Back Up Next

 POUVILLON RESTORATION PROJECT - August 2011

A hypothesis on the time line of the clock's construction

Based on the observations gained in the deconstruction of the clock as well as the literature we have so far, we have come to a hypothesis about the time line in connection with the construction of this clock. It has been widely quoted in prior literature and in particular the two auction house catalogs through which this clock has passed that the movement was a masterpiece made by Mr. Paul Pouvillon between 1930 and 1939. In the previous installments we have observed that the clock when stripped of all of its complications looks to be a complete clock. One that easily could have been conceived and executed as is and with no thought of future alterations. This clock was a two train, hour striking clock sharing many design elements seen in tower clocks; these are:

A. The frames are made of wrought iron rather than brass.
B. The rear frame pillars are turned, not unlike those seen in the tower clocks which Pouvillon was familiar, specifically the church clocks in Strasburg and Beauvais, France.
C. The dial setting clutch mechanism for the time is identical in design to those used in smaller tower clocks of the day by other French makers, Schwilgue, Ungerer, Lepaute and Gugumus
D. The oil holes used in the frame members are seen only on tower clocks; not on domestic movements
E. The massive pendulum bob at 7.085kg or 15 lb 9.5 oz is what would be expected in a regular sized tower clock.
F. As a result of the heavy bob the entire pendulum suspension system takes on the robustness of a tower clock.
G. The fine pendulum adjustment is turned using a Tommy bolt due to the weight of the bob.
H. There is a pendulum safety catch to prevent the pendulum crashing to the floor in the event of the suspension spring breaking; a feature unheard of in domestic clocks, but quite common in tower clocks.
I. The construction of many of the main frame wheels have integral collets which are then pinned to the arbor. The collets are not permanently attached to the arbor.
J. Many wheels are bolted to their collets as often seen in tower clocks, but not in domestic clocks.
K. The bevel wheels are cast in the manner of tower clocks with the interior spokes coved in relation to the rims. A few of the larger steel pinions are also made this way.
L. The construction of the going and strike barrels are exactly as those found on smaller tower clocks, with removable, screw down collets that act like nuts on a thread cut into the arbor.
M. The pins on the escape wheel are each set within collets. Something sometimes found on large tower clock pinwheels. On domestic clocks the pins are set directly into the wheel rim.
N. The escapement pallets are removable and are held with a screw. The pallet depthing is adjustable via a screw acting upon a slit along the vertical axis of the escapement. While this second arrangement is not unheard of in domestic clocks, it is quite common in pinwheel escapement tower clocks.
O. Lathe turning centers left in the ends of the winding square arbors, a common feature in tower clocks, but unacceptable in domestic clocks.
P. The strike hammer lift cam is fabricated with separate components the same as some French tower clocks, in particular Collin-Wagner.

We believe Pouvillon constructed this clock before 1930 as a standalone clock without consideration of later adding any complications. This is supported by the fact that none of the attachment points for the complications were special-made within the frame. All attachment points were spots that were filed flat and holes that were often awkwardly placed. In many instances we see parts of the frame that were removed to allow for the positioning of a complication or peripheral component. All of the drives to the complications mostly in the form of bevel wheels are attached to the phase one arbors in what appears to be an ad hoc fashion. There are no provisions made for correct fitting of these later bevels on the tapered arbors. In the phase one clock, all wheels had proper parallel mounting points on the otherwise tapered profile of the arbor. None of these conditions would have been present with a clockmaker of Pouvillon's talent if the clock was pre-planned with the complications. Furthermore, the components of the clock have the wear one would expect from a much older movement and that has been re-polished on several occasions; all of wheel spokes and teeth corners have been slightly rounded. None of these conditions are present in the various complications. In addition the two sandwiched plates under the tellurian/orrery structure which support those systems is a clear example of making the upper plate which was original to the tellurian assembly and was not built by Pouvillon, fit to the existing wrought iron clock frame via the lower plate. See A discussion about the origin of the tellurian assembly. The fact that the phase one clock stands as a beautiful example in its own right when completely stripped of all complications stands as testament to this.

Below is the time line we think the clock took based on our forensic disassembly and the written literature we have to date.

Phase one is the original clock, going, and strike trains – no orrery. We believe this is pre 1930 and when built was not intended for use as a base for the later complications. In other words it was built as a complete, stand alone, clock. Probably in the very early 1900's.

Phase two is the addition of the "borrowed" tellurion (not to be confused with the later orrery of the outer planets which was added during phase three. That is below the superstructure holding the tellurium). We believe this to be the first complication as many power feeds to other complications come from this system. This is the only complication that was not entirely made by Pouvillon, the only other parts Pouvillon did not make were a few watch-sized bevel wheels in the mystery dial.

Phase three is 1930-1939. Here he begins his project to install the several complications on the front and sides of the phase one clock frame. These may not be all of the same complications we see today. From a date stamped on the back of the Easter calculator's center calendar disc, it appears the calculator was completed sometime in 1946 and from the literature we know he had made a final adjustment to this system in 1953 (to make the calculator perpetual). So either an earlier iteration of the calculator was there or something else may have occupied this space when the clock was exhibited in 1939. It's unlikely Pouvillon would have left this area empty. Therefore the Easter calculator could have been either in the phase three or phase four sequence. We think that probably an earlier, non perpetual form of the calculator was in the clock when Pouvillon was receiving his various acclamations in 1939. It certainly is the one complication that makes this clock stand out from from most others. Considering our assumption that WWII disrupted his work, we feel that it is unlikely he could have designed and built the entire calculator in one year or less.

1940 through 1945, WWII could have disrupted Pouvillon’s work and he may not have returned to the clock until 1946.

Phase four is between 1948 through 1953, here he adds the seven small white dials around the base of the orrery wheel pack. The leap year, the unequal hours-state of strike, the day and its zodiacal sign, the month and its zodiacal sign and the season.

Phase five is 1953-1954. In a newspaper article dated 1953 it is reported that he conceives of the perpetual piece for the Easter computer.1  We know the Easter calculator has to be tripped annually, and this would have been from the missing assembly, the 'annual cam pack' that we have reproduced which drives the mystery dial indicating 'sun time', as well as the sun rise, sunset shutters and length of day, length of night dials. We also know that for the Easter calculator to operate correctly the Epact dial needs an extra feed every 19 years. This will allow it to be perpetual for 400 years or until an extra leap year needs to be inserted and we believe Pouvillon used the fact that the Golden Number has a 19 year cycle to give that extra feed to the Epact dial.

 In the 1955 newspaper article it is reported that Pouvillon was working on a final indication, the rise and setting of the moon. 2 However, we see no evidence for this and in this effort we must assume Pouvillon failed and had added no other features past this date. By this time he was 77 years old.

Back Up Next

1. Newspaper clipping written in France, presumably Paris and hand dated 1953. 2. Newspaper clipping dated February 20, 1955, Le Parisien.