Chap. XVII.
1183
piston be 27 or 28 feet, the atmospheric weight will be sufficient to raise the water in thebarrel with a velocity much greater than the piston can have.
This valve sometimes fits the cup so tightly that it ceases to play: M. Amontonshaving constructed a forcing-pump, dipping 6 feet in water, was surprised on observingthat the valves of cast-iron, and perfectly fitted, suddenly ceased to act; he several timestook the pump to pieces, but could not discover any sufficient cause for the defect. Thevalves were placed horizontally in the barrel, and had they been pressed by the atmosphericweight, they might have been imagined in the condition of two well-polished surfaces,which can only be separated by a very great weight, but there being no air between thevalves and piston, they were only pressed from below upwards by the water which the pistonforced.
The only one cause to which the union of the valve and cup can be attached is theirbeing maintained by the water, whose particles attract each other so forcibly as to excludethe air between, them, and so require a great effort to detach them, which can only beeffected by evaporation or violent friction.
The conical valve, fig. 1807., consists of a truncated cone E resting in a cup 13 C, similarto the preceding, except that it has no ring in the middle, the stalk being very short: atits extremity is a plate II G, which prevents the valve from escaping; its great circle shouldcorrespond with a convex head, whose edges should have sufficient projection to close theaperture when the valve is down ; for there being nothing to keep the axis of the cone inthe centre, it may, by moving right or left, leave a space through which the water in thebranch may escape into the barrel. This valve has the same disadvantages as the other,that of .contracting the passage of the water.
The spherical valve, fig. 1808., is much more simple, consisting only of a sphere E, whichfalls into a cup 13 C when the piston sucks, and would certainly be superior to any otherif it did not contract the water-way, for, once placed at the base of the pipe, it would play anumber of years without needing any reparation : it is true the branch may be enlargedabove and below the valve, so that the whole of the cup, and the passage of the wateraround the sphere, may be equal to the circle of the piston, and the water have a uniformvelocity, which would render this valve infinitely more perfect. Care must be taken notto make it too light or too heavy ; if too light, and the branch is of the same calibre with
Fig. 1807. CONICAL VALVE.
Fig. 1808. SPHERICAL VALVE.
Fig. 1800. CLACK VALVE.
the barrel, the impulse of the water would raise it to a great height, and the valve wouldnot close soon enough to prevent the water coming back : it might even happen that thesame water would pass continually from the barrel to the branch, and back again as thepiston forced or sucked, for if the passage be not closed at the moment the impulse ceases,no fresh water will rise in the barrel nor in the reservoir : if, on the contrary, the valve betoo heavy, the power will have to raise it as well as the water. To obtain the proper meanthe specific weight of the valve must be regulated by the velocity of the piston, that it maynever go further from its cup than just sufl^ient to allow the water to pass, unless it be re-tained by a chain.
'Die clack valve, fig. 1809., is the least imperfect, leaving a free passage for the water : thevalve A D consists of a piece of leather, C D, tightened between two copper plates, A 13 andE F. The diameter of the first is two or three inches greater than that of the pipe L M :that of the second E F is rather less, in order to enter within it: these two plates arc heldtogether by a screw S II and nut G H; the leather has a tail, D 13, serving as a hinge, as