Repeating. A repeating watch or clock is one that strikes the hours and quarters only when a plunger is pushed, or a cord is pulled.

This is an English invention: in 1676, towards the end of the reign of Charles II, a certain Barlous [Barlow (Edward)] invented repeating clocks. This novelty spurred the emulation of most London clockmakers, who strove to make clocks of this kind, which soon produced a large number of them, most varied in construction. These clocks were still being made when, at the end of the reign of James II, the same Barlow started making watches of this kind; he had one made by Mr. Tompion, and rumor spread among watchmakers that he wished to present it at Court, in order to obtain the exclusive privilege for this kind of watch. Then some of them learned that Quare, one of the most skilled watchmakers England ever knew, had invented something similar; they urged him to oppose granting a privilege to Barlow. So they both addressed the Court, and a watch of each design was presented to the King in Council. Having tried each, the King gave preference to Mr. Quare, which news was published in the London Gazette .

The difference between these two watches is that Barlow’s repeats when two buttons one on either side of the case, are pressed: one strikes the hour, the other the quarter. Quare’s has a single plunger near the pendant for both functions: as is still the case today, pressing the plunger strikes the hour and the quarters.

Repeating clocks and watches have been made in so many designs that it would be a huge task to try describing each one separately; we shall only speak of those that are the best and most used.

Since repeating clocks are larger than watches, and their mechanism is easier to see, we shall begin with them.

Description of a repeating clock. The Horology plates and figures illustrate a repeating clock with the dial removed, showing all the parts of the motion work. [Plate VI] fig. 31 shows the caliber of this repeater . ABCDE are the wheels of the time train, as in ordinary clocks, and FGHI the repetition train. Wheels GH and the fan, as in all striking trains, merely control the speed of the train. See Strike.

Circle 79, showing one side of the great wheel, has 12 pins, 1, 2, 3, etc . ; on the other side, which is not shown, there are only three.

These 12 pins strike the hours, and the other three the quarters. Ratchet F is fitted to the squared arbor of a barrel, which extends through pillar plate PP ( fig. 32 ) and carries pulley C . [1] This arbor is perpendicular to the face of plate DD, and it enters a barrel mounted on pillar plate PP . The barrel contains a spring, which as explained in the article Barrel is hooked to the arbor and the barrel: turning the arbor or ratchet in direction 3, 2, 1 ( fig. 31 ), winds the spring. Ratchet F is connected to great wheel 79 as the fusee of a watch is to its great wheel, so by means of the click, when the spring is wound, it can turn from 3 to 2 under the wheel. When the spring unwinds, turning in the other direction from 2 to 3 , it drives the wheel as well, and thus pins 1, 2, 3, etc ., lift lever K , which causes the hammer to strike. K is only a diagram of the lever, which is shown more clearly at A ( fig. 32 ),  [2] where it and the quarter levers are mounted on their arbors. Now for the motion work.

It is shown in detail in [Plate VII], figs. 33 and 34 . T ( fig. 33 ) is the cannon or cannon wheel. This wheel, as explained in the article Cannon, makes one rotation per hour and carries the minute hand. On wheel TQ [3] is mounted the quarter snail, K R ; this snail engages the surprise L M , [4] retained by a collar ( 4 4, fig. 34 ); its function will be explained below. TQ is the minute wheel, and A is the star, which rotates once in 12 hours; its profile is at A ( fig. 34 ). Z z is the jumper or jack that advances the star one tooth each hour. See Jumper. Star A carries the hour snail, B; D is the rack driven by pinion e on pulley G ; its profile is shown at G E I . ML is the comb, with its profile at M L .

Figure 34 shows the plate with all the parts removed, showing only their locations; fig. 34, 2 shows the same plate seen from the side, with the studs for the various parts. Each part has a broken line to the stud on which it is to be fitted; 3 and 4 (fig. 34) are two springs. Assuming all the parts reassembled on the plate, as in fig. 32 , we shall now explain their functions.

Before taking up the topic in detail, it is well to recall what the repeating clock is expected to do when the cord is pulled. The four functions are 1) striking; 2) striking the hour; 3) striking the quarters, if so intended, as shown by the hands; and 4) having repeated the correct time, continuing to do so as long as the clock goes. We shall show how the parts we have just described, in their design and arrangement, produce all these functions.

When cord VV , attached to pulley G , is pulled towards D , since the pulley is mounted on the squared barrel arbor, it cannot turn without turning the latter in the same direction, from 3 to 2 ( fig. 31 ). This direction winds the spring, which can turn independently from wheel 79 ( same figure ). Therefore, the wheel will remain stationary, and the spring will be wound in proportion to the arc traveled by the pulley; so the greater the arc, the greater the tension of the spring. Now, if the cord is released, when the spring unwinds it will turn the barrel arbor in the opposite direction; as a result, the wheel will travel an arc equal to that of the pulley when the cord was pulled, in the opposite direction. The pins encounter the tail of the hour hammer and make it strike the bell. Thus we see, first, that pulling the cord makes the clock strike; next, to understand how it strikes a determined number of blows, we observe that rack D engages pinion E fitted to the pulley. The pallet cannot turn without moving the rack as well, and the arc described is always proportional to the rotation of the pulley. Consequently, if the arc described is great, the pulley will rotate a long way; the spring will be greatly tensed, and upon unwinding it will make wheel 79 ( fig. 31 ) describe a large arc; this will bring a greater number of pins under the hammer tail, and will therefore always strike a number of blows proportional to the arc traveled by the rack. Therefore, to make the number of strokes differ and always agree with the hour shown, when the cord is pulled, tail HH of the rack drops on step B of the hour snail. For example, when it drops onto step DD , with the greatest radius, the pulley has described a small arc, the spring is slightly tensed, and unwinding it will drive the wheel over an arc such that only one pin will encounter the hammer tail, which thus will strike only one blow. If, on the contrary, we assume the rack in another position, for example with the tail dropping onto step o o of the smallest circle, the spring will be wound to the limit, and unwinding it will drive the wheel so that all 12 pins pass under the hammer tail, and consequently 12 strokes will sound. It is thus clear 1) that when the cord is pulled, the clock will strike; and 2) that it will strike a certain number of times, determined by the hour snail. For the number of strokes always to be equal to the hour shown by the hand, the star jumps one tooth every hour, by means of pin K mounted on the surprise. So, assuming the clock shows 12:30, and it goes a half-hour, the surprise will jump the star one tooth, or one-twelfth of its rotation, and will change the step for tail H of the rack; now it will meet step DD , of the greatest radius, and the clock will strike only once. Thus, once the rack stands so that the clock repeats the precise hour shown by the hands as long as it goes, it will constantly repeat the correct hour. [5]

Thus, when the cord is pulled, we see 1) how the clock strikes; 2) how it strikes a determined number of times; 3) how this number always agrees with the time shown by the hands. Now we shall see how it strikes the quarters.

The comb, or quarter piece, M [Pl. VII, fig. 34), rotates on pivot N , and by means of spring 4, as soon as it is released, the tail drops onto quarter snail Q ( fig. 30 ), assumed here to be directly above the surprise. As tail 4 approaches the center, teeth I ( fig. 33 ) move away from center E ; between the teeth is a pin mounted on the pulley. Thus, when the cord is pulled and the pulley turns, the pin leaves the teeth; the comb is then free, and its tail, L , drops onto the steps of the quarter snail in position PC [?], so the clock strikes as we have said. But once it has struck the hour, the pulley pin encounters one of the slots in the comb and moves it. If it enters the first slot, d , it returns the comb, and encountering the end of the slot, it is blocked and the pulley turn no further; the clock does not strike a quarter. If, on the contrary, the tail of the comb rests on the lowest step of the snail, since teeth I are far from the pin after the hour is struck, the pulley can turn further and the wheel as well, which strikes three-quarters. Thus, depending on which slot in the comb the pulley pin enters, the clock strikes no quarter, or one, or two, or three; and since the quarter snail makes one rotation in an hour, and its position changes from quarter-hour to quarter-hour, the clock will strike the quarter shown by the hands. If this has been fully understood, it has been seen how the repetition performs all the required functions: 1) how it strikes when the cord is pulled; 2) how it strikes a determined number of times; 3) how this number always agrees with the hands; and finally, how it strikes the quarters.

The repetition we have just described is the old French-style repetition ; its chief defect is that whether the cord is given a long or short pull, it always strikes; if it is not pulled far enough for the tail of the rack to contact the hour snail, it will not repeat the correct hour. In fact, the clock will always strike, but it will be several hours less than shown by the hands. Clockmakers call this kind of clock a “ repeater without all-or-nothing;” clocks that always strike correctly are “ repeaters with all-or-nothing.”

Description of a repeating clock with all-or-nothing : Plate VII, fig. 33 shows the motion work of this type of clock. It differs from other repeating clocks in having the motion work on the rear plate, as observed through the lens, instead of on the dial plate, as in the clock we have just described. This arrangement was devised by the clockmaker M. Le Roy in 1728, to make the motion work parts larger and their workings easier to see. Rack AA [not shown] belongs to the repetition we have just described; similarly, it engages a pinion hidden by ratchet F [not shown], mounted on the arbor of the great strike wheel. This wheel is fitted to the barrel in the same way as in the repetition just described, so by pulling the rack from A to q the spring is wound, etc. Ratchet F is mounted on the same arbor, so by rotating the pinion the arbor is turned as well, and the teeth of the ratchet encounter the lifter of the hour hammer. The lifter is arranged so that when part CGT [not shown] is at rest, as in the figure, the ratchet rotates without encountering it; so as long as part CGT remains in this position, the clock does not strike. When the cord is pulled, tail q of the rack drops, as in the previous repetition , on hour snail B . Here, however, this repetition is different, and either strikes the right hour or does not strike at all. The star rotates on a pivot that is not mounted on the plate, as in the previous repetition , but on screw P , [6] extending through the all-or-nothing, IV . This part rotates around point P and is pressed towards pin L by spring R , which bears against the pin on jumper Z ; [7] however, it can drop, describing an arc whose length is determined by the diameter of the hole for pin L , which does not allow it to drop beyond a certain point. Part LMN , [8] the quarter piece, rotating around point P , functions like the comb; it is held at rest, or in the position shown in the figure, by two parts: 1) the finger, d , fitted to the squared arbor of the ratchet, which for this function bears against the pin, o , mounted on this piece; and 2) by the tip, M , of the all-or-nothing, which holds the tail, X , of this piece. When released from finger d and tip M , it rotates from G to T by means of the spring, rr , and T rests on part H , which is the quarter snail, also making one revolution per hour.

Here is how these parts function: when the cord is pulled and the rack raised, ratchet F is turned; since finger d turns from o towards C , the quarter piece is only held by tip M of the all-or-nothing. If the rack does not descend far enough for tail q to rest on the steps of the snail, since the hammer escapement is not free because the quarter piece is holding it out of reach, the ratchet returns without encountering it, and the clock does not strike. If, on the contrary, the rack rests on the snail and lowers the all-or-nothing slightly, so that its tip M no longer holds tail X of the quarter piece, the piece drops, releases the hammer escapement, and rests on the quarter snail, making the hour hammer strike as many times as ratchet teeth passed. Once the hour has been struck, the quarter piece is returned by finger d engaging pin o on the piece, and as its teeth encounter the hammer escapements, they strike the quarters. Clearly, here the rack and the quarter piece are arranged as in the previous repetition , namely: depending on the height of steps of the snail, the clock strikes more or fewer times; and depending on steps 0, 1, 2 , etc. of the quarter snail, the clock strikes the hour only, or one or two quarters, etc.

1. The original has “b” (Translator’s note).

2. The original has “BB,” which is not in the figure (Translator’s note).

3. The original has “T t” (translator’s note).

4. The original has “R & r” (Translator’s note).

5. Figure 32 shows the snail for the hours, but not the rack; the quarter comb, but not the snail (Translator’s note).

6. The original gives “V” (Translator’s note).

7. The original gives “E” (Translator’s note).

8. The original gives “CGT” (Translator’s note). In this last part of the text, the letters do not correspond to the Plate. It may be that a different Plate was intended.