gradual diminution. We have already shown this in the formation of swells by changes of slope in streams. We have another experiment. Slowly close a faucet till the liquid vein flows in a mere thread (Fig. 11, xv). The orifice in this case being perfectly cylindrical, the contraction is due on one side to the adherence of the liquid to the sides of the hole, and on the other side to acceleration, and takes place equally in all directions. Intercalate a solid body or a liquid surface at a few centimetres from the hole. The pressure at the base of the column augmenting by resistance, we might expect the vein to take the continuous form of two reversed cones (Fig. 11, xvi). It does not, but takes the form of swellings, of knots regularly placed, which give the vein the appearance of a chaplet (Fig. 11, xvii). This phenomenon, interesting as a case of
Now that we know the details of the form and constitution of the upper part of the liquid sheet, let us see what occurs lower down. On issuing from the canal, the lateral molecules of the sheet are subjected to the action of a component of horizontal centripetal velocity which turns them toward the middle. The particles in the middle, on the other hand, thrown toward the outside, acquire also a horizontal component, but centrifugal (Fig. 8, xiv). The point of stable equilibrium is where the jet has a cylindrical form. But there are various things to be observed.
As the rope-walker does not cease oscillating when he reaches the point of stable equilibrium, but by virtue of his inertia goes as far beyond it on the other side, so the horizontal oscillation of the molecules of water, repeating itself, produces a new sheet in a plane perpendicular to the first one, but this time in the form of a double tongue—that is, of a disk or lentil. The phenomenon repeating