The pump of which we speak is shown in Figs. 16 to 21, and is the invention of Messrs. Laurent Bros. & Collot. It may be described briefly as follows: The apparatus, as a whole, has for base a cast-iron reservoir; A, to the top of which is fixed the pump properly so-called, B, as well as the clack box, A, and safety valve. The pump is placed opposite an upright, D, whose top serves as a guide to the prolongation, E, of the piston rod. This latter is traversed by a pivot, a (Fig 19), on which is mounted a lever, F, whose outer extremity is articulated with a connecting rod, G, which is itself connected with the cranked shaft, G¹. This shaft has for its bearings two supports, b, attached to the reservoir, and carries the driving pulleys and a fly wheel. The beam, F, having to give motion to the piston in describing an arc of a circle at the extremity attached to the connecting rod, must, for that reason, have a fixed point of oscillation, or one that we must consider as such for the instant. Now, such point is selected on a piece, H, having the shape of the letter C, and which plays an important part in the working of the pump. This piece is really a two-armed lever, having its center of oscillation in two brackets, c, at the base of the reservoir. Fig. 17 shows the relation of the beam, F, and lever, H. The upper extremity of this latter is forked, and embraces the beam, F, whose external surfaces are provided with two slots, d, in which to move slides, e, attached to studs, f, which are perfectly stationary on the extremities of the forks of the lever, H. One of the slots is shown in section on the line 1--2 in Fig. 20, and on the line 3--4 in Fig. 21. Things thus arranged, if we suppose the piece, H, absolutely stationary, it is clear that, as the oscillation of the beam, F, is effected on the studs, f, as centers, the piston of the pump will perform an invariable travel whose extent will be dependent upon its position between such point of oscillation and the point of articulation of the connecting rod, G. But we must observe that even according to such a hypothesis, the point, f, would not be entirely stationary, because the point of articulation, a, upon the piston rod being obliged to follow an invariably straight line, the slots, d, will have to undergo an alternate sliding motion on the slides, e, save, be it understood, when the latter are brought to coincide exactly with the center of articulation, a. Now we shall, in fact, see that the point, f, can move forward in following the slots, d, and that it may even reach the point of articulation, a, of the beam, F, on the rod, E, that is to say, occupy the position shown in Fig. 18, where the oscillation of the beam, F, being effected according to the point, a, the stroke of the piston has become absolutely null.