Short URL for this page:

[image ALT: Much of my site will be useless to you if you've got the images turned off!]
Bill Thayer

[image ALT: Cliccare qui per una pagina di aiuto in Italiano.]

[Link to a series of help pages]
[Link to the next level up]
[Link to my homepage]

[image ALT: link to previous section]
This webpage reproduces a chapter of
Roman Era in Britain

by John Ward

published by Methuen & Co. Ltd.
36 Essex Street W. C., London

The text, and illustrations except as noted,
are in the public domain.


[image ALT: link to next section]

 p229  Chapter XIII


Padlocks and Fixed Locks — their types and mechanism as indicated by their keys

That locks were in general use is proved by the keys found on most of our Roman sites. Of the actual locks, few remain, and these are of two kinds, small fixed locks with hasps for boxes and caskets, and padlocks of a peculiar type which have survived, but not without change, in some Eastern countries. Locks suitable for doors have not been identified, yet they must have been common enough, for many of the keys could not have been used for padlocks and are too large for the hasped locks just referred to. It is probable that these larger locks were wholly or partly of wood. For the exact mechanism, with the exception of that of the padlocks, we have to rely more upon a comparison of the keys with those of old forms of locks that have survived, than upon their actual remains.

[image ALT: missingALT]
Fig. 65. — Padlocks and their Keys. (All ⅓)

The padlocks shown in Fig. 65 are all upon the same principle, and may be termed spring padlocks. A is a typical example of the larger sort found at Great Chesterford in 1854. It is of iron, and consists of two separable parts, — (1) a rectangular box, a, with a long rod attached to the upper surface and bent back as indicated, and (2) a bolt, b, one portion of which inserts into the case and is provided with catch-springs, the other having two arms ending with eyes. The rod served the purpose of the shackle on the modern padlock; and the bolt pushed home, sliding its eyes upon the rod in so doing, the catch-springs prevented its removal until compressed by a key. B is a transverse section of the case to show its construction. The sides are a  p231 continuous piece of iron, and the whole is secured by rivets passing through the top and bottom. D is a bolt of simpler construction from Llantwit Major. The portion to enter the case is doubled and each piece is pointed, the upper having two lateral springs arranged like the barbs of an arrow, and the lower a single one on the under side. In E, we have provided this bolt with a case and rod of the usual form, the former shown in section. Two examples of keys for these padlocks are given (F), and one to compress the Llantwit Major springs would require two rectangular openings in its foot as in the second. The foot being introduced into a narrow slit in the end of the case opposite the bolt-hole, was wholly inserted by a movement indicated in E. In pushing the key forward, its openings invested the springs and compressed them, when the bolt could be withdrawn by hand. The bolt of the Great Chesterford padlock is more complicated, having two parallel arms and eight springs, and thus required a key with intricate openings (C). Padlocks of the above type have been found at Silchester, Caerwent, Irchester, Cirencester, and elsewhere, but the keys are more numerous.

At Great Chesterford, two padlocks of a more compact form were also found, of which H is one. The rod ends with an eye, and is turned down so that the latter faces the bolt-hole, but with an interval to allow of the passage of the links of the chain to be secured. The bolt is straight and sufficiently long to project through the eye and so close the interval. Padlocks of this type were in use in medieval times and still survive in the East.​1 The key, G, found at Swanscombe, is almost certainly that of a padlock with a keyhole of the shape of its foot.

[image ALT: missingALT]
Fig. 66. — Old Scottish Tumbler Lock and Keys, and Roman Keys.
(The latter ⅔)
Turning now to the fixed locks: it is probable that locks akin to the well-known wooden locks of Scandinavia and Scotland were used in Roman Britain. The essentials of these locks are a bolt and two or more falling pegs, known as tumblers. These freely move in vertical grooves in the back of the case, and when the bolt is shot they fall into notches in it, and it cannot be drawn until they are raised by a key. The key has two lateral teeth, and upon being inserted into a horizontal groove in the back of  p232 the case end of the tumblers, the teeth coincide with the latter, and by lifting the key these are raised, when the bolt can be drawn by hand. Fig. 66, A, presents the front, side, back, and longitudinal section of one of these locks, and B is its key. C is the key for a similar lock with three tumblers.

An old improvement upon these locks consists in the tumblers falling into holes in the bolt, in which case the key is inserted below the latter. The key being lifted, its teeth, enter the holes and push up the tumblers, thus taking their place, and the bolt is drawn by a lateral movement or slide of the key. Locks of this type are in use in Egypt and elsewhere in the East. We thus have two types of primitive locks. In the one the bolt is hand-drawn, and, in the other, key-drawn. In either case, all the tumblers must be raised before the bolt is free, hence the key must have a corresponding number of teeth and arranged in the same manner. Keys of precisely the same form as the above have been sparingly found on our Roman sites, and imply that locks of the principle just described were in use in Roman Britain. D is a bronze example in the British Museum from Kingsholm, Gloucestershire, and a portion of a bone one has been found at Gellygaer.2

The defect of the above locks is that in order to allow of the key being inserted they have to be attached to the front of the door; but before the Roman conquest of Britain, keys had been devised to operate locks placed on the back. Keys of the form of E and I, which are found on both Roman and Anglo-Saxon sites, were adapted for this purpose. Similar keys are still used in Norway for wooden spring locks.​3 The spring is nailed to the door and its free end, which is towards the staple, strikes into a recess in the bolt when shot. The keyhole is a horizontal slit passing through door, spring, and bolt, and the key, pushed far enough through, is given a quarter-turn, then pulled forwards so that its teeth enter two holes in the back of the bolt and press down the spring, and the bolt is then drawn by sliding the key along the slit. It is not unlikely that these  p234 Roman keys operated locks of this principle. On the other hand, it has been thought that they operated tumbler locks similar to the Scandinavian, only placed on the backs of the doors. In this case, the keyhole would be vertical,​4 and giving the key a quarter-turn as before, its teeth would be drawn into holes in the tumblers, and so allow of them being raised. The bolt now free would have to be drawn by some simple contrivance, as a cord passing through the door. The keys, F, G, H, J, K, L, are found on both Roman and post-Roman sites, and are evidently variants of the above.

The keys of the next group, M to P, belong to a large and distinctively Roman class, and they differ from the modern in their bitts being transverse to their handles. They operated tumbler locks of the second type referred to above, but with keyholes in the front instead of the side, hence their peculiar form. The keyhole was not immediately below the tumblers, but on one side, so as to render them less accessible to the lock-picker; and it was L‑shaped as in the bronze lock-plate and hasp of a box from Colchester, shown in Fig. 68, A. The bitt was inserted sideways through the wider end of the hole. In this position the teeth were downwards; but by turning the key they were brought upwards in a horizontal position to the left immediately below the tumblers. The key was then raised, and in so doing the teeth displaced the tumblers; and finally the bolt, now free, was drawn by sliding the key to the right.

The first key of the group has no teeth, but the whole projecting side of the bitt may be regarded as a single tooth, which, of course, would fit an oblong hole in the bolt. As it would be easy to raise a single tumbler with a piece of bent wire or a smaller key, it is probable that there were several tumblers, so that unless all were raised together the bolt could not be drawn. The teeth of the other keys fitted a corresponding number of holes and raised a corresponding number of tumblers, those of the fourth key being in two series and of different shapes.  p235  Fig. 67, A, presents the upper surface of a metal bolt in the Guildhall Museum, and B, the under surface of a more complex bolt from Caerwent, which would require a key similar to the last to fit it. Most of the keys of the present type were for the locks of caskets and boxes, but the larger were apparently for doors, as certainly were the large iron keys with their teeth arranged in a zigzag, as in C.

While the tumblers in the door-locks may have been simply falling ones, it would be necessary for those in box and casket locks to be pressed down by springs. In restorations of these locks, this is shown as accomplished by a single spring.​5 This could hardly have been the case, for by raising one tumbler — no difficult matter — the spring would be released from the others, and, by turning the box upside down, these would fall back from the bolt, which could then be easily drawn. To be really effective, each tumbler should have its own spring.

[image ALT: missingALT]
Fig. 67. — Tumbler Bolts, and Keys of several types. (A, B, ½; the rest, ⅓)

Our next examples, Fig. 67, D to F, also belong to a large class of Roman keys which have a familiar look to modern eyes, but are more akin to medieval keys than ours. They are true revolving keys, and mark an advance in the locksmith's art, as a simple revolution one way or the other shot the bolt or drew it. It is not surprising that this advantage, combining simplicity of movement with expedition, should have secured the eventual victory of the revolving-key lock over its rivals. This lock had already reached a stage that persisted far into medieval times — until, in fact, the 15th century, when the craft of the locksmith attained an unsurpassed perfection in Germany and France. It was not a tumbler lock, and it was not until about a century and a half ago that the revolving key was made to operate tumblers, but of a different form than the ancient.

The Roman revolving keys, like the medieval and modern, are of two varieties, the 'pin' and the 'pipe,' the one having the stem solid and projecting beyond the bitt, and the other having it tubular. The principle of the lock is simple. The outer or fore-edge of the bitt presses, during part of its revolution, against a projection or stop on the bolt, and so propels it for a short but sufficient distance — the movement is that of the rack and pinion,  p237 the key being a pinion with one cog. If the mechanism of the lock went no further than this, it is obvious that any key with a bitt of the right length and sufficiently narrow to turn in the case would operate the bolt. To render this difficult or impossible, obstructions or wards were introduced into the case, which could not be passed by the bitt unless it had corresponding slits or openings. The key, Fig. 67, F, from Silchester, has two of these slits which would correspond with two little pegs or curved plates, the one attached to the front of the case and the other to the back. The keys D and E, from Caerwent, have, in addition to these slits, a number of notches on the fore-edge of the bitt. These imply a toothed ward-plate which would bar access to the bolt unless the notches corresponded with the teeth. Roman wards rarely went further in intricacy, in this respect contrasting with those of the later medieval locks. The curious iron key, F′, from Bath, is almost certainly Roman.

The keys of the next group, G to K, are less frequently found than the last. They chiefly differ in the fore-edge of the bitt having a right-angled flange or one or more teeth. It is probable that they answered to the modern latch-keys, that is, that by a half-turn a key of this type pressed back a bolt that was shot by a spring. This is confirmed by the fact that many of these keys have 'island' ward-holes. It is obvious that if a complete revolution was intended, the ward such a hole was designed to pass, could have had no support. If, on the contrary, only a half-turn was necessary, the ward could be fastened by its end to a transverse plate which would serve also as a stop to the bitt. The most feasible explanation of the right-angled flange or teeth is that they caught against a stud or studs on the lower surface of the bolt and so propelled it beyond the plate, there being a corresponding notch or notches in the top of this to allow of their passage. If a key failed to carry the studs beyond the plate, the bolt, of course, would be only partly drawn. The examples shown are from London, and it is noticeable that keys of the type usually have long loop bows, as in G.6

Our next two keys, Fig. 67, L and M, resemble those of the  p238 French latches which were in vogue until a generation ago. The keyhole of the French latch is of this shape — 
[image ALT: missingALT]
	. The key is inserted in the bottom slit and is then raised, the short stem sliding up the vertical slit. In doing this, the bitt has to pass a horizontal plate-ward, as also a narrow vertical plate to the foot of which the ward is riveted. This vertical plate is just within the vertical slit, and it serves as a screen to prevent access to the lock above the ward. The key, having passed the ward, comes into contact with a descending arm from the latch, and so raises the latter. The bronze plate of a hasped lock in the Guildhall, Fig. 68, B, would require a key of this form. There is no doubt that the movement of the Roman keys of the type was identical with that of the French latch-keys, but it is doubtful whether they lifted latches. It is more likely that their locks had bolts, and that in lifting the key the bolt was freed from tumblers of some special form. The key, however, would not be competent to draw or shoot the bolt, and the horizontal hole above the keyhole in the lock-plate just referred to indicates how this may have been accomplished. If the bolt had a small knob protruding through it, it could then be moved with the one hand while the key was raised with the other. The keys are rather rare, and the two shown are Guildhall examples.

Our next key, Fig. 67, N, is a rarer form of lifting key, which differs from the foregoing in having a marginal row of long teeth. The teeth seem to have raised tumblers that passed through the bolt; but beyond proving that the lock had both wards and tumblers, it is difficult to understand its operation. Probably it moved the bolt by a sliding movement.7

[image ALT: missingALT]
Fig. 68. — Lock-Plates, Keys, Needles, etc.
(A, B, C, ⅓; the rest, ⅔)
There yet remains another key of a very unkey-like appearance, which, although frequently occurring on Roman sites here and in France, is also found with Late-Celtic remains in both countries, for which reason it has been called the Celtic key. It is a bar of iron bent somewhat into the form of a sickle, with a flat handle. Fig. 68, C, is a typical example from Rushmore, and is 12 ins. long, which rather exceeds the average. There have been several suggestions as to how it was used, one being that it  p239 worked a bolt with a single tumbler, and another that the hand was thrust through a hole in the door and the key was used as a hook to pull the bolt. Neither, however, accounts for the curious shape of the key. We offer another suggestion — that it was pushed through a small hole in the door at a distance above or below the bolt equal to that between the point and the turn or 'neck' at the foot of the handle, the point being inserted in a hole in the bolt, as indicated in the illustration, which presents the section of part of a door. In this position the key became a level with the key hole as the fulcrum, and by pressing the handle to the left or right the bolt was moved in the contrary direction. It is a simple contrivance, and may seem to afford little security; but it is evident that only a key of the right length from neck to tip would be effective.

Whether plain or ornamented, the locks and keys that have survived almost invariably exhibit the good workman­ship common to all the productions of the metal-worker of the era. The hasped lock-plates are comparatively plain, but were often held by bronze nails with more or less ornamented disc-shaped heads. One found at Rushmore had a hinged keyhole-cover decorated with a youthful head in a Phrygian cap, the covers usually being internal and turned by a small external lever as will be observed in Fig. 68, A and B. The keys were sometimes elaborate, as the two typical bronze bows, Fig. 68, D, E, indicate. The keys of small trinket boxes were often in the form of ring-keys to be worn on the finger, of which two examples are shown in Fig. 76, P, Q.

The Author's Notes:

1 Examples are shown on plates v to ix, Pitt-Rivers' Primitive Locks and Keys.

2 Two similar bone keys are shown in the Limes Report on Zugmantel, plate xx.

3 One is figured in Primitive Locks and Keys, plate iv.

4 A lock-plate with an I‑shaped hole from Rushmore is figured in Pitt-Rivers' Excavations, i, plate xxiv; and another by Liger with a horizontal slit, perhaps for a small knob by which the bolt was drawn when released by the key.

5 There is such a restoration in the Guildhall Museum.

6 Pitt-Rivers, in Primitive Locks and Keys, considered that these keys raised tumblers, but it is inconceivable how they could have done so.

7 Liger figures lock-plates with 
[image ALT: missingALT]
	-shaped holes, probably intended for keys of this type.

[image ALT: Valid HTML 4.01.]

Page updated: 14 Dec 20