The Collection of Peter Lin, Part II

Antiquorum Online
©2003

An Extraordinary Tiffany & Co.Clock:

Conant’s ingenious

“Isochronal Regulator”

Philip Poniz
Director
Horological Expert

The history of horology is marked by a number of inventions that only the most able of minds could have conceived. Perhaps the best known example is Breguet’s tourbillon, a device which enabled the positional error of a watch to be averaged out.

Few people are aware that an analogous device was invented for clocks. Of course, the device was not invented with the purpose of averaging positional error, since this does not apply to a stationary clock Clocks, however have another inherent problem – the isochronism of the pendulum. In theory, the length of duration of a mathematical pendulum’s swing, (called the period) is always the same regardless of the amplitude

of the swing. However, in practice this is not the case. Christiaan Huygens himself, the 17th century scientist and theoretician who discovered the laws governing the pendulum, was aware of this problem and in an effort to eliminate it, he implemented cycloidal cheeks at the pendulum’s suspension. However, such cheeks, though mathematically and theoretically correct, were difficult to manufacture, and in practice often created problems greater than those they were supposed to solve.

The transfer of energy between the weights and the pendulum, the differences in barometric pressure, and irregularities in the train are all factors which cause variations in a pendulum’s amplitude. To these inherent problems are added the vibrations… caused by the motion of footsteps and traffic of a busy household.

Even if one were to make a clock with a very smoothly running train, and if this clock were placed deep in the cellar of a house where there are practically no vibrations, the phenomenon of transfer of vibrations between the weights and the pendulum would still exist.

Multiple pendulum clocks based onsympathetic “resonance”

Around 1700, clockmakers first observed the peculiar behavior of the pendulum in long-case clocks. They noted that when the weight dropped to the level of the bob, it started to swing in the opposite direction to that of the pendulum, and the closer it came to the bob, the more its amplitude increased, while that of the pendulum decreased and eventually stopped.

These observations led the clockmakers, active during the first half of the 18th century, to attempt to construct clocks to solve this problem.

may be considered Pawtucket’s leading manufacturer. By his energy, foresight and ability the great thread mills, so long known by his name, were originated and developed, and until today they form the largest industrial establishment not only in Pawtucket but also in Rhode Island.

He received his education at Nichols Academy in his native town, attending it in winter, usually, and working on his father's farm in summer. Afterwards he worked in a printing and

machine shop where he spent his evenings in learning mechanical drawing and studying mechanical engineering. As a result he became a very expert mechanic, and developed great ability as a mechanical engineer and inventor. He then worked in Hartford, where he soon was engaged at Colt's firearm manufactory. While in Hartford he made drawings for Christian Sharp, the inventor of the Sharp rifle, and assisted him in constructing machines for making projectiles. In 1856 he invented and patented an improvement on the Sharp rifle, known as the 'gas check', which was considered so important that the United States and British governments immediately ordered its application to all arms manufactured for them by the Sharp Rifle Co. The same year he constructed a machine for Samuel Slater & Sons of Webster, Mass., for sewing the selvedge on doeskins.

About this time Conant first turned his attention to the improvement of machinery used in thread manufacture. He was very successful. In 1868 Conant moved to Pawtucket, where he organized the Conant Thread Co., and became the treasurer of the corporation and the manager of the works. For the next thirty years he developed the business, which in 1897 employed over 2000 people and had invested capital of over $4,000,000.

He was president of the Pawtucket Institution for Savings; President and Director of the Pacific National Bank; Vice-President of the Pawtucket Safe Deposit Co.; and a Director in the First National and the Slater National banks of Pawtucket.

(From the Illustrated History of Pawtucket by Robert Grieve, 1897, pp 274-276.)
Fig. 2 Photograph of the clock.

Fig. 3. The movement of the clock as
drawn for the patent specification.

Fig. 1. The front elevation of the clock.

The result was the double pendulum clock. Double pendulum clocks were highly complex pieces; they are consequently extremely rare and very
expensive.

That being said, even the most eminent of clockmakers could not require his clients to install their clocks 26 meters below ground. Therefore, some clockmakers employed two pendulums on the assumption that if a single pendulum clock generated vibrations affecting its performance, perhaps the adjunction of another symmetrically acting system, would create the same vibrations but in an opposite phase, and the two sets of vibrations would therefore cancel each other out. This theory called for the construction of a clock with two separate pendulums, which worked in a crossed swing motion. Few such examples were made, notably two by Breguet and three by Janvier at the turn of the 18th century. Antiquorum auctioned one such example for over 1,500,000 Swiss Francs, and the other for slightly less. These were lots 44 and 45, sold on November 11, 2001 in the sale “A Tribute to Precision and Complicated Timepieces”.

A single-pendulum long-cased clock with a heavy weight will tend to stop when not properly fixed in position and a double-pendulum system, with pendulums well synchronized and swinging in opposition, will exert almost zero reaction on the combined support.

Therefore most of the transitory disturbance caused to one pendulum will have less effect in a double-pendulum system.

Abraham-Louis Breguet observed this phenomenon and described it in an unpublished manuscript, noting that his well running double-pendulum clock would stop when the suspension bracket was changed for a very rigid one. Apparently, the transfer of energy was not sufficient to create enough amplitude. Breguet also noted that the less rigid the support, the easier it became for the pendulums to run, despite the variations in their frequency, in this case 21 seconds a day.

Conant’s Isochronal Regulator

Hezekiah Conant (1827- 1902) took this idea to the next level. He used the idea of sympathetic resonance, and also invented an ingenious differential mechanism. patented on August 23, 1887, both in England (pat. 11465) and in the USA (pat. 36881 – Figs. 1 and 3).

Conant’s invention averaged out the error in a clock by the use of multiple pendulums. His invention went one step further as it also decreased such
error in a ratio which was linear to its number of

Fig. 4. Tiffany’s manufacturing register detailing the work performed on Conant’s clock.

Antiquorum sold such a clock on November 11, 2001, lot 44, for over 1.5 million Swiss Francs. The first recorded experiments with double pendulum clocks were undertaken by John Ellicott (1706-1791) in the 1750s. Ellicott observed that the pendulum of a clock "communicated" motion to the case and to whatever the case was touching, which in turn negatively affected the rate. He reported his results to the Royal Society in a talk called "Influence which two pendulum clocks were observed to have on each other".

Ellicott's experiment made it clear that in a single- pendulum clock, the pendulum causes the case and all the elements, including the weights, to vibrate, thus affecting its accuracy. For precision regulators, this presents a great problem. In order to avoid any such resonance interference, the constant-pressure regulators which were still in use less than 40 years ago at the Paris Observatory were set in the Catacombs, 26 meters below ground level, and in separate rooms.

Fig. 6. One of the differential mechanisms joining two escapements.

pendulums. In summary, a clock with two pendulums had a rate two times better, and one with four pendulums had a rate four times better. Conant’s clock not only employed an extremely complicated sympathetic resonance multiple pendulum mechanism, but improved it considerably.

Such an invention attracted considerable attention. For instance, Theodor Gribbi, the Swiss representative at the 1876 Centennial Philadelphia Exhibition and the 1893 Universal Exhibition in Chicago examined the invention and
the clock. He also interviewed Conant; a long and informative article based on this interview was published in the January 1895 issue of the Journal Suisse d’Horlogerie (pp. 214 – 218).

However, Conant’s clocks were very expensive to build. He selected Tiffany Clock Makers to build his clock and was charged a small fortune of $1,369.00 for the task. The clock was delivered on February 28, 1887 (Fig 4).

Tiffany Clock Makers

Tiffany Clock Makers was an exclusive company created by Tiffany with the purpose of producing the highest quality regulators. To accomplish this, the company chose the best clockmakers, many of whom were European, and set up a clock shop around 1879-1880 that remained active only until 1891. Undoubtedly one of the best companies in the world at the time, they built precision regulators for observatories, universities, and wealthy patrons such as Vanderbilt or Morgan and Louis Tiffany himself.

As builders of astronomical clocks, striking regulators wound with a single mainspring and other examples, it is no wonder that Conant selected them. Browsing through Tiffany’s registers leaves little doubt that Conant was their most challenging client, who ordered the most complicated and expensive clocks in the company’s entire history. As mentioned earlier, Conant’s clock cost a whopping $1,369 to produce. By comparison, a complicated clock ordered by Vanderbilt cost less than $300.

All questions of cost put aside, it was difficult to clean and maintain such a clock, let alone to understand the principles on which it worked. Finding a clockmaker capable of looking after the clock, cleaning it periodically and reassembling it correctly presented additional challenges.

Conant’s clock combines the sympathetic resonance principle with an ingenious combination of three differential mechanisms (Figs. 5 & 6).

The clock has four dead-beat Graham escapements, arranged in pairs of two. Each pair is joined through a differential mechanism (see Fig 3), which reduces the error by half. Both of these mechanisms are joined by a third differential mechanism situated in the center which cuts error in half yet again. This third mechanism carries the clock’s central seconds hand.
Since the error produced by each pair of the escapements is largely canceled out by the sympathetic resonance of their pendulums, the end error, further diminished by the differential mechanisms, is extremely small.

Conant had his clock equipped with a very precise and complicated mechanism which sent a series of impulses every 12 hours, presumably at noon and at midnight.

The impulses start in two-second intervals and end in one-second intervals, signaling that the final impulse is coming in a few seconds. Such devices are usually employed in observatory clocks.

This clock has an interesting history. In 1946 "Smiling Jack" Willey of California discovered the clock near Providence, Rhode Island (see NAWCC Bulletin Vol. 26 p. 539). During the 1960s, Willey sold the clock to Roland S. Stevens, the eminent collector of Howard clocks, who bought it on the assumption it had been made by Howard. (Howard did make a four-pendulum clock for Conant, but not until the 1890s.) Stevens eventually discovered that his best clock was made by Tiffany, not by Howard, and, perhaps out of frustration, punched its dial with Howard’s mark. (The tool was readily available to him, as his uncle was the superintendent of the Howard Clock Company.) When Seth Atwood, owner of the Time Museum, learned of Conant’s clock, he desperately wanted to buy it from Stevens, who was unfortunately unwilling to part with it. Atwood then purchased two prototypes of self-winding regulators with cast-iron bases by Howard from Yale University, which he offered to Stevens as a bargaining tool. Stevens’ passion for collecting Howard prevailed, and he parted with his four-pendulum clock, which Seth Atwood then proudly displayed in the Time Museum.

Over time, the true history of Conant’s clock was forgotten, and when it was sold during the third auction of the Time Museum Collection, it was described as a Howard clock. Stevens, who had passed away years before, would have smiled at this. Conant's isochronal clocks are extremely rare. The present example may be one of only two to have survived. The second— with two pendulums— may be viewed in the Ladd Observatory at Brown University in Rhode Island. We do know that Conant had one more clock made by Tiffany and another by Howard. However, the whereabouts of these clocks is unknown. Considering that the clocks’ mechanism is highly unconventional, and that it would have been very difficult to find a clockmaker able to service them, it is very likely that neither example survived, making the present clock possibly the most important existing precision clock built in America, as well as one of the most important precision regulators in the world.

BIBLIOGRAPHY:
C. Huyghens, "Horologium Oscillatorium", Paris, 1673, pp. 18-19.
Gallon, "Machines et Inventions Approuvées par L'Académie Royale des Sciences", Paris, 1778, Tome I, pp. 147-148, and Tome III, plate 332.
Jean-Dominique Augarde, "Nobles Seigneurs and Scientific Instruments in 18th century France: Louis-Léon Pajot d'Ons-en-Bray (1678-1754)", communication at the symposium "Origins and Evolution of Collecting Scientific Instruments", Boerhaave Museum, Leyde, 1994.
Lecture given by Anthony G. Randall at the Museum in La Chaux-de-Fonds.
Norman L. Fritz, "Pendulums driven of resonant frequency", Horological Science Newsletter, No. 2, 1997.
Ken Friedenthal, "Energy Analysis of Pendulum Perturbations", Horological Science Newsletter, No. 1, 1996.
"Two Pendulum Clock", by Ned Bigelow, Horological Science Newsletter, No. 1, 1996.
Antiquarian Horology, "Irish Section", Vol. XVI, March, 1987, pp. 488-89."Antide Janvier – Mécanicien-astronome, Horloger ordinaire du Roi", Jean-Dominique Augarde and Jean Nérée Ronfort, Paris, 1998.
"Double the Excitement with a Double-Pendulum Clock", by Stephan Gagneux, Horological Journal, November 2000.

Fig. 5. The differential mechanisms
diminishing the error four times.

Fig. 7. Detail of the case.