ment to their "horses" chests. Indeed upon consulting the sculptures of Trajan's column at Rome, several examples will be seen in which the war-horse of the Emperor is represented as adorned with one or more crescent-shaped ornaments hanging upon his chest from a sort of necklace (monile). (Pl. iv. fig. 13.) It is possible that these may have been made of other material than boar's tusks.1 But it is evident how well such curved tusks lend themselves to adaptation to the form of a crescent, white and polished, of this kind. It is therefore not unlikely that they were frequently employed for this purpose when obtainable, as they would be in countries abounding with wild boars, such as our wooded hills of North Wiltshire. Indeed more than one other pair of tusks found in the Wraxhall excavations present marks by which it would appear that they had been connected by a sheath or mounting of this kind. Mr Akerman speaks of some found together with Roman relics, which seemed to have been mounted singly for suspension round the neck, as a charm perhaps against the evil eye, in which way it is well known that horns are still worn in Italy, and were formerly by the Romans, as an amulet or charm. In regard to the Arab horse ornament I have lately been favored by a communication from Mr. Churchill, a gentleman now residing at Beyrout, who states that such crescents are frequently so worn by the higher class Bedouin Chiefs, with the object of averting, not the evil eye, but a skin disease to which the horses are liable in that climate. If so, however, it can only be as a kind of charm that the boar's teeth are supposed to operate. Mr. Churchill speaks of blue beads being worn to preserve from the evil eye, and the Arab crescent belonging to Mr. Akerman has a blue glass bead strung upon its suspending cord. Both charms are therefore probably united in this instance for the double purpose. Whether any superstition of the kind was attached by the Romans to the crescent ornament does not appear. But it is 1 We read in the classic authors of the rich 'phalere' and 'ephippia' worn by the Roman Cavalry-thus Virgil tells us 66 aurea pectoribus demissa monilia pendent." And Claudian "Dumque auro phaleræ, gemmis dum frena renitent." And Aulus Gellius "Equitatus frenis, ephippiis, monilibus, phaleris, profulgens." Lib. 5 cap, 5.

worth mentioning that a blue glass bead was found in the excavations not far from the tusk-crescent, and might have been worn with it as in the Arab example. The Pet Stag of Calpurnius Siculus was also decorated with "vitrea bullæ," employed in a belt or girth around him.

a dorso quæ totam circuit alvum, Alternat vitreas lateralis cingula bullas."

With regard to the original character of the buildings above described, they may be conjectured to have formed the Villa Rustica, or country residence, of some Roman personage of Civil or Military importance; the site being chosen perhaps for the advantage of the chase, since the surrounding hills, no doubt, formed part of an extensive forest in early times, and the number of deer-horns and boars' tusks found in their rubbish would support this opinion. The servants and followers of the proprietor, perhaps also a small military detachment, stationed there for the defence and security of the neighbouring great Military Highway, the Foss-road, occupied probably the buildings adjoining to the principal habitation. The abundance of bones, fragments of pottery and other relics attest the continued inhabitancy of this station for a period of several generations.

Other spots in the vicinity shew vestiges of Roman occupation. Indeed on the opposite point of hill to this villa, across the adjoining glen, but within the parish of Castle Combe, the labourers digging the ground for the plantation now growing there, some years since, met with a small stone slab having the figure of a hunter or huntress spearing a stag rudely sculptured upon its face, (perhaps a votive altar to Diana), (Pl. iv. fig. 15.), together with a heap of about three hundred brass coins mostly of the Lower Empire. And in the parish of Colerne, between North Wraxhall and Bath, the remains of another Roman villa, having several tessellated pavements, were opened a few years back. Similar vestiges are, indeed, frequent along the whole range of heights traversed by the Foss-road from Bath to Cirencester. Upon or near the western escarpment of these heights (the Cotswolds) several camps are still to be seen, which formed a series of posts protecting the country lying to the east-ward

from the side of Wales. And it seems probable that the officers or chief men commanding the legions stationed in these camps, or at Bath and other towns upon this military road, possessed their villas or hunting-boxes on favorable sites in the neighbourhood; and of such these remains at North Wraxhall may perhaps present us with one example.

The spot in which it occurs is distant a good mile on the N.E. of the Parish Church of North Wraxhall, and about the same distance from Castle Combe, and also from the nearest point of the Fossroad. It is rather difficult of access for carriages. But a lane from Wraxhall leads to within a couple of fields of the spot. On the side towards Castle Combe it can only be reached on foot or on horseback, the deep glen which bounds the site on that side having no practicable road. I may add that the good state of preservation in which the greater part of this villa was found is owing to its having been over grown with wood perhaps ever since its original destruction. The wood indeed, was grubbed up only within the memory of living

men.