The dilemma or. Automatic watches with verge escapement, but without fusee!

Transcription

1 The dilemma or Automatic watches with verge escapement, but without fusee! by Joseph Flores Comparative study of two automatic movements, of identical design, but signed and finished differently The verge escapement At the time of the first attempts in the 1770s to produce watches which are wound by the movement of the persons who carry them, according to the expression of the time, watches now known as automatics, the active brains of certain watchmakers were faced by a dilemma, whose explanation can make modern technicians smile. The question can be summarized as follows: Is it possible that a watch does not run while it is being wound? The answer these days is obvious, is it not, and yet at the time it was the case for the large majority of watches. Because practically all watches were equipped with the verge escapement, which is an escapement with recoil, and which could only function well with a part that has now disappeared: the fusee. And it is this fusee, which was essential, that caused the problem, Except by modifying it or finding a device to replace it Two questions are raised, and my answers given here: 1. Why does the verge escapement require a constant energy? 2. What is the problem arising from the fusee? Here are the explanations: The principle of operation of this escapement, the oldest of all, is known as with recoil. It should be understood that the phase of recoil we are considering occurs after impulse, and this kind of escapement has only 2 phases of operation: Impulse - Recoil. Whereas the escapements which followed, those known as at-rest and others known as free, comprise three phases: Rest - Release - Impulse. Thus the recoil of the verge escapement should not be confused with the recoil of lever escapements, due to draw, and which occurs before the impulse. The drawing and photograph in Figure 1 show this escapement. The verge is made up of 2 pallets, which successively operate the teeth of the escape wheel. These pallets are made up of a flat face, the plane of impulse. When a tooth of the escape wheel touches this face, it pushes back the verge, providing the energy to ensure the oscillation of the balance or the pendulum. It should be noted that the escape wheel always has an odd number of teeth, so that when a tooth acts on one pallet, the other pallet is between two teeth. [159] The photograph presents one of these positions in which tooth 1 transmits the impulse to the pallet (in the direction of the arrow) and the other pallet passes between teeth 2 and 3. But why does this escapement require a constant energy? Everyone knows that in modern escapements, the angle traversed during the phase of impulse is always identical. In this escapement, and this is the problem, it is not the case. 1

2 Figure 1: The verge escapement Indeed, it is seen that the angle of recoil produced by the impulse face of the pallet, is added to the angle of impulse. As this angle of recoil varies according to the intensity of the energy, the angle of impulse varies as much, and obviously isochronism suffers as a result. It is this detail of operation which is the reason that the very great majority of the old watches have a fusee, the element which creates the constant energy that this escapement requires, making equal, as far as possible, the angles of recoil and therefore the angles of impulse. But in addition, the fusee has a disadvantage which made the design of automatic watches complex. This is that it needs to run during winding, because it is precisely this that is the disadvantage of the fusee: the watch does not run when it is wound. Here the explanation. The fusee Figure 2 shows the traditional arrangement for providing the energy for the oscillations of the balance. It is made up of a barrel containing the mainspring, like now, but which was connected by a chain to the part which has disappeared from modern watches: the fusee. More precisely the body of fusee, a cone with a helical groove in which the chain is rolled up. [160] One can view the fusee as a pulley with a variable radius, and in addition remember that the spring in the barrel provides a variable energy according to its degree of winding. If it is fully wound it produces the greatest energy. As its unwinds, the energy produced decreases. Figure 2 The assembly in Figure 2 shows this arrangement. When the spring in the barrel is wound, it pulls on the chain making the fusee turn. The first wheel, on the base of the fusee, gears with the traditional train of the watch to make it function. Thus, with a fusee, when spring is wound to the maximum, the chain is at the top of the fusee, and pulls on the smallest radius. As the watch runs, the spring unwinds and loses strength, but the chain then draws on an increasingly large radius. And according to the principle of the lever, one gains on the outlet side of the fusee what spring loses in strength. But then, where does the problem lie? Why doesn't the watch run while it is wound? 2

3 The problem for the researchers of automatic watch construction, is inside the fusee, namely: not running simultaneously with winding. In Figure 3, the fusee has been opened up, with the body of fusee seen from underneath at the left, and the first wheel at the right. Under the body, one sees ratchet teeth 1 and on the wheel a click 2, acting under a light spring 3. When the fusee is assembled, the click acts on the teeth in the body of fusee, with 2 consequences: 1. While the watch is running, the body of fusee, drawn by the chain, makes the first wheel turn via the click catching in the teeth. 2. On the other hand, at the time of winding, which is done by turning the arbor of the fusee by the square (see Figure 2), one makes the body of fusee turn in the opposite direction, to roll up the chain, so that the click slides over the ratchet teeth. And because of this the first wheel is no longer activated. Thus the watch does not run during this period of time. Figure 3: Separated fusee; body of fusee left, first wheel right This point being understood, it should not be thought that it had a great importance in watchmaking at the time. Indeed, in the second half of the 18th century, the accuracy of watches was not very high, and it is only in the more elaborate mechanisms which required a higher precision, such as chronometers for the Navy, that a device, invented by Harrison around 1754, called maintaining power, was generally used to neutralize the detrimental effect of the fusee. Nevertheless this maintaining power is not suitable for a device which winds continuously, as is the aim of automatic systems. Solutions As stated at the beginning of this article, the fertile brains of the watchmakers of the time obviously found solutions. But these had no long term future because at-rest and free escapements were mastered by the decade of 1780s, and the problem no longer arose. It is thus, in my opinion, in the decade of 1770s that the methods, which will be discussed, were used. There are three known attempts: 1. The best known is that of December 1778, due to Hubert Sarton when he deposited a watch with the Academy of Science in Paris; the repercussions of which on modern watchmaking are incalculable. It is described in my article Automatic Rotor Watches, Comparison of The Report of the Academy with Known Movements. 2. Another solution cannot be attributed, because it exists in only 2 movements signed differently (Figure 4): Breguet à Paris and Papillion à Paris. Without documents it is impossible to be sure who is the inventor. These watches are presented below. 3. Then a patent granted to Recordon, n 1249 dated March 18th 1780, but we do not know of a watch using this method. Option 2 concerns the adaptation of an automatic device to a watch with a verge escapement but without a fusee; which seems improbable and yet two exist. Three questions arise: 1. What date are these 2 movements? Can they be made before December 1778, the date of the deposit by Sarton of his modified fusee watch? For the moment, there is no confirmation, because to my knowledge there is no documentation on these watches. 2. Although they have differences in finishing, these two pieces are built on exactly the same caliber. Who is the originator of this caliber? One of the two signatories or a third person? Again, only a document will tell us. 3

4 3. This design uses a verge escapement, which obliged the originator to find a solution so that winding is done jointly with running, therefore needing a very complicated device having an obviously a high cost. Can one imagine that it was done when automatics with escapements with less need for a constant energy source were being made? Yet again, only a document can give an answer. Figure 4: Breguet à Paris (left) and Papillion à Paris N 2069 (right) Supposing that the Breguet watch was made by the great Breguet, and that it can be dated between 1778 and 1780, why would he have chosen the verge escapement, which involved him with such a complicated and expensive construction? A plausible answer is that Breguet made this piece before he had sufficient experience with the new escapements, which certainly were essential shortly after. And if Breguet did not have this control, a fortiori others? We would thus have a confirmation that the innovators of the automatic system had, before 1780, to adapt their devices to a watch with a fusee with the problems that posed. Overall Description Although functioning like almost all of the automatic watches designed at the end of the 18th century, that is, with a laterally oscillating weight, the two movements presented here have only that in common with the other old automatics listed or described in the texts. Moreover there are currently only two known specimens. Their general characteristics are: 1. A unique automatic device, never found in the watches of Breguet or others. 2. The locking system for the weight is also new. 3. The fixing of this weight also was not used before. 4. These movements are equipped with verge escapements which, as we know, require a fusee and which we also know is not suitable for automatic winding. 5. They are astonishing because, in spite of the escapement, these two movements do not have a fusee. 6. We find, on the other hand, two barrels. They have nothing to do with the purpose for which Breguet used two barrels, known as in parallel, which was to enable the use of longer, weaker mainsprings. 7. This arrangement constitutes a remontoir (a remontoir d égalité ) of a design worthy of a very great watchmaker, ensuring the escapement has a constant energy coupled with simultaneous winding and running of the watch. These are the principal visible elements, which will now be described in detail, in the order of the following synopsis: Synopsis Items 1 to 10, general presentation of the movements: 1. Preliminary Questions 2. Collections: Geneva and Cluses 3. Complete Movements weight side 4. Top Plates without the weight 5. Arrangement of the mobiles on basic plate 4

5 6. Arrangement under the dial: dial-work 7. Both sides of the weights 8. Energy: the barrels A and B 9. Thickness 10. Improving the rate Items 11 to 24, detailed presentation of the functions: 11. The automatic device 12. Winding of the barrel A 13. Function of the barrel B 14. Regulating Train 15. Device for rewinding barrel B 16. The lever 17. Preliminary data 18. The watch running 19. Position of the lever 20. Rewinding of barrel B 21. Remark on the rewinding mechanism 22. Locking of barrel A 23. Damping the weight oscillations 24. Equilibrium of the weight 1: Preliminary questions Obviously they must be subjective, given the complete lack of documents from the time. They reflect only my personal convictions, based on assumptions made in answer to three questions: Who? When? Why? Who? If we only knew of the piece signed Breguet, or only that signed Papillion, any historian would attribute the origin of the caliber to the signatory. It is what I did in 2001, when I published my work Perpétuelles à roue de rencontre when I did not know about the Papillion movement. Concerning the origin here is what I wrote: For me, it is without doubt the great Abraham-Louis Breguet, even if the Master never spoke about it, or at least in what we currently know of him. Will we one day find documents other than those already known and presented in many books? It seems to me, unfortunately, that it is unlikely! His career was examined with a fine tooth comb and there remains, as we already know, only the black hole of his beginnings, that is to say between 1775 and approximately 1782 The hope for this period is quite small, but one never knows. I can do nothing but believe, but my convictions are great, that such ingenious work on an automatic watch, when one knows that this kind of device was part of his very first concerns, cannot be the work of another who, moreover, would not have disappeared without trace, which would be quite as astonishing In 2003, when the Papillion piece made its appearance, I obviously made my mea culpa, but nevertheless, continuing to think that Breguet is there for some reason, here what I wrote: In 2001 I thought that such designs could only be born from brains whose capacity for reflection is higher than the average Is it true that the question remains, that such brilliant devices can be conceived in brains whose owners remain to us unknown, or that, although one finds the name of Breguet on a piece, he did not take part in its development? I have, for my part, great doubts Nevertheless it is necessary to discover who was this Papillion, if he had a relationship with Breguet, etc. Assumptions are not answers. A Few years passed and for my part I still did not have an answer concerning the originator of this caliber, but obviously I continued to think that only a descriptive document would give us the answer. But definitely this originator was a brilliant mechanic. 5

6 When? Whether it is one or the other of these two signatories, or a third, I will locate this work in what I name the black hole in the career of Abraham Louis Breguet, i.e. between 1775 and approximately Although subjective, I have this impression because of the design of this movement, compared with what Breguet did later. The Breguet, just like the Papillion movement, is of relatively modest finish, unusual for the great Breguet. It is one of the reasons which could make one think that it is not him. It is this reason which makes me think that it was one of his early achievements, and in my opinion an attempt without a future. And so between ? Why? It is obvious that the design ensuring constant energy and simultaneous running was, in the approach of the initiator of the system, a priority. But why didn t he decide to transform the fusee? He knew without any doubt the principle of the fusee, that Breguet used anyway, and which is at the heart of the design of Sarton, even if he did not know the adaptation of the latter. Well no, he created his own system, which in my opinion, perfectly reflects the temperament of a great watchmaker. But why chose this escapement, which requires a constant energy, whereas one of the others, such as the cylinder or the virgule escapement much simplified things? (Of course he knew of them, since they had been invented at this date, and even some times used.) It is obviously the most significant question. Still, some think that it could be a question of later manufacturing, around 1790, for the Turkish market for example. But then at that time, if it is Breguet, he kept his books perfectly, so why did he never mention it? I strongly doubt it! Then if the originator chose this escapement, I imagine that for him it was the best and the only solution. Quite simply because he had at the time, I think (about the dates where I locate this work, ) neither control of, nor confidence with the other escapements. It is besides necessary to identify existing automatic pieces equipped with at-rest or free escapements, and going back with certainty to before They are not legion; in truth I do not know of any I note in addition that even Recordon in 1780, in his patent application in England, did not use one of the new escapements, since he made a piece with the verge escapement and another with a little known escapement, probably of his invention I rely on the assumption of the lack of control and confidence with the new escapements, which those who imagined this automatic system of winding, without doubt did not yet have So as not to rely on subjective assumptions, I hope more concrete data will come to confirm or contradict, and history will always benefit from it. 2: Movement signed Breguet à Paris and movement signed Papillion à Paris, n 1069 Under the number CS 436 in the inventory of the old musée d horlogerie et de l émaillerie in Geneva, hides the automatic movement of a watch which presents all the marks of an important piece, if not outstanding, in the history of the automatic watch; which is at least my deep conviction. This movement, if it does not look much, because it is unfortunately in rather bad condition, conceals such a brilliant design, that it appears very difficult for me to attribute it to any other than he whose name is inscribed on the oscillating weight: Breguet à Paris without any serial number (Figure 5). But what of this signature Breguet or not Breguet? Some dared to suggest that it could be a fake signature! I cannot accept this version, nor contradict it, knowing nothing to do so with certainty. Now if it is made by another watchmaker, in my opinion he would not have had any need to envy the Master, because he would have had all the opportunities to become one himself Before continuing, it should be added that this movement has not yet been assigned its full value, and that what was written before by the historians of watchmaking does not correspond completely with reality. It was indeed necessary to have the opportunity of disassembling it, which I did with the greatest pleasure, to understand all subtleties of it. 6

7 Figure 5: Breguet Signature It is under the inventory number MM 431 of the collection of the musée de l horlogerie et du décolletage at Cluses, that hides another automatic movement, whose examination raises as many questions, in addition to those posed by the movement signed Breguet. This lets us foresee that the history of watchmaking still has beautiful times in front of it, and researchers to come will work on discoveries yet to be made. This movement is intriguing because its resemblance to the movement signed Breguet à Paris is indisputable, as the components presented hereafter will convince you. But its signature raises questions, because this signatory is, one can say, practically unknown. It is Papillion à Paris engraved on the weight, also with a number, n 1069 (Figure 6). Figure 6: Papillion Signature By limiting research to the time at which this piece was made, that is to say almost with certainty between 1775 and 1785, here is what one finds in various dictionaries. It is seen that it is rather thin and, moreover, it should be noticed that none of these Papillon are written with an i, as is the case on the movement. But nevertheless the author could be well Jean Pierre Papillon who is the only one given as working in Paris. Baillie, Watchmakers and Clockmakers of the World, 1947, p. 242: Papillon, Jean François. Geneva. Ca Patrizzi, Dictionnaire des horlogers genevois, 1998, p. 301: Papillon, Jean-François, master watchmaker, Papillon, Jean-Pierre, son of Pierre, Paris (France), watchmaker, recorded living in

8 3: The movements weight side Here are these two movements shown from the top side, where the weight is mounted (Figures 7 and 8). I will return to these weights whose form is not common, but it is seen that the point of pivoting is on the edge of the plate, for side oscillations. Moreover, in both cases it bears a name, doubtless that of the finisher, which does not mean it is of the inventor. Once more, with the signature on a movement we cannot tell if it is the inventor, the finisher, or why not only the salesman? I repeat that only a document can provide the key element for attribution. The view of each movement does not show much, the form and size of the weights practically hiding all of the top plate. These weights on the other hand, each carry the signature indicated, but until now I have not been able to establish a bond between these two signatories. The very first technical information: the diameter of the plates is slightly less for Papillion, 47 mm, compared with 48.2 mm for Breguet. The construction is known as in frame ; that is the movement is composed of two plates, mounted on pillars, and therefore without bridges, and the assembly has a hinge seen at the bottom of photographs 7 and 8. Figure 7: Movement signed Breguet in Paris Figure 8: Movement signed Papillion in Paris n : The movements without weights The weights having been removed, you see in Figures 9 and 10 that there is a consistent, identical arrangement of the various mobiles. Figure 9: Breguet movement without weight Figure 10: Papillion movement without weight 8

9 Initially the angle of the cock and the balance is a little different. Obviously, it is the dial plate carrying the hinge which was not placed identically. Then for even more legibility the cocks and balances were removed; see Figures 11 and 12. This makes it possible to better see the part 2. It is the rack for regulation (advance/retard) which, on the Breguet watch, is hidden under a triangular plate, but otherwise the arrangement is the same. Then 3 is the arbor of the weight, with a small spiral spring visible on the Breguet watch; it is the equilibrium spring to balance the weight. This spring is missing from the Papillion watch, but we will see later that there was certainly breakage and modification. The cocks are practically identical, and the end stones for the balances, composed of a dovetailed metal strip, are completely similar. Lastly, in the opening in the plate 4 is seen the verge escape wheel. Figure 11: Breguet Movement without weight, cock and balance Figure 12: Papillion Movement without weight, cock and balance 9

10 5: Arrangement of the mobiles on the plate Figures 13 and 14 show each movement with the top plate removed, displaying the arrangement of the mobiles which are between plates. In these views the escape wheel is missing, as in all verge escapement watches it is fixed under the top plate. Figure 13: Breguet movement arrangement on the plate Figure 14: Papillion movement arrangement on the plate An enlargement of the pillars makes it possible to note a difference in form, but not indicating anything. 10

11 In any case the similarities are striking, not to say convincing, and undoubtedly indicate the same origin of the ébauches. And obviously this similarity is all the more striking in that this construction is completely new, at least in regard to what is known. An arrangement of two barrels in series 1, one rewinding the other, seems to me far from widespread. In fact, that is one of the devices that are called remontoirs or remontoir d égalité. We will understand this function later. Another characteristic, the provision of a small train 3, identical to those used in repeater watches, and one can wonder what it is used for on these movements which are not repeaters Here is the list of the various functional parts: 1 Two barrels providing the energy for operation. 2 Three wheels of the time train. 3 Regulating train for winding. 4 Pinion connected to the automatic device responsible for winding this barrel. 5 Arbor of the oscillating weight. 6 Arbor for the regulator. 7 Device for locking the weight. 6: Arrangement under the dial, dial-work Figures 15 and 16 are views of the dial-work of each movement, which is invisible when the watch is finished. Figure 15: Breguet dial-work 11 Figure 16: Papillion dial-work The first remark is that unfortunately the Breguet movement is largely incomplete, but obviously it was made exactly like the Papillion movement. Here is what one finds on the latter, which one can thus imagine on the former. Most striking is a relatively large wheel, very finely cut, and surmounted by an articulated lever which is hinged at its left end, where the lever is fixed onto a square arbor, 1. Even if the operation is absolutely identical, the end of each of these levers is different. That will be explained later. Then in the center of the Papillion movement is the canon pinion and 2 wheels (the motion-work) which carrying the hands, as in all watches. However there is a third wheel meshing with a toothed rack. This rack is also mounted on a square arbor 2. On the Breguet only the arbor remains at 2. Then a rather mysterious bar 3, which in fact is part of the device used for locking the weight. Lastly, because all will be explained later, a spring mounted on a screwed base 4, of which there is no trace on the Breguet movement.

12 As we might guess, the large wheel with fine teeth is part of the device which transmits the oscillatory movements of the weight to wind the mainspring of the watch. 7: The weights on each of their faces As the views show, these weights are particularly similar (Figures 17 and 18) Figure 17: Breguet weight both faces Figure 18: Papillion weight both faces Very logically they are pierced with a square hole 1 for the arbor, also finished square. Figure 19 is the arbor of the Papillion watch much enlarged: Figure 19 On the square on the right of this arbor is attached the articulated lever connected to the large wheel with fine teeth, all of which will be explained later. The weight is placed on the square below the screw thread. Important is the oval form, identical for both, but which is novel for automatics of the time, of which the form is usually that of the piece presented below, Figure 20. Could this be an indication of an earlier date? It is for me very likely. The resemblance continues underneath these two weights, which have a piece 2 held by two screws with a third screw near a hole. The purpose of this will be developed below, but it is used for locking the weight. 12

13 Figure 20 Then, near the square hole is a steel part 3 held by a screw, a part which is obviously broken on the Breguet watch. The purpose will be explained later, but it is the banking spring. A final remark on these weights, before seeing how their oscillations are transmitted to ensure winding, is an observation on the differences in their dimensions. It was noted that the Papillion weight is smaller, and the photographs below, Figure 21, taken on graph paper, show the Papillion is 36.6 mm long and weighs 22 grams, against 42.7 mm for the Breguet and 35 grams. Figure 21 8: The energy: 2 barrels, A and B These parts are the really outstanding characteristic of these automatic movements. Being equipped with the verge escapement, the energy must necessarily be constant. All the watches with this escapement have a fusee, but it is known that an ordinary fusee, which does not ensure the transmission of energy during winding, cannot be used with an automatic winding device which winds continuously. Now here the originator, Breguet or Papillion, if it is one of them, used a very imaginative system. But obviously the question, already asked, can be repeated: Why wouldn't they have used another escapement which would have avoided everything that will be shown? To avoid any confusion it should first be stated that Breguet used two barrels in a different way, placed on either side of the center wheel. Each of these barrels meshes separately, in a synchronous way, with the pinion of this wheel, and without any bond one with the other. All his biographers explained the reasons for this arrangement as in parallel. Here we have a very different arrangement, and here is how it works. Asynchronous barrels in series This is absolutely not the arrangement which has just been described. The objective is twofold: To ensure a constant energy and simultaneous running of the watch. For this purpose, the barrels are laid out so that one gears with the other by a set of ratchets. This arrangement is known as in series but in a particular way. It is to my knowledge unique, at least on automatics. This assembly is visible in photographs 22 and 23. The first barrel, labelled A, is wound continuously by the automatic device, except that it is limited by a stop, locking the weight when the spring of this barrel is wound to the maximum. 13

14 Figure 22: The two barrels of the Breguet watch Figure 23: The two barrels of the Papillion watch The second barrel B is rewound at regular intervals and by an equal amount, from which the term remontoir ( remontoir d égalité ) comes. This rewinding is provided by barrel A every three hours. So the energy of barrel B is practically constant, even completely if the watchmaker carefully chooses the most favourable possible part of the curve of unwinding of the spring in barrel B. If things seem simple in principle, they are particularly subtle in their realization. Drawing of the arrangement To try to give the clearest possible explanation of this very particular arrangement, it is shown not only in a drawing (Figure 24), but also in a diagram of its action (Figure 25). It will undoubtedly be necessary to refer from time to these pages, but all the enumerated points will be the subject of detailed descriptions. 1) Barrel A, of a diameter of 18.2 mm, is composed of a drum and traditional lid. The spring placed inside is hooked, on the one hand, to the wall of the drum and, on the other hand, to the hook on the arbor, as in any barrel. 2) Barrel A has a winding wheel which is connected to the pinion of the wheel with fine teeth, actuated by the oscillating weight. 3) The teeth of this barrel are connected, on the one hand to a regulating train, and on the other hand to the winding wheel of barrel B, called the rewinding wheel. 4) Between this rewinding wheel and barrel B is a particularly subtle device called the release mechanism. 5) A lever called the rewinding release lever connects the release mechanism to the regulating train, connected to barrel A. 6) The teeth of barrel B mesh classically with the pinion of the center wheel, the first wheel of the time train. 7) The winding wheel of barrel A carries a system of stop-work, allowing a little less than seven turns of winding. 14

15 Figure 24: Drawing of the arrangement of the barrels Diagram of action Figure 25 (1) BLUE, left-hand column: The watch being carried, the automatic system winds the spring in barrel A via its wheel. (2) RED, right hand column: The spring of barrel B, having been given a certain quantity of energy by the watchmaker during assembly of the watch, makes the watch run via the train and the escapement. (3) GREEN, line top: Barrel A is motionless as well as the rewinding wheel B. (4) YELLOW, in the center: They are locked by the regulating train on the one hand, and the locking/releasing mechanism on the other hand, connected by a lever (yellow). 15

16 At intervals the lever releases the regulating train and then barrel A rewinds the spring of barrel B via the wheel B. The spring of barrel A is unwound and spring of barrel B is wound. The lever, re-locking the system, the watch continues to run due to the rewound barrel B, and the automatic mechanism continuous to wind barrel A. And, at a later time, the cycle starts again. 9: The movements: thickness The thickness of these two movements is an important point, because it is known that the great Breguet (without saying that this movement is his) was attached to producing relatively flat watches, probably under the influence of Lépine, from whom he seems to have taken some council. And by this one could still think that it was not his work, even though the design is brilliant, because these two movements have a thickness of 17 mm! Figures 26, 27, 28 and 29, present side views of the Papillion movement from various angles. For comparison, I point out that the movements with a rotor, built on the basis of that deposited by Sarton with the Academy in 1778, have a thickness of 15.5 mm thus to make a flat caliber was not the first aim of the originator of these movements Whereas the distance between the plates is a small 6 mm for Papillion, and 6.4 mm for Breguet. We note that of this space the weight takes up 7 to 8 mm. The balance placed on this side is really unfavourable for the thickness, because it requires the weight to be moved away from the plate. Figure 26 clearly shows this space. As already known, this movement is equipped with a verge escapement, with a traditional train, described elsewhere. The various side views of the movements are: Figure 26: Section of the movement with the weight arbor, and the weight in position. On the left of the arbor is the contrate wheel. Figure 26 Figure 27: The arbor used for the locking the weight at the end of the winding of barrel A, and, on either side of this arbor, barrel A on the left and barrel B on the right. Figure 27 Figure 28: Protruding from the top, the arbor for the weight and, on each side, the banking pins. One also sees, between the plates, the verge escape wheel and the screw to adjust the functions of the escapement. 16

17 Figure 28 Figure 29: In the center is the arbor of the regulator. On its left is barrel B and on its right the contrate wheel. Figure 29 10: Final adjustment of the rate The advance/retard function of these watches, which is done by what is called the regulator, is completely identically on these two movements, and in a particular form. As was shown in points 3 and 4, the assembly regulating the balance spring is under the weight. So there is no possibility to access it to carry out final adjustments, the weight hiding all of the top plate. This detail has not deterred the designer, who made a relatively complex arrangement, which indicates, if it still has to be demonstrated, that at the time this researcher did not stop with these details, though it took some time and some additional cost. Figure 30: The regulator racks, top plate Figure 31: The regulator rack under the dial Thus, not having access to the side of the weight, it was necessary to allow the final adjustment to be made on the other side of the movement, where the dial is That s no problem, the regulator will be on the dial, with its advance/retard indication, and here is how this was done. 17

18 On Figure 30, indicated by 2, is a rack. It gears with the second rack 1, carrying the index pins, between which the last turn of the traditional balance spring passes (white arrow). To bring this regulator to the other side, rack 2 is mounted on the end of an arbor finished with a square (red arrow, figure 30), which passes through the movement (see Figure 29). On the opposite side, under the dial, the arbor is also finished with a square (red arrow, Figure 31) on which is another rack 3 Figure 31. But it was still necessary to bring the indication onto the dial, which is why rack 3 meshes with a wheel placed at the center 4 and an indicating hand is placed on its canon, which is formed into a pentagon. This wheel has a blank section without teeth so that it can only turn about 270 ; it is also shown separately in Figure 32. Figure 32: The regulator hand To show what the dial would look like, (but it is practically 100% certain to be correct), the movement was set up with another dial, on which was painted a half circle above the figures 5/7, with the traditional letters A/R for advance/retard (Figure 33). It is clear now that by turning the hand to the wanted angle, one moves the index pins shown in figure 30, and the watch is advanced or retarded. On the other hand, to show that this watchmaker did not simplify his life, Figure 33a shows that if he had simply extended the square to the dial, it would have the same result Figure 33: Assumed dial Papillion movement 11: The automatic winding train Figure 33a Like the overall construction of these movements, except for the train and the escapement, the automatic winding mechanism is completely new; at least, for my part, I have never seen it before. As already specified, the weight is arranged classically, the principle of winding which one finds on all automatics of the time, except for the 5 known pieces with rotors. The arbor is thus on the edge of plate, but as the weight is on one side of the watch, and the automatic train is on the other, under the dial, as we have just seen with the regulator, this arbor crosses the movement. The arbor of each one of these movements, is shown respectively (Figures 34 and 35) that is to say, on the left Breguet and on the right Papillion. It is on the square end that the weight is placed. Figure 26, where the weight is in place, shows this perfectly. On the other hand a detail of the fixing of the weight should be noticed. It is seen that the top of the arbor of the Breguet watch displays a rupture (at the arrow Figure 34), which leads us to think that the weight was held by a pin, and that a repair was carried out by adding a side screw (Figure 36). Whereas 18

19 the Papillion weight is held by a small nut, the end of the arbor being threaded (Figures 35 and 37). Figure 34: Breguet Figure 35: Papillion Figure 36 Figure 37 Now let us look at the other side of the movements, the dial-work where the train of the automatic mechanism is located, which, as shown in Figures 38 and 39, has a large wheel with very fine ratchet teeth. Figure 38 Breguet The Breguet watch has 150 teeth inclined to the left, and the Papillion watch has 120 with the same slope. On each of these views, the end of a square arbor is seen in the red circle. It is of course the arbor which crosses the movement, on which the weight is mounted on the other side (see also Figures 26 and 28). So it is easy to understand that the oscillations of the weight, while the watch is carried, make the small part mounted on the visible square oscillate. This part is connected by an articulated joint, visible in the yellow circle, to another part, under which the wheel with fine teeth rotates. Indicated by arrows, this part of the articulated lever carries 2 clicks known as winding clicks. They are laid out differently in the Breguet and Papillion watches. Two other clicks, also marked by arrows, are mounted on the plate. These clicks are known as retaining clicks. 19

20 Figure 39 Now is the time to understand how the oscillations move the wheel, and then transmit the movement to wind the mainspring. We see by the computer editing of Figure 40 that when the weight oscillates, the part which is attached to the square on its arbor also oscillates. These oscillations are simulated by the triple reproduction of the part. Figure 40: Articulated Joint If we now consider only the double articulated part and draw the wheel and the clicks (Figures 41 Papillion and 42 Breguet) as is shown, then: Figure 41: Arrangement of the system for winding on the Papillion movement Considering the Papillion watch (Figure 41), when the weight descends (the red arrow) the small end of the articulated lever also goes down, and the 2 red clicks make the wheel turn anticlockwise. The 20

21 retaining clicks (blue) slide over the teeth. Conversely, when the weight goes up (blue arrow) the winding clicks (red) pass over some teeth, while the retaining clicks mounted on the plate ensure that the position of the wheel does not change. Concerning the Breguet watch (Figure 42), even if the clicks are positioned differently to the Papillion watch, the action is absolutely identical. Indeed it should be understood that when the weight goes down, (red arrow) the articulated lever also goes down, but as the winding clicks are on the other side of the pivot, that end rises and these clicks turn the wheel anticlockwise. Figure 42: Arrangement of the system for winding on the Breguet movement The enlargement (Figure 43) of the clicks of the Papillion watch makes it possible to explain why there are two winding clicks and two retaining clicks Besides providing some safety, one clearly sees that only one click is completely in mesh with the teeth of the wheel, whereas the other is only half way. It is the same for the winding clicks, and the explanation is in the fact of a greater effectiveness, because each click takes a tooth again with each half movement. Figure 43: View which makes it possible to see the shift of the clicks 12: The winding of barrel A Barrel A is not that which runs the watch, but that which is charged with rewinding barrel B, which is the barrel which runs the watch, the function of which will be described later. As barrel A is between the plates, it is necessary that the pinion of the wheel with fine teeth penetrates between these plates. For that this plate has a relatively large round opening, adjacent to the barrel. It is through hole that the pinion passes as shown in Figure 44. In this position, one sees that this pinion meshes with the teeth placed under barrel A. These teeth are those of the winding wheel for this barrel. 21

22 Figure 44: Pinion of the wheel with fine teeth, placed on the other side of the plate It should now be seen how barrel A is constructed. In Figure 45 the barrel is open, its spring inside is of a beautiful blue, as the spring makers could make them in the 18th century. On the other hand, its inner end is not fixed to the arbor of the barrel as usual. This barrel has a simple stem in the center, whose only function is to center the barrel lid. Figure 45 Figure 46 This lid (Figure 46), which covers the barrel, is in fact the winding wheel for this spring. One sees teeth on its circumference, and it meshes with the pinion of the automatic wheel with fine teeth (see Figure 44). Moreover it is this ratchet which holds the inner end of the spring with its hook, indicated by the arrow. 22

23 So when the winding ratchet is put on the barrel, on the one hand it is centered by the stem which enters the central hole of the ratchet, and on the other hand, the hook on its central boss holds the end of the spring which is pierced with small hole. Thus the oscillations of the weight will turn this ratchet and will wind the spring. 13: Function of barrel B We pass to the other end of the arrangement of the two barrels, which constitute the energy source of these two automatic movements, of the same construction. This other end is of course barrel B, which in fact is the traditional barrel of a watch, that which provides the energy used for the oscillations of the balance and balance spring via the escapement which, you will recall, is a verge escapement that requires a constant energy. In the traditional way, it is the spring in this barrel B which makes the watch run, and the teeth of the barrel mesh with the pinion of the center wheel, Figure 47), the first wheel of the time train (Figure 48). Figure 47: Meshing of barrel B with the pinion of the center wheel Figure 48: 3 wheels of the time train, from right to left: Center wheel, Third wheel and Contrate wheel Figures 49 and 50 show barrel B, respectively that of Breguet and that of Papillion. Their particularly small diameter may be surprising, less than 14 mm, whereas barrel A is more than 18. But it should be understood that the energy in this barrel is renewed regularly and after a relatively short time, 3 hours for Breguet and 4 hours for Papillion. Two things should be noted: There is a pin planted in the lid and the square on the arbor is very long. 23

24 Figures 49 and 50: Barrel B of Breguet (left) and Papillion (right) Figure 51: The disassembled barrel B. 14: The regulating train for rewinding Figures 52 and 53: The regulating train for rewinding Several views show a mysterious little train, which makes one think of the small trains in repeater watches, because it is composed of 4 small wheels and a pinion. 24

25 It is given here from 2 different angles (Figures 52 and 53). They clearly show the 4 small wheels and the pinion. On the other hand, the view in Figure 52 makes it possible to see the meshing of the pinion of the first small wheel with the teeth of barrel A (arrow). This train has two purposes: 1) As in repeaters, it is used as regulator for the unwinding of barrel A. 2) It is by this train that the function of unwinding is locked and released. These points will be explained after looking at the complex system which is on barrel B. 15: Mechanism for rewinding barrel B The first view of this barrel (Figure 54) was taken in profile in order to clearly show that on the relatively long square arbor, as was indicated in Figures 49 and 50, is placed a particularly elaborate mechanism. Superimposed on the arbor and at the top of the barrel, there is a disk and then the winding wheel, of which we will see how they are made, noticing that the disk carries 2 pins. Figure 54: Barrel B complete, seen in profile Figures 55 and 56: The two sides of the disk This disk (Figures 55 and 56), seen from both sides, has 2 notches opposite each other and a pin on each side. It is bored through the center. Let us pass to the winding wheel in Figures 57, 58 and 59, which one can consider the interior views since, when assembled, what is shown is not visible. On a toothed wheel is fixed a ring notched like the disk, that is two notches placed opposite each other. Also, there a square recess and a hole. Finally the central hole is a square opening in a boss. In the recess formed by the ring, is placed a spring, whose end is put in the square recess in the ring. 25

26 Figure 57 Figure 58 Figure 59 We will continue the assembly of this ratchet; to my knowledge no other watch has a device of this kind, but it should be said that these movements are really unique achievements, and never described. The history of the automatic watch grows richer To finish the assembly, the disk is placed on the boss and a small steel collet holds it in place so that it can move freely (Figure 60). Figure 60 Figure 61 26

27 Placed thus, note that the notches in the wheel and the disk are not opposite each other. This position is imposed by the internal spring, because the pin of the disk fits in the square recess of the ring, where is it held by the spring. The view showing the disk transparently, gives the position of the pin with the return spring (Figure 61). The disk is thus held by this spring. 16: The locking/unlocking lever It is still not finished, and before passing to the action, it is necessary to describe an important part, which causes the releasing and locking of the rewinding mechanism of barrel B. It is a large lever with 2 arms, practically forming a right angle and whose pivot is at the apex. Figure 62 shows the lever in its position in the movement, one of its ends is on the level of the pinion of the regulating train 1, and the other on the level of the rewinding wheel 2 on barrel B. Figure 62: The locking/unlocking lever Figure 63 is another view with the trains in place, and Figure 64 is a view of the placement of the lever and its spring when the movement is assembled, a marvellous arrangement Figure 63: The trains and the lever in place 27

28 17: Preliminary data Figure 64: Position of the lever under the top plate Before passing to the explanations of operation, it is necessary to specify important data in the construction of these two movements, data which proves that at the beginning these two movements were based on the same caliber. It was not what is called a blanc roulant, which has the train already cut, because, at least for the energy train and that of the automatic system, each finisher made them differently. Here are the various counts: TIME TRAIN BREGUET PAPILLION Center wheel Third wheel Contrate wheel Escape wheel Concerning the time train we note that their various counts are identical, and knowing that the center wheel makes one turn per hour, they give identical time frequencies for the two movements, that is to say: 60/6 x 50/6 x 50/6 x (13 x 2) = beats/hour ENERGY TRAIN BREGUET Barrel A wheel Teeth barrel A Barrel B wheel Teeth barrel B Center wheel pinion PAPILLION It is different for the energy train which decides, on the one hand, the running reserve, and on the other hand, the frequency of rewinding barrel B by barrel A. We will see later the consequences of these various counts on the running reserve, but here, I indicate the differences in frequency of rewinding barrel B, determined by the relationship between the duration and the revolutions of barrel B. We note:

29 On Breguet barrel B makes one turn in 72 : 12 = 6 hours. On Papillion it is 80: 10 = 8 hours As the release of rewinding is produced by the winding wheel of barrel B, at least by the ring and the disk which it carries, both with 2 notches, we can already say that in the Breguet the rewinding occurs every 3 hours (6: 2) whereas in the Papillion it is every 4 hours (8:2). The effects on the effectiveness of the system seems to me to be negligible. 18: The watch when running To describe the operation with the greatest possible clarity is no easy thing, just as the description of the whole mechanism was difficult. To explain the operation I will present things in the opposite direction to the flow of energy, that is to say, starting with barrel B which transmits its energy to the time train and going back to the automatic mechanism which winds the spring in the first barrel A. So that the watch runs, barrel B was set up with a certain quantity of energy by the watchmaker who assembled the movement. Knowing that this barrel will be rewound every 3 hours for Breguet, and every 4 hours for Papillion, this watchmaker will have known to take the flattest part possible of the curve of unwinding, so that the energy which arrives at the escapement is as constant as possible. This first part done, it is imagined that the watch runs and barrel B turns. To know what will occur, it is necessary to return to the arrangement of the device on the arbor of this barrel, which was named the rewinding mechanism. Planted in the lid of this barrel is a pin (see Figures 49 and 50 and shown again Figure 65). Placed on the square of the barrel arbor is the disk which also carries a pin (see Figures 60 and 66). These two pins are at equal distance from the center of the barrel, so that one cannot turn without meeting the other at some time. This time is every 3 hours and every 4 hours each of these movements. Figure 65: Pin on barrel B Figure 66: Pin on the disk Figure 67: Meeting of the 2 pins Let us suppose that this time has passed and the pins arrive in contact (as is the case in Figure 54, repeated in Figure 67). 19: Position of the lever Figure 68: The mechanism locked by the lever 29

30 It is necessary to go back a little to note the position and action of the locking/unlocking lever during the time that barrel B rotates. This lever (Figures 62 to 64) has one of its ends in the pinion of the winding regulating train, and this pinion is thus locked. Here is why. Barrel B turns in the direction of the arrow and makes the watch run (Figure 68). Superimposed on its arbor is the winding ratchet 1, and under this ratchet is the disk 2. The pin placed under this disk is indicated 2a at the end of the red arrow. As for the pin placed on the barrel, it is at 3a at the end of the red arrow. In this position, one of the ends of the lever is lodged in one of the two notches (text on the photograph). And the other end of the lever is between the leaves of the pinion of the regulating train, which locks everything except barrel B which turns in the direction of the arrow (Figure 68). Figure 69: pins in contact Thus barrel B turns and the disk 2 does not turn, so that the pin 3a, approaches the pin 2a, until they meet (Figure 69). At this moment the locking mechanism is released. Figures 70 to 75, show the various positions of the mobiles during this action. Figure 70: The end of the lever (marked with blue arrows) is in the notch (red dotted line). The red pin (on the barrel) pushes the pin of the disk in the direction of the arrows. The train is blocked (Figure 74). Figure 71: The end of the lever starts to slide up the slope of the notch in the disk (blue arrow), the lever turns and its other end starts to release the pinion of the regulating train. Figure 70 Figure 71 Figure 72: The end of the lever continues to slide up the slope of the notch in the disk and its other end releases the pinion of the regulating train. Figure 73: The end of the lever is on the circumference of the disk, and its other end has completely released the pinion of the regulating train (Figure 75 below). Figure 74: The regulating train is blocked by the end of the lever which enters the leaves of the last pinion (another end position 78 above) Figure 75: The regulating train is released because the end of the lever moves from between the leaves of the last pinion (another end position 80 above). 30

31 Figure 72 Figure Rewinding of barrel B Figure 74 Figure 75 We are in the position where the watch runs and the spring of barrel B is partly unwound. So that it preserves a constant rate of energy which it distributes to the regulator of the watch, it should be rewound. This function is automatic. In parallel, the spring of barrel A is wound by the movements of the person carrying the watch, caused the oscillations of the weight. Therefore, as soon as the lever releases the pinion of the regulating train, the rewinding of barrel B occurs. As shown in Figure 76, the last pinion of the regulating train is free, because the lever turned (blue arrow) as specified above. As the first pinion of this regulating train is in mesh with the teeth of barrel A (red arrow), and its spring is wound, barrel A starts to rotate, not rapidly because the regulating train slows its rotation. As the teeth of barrel A mesh with the wheel on barrel B (black arrow), it rotates it. This ratchet is mounted on the square arbor of barrel A, which is hooked to the spring, and winds it. But how much? During the rotation of the wheel on barrel B, the end of the lever slides on the circumference of the notched disk, but as this disk has two notches, opposite each other, it is at the end of half a turn that the end of the lever falls into the other notch (Figure 77), and the other end of the lever, enters the leaves of the pinion of the regulating train (see Figure 74), and locks this train and, consequently, barrel A. Thus the wheel of barrel B has made half a turn and barrel B is rewound that much. Calculating the amount is simple and the result is different for each of these movements. As the barrel of Breguet makes one turn in 6 hours, half a turn of rewinding provides a running time of 3 hours. For Papillion, the barrel makes one turn in 8 hours, and so the rewinding provides 4 hours of running. 31

32 Figure 76 Figure 77 We thus see that, being rewound every 3 or 4 hours, the energy which arrives at the escapement of these movements is without doubt as constant as that with a fusee, but the system is not inevitably simpler to realize 21: Remark on the rewinding mechanism Like all remontoirs, the principle is always the same and the action identical, namely that they function very like an escapement. That is one finds the three phases of traditional operation there and these three phases also apply here: Rest Release Impulse, which becomes rewinding. Simply, here the rest lasts 3 hours in one case, and 4 hours in the other. Then as soon as the pin on barrel B, meets the pin under the disk, the release starts. It finishes as soon as the lever releases the pinion of the regulating train, and then the rewinding begins, which can be compared to the impulse in an escapement. This rewinding finishes as soon as the lever re-enters a notch of the disk, and it is the beginning of a new cycle which begins with a new rest. 22: Locking of the winding of barrel A Everyone knows, on a watch with automatic winding, if there were no device to limit the winding of the spring, and therefore to lock the oscillations of the weight, breakages of the spring, or the automatic train or both, would occur. The originator of this caliber thought of it obviously, and here is how he did this. The system being absolutely identical on both watches, it is impossible to say who is this originator. One needs several views to explain the arrangement and action of this locking system which is unlike all other automatic mechanisms. The first views (Figures 78 and 79) present the ratchet of barrel A, that of the Breguet part and that of Papillion part. It is perfectly clear that this is a relatively traditional system of stop-work, composed of a finger, mounted on the arbor, which meshes, with each turn of the arbor, with a partially toothed wheel, here of 5 teeth and 6 spaces on one and 4 teeth and 5 spaces on the other. I said relatively traditional, because on this system a difference appears compared to traditional stopwork. The small wheel has, after the last space, an inclined plane boss that the enlargement shows perfectly, and whose function is described below. 32

33 Figure 78: Stop on barrel A ratchet of Breguet Figure 79: Stop on barrel A ratchet of Papillion [218] Having seen the first part of this locking mechanism, that is under the ratchet of barrel A, it is necessary to see the second part, that placed on the plate. If you re-examine figures 23 and 24, you will note a part numbered 3 on the photographs, which was classified as a little mysterious. Here is what it is, with the explanations in legends: Figure 80: Key-shaped hole in the plate. The dotted red line represents a pin which crosses the part and holds it there freely. Figure 81: Locking piece of a particular shape which is placed in the opening. 33

34 Figure 82: Locking piece in place on the Papillion plate. The pin goes through the piece. Figure 83: Locking piece in place on the Breguet plate. The position and the fixing of this locking piece. It is stepped, which one sees in figure 82, and it is held in the thickness of the plate by a cross pin, represented by a red dotted line. Thus it can rock like a swing; if the back (the left part) drops, the front (stepped part) rises. Figures 84, 85 and 86 show 3 views of the way in which this locking lever acts. Figure 84 Figure 84: In this position the spring is completely unwound. The stop-work finger b is pressed against the stop-work wheel c. d is the position of the boss under the wheel, indicated by a dark area, and a is the end of the pin holding the locking piece. Figure 85 Figure 85: The mechanism is acting and winding the spring, the barrel arbor and the stop-work finger turn anticlockwise. With each whole turn the finger turns the stop-work wheel one tooth, and the view here is that after three turns of winding the spring. 34

35 Figure 86 Figure 86: Finally the stop-work finger meshes with the last tooth of the stop-work wheel. The boss on the stop-work wheel, indicated by the dark area, arrives at the circular part of the locking piece and pushes that part down. As the lever is hinged, its other end rises. But what does raising the end of the locking lever do? In Figure 87 one sees this end of the lever. Figure 87: The end of the locking lever which crosses the movement. Figures 88 and 89: Views of the locking piece in its low and high positions respectively. Figures 88 and 89 show the position of this stem clearly, placed beside one of the pillars of the watch. A difference between these two views: On Figure 88 the lever is in the low position, pushed down by a flat spring. In this position the stem does not protrude above the plate and the weight continues to oscillate with each movement of the wearer, winding the spring in barrel A. Then in Figure 89, the stop-work wheel having arrived in the position of figure 86, the boss makes the locking lever rock and the other end protrudes above plate. Figures 90 and 91 are the same as Figures 88 and 89, but the weight has been put on. It can be seen that in Fig 90 the stem is away from the weight, which can oscillate. Whereas in Fig 91, it enters a hole under the weight and so locks it. 35

36 Detail of the adjustment of locking Figures 90 and 91 Each weight is shown again here, (Figures 92, 93 and 94; see also Figures 17 and 18) and we can see the hole (arrow) which the locking stem enters to stop the oscillations which would be likely to break one or other part of the mechanism. But a simple feature in the construction of the weight makes it possible to adjust the moment of locking with precision. The weight is made up of two parts, a plate being held underneath by screws. Two of these screws hold the plate, but the third, which does not enter a hole in the weight, allows it to be moved away and thus to make the locking stem enter the hole sooner. For this adjustment it is necessary to unscrew the two holding screws, move the plate away by turning the other screw, and then retighten the other two screws A detail moreover which indicates that the watchmaker did not stint on his time Figures 92 to 94 23: Damping of the oscillations This is a device which one does not find on automatic watches with rotors, on which the rotor can turn complete revolutions, does not undergo shocks, and the system is relatively quiet. It is not the case in watches with side oscillating weights, and obviously when the weight strikes at each end of its oscillations, a relatively noisy shock occurs, which also risks causing damage to the system. To avoid this disadvantage, the originators of these calibers always incorporated what are called banking springs. Figure 95 Fig

37 Breguet also used them, and he explains it in some manuscripts that Alfred Chapuis had the privilege to hold in his hands. He reproduced parts of them in the work La montre automatique ancienne, and that which follows is drawn from his book. The book reproduces the manuscript of the extract concerning these banking springs, shown here (Figures 95 and 96), on which we see that Breguet (?) drew a spring. The transcription of the text is: when the movements of the watch are strong and the resistance of inertia is joined with an acquired force, the weight then meets the end of the spring (n) which carries a roller; and if the movements are very violent the part (m) of the weight strikes at (o), this jolt affects only the pivots because (m) is in the center of percussion, and is further dampened by the spring (o p) which is formed to prevent and resist all types of jolts. The part of the weight which corresponds to its center of percussion strikes against a stronger spring. This transcript of Chapuis, which it was necessary for him to say that it was particularly difficult, to me has a small error in the sentence that on my part I will read as: this jolt does not affect the pivots Which is quite different It is nevertheless clear that, as we will see, this description, which corresponds to all those made by Breguet and others, does not correspond to the pieces described here, and it is this system which we will analyze now. For this description it is necessary to again examine each of the weights, to see what they have underneath, beyond the locking plate, which has just been explained. Here is the end where the weight pivots and which have things attached (Figures 97 and 98). Still, here it is necessary to go back a little. In 2000, when I knew only of the Breguet watch, I found a part on the weight which seemed to me to be broken (Figure 97), but noting, in addition, 2 vertical pins on the top plate, one on each side of the pivot (the arrows on Figure 99). I advanced an assumption that I presented in my book Perpétuelles à roue de rencontre, published in 2001, and I give here what I said at the time, pages 146 and 147 of the above mentioned book. Figures 97 and 98 I wrote: After an examination with a magnifying glass I think, but with a strong probability, that the part under the weight was broken. I thus imagined that this part was a double spring, each one of its ends being broken and, a priori, they were formed as shown by computer graphics, on Figure 100, on which the red part is that remaining under the weight, and the grey part that which is missing Figures 99 and 100: Pins on the plate and the assumption of the form of the spring, made in 2001, That is, when the weight oscillates, the end of this spring comes into contact with one or other of the pins, planted vertically in the plate and indicated by the white arrows on Figure 99, and thus dampens the shock. The discovery, in September 2002, of another, identical movement, signed Papillion à Paris and the property of the Museum at Cluses, brought confirmation to the assumption, similar elements confirmed here by some views of this movement. 37

38 Figure 101: View of the banking pins for the weight on the Papillion piece. In this first view (Figure 101), the same two pins are laid out as in Figure 108 for the Breguet movement. Likewise under the weight (Figure 98) one sees a complete spring with a similar form to that which I had supposed. Finally a side view (Figure 102) shows clearly the 2 pins (white arrows) and the spring placed under the weight (red arrows). Figure 102: The weight showing the position of the pins and the spring. 24: Equilibrium of the weight Here is another point, which this time brings us closer to what Breguet made, but also what is found in practically all automatic watches at the end of the 18th and the beginning of the 19th century: The equilibrium spring. Breguet explained to us the purpose of this spring; he said this: Thus, when the watch is vertical, with midday at the top, the support spring must hold the arm of the weight in an almost horizontal position, a little above the center of the oscillations which it must sustain. The weight thus held in equilibrium resists by its inertia and remains almost motionless during the vertical movement of rise and fall which the watch caused by walking or the other similar movements of the body. The case and the movement only oscillate, compared to the weight, which, as a result, is the same thing as if the weight only oscillated. I do not know if that is perfectly clear for everyone and a demonstration will make certainly the thing more comprehensible. Here are 3 possible arrangements, after making a hole in the end of the handle of a hammer and having fixed this hammer to a wall with a nail in the hole, leaving the hammer free. 38

39 Figure 103 Figure 104 Figure Rotor (Figure 103): In this arrangement the weight is pivoted in the center of the movement and its mass always holds it vertical. It does not need any spring and it winds the mainspring of the watch either by swinging or by rotating. 2. Lateral weight (Figure 104): In this arrangement the weight is pivoted on the edge of plate horizontally. But if it is left free like a pendulum weight, it remains in its bottom position consistently. 3. If one adds spring, (here, for the example, a rubber band, Figure 105, which holds the weight in an intermediate position, one notes that the least movement from top to bottom makes the weight oscillate very freely. Let us see the arrangement of each piece. If one looks at the part of the plates of each, where the arbor of weight is (Figures 106 and 107), one notes a difference On the arbor of weight of the Breguet watch, one sees a small equilibrium spring attached to the arbor and whose other end is fixed to the plate by a screw (indicated by the large arrow Figure 106). Here in addition, indicated by two small arrows, are two holes in the plate and it is possible to place the end of equilibrium spring in one or the other, and thus to regulate the tension and therefore the balance of the weight. On the other hand, on the same part of the Papillion movement (Figure 107), one sees the arbor of the weight, but the equilibrium spring is absent, and there is a small plate Is this another system? But where and why? We will see shortly. Figures 106 and 107: Part of the plates showing the arbor of the weight and the equilibrium spring. But now, by placing a reduced and transparent image of the weight on the arbor of the movement signed Breguet, (Figure 108), one sees the function of the spiral spring which, according to its tension, will maintain the weight in balance, as in the demonstration of the rubber band supporting a hammer. 39

40 Figure 108: Position of the weight (transparent) and the equilibrium spring And what is on the Papillion movement? It is necessary to look at the other side of the movement, (Figure 109), that of the dial-work, to find a curved spring, red arrow, which appear to be not very catholic, but nevertheless is quite elaborate. Figure 109: The equilibrium device on the Papillion movement The blade of the spring is attached to a base, on which it is held by a screw, which also holds the base. Its other end is in contact with the winding articulated lever. Thus placed, one imagines that this spring holds the weight in equilibrium, like the equilibrium spring on the Breguet movement. Is this arrangement original? For my part I think not, and that in the beginning there was an equilibrium spring. The small plate in figure 107 that I described as odd, and also the 2 small holes which one sees (Figure 110) which were certainly intended to regulate the tension of an equilibrium spring, tends to make one think so. Figure