Concerns regarding the Conservation of Functional Horological Objects

Jim Moss
Professional Horological Conservator
Private Practice
Littleton, MA, USA
Professional Associate Member, AIC

© 2003 James Moss

This document is a response to a query in the Conservation DistList, Conservation DistList, Instance: 16:46, Friday, January 31, 2003, subject "Tall case clock". This is a start in the discussion about conservation of horological objects but is not definitive.

If the clock is to function, the following items need to be considered: Using a 10,000 year view point as a measuring stick, consider the following:


If the clock is one example of a design of some very low production number, even if the maker is not well known, then it needs to be conserved with care. The low production volume imposes a severe limit on the amount of historical information that we have available on this particular maker and that limit dictates, in the case of a requirement of functionality, a replication of the artifact.

A replication will relieve the problems associated with the replacement and replication of worn components as well as the disappearance of historical information. It will not, however, relieve the costs associated with functionality: regular maintenance, damage by handling, and wear. It will also be expensive to make the replication.

In the case of non-functionality: the development of a careful Conservation plan for the protection of this object is absolutely required. Permit me to digress for a moment and express this thought: so often we have a tendency not to protect the un-popular makers of the day as well as we protect the popular but in the grand scheme of things, they are all equally important.

If the clock is "one" of many (in other words, a production clock... easily found with many other "clones" in existence), then as long as one or two are held in "trust" somewhere on this earth, the operation of this object along with the ramifications of function is certainly acceptable. The difficulty here is determining "if" any "perfect examples" are held in "trust" and if that "trust" is irrevocable! Replication and replacement of worn components of a "production" object deemed acceptable for use as a functional object is a reasonable and acceptable position to hold in this case. But, remember that the supply of production clocks or watches is dwindling with no replacements in sight!

Loss of Historical Information

Historical information can be lost in many ways such as: handling, cleaning methods, exposure to UV, wear, and treatment techniques. But, what is the historical information that we are losing?

Some historical information can help us to identify and understand the materials that were used as well as to contribute to the larger picture of where these materials fit into a countries socio-economic development. Likewise, determination of how these materials were manufactured and used can also point to information about a countries economic and technological growth. Thus it is important to determine if the metal was hammered or rolled or cast or filed or machined or produced using cost reduced methods (such as is being done today). Replacement of components within an object may not totally compromise the historical worth of the object in the larger sense but will render interpretation more difficult in the future. If the object is a "one" of a kind, then the impact from repairs is significant: if the object is "one" of many, then the impact is lessened. Cleaning can affect ability to track an object from its generation to its present location: various "dirt" or "salt" deposits can provide precious clues to an objects past but to some observers can be visually unbearable. Cleaning can erase clues about the environment in which the clock or watch has existed. And, without documentation prior to any treatment, we've lost it all!


In all cases of functionality (this can be applied to all manner of functional objects such as airplanes or trains or machine tools, etc), wear is going to take place: wear of the bearings (or bushings if the bearings have already been replaced), wear of the pivots that rotate in the bearings and wear of the sliding components such as the escapement pallets along with their respective escape wheel teeth, wear of the conjugate surfaces of the striking and chiming levers, and wear of surfaces of the gear teeth that are in contact with one another. In fact, all wear is caused by sliding one surface over another be it sliding or rotating.

Consider, for a moment, the motion of a shaft in the "Time" portion of the clock or watch: each time the clock "ticks" the shaft will perform the following actions: "stop", "roll ahead", "stop", "backup" ( if the escapement is a "recoil" escapement), "stop", and then repeat the sequence over and over. When the shaft "rolls ahead" it tends to climb up on a wave of lubrication: when it comes to a "stop", it sinks down through that wave to rest very close to the bearing wall. The "dwell" time or "stop" time is dependent upon the escapement design but it can be sufficient to allow the shaft to settle down so close to the bearing wall that it is possible to catch a high spot of material and the next time it "rolls ahead", it will displace that high spot of material from the bearing wall: this is a form of wear.

Consider, for a moment, the motion of a shaft in the "quarter chiming" gear train: those shafts rest in one position for almost 10 minutes... plenty of time for the pivot to approach the bearing wall: far more time that then "time" train has. This resting period can cause more wear in this train than in the "time" train if allowed to cycle an equivalent amount as the "Time" train.

Consider, for another moment, the motion of the shaft in the "hour striking" train... it rests in one position for almost 50 minutes: far more "resting" time than either of the other gear trains and far more potential to produce wear as a result of penetrating the barrier layer of lubrication prior to motion.

Airborne Abrasives

Now, consider, for a moment, the introduction of airborne abrasive particulate and what would happen if a particle interjected itself between the pivot/bearing interface. Because the bearing material is usually softer than the pivot material and the abrasive particle larger than the irregularities of either surface, the abrasive particle will become imbedded into the bearing wall and act as a piece of "sandpaper" or as a "lap" and cause wear to occur on a regular basis. If the lubrication is renewed without the pivots being polished or the bearings being replaced, the wear will occur faster because the abrasive particle's cutting edges will be kept clean by the lubrication. (This is sometimes evidenced when a clock has been run for a very long time without maintenance and then is just "cleaned": it usually will not run reliably after the "cleaning". This is because the bed of gunk that has covered the abrasive particle was removed during the cleaning process. Now, not only is the particle removing materials rapidly from the pivot but the pivot and bearing are worn and don't fit together as well as they did before the wear occurred.)

Wear can take place from a variety of reasons in addition to the above noted ones. Wear can occur as the result of gears engaging with one another and applying a pressure to the rotating pivots causing the bearings to wear in one direction. Wear can occur as a result of lubrication that has failed. Wear can occur as a result of a corrosive atmosphere. And so it goes.


Part of the key to reduced wear is the type of lubricant that is used: not all clock or watch lubricants are created equal! What is needed is a long chain polar lubricant: one that can lay molecular chains of lubrication between the interface of the pivot and the bearing wall and have one end of that chain "attached" to either the bearing wall or the pivot. The same holds true for any other sliding components. But, this lubrication is not a total solution and it never will be. Lubrication will not prevent wear: it will only slow it down. At the atomic or crystalline level, it is impossible to completely separate the sliding surfaces. Clock and watch lubrication is not pressurized so the components are not always floating on a surface of lubrication and in the long run, pressurized lubrication will not prevent wear either.

A common mistake is oiling the clock or watch without proper cleaning or the addition of new lubrication to the old. Not only is there a very good chance of an unwanted chemical reaction between the two lubricants but the lubricating qualities of the new lubrication will be negated or degraded by the presence of the old lubrication's compromised components. Another result of adding lubrication is that the clock is allowed to function far beyond a reasonable maintenance period and this will cause severe wear to occur.

The older lubrications were made from sperm whales or porpoises and contain various acids as well as organic components: one will lead to corrosion of the metals and the other to decomposition of the lubricant.

Environment (Including Visitors)

As you can imagine, the welfare of the object is subject to its environment also! Having a stable temperature and relative humidity can play a large role in preventing surface corrosion and the development of micro-climates within the mechanism. The volume of visitors and the frequency of the fluctuation of temperature and RH as well as the number of particulate generating objects in the vicinity such as rugs and open windows will drastically affect the MTBT (mean time between treatments) and a short MTBT can be an expensive proposition.

A MAJOR PROBLEM: Who is qualified to conserve a clock mechanism?

Not all clock repair people are created equal: not all have the same training nor background nor the interest in preservation. There are very few trained horological conservators in the world but there are hundreds of local repair people. Most of these people either learned this trade by using text books with techniques that can be very damaging to an object or they apprenticed under someone who used those texts. Very few of the local clock repair people have a detailed understanding of the mechanics of horological objects let alone the understanding to use techniques that would be harmless to the object. There are some schools that produce very well trained and informed students in the mechanical aspects of horological mechanisms such as the British Horological Institute (BHI) but they do NOT teach conservation techniques or philosophy and they still use the same text books with the same techniques that can be damaging to an historical object. Other schools have attempted to teach the conservation of horological objects but have not been very successful at this point.

The best candidate of all as a horological conservator would be a person "program trained" in conservation and who has attended the BHI. or the WOSTEP program (Watchmakers of Switzerland Training and Educational Program). There are no serious 2 or 4 year horological schools in the United States as of this date ( there are some evening classes and a couple of schools for hobbyists).

Conservation is an attitude: they may have the best training in the world but not have a conservation view point or attitude or values and those are the most important ingredients.

So, how do you judge if the local repair person is capable of "conserving" your clock? Of necessity, you need to know something about the appropriate conservation techniques used to conserve a clock, you need to know the person's source of education, you need to know something about the mechanism, and the mechanical techniques used for the "repair" of them: you cannot rely on the recommendations of others in your field. Regardless, closely monitoring of the treatments while they are happening is absolutely necessary when hiring outside contractors: their motivation is money and yours is preservation and usually the twain never meet.

In addition, it has been my experience that detailed past treatment documentation is at best scanty or non-existent. You must demand and receive documentation on all of the chemicals used for cleaning, the names of the coatings used if any, the manufacturer of the lubrication as well as its description and the chemical analysis, a "before" and "after" report of the bearings replaced, the pivots polished, and the components replicated as well as any adjustments that were made. There must be no secrets!

To help you understand a small amount about the conservation of functional clocks, I am providing you with some conservation view points as well as some information about some repair techniques. The following information is not a full fledged treatise on the subject but it will be enough to help you. Also, be sure to use the references that I have provided to you at the end. Just a note: the conservation of non-functional clocks is different; it is less invasive.

Bearing Techniques

Bearings are holes in the front and rear plates of a clock movement into which the ends of the shafts (called pivots) are placed. The purpose of a bearing is to locate accurately the position of the shaft. Accurate location of a pair of shafts insures that the proper placement of their respective gears and insures efficient transmission of the available power. As the bearings wear, the distance between the gears increase and causes the transmission of power to become more inefficient and at the same time increases the wear of the gear teeth, pivots, and bearing walls.

When bearings need to be replaced, no more material should be removed than is necessary to retain the new bushing safely and to provide proper centering. It is usually necessary to file out the worn bearing hole until the hole is once again centered about the original center location. If at all possible, removal of material should not exceed the diameter of the original oil sink. The use of pre-manufactured bushings (readily available from suppliers) that are not of the proper length must be avoided because the length of the journals will be too short to support the loads imposed upon them. On the other hand, the use of pre-manufactured bushings that are too long is also inappropriate because , in time and through wear, they will effectively eliminate the end-shake necessary to allow the shaft to rotate freely. Stacking pre-manufactured bushings to make up for the lack of journal length must be avoided as the join between the two bushings will provide a place for abrasives to gather. It is best to make bushings to fit that to try to make do with what is on hand. When making bushings, one must make sure that the alloy is known and that information is contained in the documentation for the clock or watch.

When a bushing has been installed, the oil sink should match the shape and depth of the original but should not be blended to the point of being unable to be identified as a replacement. If at all possible, the bearing material should be of a slightly different mixture of the same alloy type so that in the future ready identification can be done with an SEM or visually.

Oil Sinks

Oil sinks are concave hemispherical cavities that surround the end of the bearing and are located on the outside of either the front or rear plates. Their purpose is to hold and prevent the spread of the clock lubrication that is applied to the bearing. Most people apply far too much lubrication to a clock or watch pivot and this usually causes the lubrication to run out of the bearing and by capillary action will cause the bearing to go dry. The proper amount of lubrication is when a micro-meniscus forms between the pivot shaft and the base of the oil sink.

Cleaning Techniques

Unlike a painting or a sculpture that can have its surface "easily" cleaned, a clock movement has to be completely disassembled, then cleaned of surface accretions, and then re-assembled (usually it is not as simple as this although we all wish it were so!). Clock movements can consist of up to 500 individual components and each component has a series of surfaces that need to be cleaned. Effective and safe cleaning cannot be done by the "dip and swish" method! And, cleaning should not be done using an ultrasonic cleaner: it must be done by hand.

The preferred cleaning solution used by most clock repairers is an aqueous ammoniated cleaner either pre-manufactured or home-made. In any case, exposing stressed brass ( most clock components contain stresses) to ammoniated cleaning solutions will produce a known and verifiable effect: Stress Corrosion Cracking (SCC). SCC may not happen instantaneously but it will happen and it is not reversible. Some repairers use a pre-manufactured "non-ammoniated" cleaner but instead of ammonia, it contains one of the following relatives of ammonia: "mono" or "di" or "tri" ethanolamine. These particular chemicals are also capable (an have been documented as causing) of producing SCC. This means that you have to insure that your object is cleaned in a manner that is harmless to your object. (Ref: Ammoniated cleaning solutions by James Moss: American Horological Times, Volume 22, Number 2, February 1998; British Horological Journal, Volume 139, Number 8 .)

For cleaning a clock movement, I presently use hydrocarbons and alcohols in combination with manual labor, fine brass and steel brushes (only occasionally), stiff bristle non-metallic brushes (all of the time) and sharpened pieces of wood called peg wood: I do not use an ultrasonic. If steel brushes need to be used, they should only be used on steel components: likewise brass brushes should only be used on brass components because if a brass brush is used on a steel surface, corrosion cells can be set up and over the long term and in the right conditions, corrosion can take place. Restricted use of the metallic brushes is necessary as they will remove any protective oxide layers even with careful use.

Pivot Polishing

The purpose of the pivot is to work in conjunction with the bearing to locate accurately the position of its related shaft. Each shaft has two pivots: one on either end. As stated in a prior paragraph, accurate location of a pair of shafts insures that the proper placement of their respective gears and insures efficient transmission of the available power. The pivots must be polished to a mirror finish (they will appear to be black if properly polished) and usually, but not always, they are cylindrical and have parallel sides. A polish less than mirror like will produce wear as it is actually a miniature rotary file!


If a component is damaged or worn to the point that simple polishing will not allow it to work properly, then the entire component needs to be replicated. This means that if there is a gear that is too damaged to work properly, then the gear and its shaft and any other of the assembly's associated components need to be replicated and the original assembly needs to be stored safely. You should be able to substitute either assembly into the mechanism without any additional work, and the clock should work properly (assuming that the part needing replication was complete and whole). Replication of only part of an assembly many times can cause more damage to the overall assembly during the process than is necessary. Replications should be made of material similar in composition as the original but sufficiently different that analysis can easily determine the difference between the original and the replication. All replications must be signed and dated and labeled as a replication.

Maintenance Period

Because the preservation of a functional object is paramount, cleanliness is one of the major keys to longevity. This means that frequent cleaning of the clock movement is imperative. From a long term preservation standpoint, the MTBT of once every year would not be unusual and would be recommended: a MTBT once every two years would be approaching the "too infrequent" mark.


Under normal circumstances, clock and watch components in functional objects should not be coated as there is a possibility that the lubricants used could react adversely with the coatings. Brass has a naturally occurring oxide that will cover the brass surfaces with a perfect "non-holiday" coating and is the best that nature has to offer. Coatings applied by spray gun or human beings are fallible and need to be replaced approximately once every 20-25 years under the best of conditions. By utilizing the oxide coating the brass, the process of removal, preparation, and re-application of an organic coating is one more intervention that will not be needed and one less chance for damage to the object.

Techniques to Correct for Wear

Many times, in an effort to retain historical information, a person will move the position of a component to an unworn section of its conjugate component. After a few moves such as this, there will be a permanent loss of information about the method of manufacture of that component: in the interest of the preservation of historical information, it would be better that the entire assembly be replicated and the original assembly be saved with the one spot of wear than to "fix" that one spot by repositioning components.


Winding the clock, setting its hands and the various dials, as well as moving the clock can expose the object to serious damage. Always consult with a responsible horological conservator for proper instructions concerning these actions before performing them. There are many variations and there are no absolute rules regarding the above processes.

Non-Functional Clocks and Watches

Even if you choose not to run the clock, this decision does not abdicate the necessity for preventative conservation and stabilization. Surface accretions, coagulated lubrication and lubrication that contains an acidic component, as well as other contaminants need to be removed from the interior and exterior surfaces of the mechanism. Protection from corrosion needs to be provided. Coatings need to be carefully determined before application.


"Ammoniated cleaning solutions" by James Moss: American Horological Times, Volume 22, Number 2, February 1998; British Horological Journal, Volume 139, Number 8 .

Conservation of Clocks and Watches edited by Peter B. Wills, BHI, ISBN 0 9509621 4 7, 1995

[Search all CoOL documents]

Timestamp: Sunday, 23-Nov-2008 15:20:17 PST
Retrieved: Friday, 24-Mar-2017 06:13:50 GMT