This escapement, always a favorite with clockmakers, has had a long and interesting history and development. Because it started with a suddenly achieved reputation, and because it is adapted to obtain fair results with the cheapest and consequently most unfavorable working conditions, it has won its way into almost universal use in the cheaper classes of clock work; that is to say, it is used in about ninety per cent of the pendulum clocks which are manufactured to-day.

It achieved a sudden reputation at its birth, because it was designed to replace the old verge, which, with its ninety degree pallets close to the arbor, and working into the crown wheel, required a very large swing of the pendulum. This necessitated a light ball, a short rod, required a great force to drive it, and made it impossible to do away with the circular error, while leaving the clock sensitive to variations in power. The recoil escapement was therefore the first considerable advance in accuracy, as its use involved a longer and heavier pendulum, shorter arcs of vibration and less motive power than was practicable with the verge; and as the pendulum was less controlled by the escapement, it was less influenced by variations of power.

In the early escapements the entrance pallet was convex and the exit pallet concave. Escapements of this description may still be met with among the antiquities that occasionally drift into the repair shop. Later on both pallets were made straight, as shown in Fig. 41. It will be seen by studying the direction of the forces that the effect is to wear off the points of the teeth very rapidly, and for this reason the pallets were both made convex (See Fig. 42), so as to bring the rubbing action of the recoil more on the sides of the teeth and do away to a large extent with the butting on the points which destroyed them so rapidly.

The rather empirical methods of laying out the recoil escapement, which have gained general circulation in works on horology, have had much to do with bad depthings of this escapement and the consequent undue wear of the escape wheel teeth and great variation in timekeeping of the movements in which such faulty depthings occur, particularly in eight-day movements with short and light pendulums. The escapement will invariably drive the clock faster for an increase of power and slower for a decrease; an unduly great depthing will greatly increase the arc of vibration of the pendulum, as the train exerts pressure on the pendulum for a longer period during the vibration; the consequence is that instead of the pendulum being as highly detached as possible, we have the opposite state of affairs and a combination of a strong spring, light pendulum and excessive depthing will easily make a variation of five minutes a week in an eight-day clock.

The generally accepted method of laying out this escapement is shown in Figs. 41 and 42, as follows: “Draw a circle representing the escape wheel; multiply the radius of the escape wheel by 1.4 and set off this as the center distance between the pallet and escape wheel centers. From the pallet staff center describe a circle with a radius equal to half the distance between escape wheel and pallet centers. Set off on each side of the center line one-half the number of teeth to be embraced by the pallets and from the points of the outside teeth draw lines tangent to the circle described from the pallet center. These lines would then form the faces of the pallets if they were left flat.”

We wonder how much information this description and the drawing conveys to the average reader. How long should the pallets be? What is the drop? How much will the escape wheel recoil with such a depthing? What arc will the pallets give the pendulum? Why should the center distance always be the same (seven tenths of the diameter of the wheel) whether the escapement embraces eight, or ten, or six teeth? As a matter of fact it should not be the same. We could ask a few more questions as to other details of this formula, but it will be seen that such a description is practically useless to all but those who are already so skilled that they do not need it.

Let us analyze these drawings. A little study of Figs. 41, 42 and 43 will show that there is really only one point of difference between them and Fig. 32, which shows the elements of the Graham, or dead beat. The sole difference is in the fact that there are no separate locking planes in the recoil, the locking and run taking place on an extension of the lifting planes. Otherwise we have the same elements in our problem and it may therefore be laid out and handled in the same manner; indeed, if we were to set off on Fig. 32, the amount of angular motion of the pallet fork which is taken up by the run of the escape wheel teeth on the locking planes, by drawing dotted lines above the tangents, T, we should then have measured all the angles necessary to intelligently set out the recoil escapement. We should have the lock at the tangent, T, the lift and the run (or recoil) being defined by the lines on either side of it, and the length of our running and lifting planes would be found for the entering pallet by drawing a straight line between the points of the two acting teeth of the escape wheel and noting where this line cut the lines of recoil and lift. A similar line traced at right angles to this would in the same way define the limits of run and lift on the exit pallet. It will therefore be seen that our center distances for any desired angle of escapement may be found in the same way (Fig. 28), for either escapement, and thus the method of making the pallets for the ordinary American clock, Fig. 43, becomes readily intelligible. The sole object of curving the pallets, as explained previously, was to decrease the butting effect of the run on the points of the teeth. This is accomplished in Fig. 43 by straight planes on the pallets and straight sides to the teeth with 20° teeth on the escape wheel; merely inclining the plane of the entering pallet about six degrees toward the escape wheel center, thus serving all purposes, while the gain in the cost of manufacture by using straight instead of curved pallets and wheel teeth is very great.

One factory in the United States is turning out 2,000,000 annually of two movements, or about 1,000,000 of each movement; there are four other larger factories and several with a less product; so it will readily be seen that any decrease in cost, however small it may be on a single movement, will run up enormously on a year’s output. Suppose the factory mentioned were enabled to save only one-eighth of a cent on one of its million movements manufactured last year, this would amount to $1,250 per year, a little over $100 per month. Thus it will be seen that close figuring on costs of production is a necessity.

Fig. 44 shows the method of drawing the escapement according to the common sense deductions given above. As the methods of laying out the angle of escapement, lock, lift, and run, were given in detail in Figs. 28 to 32, they need not be repeated here.

Fig. 46 shows the escapement frequently used in French “drum” clocks and hence called the “Drum” escapement. These are clocks fitted to go in any hole of the diameter of the dial and hence they have very short, light pendulums. An attempt is made to gain control over the pendulum by decreasing the arc of escapement to not more than two and sometimes to only one tooth. This gives an impulse to the pendulum only on one-half of the vibrations, the escape wheel teeth resting and running on the long circular locking pallet during alternate swings of the pendulum. The idea is that the friction of the long lock will tend to reduce the effect of the extra force of the mainspring when the clock is freshly wound. Such clocks often stop when the clock is nearly run down, from deficiency of power, and stop when wound, because the friction of the escape wheel teeth on the locking plane is such as to destroy the momentum of the light pendulum. All that can be done in such cases is to alter the locking planes as shown by the dotted lines, so that the “drum” becomes virtually a recoil escapement of two teeth.