and it would be left still further behind. Now, this would continue until at a certain speed the water would fly off in the opposite direction to that in which the globe revolves. At this speed no water would remain on it, and all this would occur while the solid portions of the earth would have sufficient cohesion to hold together. From this it is reasonable to suppose that there is a slight centrifugal current to the equator, and that from the time the revolving speed of the globe causes the water to bulge out at the equator gravity begins to lose control, so that at the same time that the water on the equator begins to bulge out it also begins to lag behind, which is equivalent to a current in the opposite direction to that in which the globe is travelling. The centrifugal current and the water which lags behind on the equator join, and this is the cause of the Gulf Stream, the source of which would be somewhere in the vicinity of the north-west of the South American Continent, on the equator; thence to the Galapagos Islands the bulged-out water on the equator is gradually left behind, and as it goes west it increases in volume and speed all through the Pacific Ocean until it is broken up by the islands between the Malay Peninsula and the Continent of Australia. However, a large body of this current must pass through into the Indian Ocean, and goes on increasing and accumulating all through the Indian Ocean, until it is met by the advancing Continent of Africa, which meets the current like a float of a huge paddle-wheel, and the only way in which it can escape is down the coast of Africa, through the Mozambique Channel, thence down to the Cape of Good Hope. Here it would be impossible to follow it; there will be too many eddies, surface- and under-currents, which must take place with such a large body of water working its way round the Cape. Much of it, no doubt, will be deflected in directions different from what we might expect. However, some of it no doubt does get round the Cape into the Atlantic Ocean. Its course would be chiefly to the north, to replace the water flowing down the South American east coast; subsequently joining the heaped-up water at the equator, which is also getting left behind in the Atlantic Ocean, and which is forming another westerly current. Those two currents combine, and are afterwards met by the advancing South American Continent. The stream then runs down the slanting northern part of South America from Pernambuco into the Mexican Gulf, and from the gulf onward to the north. This stream is so vast, and its temperature so high, that as it travels north it changes the climate from the north-west of Europe to Iceland. It is over fifty miles in width, and different in colour to the Atlantic Ocean. Its channel through this ocean is almost as clearly defined as

the channel of a river flowing over a continent, and in some places it attains a speed of four knots per hour. As it flows north towards Newfoundland it is divided — one branch crosses the Atlantic towards the British Islands, and this current flows into the seas, channels, and inlets surrounding those islands, complicating the high- and low-water gravity tides.

From this it will be seen that the whole of the Gulf Stream is collected on and near to the equator, and this accounts for its very high temperature. In the same way a large portion of the water which is left behind on the equator is met by the South American Continent, and the only way for it to escape is by flowing down the coast from Pernambuco to Cape Horn, and, after getting round the cape, it travels north to the place from where we started; so that there must be a continuous circulating current, and also equilibrium circulation.

When a mass of water is set in motion any attempt to follow its course must be very problematical. Even in rivers, where the mass of water is descending, there are also currents ascending; but in the oceans there are other influences, causing circulation which would lead to greater complication. However, my object is to determine the cause which sets the Gulf Stream in motion; its course in any case would be the same.

I am informed that the equatorial wind theory has been demonstrated by actual experiment. The model is described as follows: A glass trough is filled with water, then wind is blown along the surface of the water, and the result is a current similar to the Gulf Stream. But the same current would take place if the trough is moved forward at a uniform speed. Both these experiments would be futile, because in each case both the gravity and centrifugal forces are ignored. To make the gravity effect clear I will give the following illustration: Water will fly off a grindstone when driven at a moderate speed, but if a power existed in the centre of the grindstone which held the water to it, then it would require to be driven at a greater speed to cause it to fly off. So it is with the trough experiment: gravity exerts its power over both the trough and the water, but the trough has no gravity power over the water. Just imagine the effect of moving the troughful of water at a speed of fifty miles an hour, and consider that on the equator the speed is a thousand miles an hour. Then the centrifugal force acts on the particles of water furthest away from the centre, where the speed is greatest, so that the tendency would be to skim the tropical surface-water on to the equator. This centrifugal force also counteracts gravity, especially on the equator, where the

of the globe is too insufficient in proportion to the enormous weight and mass of the solid globe; and when we consider the speed at which it revolves, combined with the momentum it has acquired, the unstable fluid on its surface could not be a factor in determining its centre of gravity.

In order to get a better idea of the relative proportion which the globe and the oceans have to each other, I will suppose we have a globe 14 in. in diameter, which in round numbers would be 3 ft. 6 in. in circumference. The specific gravity of this globe is more than twice that of granite and five times greater than water. On such a globe the depth of the oceans would be represented by a thick sheet of paper. How could a globe of this density, while spinning round, have its centre of gravity influenced by such an insignificant amount of light fluid which is free to move in any direction? I am desirous of making this as clear as possible, as several statements herein may depend upon this fact.

In conclusion, we have the fact that the globe does revolve on its axis, and its axis is through its centre of gravity, so that when a tide occurs under the moon a derangement of the centre of gravity takes place, which must be adjusted in some way, and the mode of adjustment is very simple. It is accomplished by the tide which occurs at the same time on the opposite hemisphere.

From what has been said it is apparent that the tide opposite to the one raised by the moon is what I call an equilibrium tide to counterbalance the tide produced by the moon, and is governed by the same law or power which makes it compulsory for all revolving bodies to revolve on their centre of gravity, and which has also the power to adjust the equilibrium of the water when it is disturbed. This disturbance and adjustment being independent of the solid mass, consequently, for an absolute certainty, there is an equal quantity of water on the Northern and Southern Hemispheres, and, as there is a greater area of water on the southern oceans, the northern oceans must be deeper.

It is generally supposed mathematically correct that the moon has greater power to draw the water away from the earth immediately under her, but this is not correct, for it is just the opposite which takes place. To illustrate: Let us suppose the water under the moon to be on the surface of a level plain instead of being on a globe; then in that case there would be no tide of any kind, because the water would hold without yielding to the strain of holding the moon, just the same as the land holds. Under those conditions a tide would mean that the water would have to be drawn away from the earth, which is impossible; but on a globe it is quite different, as the moon does not draw the water up im-

mediately under her. The tide on the equator is only an accumulation, which cannot be very high, as the moon's influence passes too rapidly for any great rise and fall to take place immediately under her; so that it will be seen the greatest rise and fall of the tide must be on the horizon of the globe as seen from the moon.

From this reasoning I would expect — first, that the British Isles, being near to where the rise and fall of the tide is greatest, combined with the formation of the bays, seas, and inlets, together with the Gulf Stream flowing into them, would cause the water surrounding those islands to have the most complicated tides of any in the world; second, that the tides in the Hauraki Gulf would have the greatest rise and fall of any place on the Southern Hemisphere within the 40th degree of latitude, because the formation of the land in the gulf gives easy facility for the water to be drawn away by the strain of holding the moon, and the water so drawn away is not so readily replaced, and the angle is greater than it would be nearer the equator. As the distance extends the leverage increases so as to draw the water down the semicircle, and just in the same ratio the speed on the surface of the globe decreases, so that there is a longer time of exposure to the strain of holding from the minimum to the maximum. This is not quite correct, but the effect is just the same, because the speed of the globe at the equator spreads the strain of holding over a greater surface, which is equivalent to a longer exposure to the strain of holding, so that those two factors act on the water so as to produce a greater and gradually increasing rise and fall of the tide. From this reasoning we may discard Australia and all the Pacific islands as not likely to have anything like the same rise and fall that there is in the Hauraki Gulf. Then, on the Continent of South America there is no formation of the land which suggests facility of withdrawal and obstruction to immediate partial replacement; therefore I think it is reasonable to expect the greatest rise and fall of the tide within the prescribed latitude to be in the Hauraki Gulf. On the same line of reasoning I would expect that the greatest rise and fall of the tides on the Southern Hemisphere would be in some of the estuaries near Magellan Straits, because those estuaries are at the maximum land-angle, and the speed on the surface of the globe is slowest; consequently the leverage would be greatest, and the strain of holding would be longer on one place.

In the beginning of this paper I say there are five forces which govern the motion of the earth round the sun, but it should have been six. This sixth force is that which causes all revolving bodies to revolve on their centre of gravity; but