In 1674 Huygens brought out the first watch having a regulating spring in the form of a spiral; the merit of this invention was disputed by the English savant, Dr. Hook, who pretended, as did Galileo, in the application of the pendulum, to have priority in the idea. Huygens, who had discovered and corrected the irregularities in the oscillations of the pendulum, did not think of those of the balance with the spiral spring. And it was not until the close of the year 1750 that Pierre Le Roy and Ferdinand Berthoud studied the conditions of isochronism pertaining to the spiral.

AN INVENTION THAT CREATED MUCH ENTHUSIASM.

However that may be, this magnificent invention, like the adaptation of the pendulum, was welcomed with general enthusiasm throughout the scientific world: without spiral and without pendulum, no other escapement but the recoil escapement was possible; a new highway was thus opened to the searchers. The water clocks (clepsydrae) and the hour gla.s.ses disappeared completely, and the timepieces which had till then only marked the hours, having been perfected up to the point of keeping more exact time, were graced with the addition of another hand to tell off the minutes.

[Ill.u.s.tration: Fig. 160]

[Ill.u.s.tration: Fig. 161]

It was not until 1695 that the first _dead-beat escapement_ appeared upon the scene; during the interval of over twenty years all thought had been directed toward the one goal, viz.: the perfecting of the _verge escapement_; but practice demonstrated that no other arrangement of the parts was superior to the original idea. For the benefit of our readers we shall give a few of these attempts at betterment, and you may see for yourselves wherein the trials failed.

Fig. 157 represents a _verge escapement_ with a ratchet wheel, the pallets _P P'_ being carried upon separate axes. The two axes are rigidly connected, the one to the other, by means of the arms _o o'_.

One of the axes carries besides the fork _F_, which transmits the impulse to the pendulum _B_. In the front view, at the right of the plate, for the sake of clearness the fork and the pendulum are not shown, but one may easily see the jointure of the arms _o o'_ and their mode of operation.

Another very peculiar arrangement of the _verge escapement_ we show at Fig. 158. In this there are two wheels, one, _R'_, a small one in the form of a ratchet; the other, _R_, somewhat larger, called the balance wheel, but being supplied with straight and slender teeth. The verge _V_ carrying the two pallets is pivoted in the vertical diameter of the larger wheel. The front view shows the _modus operandi_ of this combination, which is practically the same as the others. The tooth _a_ of the large wheel exerts its force upon the pallet _P_, and the tooth _b_ of the ratchet will encounter the pallet _P'_. This pallet, after suffering its recoil, will receive the impulse communicated by the tooth _b_. This escapement surely could not have given much satisfaction, for it offers no advantage over the others, besides it is of very difficult construction.

[Ill.u.s.tration: Fig. 162]

[Ill.u.s.tration: Fig. 163]

INGENIOUS ATTEMPTS AT SOLUTION OF A DIFFICULT PROBLEM.

Much ingenuity to a worthy end, but of little practical value, is displayed in these various attempts at the solution of a very difficult problem. In Fig. 159 we have a mechanism combining two escape wheels engaging each other in gear; of the two wheels, _R R'_, one alone is driven directly by the train, the other being turned in the opposite direction by its comrade. Both are furnished with pins _c c'_, which act alternately upon the pallets _P P'_ disposed in the same plane upon the verge _V_ and pivoted between the wheels. Our drawing represents the escapement at the moment when the pin _C'_ delivers its impulse, and this having been accomplished, the locking takes place upon the pin _C_ of the other wheel upon the pallet _P'_. Another system of two escape wheels is shown in Fig. 160, but in this case the two wheels _R R_ are driven in a like direction by the last wheel _A_ of the train. The operation of the escapement is the same as in Fig. 159.

[Ill.u.s.tration: Fig. 164]

[Ill.u.s.tration: Fig. 165]

In Fig. 161 we have a departure from the road ordinarily pursued. Here we see an escapement combining two levers, invented by the Chevalier de Bethune and applied by M. Thiout, master-horologist, at Paris in 1727.

_P P'_ are the two levers or pallets separately pivoted. Upon the axis _V_, of the lever _P_, is fixed a fork which communicates the motion to the pendulum. The two levers are intimately connected by the two arms _B B'_, of which the former carries an adjusting screw, a well-conceived addition for regulating the opening between the pallets. The counter-weight _C_ compels constant contact between the arms _B B'_. The function is always the same, the recoil and the impulsion operate upon the two pallets simultaneously. This escapement enjoyed a certain degree of success, having been employed by a number of horologists who modified it in various ways.

VARIOUS MODIFICATIONS

Some of these modifications we shall show. For the first example, then, let Fig. 162 ill.u.s.trate. In this arrangement the fork is carried upon the axis of the pallet _P'_, which effectually does away with the counter-weight _C_, as shown. Somewhat more complicated, but of the same intrinsic nature, is the arrangement displayed in Fig. 163. We should not imagine that it enjoyed a very extensive application. Here the two levers are completely independent of each other; they act upon the piece _B B_ upon the axis _V_ of the fork. The counter-weights _C C'_ maintain the arms carrying the rollers _D D'_ in contact with the piece _B B'_ which thus receives the impulse from the wheel _R_. Two adjusting screws serve to place the escapement upon the center. By degrees these fantastic constructions were abandoned to make way for the anchor recoil escapement, which was invented, as we have said, in 1675, by G. Clement, a horologist, of London. In Fig. 164 we have the disposition of the parts as first arranged by this artist. Here the pallets are replaced by the inclines _A_ and _B_ of the anchor, which is pivoted at _V_ upon an axis to which is fixed also the fork. The tooth _a_ escapes from the incline or lever _A_, and the tooth _b_ immediately rests upon the lever _B_; by the action of the pendulum the escape wheel suffers a recoil as in the pallet escapement, and on the return of the pendulum the tooth _c_ gives out its impulse in the contrary direction. With this new system it became possible to increase the weight of the bob and at the same time lessen the effective motor power. The travel of the pendulum, or arc of oscillation, being reduced in a marked degree, an accuracy of rate was obtained far superior to that of the crown-wheel escapement.

However, this new application of the recoil escapement was not adopted in France until 1695.

[Ill.u.s.tration: Fig. 166]

[Ill.u.s.tration: Fig. 167]

The travel of the pendulum, though greatly reduced, still surpa.s.sed in breadth the arc in which it is isochronous, and repeated efforts were made to give such shape to the levers as would compel its oscillation within the arc of equal time; a motion which is, as was recognized even at that epoch, the prime requisite to a precise rating. Thus, in 1720, Julien Leroy occupied himself working out the proper shapes for the inclines to produce this desired isochronism. Searching along the same path, Ferd. Berthoud constructed an escapement represented by the Fig.

165. In it we see the same inclines _A B_ of the former construction, but the locking is effected against the slides _C_ and _D_, the curved faces of which produce isochronous oscillations of the pendulum. The tooth _b_ imparts its lift and the tooth _c_ will lock against the face _C_; after having pa.s.sed through its recoil motion this tooth _c_ will b.u.t.t against the incline _A_ and work out its lift or impulse upon it.

THE GABLE ESCAPEMENT.

[Ill.u.s.tration: Fig. 168]

[Ill.u.s.tration: Fig. 169]

The _gable escapement_, shown in Fig. 166, allows the use of a heavier pendulum, at the same time the anchor embraces within its jaws a greater number of the escape-wheel teeth; an arrangement after this manner leads to the conclusion that with these long levers of the anchor the friction will be considerably increased and the recoil faces will, as a consequence, be quickly worn away. Without doubt, this was invented to permit of opening and closing the contact points of the anchor more easily. Under the name of the _English recoil anchor_ there came into use an escapement with a _reduced gable_, which embraced fewer teeth between the pallets or inclines; we give a representation of this in Fig. 167. This system seems to have been moderately successful. The anchor recoil escapement in use in Germany to-day is demonstrated in Fig. 168; this arrangement is also found in the American clocks. As we see, the anchor is composed of a single piece of curved steel bent to the desired curves. Clocks provided with this escapement keep reasonably good time; the resistance of the recoils compensate in a measure for the want of isochronism in the oscillations of the pendulum. Ordinary clocks require considerably more power to drive them than finer clocks and, as a consequence, their ticking is very noisy. Several means have been employed to dampen this noise, one of which we show in Fig. 169.

[Ill.u.s.tration: Fig. 170]

Here the anchor is composed of two pieces, _A B_, screwed upon a plate _H_ pivoting at _V_. In their arrangement the two pieces represent, as to distance and curvature, the counterpart of Fig. 168. At the moment of impact their extreme ends recoil or spring back from the shock of the escape teeth, but the resiliency of the metal is calculated to be strong enough to return them immediately to the contact studs _e e_.

As a termination to this chapter, we shall mention the use made at the present day of the recoil lever escapement in repeating watches. We give a diagram of this construction in Fig. 170. The lever here is intended to restrain and regulate the motion of the small striking work. It is pivoted at _V_ and is capable of a very rapid oscillatory motion, the arc of which may, however, be fixed by the stud or stop _D_, which limits the swing of the fly _C_. This fly is of one piece with the lever and, together with the stud _D_, determines the angular motion of the lever. If the angle be large that means the path of the fly be long, then the striking train will move slowly; but if the teeth of the escape wheel _R_ can just pa.s.s by without causing the lever to describe a supplementary or extended arc, the striking work will run off rapidly.

CHAPTER V.

PUTTING IN A NEW CYLINDER.

Putting in a new cylinder is something most watchmakers fancy they can do, and do well; but still it is a job very few workmen can do and fulfill all the requirements a job of this kind demands under the ever-varying conditions and circ.u.mstances presented in repairs of this kind. It is well to explain somewhat at this point: Suppose we have five watches taken in with broken cylinders. Out of this number probably two could be pivoted to advantage and make the watches as good as ever. As to the pivoting of a cylinder, we will deal with this later on. The first thing to do is to make an examination of the cylinder, not only to see if it is broken, but also to determine if pivoting is going to bring it out all right. Let us imagine that some workman has, at some previous time, put in a new cylinder, and instead of putting in one of the proper size he has put one in too large or too small. Now, in either case he would have to remove a portion of the escape-wheel tooth, that is, shorten the tooth: because, if the cylinder was too large it would not go in between the teeth, and consequently the teeth would have to be cut or stoned away. If the cylinder was too small, again the teeth would have to be cut away to allow them to enter the cylinder. All workmen have traditions, rules some call them, that they go by in relation to the right way to dress a cylinder tooth; some insisting that the toe or point of the tooth is the only place which should be tampered with.

Other workmen insist that the heel of the tooth is the proper place.

Now, with all due consideration, we would say that in ninety-nine cases out of a hundred the proper thing to do is to let the escape-wheel teeth entirely alone. As we can understand, after a moment's thought, that it is impossible to have the teeth of the escape wheel too long and have the watch run at all; hence, the idea of stoning a cylinder escape-wheel tooth should not be tolerated.

ESCAPE-WHEEL TEETH _vs._ CYLINDER.

It will not do, however, to accept, and take it for granted that the escape-wheel teeth are all right, because in many instances they have been stoned away and made too short; but if we accept this condition as being the case, that is, that the escape-wheel teeth are too short, what is the workman going to do about it? The owner of the watch will not pay for a new escape wheel as well as a new cylinder. The situation can be summed up about in this way, that we will have to make the best we can out of a bad job, and pick out and fit a cylinder on a compromise idea.

In regard to picking out a new cylinder, it may not do to select one of the same size as the old one, from the fact that the old one may not have been of the proper size for the escape wheel, because, even in new, cheap watches, the workmen who "run in" the escapement knew very well the cylinder and escape wheel were not adapted for each other, but they were the best he had. Chapter II, on the cylinder escapement, will enable our readers to master the subject and hence be better able to judge of allowances to be made in order to permit imperfect material to be used.

In ill.u.s.tration, let us imagine that we have to put in a new cylinder, and we have none of precisely the proper size, but we have them both a mere trifle too large and too small, and the question is which to use.

Our advice is to use the smaller one if it does not require the escape-wheel teeth to be "dressed," that is, made smaller. Why we make this choice is based on the fact that the smaller cylinder sh.e.l.l gives less friction, and the loss from "drop"--that is, side play between the escape-wheel teeth and the cylinder--will be the same in both instances except to change the lost motion from inside to outside drop.

In devising a system to be applied to selecting a new cylinder, we meet the same troubles encountered throughout all watchmakers' repair work, and chief among these are good and convenient measuring tools. But even with perfect measuring tools we would have to exercise good judgment, as just explained. In Chapter II we gave a rule for determining the outside diameter of a cylinder from the diameter of the escape wheel; but such rules and tables will, in nine instances out of ten, have to be modified by attendant circ.u.mstances--as, for instance, the thickness of the sh.e.l.l of the cylinder, which should be one-tenth of the outer diameter of the sh.e.l.l, but the sh.e.l.l is usually thicker. A tolerably safe practical rule and one also depending very much on the workman's good judgment is, when the escape-wheel teeth have been shortened, to select a cylinder giving ample clearance inside the sh.e.l.l to the tooth, but by no means large enough to fill the s.p.a.ce between the teeth. After studying carefully the instructions just given we think the workman will have no difficulty in selecting a cylinder of the right diameter.

MEASURING THE HEIGHTS.

The next thing is to get the proper heights. This is much more easily arrived at: the main measurement being to have the teeth of the escape wheel clear the upper face of the lower plug. In order to talk intelligently we will make a drawing of a cylinder and agree on the proper names for the several parts to be used in this chapter. Such drawing is shown at Fig. 171. The names are: The hollow cylinder, made up of the parts _A A' A'' A'''_, called the sh.e.l.l--_A_ is the great sh.e.l.l, _A'_ the half sh.e.l.l, _A''_ the banking slot, and _A'''_ the small sh.e.l.l. The bra.s.s part _D_ is called the collet and consists of three parts--the hairspring seat _D_, the balance seat _D'_ and the shoulder _D''_, against which the balance is riveted.

[Ill.u.s.tration: Fig. 171]

The first measurement for fitting a new cylinder is to determine the height of the lower plug face, which corresponds to the line _x x_, Fig. 171. The height of this face is such as to permit the escape wheel to pa.s.s freely over it. In selecting a new cylinder it is well to choose one which is as wide at the banking slot _A''_ as is consistent with safety. The width of the banking slot is represented by the dotted lines _x u_. The dotted line _v_ represents the length to which the lower pivot _y_ is to be cut.

[Ill.u.s.tration: Fig. 172]

[Ill.u.s.tration: Fig. 173]

There are several little tools on the market used for making the necessary measurements, but we will describe a very simple one which can readily be made. To do so, take about a No. 5 sewing needle and, after annealing, cut a screw thread on it, as shown at Fig, 172, where _E_ represents the needle and _t t_ the screw cut upon it. After the screw is cut, the needle is again hardened and tempered to a spring temper and a long, thin pivot turned upon it. The needle is now shaped as shown at Fig. 173. The pivot at _s_ should be small enough to go easily through the smallest hole jewel to be found in cylinder watches, and should be about 1/16" long. The part at _r_ should be about 3/16" long and only reduced in size enough to fully remove the screw threads shown at _t_.