Restoration of St. Mark's Tower Clock in Venice.

Report on the Mistakes Made During the Restoration of the Clock

Dr. Alberto Peratoner

This article is available in pdf format for off-line reading and printing.

Click on this link to download the compressed file (1 MB)

The clock, in the configuration it had before the last restoration, was the result of a series of stratifications that took place in the five centuries of its history:

1495-1499: Gian Paolo and Gian Carlo Rainieri from Reggio built the clock.
1551: General restoration, carried out by Giuseppe Mazzoleni.
1613-1615: Multiple restorations, carried out by Giovan Battista Santi.
1753-1759: Reconstruction, carried out by Bartolomeo Ferracina. The original escapement, with a verge balance (or foliot), was substituted with an alternative system (i.e., an anchor escapement) regulated by a pendulum.
1858: Luigi De Lucia perfected the timekeeping system with a particular escapement and the replacement of the pendulum (made of wood; 4,15 meter in length). Also 2 large paneled wheels (called tàmbure) for the visualization of hours and minutes on the outside were introduced.
1865-1866: Complements of restoration, carried out by Antonio Trevisan.
1952-1953: Overhaul, carried out by Giovanni Peratoner, with correction of the pendulum's plane of oscillation.


It should be noted at the outset that De Lucia's intervention (in 1858) consisted mainly of additions, including new, well-localized systems, in many cases put in place with surprising respect for the previous order.

At the end of 1996 a new restoration was entrusted to Piaget. In reality, it was only a sponsorship because Mr. Romanelli, Director of Venetian Civic Museums, directly hired blacksmith Alberto Gorla (who had already restored other ancient mechanisms) and historian of horology Giuseppe Brusa.
Mr. Brusa admitted to me that he intended to eliminate the 19th century stratification of the clock that, in his opinion, had "perverted its nature," and to warmly support a presumed philological restoration. My adamant opposition to this concept brought about, in a short period of time, my total exclusion from the restoration project, even as a source of information on the state and the real necessities of the clock.

At first, Mr. Brusa and Mr. Gorla wanted to substitute the eighteenth century clock face with a hypothetical - and questionable - reconstruction of the more complex astronomical dial of the Rainieri brothers (even though the present zodiac decorations are believed to be those of the original dial). When the Direction of Civic Museums (i.e., G. Romanelli) explained to them that such an operation would be impossible, the two men doggedly concentrated on the two large paneled numerical wheels (Tàmbure). They relied on the misconception that such wheels had "substituted" the mechanism that permitted the carousel of the Magi, a mechanism which, in their opinion, was suppressed because of the wheels' clumsy presence. This contention is absolutely false: as I have amply demonstrated in my book, the cleverness of De Lucia's intervention lied precisely in his ability to put the two functions in a "dialogue" through ingenious solutions that permitted their alternate activation. For many years I personally dislocated the parts, during the periods of activation of the Magi's carousel.

Zodiac decorations

Detail of zodiac decorations.

Click to enlarge

This picture shows how Mr. Brusa and Mr. Romanelli are wrong when they say that the large panelled wheels (for hours and minutes indications) are not compatible with the Magi's carousel. The large inner toothed wheel on which the figures were mounted has not been removed. It's clear that the tàmbures were mounted over the Magi's mechanism without the need for its removal. On the top of the picture, you can see the levers which allow lifting the panelled wheels during the carousel.
Click on the picture to see a more detailed version.


When it was decided that even the removal of the great tàmbure was not acceptable, Mr. Brusa concentrated his interest on the pendulum, the escapement and the relocation of the weights' chords below the frame of the central mechanism, with the consequential loss of the vertical development which exploited the height of the tower. This time the idea passed, maybe because it did not affect any of the external (and therefore visible) features of the clock, and the institutions involved could quietly take refuge behind the appearance that, once the external appearance of the tower was saved, the rest could pass as bickering among specialists which could only be understood by few and would continue indefinitely without ever reaching an agreement.

In 1998, while the press continued talking about a conservatory restoration, the clock, taken apart and moved to a shop in Mantua, was profoundly transformed:

a) The pendulum was relocated and brought to the opposite side of the central frame (North), far from the 'Timekeeping train' (South) and beyond the sectors where the striking trains are lodged.

b) The pendulum was substituted by a shorter one, about half the size (from 4.15 to 1.90 meters), equipped with a spring suspension (the previous one had a knife-edge suspension).

c) The new pendulum was centrally located in correspondence to a pendent guide-bar connected to an arbor that crosses the whole frame to bring - as I noted - the movement to the opposite side. The original pendulum, instead, was located laterally and was moved by a horizontal arm, hinged in the points of conjunctions to the vertical guide-bar and to the pendulum.

d) As a consequence of the mutated oscillation period (from 2 to 1.36 seconds) the escapement was completely reconstructed with its gearing (realized ex novo in the pinwheel style as the previous).
Thus the heart of the clock, the most vital and qualifying part of the whole mechanism with its beautiful and rare 4-meter-long pendulum, lost its function.

The new 1.90 m. long pendulum located on the north side of the main frame, far from the time train and beyond the striking mechanism. The suspension is of the spring type. The entirely new escape wheel can be seen together with the newly made split pendulum rod. Gorla had to use this solution to accommodate the axis to the north dial (towards the "Mercerie").
Click on the picture to see a more detailed version.

Click to enlarge

Mr. Romanelli, who thought of these changes as negligible "adjustments," once said: "nothing will be lost: everything will be exhibited in a museum," adding that these operations were conceived with complete reversibility in mind, as requested by the Superintendence as a guarantee.
At the Superintendence, officials vehemently denied that, indeed, this had been a retrospective philological restoration, telling me that, otherwise, it would have not been approved. Those choices had been presented to them as "technical necessities" and improvements. As far as the supposed 'necessity' is concerned, however, it is clear and known to everybody that the clock worked fine until 1997.
The claim of "technical improvements", is also absurd because, as knowledge and science improve, every kind of ancient mechanism may be improved. Indeed, such improvements are not carried out because, even though conceptually outdated, all historical mechanisms preserve an intrinsic value in the history of horology. Furthermore, when these mechanisms are functioning, as it was in our case, the reasons to preserve them unchanged are even stronger. It was sufficient (and a lot cheaper) to rectify the worn out parts to guarantee a long, precise functioning; it is, in fact, in this shape that the clock has excellently worked for 140 years and thus no improvement was necessary. No one, in fact, has ever thought about "improving" the mechanisms of the monumental pendulum clocks in Versailles to make them more precise, or because they did not match the philological taste of the restorer; nor has anyone ever dreamed of 'bringing into focus' Impressionist paintings.

On February 1st 1999 the clock, transformed as I described, was reassembled at Palazzo Ducale and presented to the press, and there it still is waiting for the conclusion of the structural work being done to the walls of the tower.
In reality many of Mr. Brusa's contentions (included in the short pamphlet presented in that occasion) openly displayed an intention to carry out a retrospective philological reconstruction, grimly disrespectful of the undoubtedly valuable historical mechanism.

At this point, it should be noted that the last letter sent by Mr. Romanelli to the newspaper "Il Gazzettino" (on August 25) increased the suspicion that a "philological reconstruction" was what was actually carried out. This letter referred to the 19th century restoration in unflattering terms, as an operation that had radically upset the very concept of the clock and "its philosophy," even though there are archival documents that clearly prove such assertions false.

It should also be noted that this last restoration is indeed reversible as requested, and there is still time to take action in that direction.

Confronted with the accusations published by the newspapers "Il Gazzettino" (August 22) and "La Nuova Venezia" (September 12), Mr. Romanelli, responsible for having supported Mr. Brusa's choices, responded by asserting the principle of scientific authority: "we relied on the best specialists," he said. Therefore it is important to observe that the publications written so far by Giuseppe Brusa reveal a rather superficial knowledge of the clock of St. Mark Square's Tower.

In his volume "L'arte dell'orologeria in Europa [The Art of Horology in Europe]", Bramante ed., 1978, for example, he mentions the "torre delle hore [tower of the hours]" in St. Mark Square, which is first of all just a clock, with a spectacular dial and automatic bell-ringers." (Page 11). Besides the amenity of the description, torre delle hore is a denomination that does not appear in any document known to us; nor was it ever a frequently used or classic way to refer to it historically.
Later, in the same volume, Mr. Brusa states: "unfortunately the dial and the mechanics [of the clock] were substantially modified after the mid XVIII century by Bartolomeo Ferracina, an excellent clockmaker but a terrible restorer. It's not easy to establish how much of the original is left except for the wall structure and the famous Moors" (page 40). It is very easy, indeed. It is enough to know that Ferracina received the whole old mechanism, the monetary value of which was paid to him based on the weight of the recycled metal and added to the monetary retribution for the work done. It is also absurd to express a judgment on Ferracina as a "restorer." Bartolomeo Ferracina wasn't a "terrible restorer" simply because he was never a restorer: in 1757 he was paid by the Procuratia de Supra to completely rebuild the clock, with the exclusion of the Magi's carousel, on which he intervened later.

Two of the three Magis

Two of the three Magis

Mr. Brusa continues saying: "Originally, on a ledge facing the square, a procession of Angels and Magi, at the striking of the hour, passed in front of the golden-bronze Virgin Mary with Baby" (page 40). In reality, the procession of the Magi continued to function, on the solemnities of the Ascension, in May, and the Epiphany, being purposely arranged by the people responsible for the clock maintenance and constituting a noteworthy attraction for many, even during the years in which Mr. Brusa was writing his book. Disheartening, also, is the description of "a procession of Angels and Magi;" there have always been the self-moving statues of three Magi, preceded by only one angel with trumpet.

Again: Mr. Brusa concludes his discourse, that for all other aspects remains always superficial and never ventures deeply into the analysis of the mechanism's specificity, writing: "Maybe this is the oldest example left of a turret clock with dials visible from different points, as later became frequent on bell-towers and towers." (Page 41). Thus he himself, a historian of horology, forgets the older and world famous clock of Rouen's tower, erected in 1389.
We should also add that, in this book, Mr. Brusa acknowledges the rarity of pendulums with a 2 second period, that is those pendulums, one of which he took the liberty to dismiss: "two second pendulums, long about mm 3975, or even slower and longer are rare and can be found in astronomical and turret models" (page 460).
But that is not all. The pamphlet distributed to the press the day the restored clock was presented (Feb. 1, 1999, at Palazzo Ducale), contains other and more significant inaccuracies. They are more significant because, in the meanwhile, Mr. Brusa could have concentrated on studying this particular clock; circumstance that, evidently, was not sufficient to eliminate the superficiality already demonstrated in the past. The pamphlet is called "Restauro dell'Orologio della Torre [Restoration of the Tower's Clock]" and includes a short 4-page chapter (pp. 28-31) with the title "1499-1999. Il restauro della meccanica e il ripristino delle funzioni dell'Orologio di Piazza San Marco [1499-1999. Restoration of the Mechanics and Reactivation of the Functions of St. Mark Square's Clock]" written by Giuseppe Brusa.

When speaking of the hated pair of tàmbure with numerical panels representing hours and minutes introduced by De Lucia (1858), Mr. Brusa argues that "the apparatus, moved by its own mechanism, obviously needed to be connected and synchronized with the main train of the clock, with the consequence of interfering to some extent with the functioning in general and with precision in particular" (p. 29). Now, it would be nice to understand the sense of such distinction, between interference in general and that on precision. As for the second, we will see what Mr. Brusa is referring to, exaggerating the extent of such 'interference' to justify an intervention on this aspect too. Mr. Brusa continues: "De Lucia clumsily superimposed the rudimentary movement of the new apparatus to Ferracina's structure, deciding thereafter to eliminate the transmission of the time movement to the dial facing the Mercerie to lighten the timekeeping train and, in addition, to sacrifice the striking train of the 'meridiana' to reduce the overall obstruction" (p. 29).
But, a) it is not correct - according to serious research methods - to formulate gratuitous judgments such as the one "clumsily" expressed above;
b) De Lucia didn't in any way eliminate the transmission of the time movement to the dial facing the Mercerie: this is a surprising mistake, because the Northern dial continued to visibly work until the clock was taken apart in 1997; c) De Lucia didn't sacrifice the 132-strike chime that Mr. Brusa calls 'meridiana' by any means. This chime was deactivated in 1915, at the beginning of World War I, because of curfew regulations and complaints about the disturbance created by the midnight bell tolls. Mr. Brusa's argument that this action was taken "to reduce the overall obstruction" is also incredible. It simply doesn't make any sense because the sector where the striking train for this chime was lodged remained mounted, and in position. Neither is it clear what such "obstruction" should amount to in a modular portion that has its precise placement within the central frame (it makes up an entire quarter of it), without interfering with anything else.

Click to enlarge

Middle frame drawing. The 132-blow strike system is placed on the left side and clearly doesn't interfere with anything.
Click on the picture to see a more detailed version.

Mr. Brusa goes on: "He [De Lucia] longed to demonstrate his expertise in traditional horology and thus ended up causing a radical distortion of Ferracina's mechanics". The intervention by De Lucia, in reality, wasn't such as to justify these contentions; here instead we can see a real projection, as such words fit perfectly what Mr. Brusa e Mr. Gorla have done.
The description finally reaches the focal point of the controversy: "Questionable the idea of substituting the original pendulum with another very voluminous one, 410-centimeter-long, with 2-second-long oscillations, in order to obtain greater reliability of oscillations and better isochronism. Due to lack of space such a pendulum could not be located behind the clock, as the previous, and was therefore inevitable to place it frontally. To avoid interferences with the pinion that transmitted the motion to the wheels of the clock-face, it was necessary to move the new suspension laterally and that required more than one experimental solution, as recently emerged. The gears ratio had to be modified consequently". (p. 29). On the contrary, it is clearly and indisputably demonstrated by archival documents and drawings and by the number of oscillations noted in a contemporary book on the Clock Tower, that the pendulum wasn't lengthened by much and that it was previously placed in the same position of the one built by De Lucia.
It is evident that, on this matter, Mr. Brusa starts from a faulty assumption, following which he forces himself to establish a chain of consequences that lead to a preposterous, patently false conclusion. The starting assumptions are: a) De Lucia substituted the old pendulum with a much bigger one, and b) the pendulum thus had to be located in correspondence to the anchor's arbor, directly guided by the connected crutch and, therefore, behind the mechanism (North side). Consequences: I) the pendulum, doubled to a 4 meter length, could no longer maintain its previous position and had to be relocated 'frontally' (South side), where it was in recent times; II) on this side, however, the central position in correspondence to the anchor's arbor was occupied by the arbor exiting from the main wheel of the timekeeping train (i.e., the arbor that transmits motion to the astronomical dial), therefore De Lucia had to move the pendulum laterally and was forced to connect it through a horizontal transmission arm; III) another necessity was the perforation of the floor, because of the pendulum's length.

Such fantastic theory is based on unnecessary complications and totally falls apart because archival documents demonstrate that the horizontal transmission arm existed before De Lucia's intervention (1858). Specifically, the 1856 drawing published in my volume on the clock clearly shows the pendulum with its horizontal arm.

Horizontal arm

Drawings from the Technical Report by Annibale Marini and Giovanni Doria (July, 22 1856) show the original position of the pendulum assembly and the existence of the horizontal bar before De Lucia restoration, carried out in 1858. The horizontal bar used to transmit the motion to the pendulum can be clearly seen in the lower left side of the sketch and the Ferracina escapement on the top right (Celestia Historical Archive of the Venice Municipality, 1855-1859, III / 5 / 6 - Tower Restoration Works).

On the other hand, it is much easier to think that the pendulum had always been near the sector where the timekeeping train is lodged (as it is more logical and natural in any clock) and that De Lucia left it in the same position. Consequence I is also preposterous, because there are no structural reasons that would have prevented the perforation of the floor in the back position (North side) should the pendulum have been originally placed there. Deceptive is consequence II as well, because nothing would have prevented De Lucia from divaricating the pendulum near the arbor, just as Mr. Gorla was forced to do when constructing the new one. This was, in fact, necessary once the new pendulum was placed in a position where it was inevitably interfering with the arbor for the transmission of motion to the Northern dial. Reality is that, North or South, the central location of the pendulum wasn't feasible, due to the presence of the transmission arbors for the two dials.

At this point, Mr. Brusa states that "as proof of the inadequacy of the operation, one has to notice the erroneousness with which the new apparatus was synchronized to the timekeeping train: incredibly, in fact, the new movement was connected to that crucial element constituted by the escapement wheel" (pp. 29-30).
This is, once more, a gross simplification aimed at overrating the mechanical 'interference' to discredit De Lucia's work and feeling authorized to overturn his restoration. In reality, before Brusa's restoration, the arbor of the escapement wheel ended beyond the bridge with a small wheel, which moved another small wheel having a period of 5 minutes. This small wheel was equipped with a pin, whose periodic contact with a small hanging lever caused the activation of the system of automatic numbers put by De Lucia on top of the central frame. Although this connection wasn't truly unquestionable in its concept, nonetheless it was never a problem for the clock precision or for the regularity of its motion.
Further in the pamphlet, Mr. Brusa rejoices for the: "reactivation of the chime he calls 'meridiana,' and that of the hour indication on the dial facing the Mercerie, unjustifiably dismissed and undoubtedly worth being reactivated" (p. 30). Now, while Mr. Brusa is right about the first mechanism (which, however, was not unjustifiably dismissed but was dismissed following a political decision of the city council), he still doesn't recognize that the time indication on the Northern dial (Mercerie) remained active until 1997, thus persisting in a rough error, unworthy of any bystander that from the street could in fact observe its functioning.
On the basis of this belief, Mr. Brusa writes that "an extension of the arbor has been predisposed so that - once it is permanently installed - it will transmit the motion to the dial facing the Mercerie", but such arbor already existed in all its length and transmitted motion to the above-mentioned dial. This contention is quite astonishing because it doesn't take into account the parts of the clock that were existing and functioning.
Further on, another detail surfaces: "Alberto Gorla has discretely introduced in the context of Ferracina's mechanics a release device that will facilitate the adjustment of the indications, (…)". Mr. Gorla added, not so discretely, two (ugly) metal discs with perforated edges to the wheels that transmit and differentiate the motion, releasing them from their respective arbors. What is even worse, Mr. Gorla made a perforation on a spoke of each of those wheels, so that they can be locked in the desired position through one of the holes of the added discs, while he could have adopted a vice-like device more respectful of the ancient and elegant wheels.
Mr. Brusa's pamphlet closes on a covert admission: "The mechanism that moves the carousel with the Magi procession has also been restored, but the project of disengaging the paneled wheels with digits to facilitate the periodicity of the procession could not be accepted as it seemed too significant an innovation in the circumstance". Here Mr. Brusa's disappointment clearly surfaces for not having been able to completely dispose of the 19th century configuration designed by De Lucia and, with it, the admission that, if it had been possible, he would have lightheartedly removed those apparatuses.


a) The restoration had to be a conservatory one. So it is generally for all cultural, historical, and artistic items. This restoration was not conservatory, and it seems even more incredible because lately Italian Superintendences had been almost obsessive in their commitment to preserve even insignificant artifacts and items.

b) This transformation could have been presented as a necessity (i.e., as indispensable), claiming that: either the features of the clock are changed, or it will not work. Such a contention is, nevertheless, indefensible, as everybody knows that it worked well until it was taken apart.

c) This transformation could have been presented, more reasonably, as a technical improvement, to achieve more precision, but this concern should not take precedence over that of preserving an ancient clock that represents an important phase in the history of mechanics. In the light of contemporary conservatory ethics and practice, the perfectibility of an ancient artifact is not sustainable.

d) Last would be the choice of conducting a 'philological operation', which would consist of restoring a previous state; in this case, that previous to De Lucia's 1858 restoration. This was indeed the plan: erasing the traces of the 19th century restoration and substituting it with another one, arbitrary and brand new. Calling it 'philological' amounts to the presumption of having established that the previous pendulum (B. Ferracina, 1757-59) was located on the opposite side (North side) of that realized by De Lucia (1858, South side), and was, at least, half as short! This theory, however, is disproved by the fact that:

- d1) N. Erizzo, only 2 years later (1860), writes of a change to 1800 oscillations per hour versus the 1828 of the previous pendulum. Thus confirming that the difference in the pendulum's length was minimal, estimated as few centimeters, and that the previous pendulum was, therefore, about 4 meters [see Peratoner, page 37].
- d2) The pendulum, in Mr. Brusa's hypothesis (and indeed, as it has been reconstructed) should have been aligned with the anchor's arbor, thus without the need of being moved by a horizontal arm such as the one that 'pushed' it sideways in the last configuration. There is, instead, a technical report by G. Doria and A. Marini, written two years before De Lucia's restoration, that includes a drawing of the pendulum with its transversal arm (Archive of "Celestia"). It is therefore natural to conclude, with the greatest clarity, that the pendulum had always been there. [see Peratoner, page 38, and table IV on page 39]
- d3) Again N. Erizzo, writing about the 1858 work, says that De Lucia added a device to the transversal arm; but if one adds something to a specific mechanical part, it means that such a part already exists. [see Peratoner, page 38]

- d4) There is the brace of the old pendulum's suspension; an item that was no longer operational (since 1952) but is clearly recognizable in its function and, more importantly, matches exactly the position where the pendulum had always been. The way the item was manufactured is comparable to that of the other static parts of the clock and, therefore, the part belongs to the Ferracina's period. [see Peratoner, lower part of page 37]

Click to enlarge

Ferracina's suspension cock.
Click on the picture to see a more detailed version.

- d5) S. Cadel, the person in charge of the building restoration, upon its conclusion writes a very detailed report on the work carried out. No mention is made of a floor perforation (the 'hole' in the tower's living room ceiling), which would have been necessary if the pendulum had, at that time, doubled its size. The possibility of such omission being an oversight is unlikely, due to the very minute detail of that report. [see Peratoner, page 41]

Click to enlarge

The pendulum bob inside a glazed box at the first floor of the Tower, in the position where it was before the half 20th Century intervention.
(Picture from the Peratoner family private archive, dated before 1950 and never published before).
Click on the picture to see a more detailed version.

- d6) This configuration can be seen in a newly released picture that shows the pendulum oscillating in the room below the clock [reproduced in Peratoner on page 40, table V, see also pp. 41-42].

In conclusion, to have insult added to injury, it is now clear that aiming at recreating Ferracina's clock configuration, the restoration carried out by Mr. Gorla under the direction of Mr. Brusa, by moving the pendulum on the opposite side and shortening it by one half, together with the 19th century configuration that it was determined to suppress, has also erased the 18th century stratification that it wanted to recreate.

Dr. Alberto Peratoner


[Next page]