The "Laconian" type of key, described by Aristophanes1 as having three teeth, and as being a cause of consternation to the women when adopted by their husbands, was almost certainly a key of this type. The majority of Roman locks, though of a more

complicated structure, are made on the same principle, as may be seen from the ancient lock No. 397 (probably from Pompeii) here exhibited, together with a model lock of the same type, and a diagram showing its original arrangement (fig. 171a-d). Here

1 Arist., Thesm. 421 ff.

the bolt has been which remains (fig.

FIG. 173.

1

ROMAN PADLOCK, WITH KEY RUSTED IN IT (No. 400). Ca. 1: 3.

shot through the end link of a chain, part of
171c). It is secured by pins, the ends of which
fit into a series of perforations in the bolt and
are kept down by a spring. The bolt was
released by a key fitted with teeth corre-
sponding to the perforations (fig. 171d). The
key lifted the pins out of the holes and took
their place.
The bolt was then drawn
aside, as the key was moved along the
horizontal slot. Several bolts, keys (e.g.
No. 398; fig. 172c), and door plates for
locks of this type are exhibited in this
Case. Notice the projections on the ring of
key c, which were used for shooting a sup-
plementary bolt, a common device in Roman
locks.

The modern type of lock, in which the key simply moves the bolt backwards and forwards, after passing through a series of wards, was also known to the Romans. This is proved by the existence of several Roman keys solely adapted to a lock of this character (e.g. No. 399; fig. 172d).

Such keys are frequently found combined with finger-rings, a convenient method of minimising the danger of loss. We may

conclude that this type of key was a favourite one for use with small padlocks.

FIG. 175.-BRONZE STRONG-BOX, WITH COVER SEEN ON INNER SIDE.
C AND & EXPLAIN THE WORKING OF THE BOLT (No. 405). 1:2.

Padlocks of Roman date are common. In this Case three of a barrel form are shown. One (No. 400; fig. 173) has the key still rusted in it. The padlock has traces of a chain attachment at one end, and was probably kept hanging to a doorpost, while the bolt was shot into the end link of a chain attached to the door. Two other Roman padlocks illustrated (fig. 174) are more ornamental in character. One (No. 401) is in the form of a circular box with hinged handle, the free end of which was fastened by pin-bolts within the box. There is also a secret catch underneath. The other padlock (No. 402) is furnished. with a chain attached to one side of it. The last link of the free end was fastened inside the box, the lid of which was closed

with a secret catch. The head on the cover is that of a Sphinx, a hint that the riddle of opening was not easy to solve. This padlock is especially interesting because of its analogy to the seal-boxes described below (p. 167). A hole in the floor of the box makes it probable that it was fastened to the object to be secured.

Other objects deserving mention are the keys for raising latches (No. 403; fig. 172b), and the combined ward and pin keys (No. 404; fig. 172c), and also the very interesting GraecoRoman bronze strong-box from Tarentum (No. 405; fig. 175). The box (a) has a sliding lid (b), originally furnished on the inside with four separate fastenings. Two are horizontal bolts shot home by turning toothed discs from the outside; the third is the catch seen at the end, which was held fast in the slot by a pinbolt (c). This bolt was moved by a disc on the outside of the cover, and was itself locked by the turning of another disc behind it; it could only be drawn back when the slot in that disc was brought into line with the bolt, as indicated in design d of the figure. The small catch on the right at the end of the box fell into position automatically when the cover was closed, and could only be unfastened by turning the box on its side. The outside of the lid shows four similar circles, over which were the revolving or sliding discs now lost (fig. 176).

Seals. These were very closely connected with locks in ancient life, and often in fact took their place. Aristophanes in the passage above quoted makes the women complain that not only did their husbands carry the patent Laconian key, but that they also (at Euripides' instigation) carried very complicated "worm-eaten" seals, not likely to be forged. Several objects in this Case illustrate the use of seals. When a man wished to secure an object he tied it up with string and put a lump of clay over the knot, impressing the clay with his signet. Such impressions are seen on several baked lumps of clay here exhibited. One large lump (No. 406) has no fewer than eight Roman seal impressions (several from the same seal), while the knot of the cord remains embedded in the clay underneath. This Case also contains examples (No. 407) of Roman seal-locks (one in wood and several in ivory). The wooden lock, found in Egypt, is shown in fig. 177a, where its probable use is indicated. The lock was suspended from the door-jamb on a pivot passed through the small hole seen at the left end. The loop or staple attached to the door was then inserted in the groove, and the movable

1 Arist., Thesm. 421 ff.

cover slid through it, as shown in the figure. The clay or wax was next pressed into the hole behind the lid, and sealed with a signet (as in fig. 1776, top view). The door could then not be opened unless the seal or the lock was broken. Such a lock would be very useful to prevent the often-mentioned pilfering by slaves.1

Another interesting class of objects is that of the seal-boxes (No. 408). They are small bronze boxes with hinged lids, and resemble in form a pear-shaped or circular lamp. Each box has a small slot cut out on either side, and three or four holes pierced in its floor. The cover not infrequently has a design in relief (such

1 Cf. Plin., H.N. xxxiii. 26: nunc cibi quoque ac potus anulo vindicantur a rapina.