clinging tendency, between external bodies and the water is proved by the sun and moon lifting and drawing the tides westward; and, as the tides are always lagging behind these bodies, they must have a strong tendency to draw the water in a circle towards the Equator on all west coasts, westward near the Equator, and from the Equator on all east coasts, thus causing a stream to flow continually in that direction; but in reality the sun and moon (and, I suppose, the planets to a small extent) tend to hold the water, and the earth in turning eastward makes the water appear to lag behind; or, in other words, the water is retarded by these bodies while the earth rotates eastward. This does not necessarily imply that the water would flow at the same rate as the tide, but that the water will have a tendency to be drawn in that direction as if it were slightly downhill. This retarding action is continually in operation, and in the same direction. We have the lunar and anti-lunar, solar and anti-solar tides, all tending to induce the water by earth-rotation to move in a circle between the land-masses and the Equator, for the Equator is a boundary when viewed from the centre of rotation, and naturally the water will tend to return eastward in higher latitudes to replace that which is being drawn northward. These latitudes offer less resistance owing to their being nearer to the neutral line between the opposing tides, and also to the fact that the gravitational pull of the sun and moon, acting as it does obliquely to the surface of the water in these latitudes, has a much weaker effect, area for area, than it has on or near the Equator, where the pull, being mostly perpendicular to the surface, must consequently have a greater retarding action on the water while the earth moves on, thus giving the current a westerly trend relative to the land; or, to put it more briefly, the difference in the gravitational pull on these two positions causes the water to circle clockwise in the Northern Hemisphere and anticlockwise in the Southern. These conclusions were arrived at many years ago, but how to embody the conditions in a working model perplexed me for a long time, for I could not divest myself of the idea that it must be done on a globe; but at last I began to see a way out of the difficulty. If one takes an imaginary bird's-eye view of a hemisphere—say, for instance, the Southern—from a great height above the South Pole, that hemisphere will appear as a rotating disc: in fact, water will behave in the same way on a rotating disc or tray, providing gravity is acting perpendicularly to its surface, and this is easily accomplished by having the tray to rotate on a vertical axis, and if necessary we can give it a uniform speed by using a controller. So far, then, things are quite simple. All we have to do to represent the Southern Hemisphere is to provide a shallow circular tray mounted on a vertical shaft, with the different land-masses—South Africa, South America, Australia, Antarctica, and a few of the larger islands—modelled in wood, placed in the tray and fastened in their relative positions and almost submerged in water, and then to rotate the tray at a uniform speed. But by doing this we are confronted by another difficulty, for if we give the tray a uniform speed how are we to give the water the very necessary retardation without in any way interfering with its free movement? To overcome this difficulty, instead of giving the tray a uniform movement for a long period, I gave it an accelerating movement for a short period by means of a falling weight just sufficient to put it in motion. I wish the point to be thoroughly understood, that accelerating the land