Salisbury cathedral clock
The Salisbury cathedral clock is a large iron-framed clock without a dial located in the aisle of Salisbury Cathedral. Supposedly dating from about 1386, it is claimed to be the oldest working clock in the world, although a similar claim is made of the clock in the cathedral of Beauvais in France (said to date from 1305). and the clock tower of Chioggia in Italy.  The clock is one of the group of 14th to 16th century clocks to be found in the West of England. (See also Wells, Exeter, Ottery St Mary, and Wimborne Minster.) An attempt to date this clock to around 1386 was made by T.R. Robinson which has been supported by others. Most of the parts of the striking train are believed to be original. The great wheel of the going train is also believed to be original.
Other clocks from the 14th century (the first century in which the mechanical clock flourished throughout Europe), such as those at Rouen (Gros Horloge), Paris (Heinrich von Wick clock) or Dijon (the clock taken by Philippe le Hardy from Courtrai in 1382), have either been lost, destroyed, or substantially modified. The Wells Cathedral clock might have been made by the same craftsmen as the Salisbury clock, but is usually dated to around 1392, and is now relocated in the Science Museum in London, where it continues to operate.
There are some doubts that the clock displayed in Salisbury Cathedral is actually the clock mentioned in 1386, as the construction is quite advanced and more comparable to clocks made in the 16th and 17th century than those made in the 14th century. The question if this is the 1386 clock is quite important as the Wells Cathedral clock was previously dated in the 16th century, but then dated 1392 after the discovery of the Salisbury clock in 1928. Dating mistakes for old turret clocks are not uncommon. The Dover Castle clock was initially dated in the 14th century, only to be later revised to around 1600.
In 1993, Christopher McKay organised a symposium with the Antiquarian Horological Society to determine if the clock could be dated to 1386. The majority of participants voted for it to be the original, but roughly 1/3 or participants voted the clock to be of a much later date.
A clock in Salisbury Cathedral that struck the hours was mentioned in 1306. This was probably one of the precursors of the 1386 clock, one of the many early examples of mechanical water clocks that are mentioned from c. 1280 onwards.
The clock was found in the cathedral in 1928. It had a pendulum, which appeared to have been installed at a later date. The clock was restored in 1956, and a verge escapement and foliot were installed. There were no drawings or documents available, so it is unlikely that the original foliot and verge escapement looked exactly like the one now installed in the clock.
The striking train of the clock is believed to be original.
Like many of these more practical devices, its main purpose was to strike a bell at precise times. It probably did not have a dial. The wheels and gears are mounted in a four-post wrought iron frame. The framework was not held together with nuts and bolts (which had not been invented), but rather with metal wedged tenons.
The escapement was a verge escapement with a foliot, standard for clocks of this age. The power was supplied by two large stone weights. As the weights descend, ropes unwind from the wooden barrels. One barrel drives the going train which is regulated by the escapement, the other drives the striking train whose speed is regulated by the fly (air brake).
Before the weights reach the floor, they have to be wound back up again, a task that explains the presence of two large wheels shaped like steering wheels at either end of the clock.
The clock was a 'single strike' clock that struck only on the hour. It made one strike per hour of the day (e.g. 12 strikes at noon). The left half of the clock (as in the photograph above), is the striking train; the right half is the going train.
At the end of the 17th century, the Salisbury clock, like many others, was modified from verge and foliot to pendulum and anchor operation. This usually made clocks much more accurate, even though trials in the early 1990s by Michael Maltin showed that the clock was running to within two minutes a day if the rope on the barrel was kept in a single layer. As soon as there are two layers, there is more torque applied to the barrel by the weight and the clock will go faster. As a single layer of winding is enough to keep the clock going for 12 hours, it could have been kept exact to within 2 minutes per day if it had been wound twice per day.
In 1790, the old bell tower 'on the ditch of the close of the canons of the said church' mentioned in the deed of 1386 which had housed the clock was demolished, so the clock was moved to the Cathedral's central tower. In 1884, a new clock was installed and the old one was left to the side.
The clock was re-discovered in the tower in 1928 by T.R. Robinson, an horological enthusiast who went up the clock tower to see the new clock (installed in 1884). The presence of the old clock was known to many, but nobody attributed much importance to the old clock. It was only T.R. Robinson who believed that it was the clock mentioned in 1386. From photos taken in 1928, it looked to be fairly complete. Eventually its historic importance was realised. It was first put on display in the Cathedral's North transept. Then, in 1956, the clock was restored to its original condition and started working again. The pendulum and recoil escapement were replaced by a new verge and foliot escapement, thus restoring the clock to something like its original design.
Today, the escapement operates, but the striking mechanism has been disabled.
The 1956 Restoration
Messrs. John Smith & Sons of Derby received the clock in February 1956. It was taken apart for the transport. They reassembled the clock in their workshop and compared it to existing clocks in the Science Museum before deciding how to restore it.
The help of Rolls Royce was enlisted to have X-ray photographs of two of the wheel arbors taken. This confirmed that the two arbors of the going train had been lengthened when the clock had been converted to pendulum operation. Subsequent investigations revealed that the clock had actually been converted twice, as remains of an earlier pendulum escapement were discovered.
|Train||Part||Previous work||1956 Restoration|
|going train||great wheel||disc with two pins attached to great wheel as it only turned once every two hours due to conversion to pendulum||after removing the disc, the original hole for the pin was discovered and a new pin was fitted|
|going train||escape wheel||replaced with anchor escape wheel||new escape wheel forged, teeth marked out by hand, riveted onto existing second wheel|
|going train||winding barrel||new wooden winding barrel fitted|
|going train||foliot support bracket||new foliot support bracket forged and fitted|
|going train||foliot||new foliot forged and fitted|
|striking train||hoop wheel||flange of hoop wheel reversed||flange of hoop wheel replaced correctly|
|going train||turning direction, strike release||strike release moved to front of frame, rotation of going train reversed to clock-wise to fit clock dial (1613)||strike release moved to back of frame, rotation of going train reversed to anti clock-wise as it was originally|
The frame height is 1.24m, the width 1.29m, and the depth 1.06m.
Great wheel to verge escape wheel: 100 to 10, verge escape wheel 45 teeth.
The Great wheel turns once in 3600 seconds (1 hour), so the verge escape wheel turns once in 360 seconds. One full foliot swing thus takes 8 seconds, or 4 seconds per half swing.
Seen from the going train side, the great wheel with the winding barrel turns anti-clockwise, and the escapement wheel turns clockwise.
Great wheel to fly: 64 teeth driving an 8 pin birdcage; second (hoop) wheel 64 teeth driving an 8 leaf pinion on the fly. So each turn of the great wheel makes the fly turn 64 times, or 8 turns of the fly per strike of the clock (as the great wheel has 8 striking pins).
Great wheel to count wheel: 8 tooth pinion to internal 78 teeth on count wheel, with 8 striking pins on the great wheel. This directly corresponds with the 78 strikes the clock will make in 12 hours (1 + 2 + 3 + 4 + 5 + 6 + 7 + 8 + 9 + 10 + 11 + 12 = 78). The count wheel turns once every 12 hours, so the great wheel turns 9.75 times, operating the strike lever thus 9.75 x 8 = 78 times.
Seen from the striking train side, the great wheel with the winding barrel turns clockwise, the hoop wheel anti-clockwise, the count wheel clockwise, and the fly clockwise.
One particularity of this clock is that the count wheel only appears to have 11 notches, but the clock strikes 12 times each complete turn. This is because the first strike (when the clock strikes one) is executed whilst the lever is still within the first, wider, notch. The locking piece is lifted out of the hoop of the hoop wheel, which then turns once, which leads to the great wheel turning 1/8 and striking once. Then the locking piece falls back into the hoop wheel. The fly will continue to turn until it stops gently on his own as it has a ratchet mechanism, which protects the bird cage on the fly arbor.
- C F C. Beeson English Church Clocks London 1971
- R P Howgrave-Graham New Light on Ancient Turret Clocks, Antiquarian Horology, 1954
- Christopher McKay (Editor) The Great Salisbury Clock Trial, Antiquarian Horological Society Turret Clock Group, 1993
- Anthony J. Duley, The Medieval Clock of Salisbury Cathedral, Friends of the Salisbury Cathedral Publications, 1977