forms, taking care, in the method by compound ratios, to put all the drivers under the line of division, and all the driven ones above; so that, when an ascending movement is represented, the wheels may be the denominators, and, when a descending one, the pinions.

The arrangements for calculating a train, and, as such, of forming a clock movement having been examined, it may now be advisable to revert to the mode of regulating the wheel-work by what is called an escapement. The term is derived from the French eschappement, and it is employed to illustrate the action of the pallets in connexion with the last wheel in the train, the teeth of which may be said to escape at each oscillation of the pendulum.

The earliest mode of forming a pair of pallets, or verge, being exactly similar to that employed in the regulation of a common watch, will be fully examined in another department of our work, and it may be enough to say, that its application to the best constructed horological machines has been long since abandoned.

The swing or scape-wheel, represented at fig. 4. plate I. HOROLOGY, is furnished with a pair of pallets, C, D. The wheel passing in the direction of the arrow resting on the pallet C, and as the pallet's arbor, B, is connected with the pendulum by a rod or crutch, the pendulum must oscillate in the same direction. Passing down the inclined plane the tooth is now found to escape, still turning in the same direction. The wheel then rests on the pallet D, and, being aided in its operation by the tendency of gravity to bring back the pendulum, soon passes to the pallet on the opposite side, and the process is repeated as long as the train continues its impulse.

Having thus briefly examined the action of a common anchor escapement; which, from its great simplicity, has been placed first in the order of arrangement, it may now be advisable to direct the reader's attention to the pallets without recoil, invented by the late Mr. George Graham. In this escapement the seconds' hand stands still after each drop of the pallets, and hence the term dead-beat, whereas the hand of a clock regulated by the recoiling escapement is always in motion, oscillating backwards and for

wards.

Mr. Vulliamy's improved mode of constructing an escapement without recoil, is represented fig. 5, in which the pallets AA are allowed to expand or contract, by the motion of a double screw, B; and this escapement differs from all the others in the accuracy with which the acting parts are formed, nearly the whole of the pallets with their frame being executed in the lathe.

To form the pallets a ring of steel is, in the first instance, prepared of the required size, and the arms, L, M, turned with a circular groove to receive it. The ring may then be cut in short lengths, and inclined planes formed at the proper angle. The two arms of the pallet frame are held together by the collet and screws Y, Y, and the regulation is performed by the larger end of the screw B, being furnished with a coarser thread than its other extremity, so that the dispro

portion between them is capable of producing the most delicate adjustment.

The great advantages in this mode of construction are, 1. That the rests of the pallets are correct portions of circles; the centre of which circles is the centre of motion of the axis of the verge, and the pallets move in the same circles, and, consequently, there will not be any recoil in the escapement, 2. That the pallets must be of equal thickness, and consequently the drop the same on both. 3. That the pallets may be made perfectly hard, if properly treated, without risk of altering their shape: and should a pallet be spoiled by an accident in hardening, or a flaw or imperfection of any kind be discovered, another exactly similar is easily made to replace it out of the original ring. When the pallets are made out of the same piece of steel as the arms of the frame, it is difficult to preserve their shape correctly in hardening, and to retain the acting part of the pallet perfectly hard. To obviate this difficulty, the pallet has sometimes been made a separate piece, with a short arm, by which it is fixed with two screws to the arm of the frame; but this is only to exchange one evil for another, as, independent of other disadvantages, which it is unnecessary to enumerate, it is very uncertain, with the pallets fixed in this manner, whether or not the rests of the pallets are concentric with the centre of the axis of the verge. The slightest deviation from its original direction in the arm of the pallet, by hardening or any other cause, has the effect of removing the centre of the circle, forming the rest of the pallets, from the centre of the axis of the verge to some other place, the consequence of which is to render the escapement a recoil escapement. 4. That the mode here recommended of constructing the pallets, offers a great facility for making the inclined planes of the pallets equal to one another, or of altering them, as may be required; and consequently the angle which the pendulum is led by one pallet, will be equal to the angle led by the other.

con

To a person unacquainted with the mechanism of a clock, the pendulum appears but as an appendage to a very complicated machine; whereas, the fact is, that the series of wheels and maintaining power we have now been describing, are, in reality, but appendages to the pendulum, every part of the machine being co structed in subservience to its motion. From this then it will be seen, that it is the pendulum which is the efficient measurer of time, whilst the office of the wheels is to record the divisions marked by its oscillations; so far, indeed, are the wheels from contributing towards the regularity of the pendulum. that they are mostly found to disturb it. The office then of the wheels is to prevent the pendulum coming to a state of rest, which it effects by repeated impulses, at stated periods.

The honor of first applying a pendulous body to regulate the time in horological machines, has been claimed by mechanics in almost every part of Europe: indeed the Arabians, as far back as the time of caliph Haroun Alraschid, state their claims to the invention, although it is

more than probable, that, if the use of the pendulum were known in the east prior to the Seventeenth century, it was employed without the accompanying train of wheels which constitute a modern clock.

If the resistance arising from the friction at the moving parts, and from the motion commumicated to the air, were always the same, and the clock were urged by a weight, the action of the swing-wheel on the pallets would be always the same at a given place, in consequence of which, the figure of all the parts being supposed invariable, the arc of vibration would be constantly of the same magnitude; namely, such as, that the motion lost by the resistances opposed to the pendulum should be accurately equal to the motion communicated by the pallets, and the times would be equal; that is to say, the clock would be perfect, and would measure time accurately. But these conditions are not easily obtained. It is not found, however, that the variation in the resistance of the air, arising from its change of density, occasions any sensible irregularity in clocks. The most considerable irregularities in the movement arise from the tenacity of the oil applied to the moving parts. For the oil is less fluid in cold than in hot weather; and when it is less fluid, a greater quantity of the maintaining power must be lost in overcoming its rigidity; whence it must happen, that the teeth of the crown-wheel will, in that case, act forcibly on the pallets, and the vibration will be less. If the pendulum be suspended on an axis, this cause, together with the constant wear, is very injurious to the performance of the machine, but this defect is remedied by suspending it by a straight flexible spring, as is shown in fig. 6 to 14 of plate I. HOROLOGY.

The rod of the above pendulum should be made of straight grained yellow deal, which may be procured from the lath-maker's, it should be split down both ways; neither the sort which is white and spongy, nor that which is of a strong grain, and full of turpentine. The rod is a cylinder of about five-eighths of an inch diameter, and forty-two inches long; it should be dried and gilt, and if varnished it would be less subject to changes from moist weather. The rod being first roughed out, a brass ferrule (a, fig. 6 above), must be driven on its lower end, previously turned to receive it, the rod is then to be put into the lathe, the ferrule turned true, and a few other places in the rod may likewise be made round; the whole is afterwards to be planed straight, round, and smooth; a hole is then to be drilled at the bottom of the rod, to receive the wire b along the axis. This wire should be steel, and the part which goes into the rod a little taper, and rather larger than the hole in the end of the rod, the rest of the wire cylindrical, and the end conical; a screw must be cut upon the cylindrical part with stocks; the wire must be forced into the hole at the bottom of the rod, and then cross-pinned through both ferrule and rod, as at P. The top of the rod, fig. 9, is slit along the grain with a fine spring saw, to receive the spring at X, by which the pendulum is suspended; the two parts are

drawn together by a screw, and made to pinch the spring; this screw passes through the quarter part of a brass ferrule, and is tapped into the opposite quarter part; the head of the screw, with the first quarter, appears at c, fig. 9. The spring is a piece of strong watch-spring, which has not been coiled up; the upper part has two cylindrical buttons rivetted to it, opposite to each other, one of these appears at Z; these bear the weight of the pendulum during the time of adjusting its suspension, before the screws are drawn tight. The ball of the pendalum is made of lead, and consists of two parts screwed together upon the rod, so as to pinch it. Fig. 7 is the ball as it appears edgeways, and shows the section down the axis of the rod where the two parts join. The shape of the ball, when the two parts are screwed together, is the middle frustrum of a globe, as is seen by the figure. These two parts should be moulded from a neat turned pattern of wood, where the hole should be left to receive the rod; they may be cast so near their true form, as to give but little trouble in turning down in the lathe and finishing; if the pattern be made true, the axis of the rod will pass through the centre of gravity of both. Fig. 6 is the pendulum seen flatwise; two pieces of brass are soldered to the back part of the bow, and tapped to receive the screws which fasten the two parts together; one of these pieces appears at y, fig. 12. The place of the ball upon the rod being found, it is then to be screwed fast to the rod, and not to be removed to regulate the clock. On the screw part of the wire, at the bottom of the pendulum-rod, is a cylinder of brass in two parts, the screw passing through the centre of both parts. The upper part, d, d, fig. 6, consists of a milled torus, and a plain cylindrical part, both in one piece; the cylinder has numerical figures engraven on it, in the order they are represented in the plate, the lower part consists of a milled torus only, as at e, e. When the upper part is screwed to its proper place, it must be held fast, and the lower part screwed against it, so as to pinch the screw-wire, and secure it against any accidental turning. Whenever there is occasion to move the upper part (in order to regulate) the under part must first be detached till the adjustment be made, and then screwed close again, as before. This part may be called the regulator, and will perform that office with a much greater degree of correctness than where the whole ball of the pendulum is moved.

Having thus described the pendulum-rod with its ball, we may now describe the proper method of suspending it, which is by a projecting cock made of brass, and is composed of three distinct pieces, fixed together with rivets and screws. It is difficult to give the exact form without giving many views of it; but the general principle may be easily explained. Strength and steadiness are particularly sought in its formation, and the side view, fig. 9, will make it appear how these are attained in the vertical line, by the part marked a, a, above the line of suspension, and that marked c, c, below the line, as these serve as strong brackets each

way; but the part which serves as its principal support in the horizontal line does not appear in the side view, but may be seen in fig. 11. The form of the part from b to b, is the same as that seen in the side view from the dotted line e to c. The screw marked at Z, fig. 9, appears sideways in the plan; this screw goes through the two parts, which project forward to hang the pendufum upon. The right angled part d, d, b, b, fig. 11, is fixed to the flat brass plate by three rivets, as large as their thickness will admit, one at the angle near the lower b, and one at each extremity of the piece; the parts are put together by rivetting, that, when separate, they may be hammer-hardened. This cock is firmly fixed to a strong piece of wainscot, which is placed against the back of the clock-case, and the whole firmly attached to the wall. The mode of suspension should be such, that none of the lateral motion of the pendulum, as it vibrates, can be commur.icated to the other parts of the apparatus, nor should the whole be liable to be disturbed by foreign causes. The two planes of the cock at Z, which are to receive the string between them, should be filed flat when the plate T is taken off, and thus the spring may be pinched firm between them; so also should the cheeks of the slit at the upper end of the rod, so that the spring should not have the least play at either of its terminations; otherwise, its force will be very unequal. In placing the cock, care should be taken that the place where the spring bends, fig. 9, should be adjusted to the level of the verge or arbor of the pallets at A H.

When the pendulum is to be suspended upon the cock, take out the screw in fig. 9, and release the opposite one at 2, and hang the pendulumspring between the two planes at Z, fig. 9, putting the cylindrical buttons, which are rivetted at the top of the spring, into the hollow made to receive them at Z, then return the screw into its place, but do not tighten it; then tighten the screw c, at the top of the pendulum-rod, and afterwards the two screws, H Q, which will secure it in its place; and from its having hung freely, before these screws were tightened, the several parts will have been drawn into the true perpendicular line; the clock is afterwards put to it. We may now explain the contrivance by which the pendulum receives its impulses from the wheel-work. A, fig. 12, is the verge or arbor, on which the pallets are fixed; 1 1, is a round piece of brass rivetted to the collet; k k k is the stem of the crutch, seen edgeways, and in fig. 13, it appears flatways. In the centre of the upper part is a round hole, A, made to fit the verge; and at 1, 2, are two circular slits. In fig. 12, at 2, 2, is another round piece of brass, fitted rather loose on the verge; the screw at A, and another on the opposite side, go through the fixed plate marked 1, and also through the curved slits on fig. 13, marked 1, 2, and are tapped into the plate, marked 2 2, fig. 12, so that the crutch has a considerable motion round the centre of the verge, and may be fixed in any position by these screws, one of which only can appear in this view, and is opposite to A, fig. 12. At the other end of the stem of the crutch, fig. 13, is a hole to receive the screw

shank of the steel piece seen edgeways, fig. 8, and, when screwed up, appears at K, L. The sides of this piece must be filed flat, and polished, or at least a fine grain given to it; its thickness should be about of an inch; the end of the flat part is seen at fig. 14. The shoulder marked ww, at fig. 8, should be turned flat, and when the screwed shank is put through the hole B of the crutch, in order to fix it, a collet of brass should be interposed between the nut and the face of the crutch; this collet or brass plate should be turned hollow towards the crutch, and somewhat round towards the nut, which will make the fitting more effectual. The flat faces of the steel plate must be set parallel to the line A B, fig. 13. An oblong hole is pierced through the wooden rod, fig. 14, in the direction of the axis of the rod; two fine steel screws s s, are tapped through the sides of this hole. These screws pinch the flat part of the steel piece between them; the ends of the screws which bear against the plate are somewhat rounded off; the ends of these screws, and the flat part they bear against, must be made as hard as possible. The holes for the screws must be made at right angles to the flat sides of the faces of the steel piece, and must pass through the axis of the rod. These screws are of an inch in diameter, and have eighty threads in an inch. They must be forced in so as to cut their own threads in the wood, after which they must never be turned quite out. After having properly suspended the pendulum, and come to set up the clock, draw back the screws s, s, fig. 14, so as to leave room for the flat of the steel part, T, to enter clearly between them. To put the clock in beat, release the screws at A, fig. 12, and its opposite screw, (which is hidden in this view,) so as just to let the verge move stiff in the hole of the crutch. The frame containing the wheel-work must then be set into its place, carefully directing the flat of the crutch between the screws, which pass through the sides of the rod. After having screwed down the frame of the work to the rising board as usual, the crutch must be held fast while the pallets and verge are turned so as to bring the clock into beat. The screw at A, fig. 12, and its opposite must then be tightened, so as to set the pallets and verge fast to the crutch. The back frame must be cut so as to get at the heads of these screws with a key from the front of the clock. These screws have square heads, not slits, and are turned with a key, to prevent the thrusting forwards which is necessary when a turn-screw is used. The clock may be adjusted into beat with the greatest nicety, by releasing one of the screws, s, s, fig. 14, and screwing up the other; taking care not to overturn these screws so as to strip the threads in the wood. The rule to be observed is this: the artist must always hear the flat, T, strike against the screws, and if it be but heard, they cannot be too close; these parts should be oiled. O, P, fig. 10, is a piece of steel wire, which passes through the axis of the rod in order to catch it, by means of two slips of wood properly cut and fastened to the rising board, so that this wire just keeps clear of it, in the vibrations of the pendulum. The ingenious clock-maker will easily perceive, that if the above work be carefully ex