There is another kind of siphon called _the concentric_ or _inclosed diabetes_, the principle of which is the same as that of the bent siphon. As before, let there be a vessel, A B (fig. 2), containing water. Through its bottom insert a tube, C D, soldered into the bottom and projecting below. Let the aperture C of the siphon approach to the mouth of the vessel A B, and let another tube, E F, inclose the tube C D, the distance between the tubes being every where equal, and the mouth of the outer tube being closed by a plate, E G, a little above the mouth C. The lower opening of the tube E F must be so far removed from the bottom of the vessel as to leave a passage for the water. These arrangements being completed, if we exhaust, by suction through the mouth D, the air in the tube C D, we shall draw into it the water in the vessel A B, so that it will flow out through the projection of the siphon until the water is exhausted. For the air contained between the liquid and the tube E F, being but little, can pass into the tube C D, and the water can then be drawn after it. And the water will not cease flowing because of the projection of the siphon below:—if, indeed, the tube E F were removed, the discharge would cease on the surface of the water arriving at C, in spite of the projection below; but when E F is entirely immersed no air can enter the siphon in place of that drawn off, since the air which enters the vessel takes the place of the water as it passes out:—the discharge then, will not cease, for the whole of the outer aperture of the tube, where the water issues forth, is always lower than the surface of the water in the vessel, and, as one level can never be attained, all the water is drained off, attraction being exerted by the deeper column. If we do not choose to draw out the air in the tube C D by suction, water may be poured into the vessel A B until, when it has risen above C, a discharge begins through C D. In this case, again, all the water in the vessel will be drawn out. This instrument is called, as we said before, _the inclosed siphon_, or _the inclosed diabetes_. It is evident from what has been proved above that as long as the siphon is stationary the stream through it will be of irregular velocity, for the result is the same as in a discharge through a hole pierced in the bottom of a vessel, where the stream is irregular from the pressure of a greater weight on the discharge at its commencement, and, of a less, as the contents of the vessel are reduced. In like manner, in proportion as the excess of the outer leg of the siphon is greater, the velocity of the stream is greater; for a greater pressure is exerted on the discharge than when the projection of the outer leg below the surface of the water in the vessel is less. Therefore we have said that the discharge through the siphon is always of variable velocity. But we must contrive a siphon in which the velocity of the discharge shall be uniform.