How early turret clocks worked
This document has three sections. Click on a heading to jump to the relevant section.
- The first, Principles, outlines the problems involved in building a machine to ‘count’ the time of day. Just what, exactly, do mechanical clocks do?
- The second, The first turret clocks, shows how large mechanical clocks regulated by a ‘verge and foliot’ worked. This was the most important form of mechanical timekeeping between their first appearance in the late-thirteenth century, and the new technologies available after the mid-seventeenth century.
- Finally, After the ‘horological revolution’ concerns a complex of new technologies that transformed the technical and practical availability of much more accurate timekeeping. These using a pendulum as the clock's regulator (spreading from the 1650s) and to one of several new types of escapement, the best-known among them being the anchor escapement developed during the 1660s and 1670s.
1. Principles
Mechanical clocks work by controlling the release of energy to create a regular and consistent series of movements; to keep a count of those movements; and to signal the counted movements in the familiar units of hours and minutes). The signalling can take various forms, of which visual displays and aural signals have historically been the most common.
Modern clocks have long been largely self-contained devices, capable of running for extended periods without close human supervision, but the late-medieval and early modern situation was very different. These early clocks relied on considerable amounts of ongoing human activity that has, in the long run, been progressively incorporated ‘into the machine’.
The two figures summarise these changes, and the narrowing of human activity in the ordinary running of a public clock.
 |
| Figure 1: Workings of a verge-and-foliot turret clock |
 |
| Figure 2: Operation of verge-and-foliot regulator |
Although we habitually think of an alarm as a subsidiary function of a clock or watch, and this seems a general apprehension in the modern West, in understanding the history of timekeeping it is helpful to see the alarm as the older device, and the first mechanical clocks as derived from alarms.
In a medieval clockwork alarm, a weight was raised, and then prevented from falling under gravity because it was restrained by a gearing system. This slowed weight’s descent, greatly extending the duration for which the machine was ‘active’. The weight’s descent turned the gearing which, at a pre-set point, tripped an audible signal, usually the striking of a bell. The mechanism now remained inert until the trip device had been set again and the weight re-wound.
A successful alarm device needed to run for a roughly predictable period, perhaps of several hours – as where they were used in monasteries to alert a monk to ring the bell signalling night-time services. A clock needed to do more than this, specifically to run consistently in relation to the 24-hour cycle of day and night, coordinating the counting of twenty-four hours with solar and other celestial movements.
In an account written in 1271 that has become central to the early history of clocks, the monk Robertus Anglicus reports that constructors of horologia (timekeepers) have been attempting to build a weight-powered device that would turn a wheel once in exactly twenty-four hours. That is, the turning of the wheel would be in phase with the apparent turning of the celestial equator. “But they cannot quite complete their task which if they could, it would be a really accurate horologium” (quoted by Thorndike 1941: 242). In other words, clocksmiths’ key goal was a way of releasing the potential energy of a suspended weight, in controlled and regular amounts, to drive a clock at a constant rate coinciding with sidereal time, and the apparent motion of the fixed stars.
Very soon after this account, mechanical timekeepers were clearly in use in monasteries and other religious institutions in various parts of Europe. Documentation is far too patchy for a complete picture of exactly where and when the first successful timekeepers were in operation, but there were clearly several in late-thirteenth century England. Although several problems remained to be overcome in early clocks, and despite their reliance on active human participation in winding, setting, and cleaning, to continue running, the broad technology was soon widely established.
Figure 1. New Winchelsea, Sussex
2). The first turret clocks
Turret clocks underwent major technological changes over time, but from the late-thirteenth century until the late-seventeenth century, the key mechanical mechanism was the verge and foliot escapement. Though other escapements existed, if only briefly, the verge and foliot was by far the dominant escapement mechanism.
|
| Figure 1: Workings of a verge-and-foliot turret clock |
|
| Figure 2: Operation of verge-and-foliot regulator |
The ultimate source of energy powering the clock was the potential energy of a weight suspended from a rope wound around a barrel. The descent of the weight turned the barrel and the axle on which it was mounted. A series of geared wheels transferred the rotation to an axle on which were mounted the crown wheel and (not shown here) further gearing to the going train of the clock. Rotation of the crown wheel axle was ultimately powered by the force of the clock-weight falling under gravity, but the rotation is limited by the elements shown.
Two facets of verge and foliot escapements restrained the escape of the suspended weight, and controlled its release in very small, and relatively slow increments. First, the gearing between the barrel on which the rope was wound, and the turning of the crown wheel ensured that the rope unwound very little with each movement of the crown wheel. Second, the turning of the crown wheel was slowed because each tooth of its rotation switched the engagement from one pallet to the other. Each engagement between pallet and crown wheel decelerated and then halted the rotation of the foliot, before accelerating it in the reverse direction until halted by the next engagement with the other pallet.
Because the pallets engaged the crown wheel at two points across the diameter of the crown wheel, and were offset from one another by 90 degrees, the crown wheel’s engagement with each pallet caused the verge, and with it the foliot bar, to reverse direction. This took time because of the weights at either end of the foliot, so rotation of the crown wheel was slow.
Despite the reversals of direction of the verge and foliot, the crown wheel and the drive to the going train always rotate in one direction. This combination of oscillatory and unidirectional movement was the great technical achievement of medieval clockmaking. If a clock ran too fast or too slow, it could be adjusted by adding or removing material from either the suspended weights, or the foliot, or by moving the weights towards, or away from, the ends of the foliot bar.
The reversing rotations of verge and foliot were important because they slowed down the crown wheel, and the clock-weight’s fall, and hence reduced the frequency with which the clock-weights would require re-winding. The situation of most church clocks in towers and steeples was as much for the increased height from which their weights could be suspended and lowered under control, as for increasing the audible range of their hour-striking (which involved the height of the bells rather than the clock). Most sixteenth-century English church clocks were wound once a day, though a few required more frequent attention.
3). After the “horological revolution”
Late in the 1650s, the Dutch mathematician Christian Huygens completed the project initiated by Galileo in the 1580s, to successfully incorporate a pendulum as the regulator for a clock, and to have it built. Galileo noted that pendulums of any particular length were close to isochronous: that is, the period of swing of an oscillating pendulum of a given length varied relatively little regardless of whether the swing was large (and fast) or small (and slow). Huygens calculations showed that this unless the arc was small, a pendulum’s behaviour was only isochronous if the path traced out by its bob followed the curve of a cycloid, when seen square-on from the front. The curve of the cycloid was slightly flatter than that of a semi-circle of similar radius, and Huygens’ clock design used metal ‘cheeks’ to force the pendulum bob to follow a cycloidal path. In 1658, a Dutchman working in London, Solomon Coster, completed the first pendulum clock. This was a relatively small household clock for domestic use, but the first turret clocks with pendulums appeared within a very few years.
The first pendulum clocks retained a verge escapement, though the crown wheel was now horizontal rather than vertical. However, the verge was very soon superseded by the anchor escapement, which rapidly became standard for new turret clocks. Among the advantages of the anchor escapement was that it required the pendulum to oscillate through a considerably smaller arc than a verge escapement, an arc within which the difference between the arcs of a circle and a cycloid were negligible, so the pendulum was isochronous without needing Huygens’ ‘cheeks’ to perform consistently.
|
|
Alternate oscillations of the pendulum caused the anchor to engage with the escape wheel so as to first impel it and then restrain it. The shaping of the gear teeth also served to transfer energy to maintain the impetus of the pendulum with successive contacts, cutting energy losses from the mechanism as a whole. Overall, the mechanism was comparatively delicate, though still sizable, but rather simpler in the range of contacts it involved.
Comparison of the two diagrams shows a clutch of related design changes.
First, the foliot was replaced by the pendulum, and dispensed with. Second, the pendulum oscillated in a vertical plane rather than the horizontal plane of the foliot bar. Where a verge escapement was retained, the orientation of both the verge and the crown wheel changed, too. The verge was now horizontal (thus maintaining its perpendicular position relative to the axis of the regulator), as was the crown wheel.
In most cases, however, pairings of a pendulum with a verge escapement were short-lived, because of a third change, with the verge being superseded by the anchor escapement, involving the pendulum with a crown wheel that remained vertical (and here referred to as the escape wheel).
Fourth, the role and distribution of mass among the parts of the mechanism changed significantly. In the verge and foliot, the pace of the escapement was largely dependent on the weights suspended from the foliot bar, and in their position. The heavier the weights relative to the force transmitted through the pallets, the slower the escapement. Sheer mass was less important for pendulum-regulated clocks, though they still required considerable height for the descent of the weights between re-windings.
Fifth, rather than weight, the length of an oscillating pendulum was its most critical feature - more precisely, the distance between the pivot from which the pendulum was suspended, and the centre of gravity of its bob, so long as the swing was through an arc short enough to be isochronous.
Finally, though, anchor escapements resembled their verge predecessors, in that the oscillating pendulum directly restrained the crown wheel, whose turning was again controlled by two points of contact from the escapement, and the inherent braking effect of the regulator’s oscillation.
The performance of pendulum clocks, especially combined with an anchor escapement, was substantially improved compared with their verge-and-foliot predecessors. They were more expensive, not least because initially there were few craftsmen able to build and maintain them, whereas there were considerable and quite widely distributed bodies of knowledge about the older technology. From a long-run perspective, though, they spread relatively rapidly, and the practical knowledges involved in their running and repair became more widespread. Both encounters with the devices themselves, and formal publications were significant contributory factors in these new knowledges.
Despite their improved performance, pendulum clocks were not self-contained. Sundials continued to be important in maintaining sun-time and clock-time. Indeed, people quite quickly came to have new and higher expectations of sundials, the accuracy expected from them, and the refinement in their time-telling.