formed by the outflow of waters which have been forced into the Caribbean Sea by the trade-winds; so that the pressure of these winds on the Atlantic Ocean forms, according to Dr. Franklin, the true motive power of the Gulf Stream machinery. According to Maury, this theory has ‘come to be the most generally received opinion in the mind of seafaring people.’ It supplies a moving force of undoubted efficiency. We know that as the trade-winds travel towards the equator they lose their westerly motion. It is reasonable to suppose that this is caused by friction against the surface of the ocean, to which, therefore, a corresponding westerly motion must have been imparted. There is a simplicity about Franklin’s theory which commends it favourably to consideration. But when we examine it somewhat more closely, several very decided flaws present themselves to our attention. Consider, in the first place, the enormous mass of water moved by the supposed agency of the winds. Air has a weight—volume for volume—which is less than one eight-hundredth part of that of water. So that, to create a water-current, an air-current more than eight hundred times as large and of equal velocity must expend the whole of its motion. Now the trade-winds are gentle winds, their velocity scarcely exceeding in general that of the more swiftly-moving portions of the Gulf Stream. But even assigning to them a velocity four times as great, we still want an air-current two hundred times as large as the water-current. And the former must give up the whole of its motion, which, in the case of so elastic a substance as air, would hardly happen, the upper air being unlikely to be much affected by the motion of the lower.