Back Up Next

 POUVILLON RESTORATION PROJECT - October 2011

Begin restoration of the Easter calculator, the annual trip snail cam, cam release lever, warning lever and calculator release lever.

Note that the format of this section of the web site is designed for the newer wide screen monitors, those using the conventional 3:2 ratio monitors may need to move the horizontal scrollbar at the bottom of the screen to see the entire page.

I have asked Buchanan to provide what I call a 'forensic' report. That is to record his observations as he goes along. I will provide the .MP3 audio file for each segment but just in case your security settings will not allow you to open this file I have also transcribed each session. My additional comments will be inserted into the text from time to time and this will be in red text. Buchanan refers to each photo by the number of that photo which can be followed by each photo above the captioned text. The .mp3 audio file will appear in blue text.MP3. Click on this text and you can then follow along with the audio file by scrolling downward through the photos as they are narrated one by one in the voice of the restorer.

At this point in the project we are beginning to examine and explore how the Easter calculator was originally made, and consider the changes that occurred after Pouvillon’s death; the missing parts and the tortured additions from later amateurs, and we went through many thoughtful exercises  to correct the damages done to this sub-system of the clock in the past. While we have already pointed out within the following text at many points, let me say here for the record that we have done nothing to this artifact that is not fully reversible. We have not altered the artifact in any way. There have been no holes drilled, no original parts altered or discarded. For any parts we reverse engineered we took pains to try to make those parts as best as could to be faithful to the original design of Pouvillon. In most cases we are confident this has happened. In a few others, lacking any historical photographic evidence we can only do our best. To simply throw up hands and say we must do nothing when we are not 100% sure of how the part was originally made, or worse, simply adhere to the strict restoration school of thought where nothing is added regardless of reversibility would be to have relegated the clock to be largely non-functional. This is the condition the clock has been in for at least the past 30 years and during this time no effort was made to prevent further deterioration yet alone preserve what was there. A non-functional clock, especially one like this which cries out to the viewer to 'Look at me, see my inner mechanism working!' is an unappreciated clock and will not last.

Pouvillon-47-010.MP3. Photo 47 010 shows the pillar mounted on the base of the clock which has no logical function. Of interest to note is that the construction of the components mounted on this pillar are extremely crude and don’t match the style of work done by Mr. Pouvillon at all. Photo 47 011 is a clevis or a rod-end on this mechanism showing the crude and bad workmanship as the flats on the cylindrical tip portions, bad filing and off-center drill holes. Actually, upon closer observation what we have here is a basic “off the shelf” hardware part not unlike that which would be used in a radio controlled model today. The only difference is that this part is of a heavier gauge. This clevis, as shown in the first photo is attached to a lever which is pivoted on the uppermost horizontal pin.  Photo 47 012. Another view of the pillar and the attached lever. The thick rod through the main part of the pillar has just been cut off or filed very roughly. The back end we also see in an irregular finish the taper portion of the pillar.

                                                  

Photo 47 013. Another view of the extremely crude workmanship of the attached components to the pillar. Also interesting to note is the wavy finish on the upper tapered end of this pillar which is very unlike any of Mr. Pouvillon’s other work. Also of interest is the small-headed screw that attaches the pillar to the base which is very different to the other substantial hex-head nuts that hold all of the other components of the clock to the base. Photo 47 014. Another view of pillar and its detritus, not shown. This pillar is our most difficult artifact to surmise as to its intended function. At this time we have no answer. We are confident that Pouvillon did make this pillar as a later addition to the clock to serve some function. As received, it is obvious someone had tried to give it a purpose by using it as a trip platform to part of the Easter calculator. A longer pin, different from all the rest, was driven into the face of the strike main wheel which is adjacent to the pillar and is designed to contact the crude lever arrangement attached to the pillar. This, in turn is connected by a foreign-looking clevis, rod assembly that is connected to a pin on the ratchet feed to the Epact feed mechanism, blue circle photo 49 002 below. This means that the rod would feed the Epact wheel once per day, since the strike main wheel rotates once daily. This makes no sense as the Easter calculator mechanism is designed to be tripped as a whole once per year only and all calculations are done internally at that time. Simply feeding the Epact has no effect on the rest of the mechanism and so it would permanently lay dormant. We surmise that someone was looking for a purpose to this pillar as we are, and chose the “path of least resistance” since one can run a connection straight up from the pillar to the feed of the epact dial. But it is certainly not correct.

The pillar has a hole drilled into the top surface, in a similar fashion to an earlier pillar that facilitates the strike fly fan assembly. In that pillar the top surface hole serves as a pivot for a vertical drive to the fly fan. However, we have not been able to deduce any such function for this later added pillar. This hole is twice the diameter and so may have been used as a mounting point for another part. At first we thought that maybe the Easter calculator annual feed was derived from a vertical arbor driven from the cam pack to the Easter calculator, but there was no way we could reproduce this connection without having to make changes to the existing artifact. So at this time we have no answers for this later added pillar.

Pouvillon-49-000.MP3. Photo 49 001. We are reducing the thickness of the warning piece. It is attached by adhesive to a block; again attached by adhesive to the table of the jig borer and we are bringing it down to match the thickness of other plates in the Easter calculator mechanism. Photo 49 002. We have in the center of the photograph the thirty tooth ratchet wheel which is read by a sector gear and this is the ratchet wheel that carries the hand for the Epact dial. It requires an eleven tooth feed once a year and every nineteen years it requires a twelve tooth feed. Now we are considering the possibility of Mr. Pouvillon intending to incorporate a nineteen year correction mechanism to give us a perpetual Epact calendar. This is referred to in both of the newspaper articles written in around 1953. Later in this installment we will see that Pouvillon probably did not complete the Easter calculator until 1946, after WWII. It’s possible that he did not even have the calculator in his clock when it was displayed and awarded the honor of Meilleur Ouvrier de France in 1939. More on this in the following photographs. Photo 49 005. We have a view from the opposite side of the thirty toothed ratchet wheel for the Epact dial and in the foreground you can see the eccentric stop pin for the feed pawl lever. This stop pin is in its natural brass state. It is not gilt like the rest of the mechanism and we are wondering whether it is part of an incomplete mechanism. If one were to rotate this stop, one would have a twelve or  eleven tooth feed that can be selected. This may have been Pouvillon's first attempt at the Epact correction - a manual 'change switch' activated every nineteenth year. But it's hardly a practical or desirable solution. The pin circled in the prior photo we think is part of what Pouvillon referred to as his discovery to make the calendar perpetual in the 1953 newspaper article. We explore this now.

Photo 49 006 is the Golden Number arbor. This arbor on the left hand side has a gear which transfers its motion through to the disc on the center of the Easter calculator which gives the actual date of Easter. And just behind the vertical feed wire we have a nineteen tooth ratchet which gets one feed per year. We’re looking at the collet on the transfer gear,  blue circled area. We can see, of course, the gear itself and the grub screw through the center of the collet and then we have a reduced shoulder which is unusual in this clock as it does not match any of the other standard collets on other ratchets or gears in this mechanism. And when one considers the need for a nineteen year correction factor on the Epact dial and here we have an arbor which rotates one in nineteen years and an unused shoulder on a collet, we’re wondering whether this was not a provision Mr. Pouvillon had made intending to fit a correcting lever to the Epact dial. Photo 49 007. We have again another view of the Golden Number arbor and side-by-side we have the collet for the transfer gear and the collet for the ratchet wheel which shows the two types of collet and the unusual nature of the transfer gear collet, not shown. Photo 49 009. If we can look at the portion in the center of the photograph that is focused you’ll see we have a plate pillar with a flat machined on it in the position which would require clearance if there was an arbor passing from the Golden Number dial to the Epact dial as you will see in the later photograph. Pouvillon-49-010.MP3. Photo 49 010. We have a rear view of the Easter calculator and I’ve place a piece of blued steel wire across the frames connecting the Golden Number mechanism on the lower left hand side, to the Epact dial on the upper right hand side. And if one imagines a correction lever passing across the frame in this manner, you’ll see that it is almost touching the pillar shown by the white arrow, and it is this pillar that has the flat machined on it. The same flat shown in the prior photograph and indicating that something indeed was passing close to this area.

 

Photo 49 011 is the warning piece fitted to the existing mechanism and you’ll notice how well placed the arbor of the locking lever is for the fitting of this warning piece. The attachment is fully reversible, blue circled part. Photo 49 012 is a new possible position for the manual trip lever and its bracket. As received the bracket was mounted on the opposite side of the frames projecting the long wire from the front of the calculator dial and we’re looking at whether or not it was mounted elsewhere on the movement to provide a trip coming from the one year arbor with all the cams feeding the sunrise/sunset and also the equation of time dials. Photo 49 013. Another view of the manual release, or trip bracket in its new possible position. Here we simply flipped it over doing nothing else to the piece to see how this might work. We believe this manual release was added later on after the ability for the clock to automatically activate the calculator was lost.

Photo 49 015 is the warning piece on the locking arbor and you’ll notice that the bluing on this arbor has been removed for some reason. This witness mark later seems to confirm our observations that a part was mounted to this area of the arbor in the past. Photo 49 016. Another view of the warning piece on the locking arbor, and we’re just trying to point out the positioning of the rocking lever arbor and how it lends itself to, an existing warning pin on the worm gear. If one looks carefully it is not hard to see that Pouvillon borrowed the worm, the worm drive gear, and fly fan from another source; most probably a music box movement. This is another example of Pouvillon’s willingness to use expediency to get the job done. However, to give credit, in this case the making of a worm gear is a very different from the skills used to make a conventional gear and the making of a worm, especially of this small size could prove to be very difficult. It is likely that Pouvillon added the small detent pin on the face of the worm drive wheel since most music boxes use a detent directly off the fly fan assembly, see photo 52 028. Pouvillon-51-000.MP3. Photo 51 001. We have an arrow pointing to an empty tapped hole in the drop down pillar for the one year cam pack. We’re proposing to use this pillar to mount the release latch for the Easter calculator. There’s a corresponding tapped hole on the opposite side of this pillar which holds the cross arm that carries the two length of day and length of night levers. And this hole which is a smaller tapped hole carries a clamping screw. Again we are doing nothing that involves the alteration of the original artifact. We believe we are replacing parts that were lost in the past, and are reverse engineering, these parts to fit exactly within the exiting architecture of the clock.

Photo 51 002. Another photograph showing the drop down pillar and the tapped hole we are proposing to mount a stub arbor to carry the release mechanism for the Easter calculator. Photo 51 003 is the dial carrying the calculated Easter dates that goes behind the center dial of the Easter calculator, and each number appears in succession in the little window. You’ll see there are two rows of numbers; the outer row pertaining to April and the inner row pertaining to March. Photo 51 004. The rear of the Easter date disc and here another lot of numerals. Interesting on the left hand side on the inner edge is the 1946 date and the sequence of numbers on this disc are the dates of Easter from 1946 which I think really dates from the Easter calculator mechanism as being completed either on or  just before 1946. This sequence of numerals on this side of the disc is, in effect, numbered incorrectly. They were in reverse order. I believe they were the first numbers imprinted on this disc and they were in the incorrect order. So he reversed the disc and punched in the new set of numerals in the correct order on the opposite side. We think this firmly dates the completion of the Easter calculator to 1946. The function of this disc is the result of the Easter mechanism’s calculations and so would have been one of the last pieces fitted to the mechanism. It is doubtful Pouvillon would have made the reverse numbering order mistake if he was making a second or third disc. Remember each disc spans 19 years. So even assuming Pouvillon had completed the calculator in 1939 he would only have used seven out of the nineteen years on the disc. There would be no reason to make up another one in 1946.   

 

Photo 51 005 showing the two identification or alignment punch marks in the collet as well as on the Easter date disc. Photo 51 006. A clearer picture showing the two locating, identification punch marks on the disc and on the collet. Photo 51 007. A photo showing the single center punch mark or dot on the collet which corresponds with a mark on the opposite side of the dial where he originally had mounted before, I think, he discovered he had incorrectly numbered the side of the disc. See Photo 51 009 below.

 

Photo 51 008. Another photograph of the two identification punch marks. Pouvillon-51-009.MP3. Photo 51 009. Just below the '2' on the left hand side of the dial right up near the screw we can see the alignment center dot which was originally used when he had the disc mounted the opposite way around. Also note the correct side has the appropriate silvering on the face; however one can distinctly see the area that was originally covered by the collet. The reverse, which had the incorrect numbering, never made it to the stage of final silvering as the mistake was discovered beforehand. The empty holes on both parts also attest to the repositioning of this piece. Photo 51 010. In the next few photos we’re going to show the scribe marks that remain on the Easter calculator plates and I think show the work of an older man who’s losing some of his physical ability and the finish on this mechanism shows many short cuts or quicker and easier methods of construction which don’t match the earlier portion of the clock. Although the design features, frame designs are clearly the same person. In the background we can see a vertical brass rod that has just been squashed flat and deformed to produce a section to draw a pivot hole through. Also we can see a click for one of the ratchet wheels which is nothing other than a flat piece of sheet brass and bent to provide a right angle.

 

Photo 51 011. Another view showing scribe marks on the frame where one can clearly see the very slender frame design. But the filing has stopped long before the scribe marks are reached. Also of interest to note on the far right hand portion of the frame is the off-center pivot hole for a ratchet pawl. Photo 51 012. Another view of the very slender set of scribe marks and filing that is far from approaching it. Again in the background we can see brass wire that has been squashed flat to provide pivot points.

 

Photo 51 013. Another part of the frame showing scribe marks and also on the left hand side we can easily see a ‘bent up’ Click for the ratchet wheel. While it is clear at this point that Pouvillon was taking short cuts, we must remember that he was, if are to believe the 1946 date on the Easter date wheel, at least 66 to 67 years old at this time. While some people retain their skills well into old age, many do not. In all fairness, the scribe marks not being met exactly is not completely a matter of Mr. Pouvillon being sloppy, although there is ample evidence of this in many places on the calculator frames, see green circle. Careful examinations of these lines show many of them being impossibly too slender if Pouvillon had tried to meet them. In other cases the lines intersect in such a way as to make meeting the outlines impossible. In many places the lines were re-scribed one or more times, see red/black arrows above. If one looks carefully on the ‘wagon wheel’ in the lower left hand corner of the photo one can see the center scribe on the near nine o’clock position spoke is way off center compared to the rest which are fairly centered. On the other hand, the curve on the lower right hand side could have easily had the lower scribe mark filed to the line, left green/white arrows. What I think happed was that after Pouvillon began to file out the frames he realized the original scribe line mistakes and then tried redrawing them. This may have gone badly and so he then began to cut the frame simply ‘by eye’ to avoid having to do the entire assembly over again. 

 

Photo 51 014. Some more scribe marks. Photo 51 015. Another view of the Easter calculator frames we can also see on the top right hand side a little bracket again just formed by bending; not up to Mr. Pouvillon’s usual standard of work. There is no question that the smaller parts like the ratchets and some brackets were made in a way that perhaps Pouvillon thought he might later correct or make to a finer form after they were tested. It’s hard to say. Photo 51 016. We have the chapter ring for the Dominical Letter and of interest is at the 6 o’clock position we have engraved ‘FD’. Which should actually read ‘ED’. This just an engraving error. The whole sequence of operation is correct it just doesn’t match the correct order that the Dominical Letter proceeds in. Again we have to wonder what was going on here. We have to assume Pouvillon was conversant with the Dominical Letters as he designed a calculator to compute them. If this was a simple mistake by the engraver why was it not corrected? It was an incredibly easy fix to make the ‘F’ an ‘E’. One can only assume Pouvillon did not catch this due to his failing capacity or felt that he would correct this at a later date.

 

Pouvillon-52-015.MP3. Photo 52 015. We have the embryo release mechanism for the Easter calculator, see arrows. On the center arbor on the one year cam pack we have the blank for the snail which obviously releases the lever on the 1st of January or on the changeover from the 31st of December to the 1st of January. And we have the actual latch or operating lever. It’s mounted on the stub arbor connected to the drop down pillar. Again here we have used only that which was present on the artifact. The attachment point to the drop down pillar was already present upon which we connected our reverse-engineered part. Photo 52 016. Another view of the release mechanism for the Easter calculator, not shown. Photo 52 018, not shown. Photo 52 024. A view of the catch or release lever and we have a spade tip, see arrow, on the end of this lever to maintain contact with the snail in the event of any side-to-side wear taking place. Here one may ask why we did not adopt Pouvillon’s correction for any side-to-side errors of the cam reading lever as he used on the equation of time cam. However, in that application there is a constant contact between the cam and the reading lever at all times. So Pouvillon could use a small set of ‘whiskers’ on each side of the end of the reading lever to keep it aligned upon the rim of the cam. In this instance, the reading lever does not ride along a contiguous smooth cam surface. It rides upon a snail cam, which means that at one point in the cam rotation it will meet a discontinuous point on the cam surface, that is, a drop off. If there is any error in the side-to-side alignment of the reading lever as it drops onto the cam face it will misalign and jam up. For this reason we use the spade device. We cannot know exactly how Pouvillon would have made this lever look absent original photographs which we have tried desperately to locate and so far with no success, but we are confident that the Easter calculator would have been tripped once per year from this location given the exiting mounting points, the space on the annular cam pack for the trip snail and the fact it is in the correct alignment from this cam to the Easter calculator’s trip lever location directly below. We did try to mimic Pouvillon’s existing equation of time lever which is located directly to the right. The damascened surface on that part will be copied on the new release lever.

 

Photo 52 025. This is the lower part of the Easter calculator mechanism and roughly in the center of the photograph you will see a horizontal, nude arbor with the new latch and warning lever for the Easter calculator release white circle. And the long, extended, inner, right angle bent lever along the top right-hand frame of the mechanism which is to finally connect with the release mechanism on the one year cam pack. This arbor is original to the calculator and is used for another exiting locking lever as outlined by the arrows. Photo 49 015 shows the blank arbor and the witness mark that shows something was mounted on this area of the arbor, we cover that mark perfectly. Photo 52 027. This is looking at the Easter calculator from the rear and you can see just behind the worm wheel gear we have a lever which engages with a pin in the worm wheel to provide suitable warning. Obviously, we are loading the Easter release cam during the course of a year and as it drops off we want the calculation to take place immediately. The white circle shows a bent portion from our locking lever mounted to the arbor described in the prior photo. With our new warning lever, this part engages our new lever which supplies a critical signal to the existing locking lever described in the prior photo. As found that lever would have lain dormant; preventing the proper functioning of the calculator.

 

Photo 52 028. Here we have the new release lever in the warning position. The white circle shows the existing pin in the worm drive wheel and the arrow points to the new detent that engages that pin and is a part of our new warning lever. Photo 52 029. We have another view of the new Easter warning mechanism. The green circle shows the two new parts mounted to the existing arbor and within the area indicated by the witness mark on that arbor. Once again we have made no alterations to the original mechanism and all additions are fully reversible.  The blue circle shows the area where the existing locking lever is attached. The new release lever is that entire angled, ‘L’ shaped inner piece. We chose this shape for the lever since it is the most inconspicuous, and tucks neatly behind the existing rectangular frame bracket. Without historical photos there is no way to know how Pouvillon would have made this look; it could take any variety of shapes. With the annular cam pack, we had some historical evidence to go on and a clearly defined environment from which to reverse-engineer our parts giving us a firm idea of what these parts would look like. With the Easter calculator levers, we can only guess at the best possible shapes. We do not doubt their necessity to carry out the intended functions Pouvillon had originally designed for the clock nor do we think the parts are functionally different from Pouvillon’s original design.

   

             Loop:

Pouvillon-41-005.mpg 

For those who cannot use the built-in viewer above, I have added the YouTube viewers below. You can also click on the individual video links above and a clip will open on your computer's built in video viewer. The videos showing the cycling of the Easter calculator are very large and can be streamed from the YouTube videos below. These two videos show the calculator   

 

Back Up Next