It is found that the escaping jet continues to contract until at a distance from the orifice about equal to the diameter of the orifice; this part of the jet is called the _vena contracta_ or contracted vein, as explained on a previous page. _Influence of tubes on the quantity of efflux._—The result before given has reference to an aperture in a thin wall. If a cylindrical or conical efflux tube is fitted to the aperture, the amount of the flow is considerably increased. A short tube, whose length is from two to three times its diameter, has been found to increase the actual efflux per second to about 82 per cent. of the theoretical. In this case the water on entering the tube forms a contracted vein, Fig. 101. just as it would do on issuing freely into the air; but afterwards it expands, and, in consequence of the adhesion of the water to the interior surface of the tube, has, on leaving the tube, a section greater than that of the contracted vein. The contraction of the jet within the tube causes a partial vacuum shown in black in the figure. Now, if an aperture is made in the tube, near the point of greatest contraction, and is carefully fitted with a vertical tube, the lower end of which dips into water, Fig. 101, it is found that water rises in the vertical tube, thereby proving conclusively the formation of a partial vacuum. If the nozzle has the form of a conic frustum whose larger end is at the aperture, the efflux in a second may be raised to 92 per cent., provided the dimensions are properly chosen. If the smaller end of a frustum of a cone of suitable dimensions be fitted to the orifice, the efflux may be still further increased, which will fall very little short of the theoretical amount. [Illustration: FIG. 102.] _Velocities of streams._—The velocity of streams varies greatly. The slower flow of rivers has a velocity of less than three feet per second, and the more rapid, as much as six feet per second, which gives respectively about two and four miles per hour. The velocities vary in different parts of the same transverse section of a stream, for the air upon the surface of the water, as well also as the solid bottom of the stream, has a certain effect in retarding the current. The velocity is found to be greatest in the middle, where the water is deepest, Fig. 102, somewhere in _m_, below the surface; then it decreases with the depth towards the sides, being least at _a_ and _b_. [Illustration: FIG. 103.]