Books, Prehistory, Archaeologia Volume 64 1913 Section XI
Archaeologia Volume 64 1913 Section XI is in Archaeologia Volume 64 1913.
Before beginning my report on recent excavations I must, with great regret, express on behalf of the Committee and of all connected with the Caerwent Exploration Fund the great and irreparable loss we have sustained by the decease of its President, Viscount Tredegar, F.S.A. From the beginning, in 1899, t° the end, his Lordship had been one of the most enthusiastic and the most generous of our supporters. To him was largely due the success of the exploration, and without his aid it is doubtful whether it could have been undertaken. We have also lost by death since the date of the last Report another of our most valued supporters, Lord Llangattock, F.S.A., who was one of our Vice-presidents, and had taken much interest in the work, especially during its early years.
Since the autumn ot 1910 no very extensive diggings have been undertaken by the Committee of the Fund, but sundry small excavations, especially in the churchyard, have enabled us to add considerably to the ground-plans of some houses and buildings previously only partially explored and described. We have, also discovered two more interesting Houses, nos. XXV and XXVI s, one of which has been completely worked out, the other only partially, as its northern portion is under the chancel of the church. Our thanks are due to the Vicar of Caerwent, the Rev. W. Coleman Williams, M.A., for permission to make these excavations in such parts of the large churchyard as have not been used for burials for ages, and it is hoped that these diggings may be continued so long as funds hold out, and that more valuable remains may be discovered.
The most interesting and most unexpected find of the past year was made in September, outside the east city wall, and will be described later. It was at first supposed to have been an amphitheatre, but turned out to be the ruin of a large round temple, with very curious features not yet thoroughly explored.
No new work could be undertaken at Caerwent in 1912 until the extensive excavations of 1910-11, in the field north-west of the church, had been filled in. The southern portion of this field was so full of human remains that Viscount Tredegar decided to leave it unexplored; more than a hundred skeletons had No new work could be undertaken at Caerwent in 1912 until the extensive excavations of 1910-11, in the field north-west of the church, had been filled in. The southern portion of this field was so full of human remains that Viscount Tredegar decided to leave it unexplored ; more than a hundred skeletons had been exposed, and with the exception of a couple of iron spear-heads, found near two of them in the western part of the field, not a coin or relic of any kind had been discovered to throw light on their date. There can be little doubt, I think, that these were the remains of the Romano-British inhabitants of Caerwent, who had been slaughtered by some band of marauders from the Bristol Channel at some time during 'the dark period', between the middle of the fifth and the end of the ninth century.1 These marauders did their work very effectually, not onlv by stripping their victims of everything worth taking away, but by raiding the town so thoroughly that not a single bit of gold and hardly any silver has been left for us to find. They also appear to have set fire to the place and to have nearly destroyed it, as is indicated by traces of burnt material, charcoal, etc., sometimes several inches thick, left in most of the buildings we have excavated. Some of the houses which were undergoing enlargement or rebuilding, probably at this period, were never completed, and it seems probable that the place was for a time quite deserted.2 One of these late, unfinished walls is to be seen in the local Museum, which stands upon the site of a late Roman or Romano-British house. On another similar wall in the south diggings a number of unopened oyster-shells were found near the rough unfinished end of the wall, probably the uneaten dinner of the masons who were employed on the work and who were apparently driven off or slain by the marauders. Several skeletons, similar to those above mentioned, were found near the wall, some of which had been thrown into a well or pit in the neighbourhood. Other similar skeletons have been found recently in excavating the round temple outside the east wall.3
Note 1. Dr. Ashby does not agree with this theory. He writes: 'The burials were made long after the Roman houses had been razed to the ground or nearly so, and covered over with soil ; for the graves were obviously dug from the top, and in digging them what little remained of the walls was destroyed and broken through in such a manner as to make it clear that the diggers did not know where the walls were — otherwise they certainly would have avoided them. This is what happens in the churchyard at the present day, and it proves that the burials in the field cannot be so very early.' To this I reply that there can be little doubt that many of the Roman houses had fallen into ruin long before the date suggested for the raids, and that in some cases there is evidence that the walls were standing to a height of four or five feet at the time of the slaughter, as for instance at the round temple, to be described later.
Note 2. Early in the sixth century the coasts of the Bristol Channel seem to have suffered very severely from piratical raids, and many of the inhabitants sought refuge with their relatives in Brittany. According to the Rev. S. Baring-Gould, 'About 514 a. d. Rhiwal arrived from S. Wales with a large fleet in the Bay of St. Brieuc and founded the principality of Domnonia. Another swarm came from Gwent' (Caerwent), 'where existence had become insupportable owing to the incursions of the Saxons. This Gwentian colony planted itself NW. of the peninsula, and called it Leon or Lyonesse, after Caerleon that had been abandoned' (Baring-Gould, Book of Brittany , p. 5). See also Lives of the British Saints, vol. i, Introduction, p. 41 et seq.
Note 3. It has been suggested that these are the remains of the victims of the plague which ravaged the country in the sixth century, but in that case they would hardly have been buried separately all over the site.
The considerable expense of filling in the excavations of 1910 and 1911 was generously undertaken by Lord Tredegar, who much regretted that his wish to keep some of the buildings open permanently could not be arranged. The little temple marked on the plan 'House XIX'1, with the slab-covered courtyard or atrium to the north, between it and the main Roman street, would have been quite worth preserving, as also the well-built cellar or vault described in our last report, which was unlike anything found at Caerwent before or since. On the completion of the filling-in the field was restored to the owner in March, 1912, and there being no other land within the city walls then available, excavation ceased for a time and the workmen were dismissed. The land inside the walls not already explored is mostly covered with the houses and cottages of the village, with their gardens and orchards, but it is hoped that some day it may be possible to excavate part of this, especially the garden which covers the remains of the Public Baths, only a small part of which has yet been exposed.
Note 1. Archaeologia , vol. LXII, pl. LX. There can be no doubt that this, as suggested by Prof. Haverfield, was a temple, which had been erected on the site of an earlier house. Ibid., p. 439, note ; ibid., p. 424, pi. LXII.
Under these circumstances it was proposed by the Vicar, the Rev. W. Coleman Williams, that some little exploration might be made in the parish churchyard, which we knew to contain remains of several Roman buildings that had been partially excavated outside its southern and western boundaries. The churchyard is a large one, and no interments appear to have been made in some parts of it within living memory. On three sides a laurel hedge runs parallel with the walls at a distance of seven or eight feet, and a considerable portion to the south-east has long been used for the storage of loose stones (mostly Roman building material) which turn up nearly every time the ground is opened. The Vicars proposal having been sanctioned by the Bishop of the diocese and other ecclesiastical authorities, and approved by Viscount Tredegar, work was begun to the south of the church, and has been continued ever since, with the results that will now be recorded. Enough space remains within the enclosure to occupy us for some months to come, until more extensive excavations can be undertaken.
Books, Prehistory, Archaeologia Volume 64 1913 Section XI House No IXs
The greater part of this house was excavated in 1902, and the remains are described in Archaeologia2 From the plan there given3 it will be seen that the north-east portion of the house is under the ground inside the churchyard wall part of the churchyard very little digging could be done here, and we were only able to trace a few feet of the north wall of the house, and of the street to the north of it, which is here in very good preservation, although a little farther to the south the courtyard of House no. IX s seems to have encroached on it. It may be noted, by the way, that nearly all the streets running south from the main road (the so-called Via Julia) seem to have been more or less blocked in later times by houses or courtyards built over them. Thus the first street from the west wall is partially blocked by the long narrow courtyard of House no. V s. The next street eastward, that above named, is almost entirely blocked by House no. IX s. The third street, passing south between Houses nos. XII and XIII s, is greatly narrowed by the western wall of House no. XIII s being built upon it, and the fourth street seems to have been completely blocked by the baths excavated by Mr. Octavius Morgan in 1856. Also, the large house excavated by the same explorer quite blocks the east end of the street running east and west. So far as we know, all the cross streets running north from the main street, except the one destroyed at a late date by the construction of the amphitheatre, were carried right through to within a few feet of the north wall. The street, however, which leads from the west side of the forum is very narrow in its northern portion, between Houses nos. II and 1 1 1 n, and does not now lead direct to the North Gate.
Note 2. Archaeologia, LVIII. 404-5.
Note 3. Ibid., pi. XXVI.
Books, Prehistory, Archaeologia Volume 64 1913 Section XI House No Xs
Of this house, the southern part of which was excavated in 1904, nine rooms and a double courtyard have been previously described.1 From the ground- plan2 it will be seen that five of the walls are continued northward under the south wall of the churchyard. Our recent exploration, although it has almost doubled the size of the plan of the house, has added only one extra room. This plan is somewhat peculiar, there being corridors on the east and west, and two open areas, probably yards or gardens, to the south. Room 3 is continued northward from the churchyard wall to the main street, a distance of 36 ft. The north wall of Room 7 is just inside the churchyard, the new room, no. 10, being north of it. Room 3 is a large one, 43 ft. from north to south and 17½ ft. east to west.3
Note 1. Archaologia, LIX. 290-1.
Note 2. Ibid., pl. LXVI.
Note 3. The dimensions of the southern rooms of House X have been given previously. Those of the northern rooms are as follows: (1) 43 ft. by 6 ft. 3 in. ; (3) 43 ft. by 17 ft. 6 in. ; (7) 43 ft. by 9 ft. 3 in. ; (8) 9 ft. 6 in. wide at north end, 4 ft. at south end ; (10) 30 ft. by 8 ft.
Some fifty coins were found in the northern rooms, ranging from Diocletian and Constantine the Great to Arcadius, and a considerable quantity of common pottery, Samian being represented by a few fragments only. On the street north of the house a small silver coin of Henry III was found.
Room 3 had a concrete floor, of which we found traces 3 ft. 6 in. below the grass level ; over this was a mosaic floor, of which a fragment remained near the east wall, with red and grey tesserae, about 3 ft. 9 in. below the grass. To the west of Room 1 a number of stone roofing-tiles was found. East of Room 8 a large covered drain1 led from the street to the south-west ; several of the covering stones remain in situ.
Note 1. Archaeologia , LIX, pi. LXVI.
Books, Prehistory, Archaeologia Volume 64 1913 Section XI The Street
The street running east and west to the north of House no. X s has now been traced from the west city wall more than half-way to the east city wall. To the south of this street, occupying the whole width of the Insula, was the fine House no. II s, the largest we have found at Caerwent, and north of it were Houses no. VII and XIV s. Further west, on the south side of the street, were Houses nos. IV, VI, and IXs, and between the latter and House no. X is another recently discovered house, no. XXV s, shortly to be described, and other buildings. The north side of the street seems to have been mostly bordered by yards and gardens belonging to houses fronting the main street to the north, the so- called Via Julia.
The large covered drain above mentioned doubtless carried oft the rainwater from the surface of the street, near the north-east corner of House no. X; there was no small surface drain on the street, as there was on the street north of the basilica, but the street slopes down here to a lower level, and part of a gutter-stone was found which no doubt had served to carry the surface drainage into the large covered drain.
Books, Prehistory, Archaeologia Volume 64 1913 Section XI House No XXVs
This house is situated to the west of House no. X s, from which it is separated by a narrow space from 3 ft. 6 in. to 4 ft. wide. It is a building ot considerable size, and is entirely within the south-west portion of the churchyard. The remains consist of buildings of three or four different dates, as will be seen from the ground-plans (figs. 1-3). The earliest building consists of a rectangular chamber measuring 25 ft. from east to west, and 18 ft. from north to south, probably with an entrance in the north wall to the street. Little more than the foundations of the walls remain, and some of these have been made use of by the later builders. Near the west wall of this chamber, and 4 ft. below the turf, were found a base metal, or plated, coin of Carausius, and a small brass of Constans ; also a small iron knife, and a fragment of bone sharpened at one end, apparently for use as a toothpick.
The building next in date was a house of considerable size, containing six rooms, one with a southern apse or bay-window, added later, and with a large walled garden or courtyard to the east. The northern portion of this courtyard (no. 2 on plan, fig. 4) could not be excavated on account of modern interments, and for the same reason it was not possible to find the doorways opening into the street through the north wall. There was a gateway in the south wall, of which the sandstone doorstep with three circular holes for the bolts remain.
The third and last state of the house is shown in fig. 3. Several of the rooms of the previous house were retained, but a projecting porch was added in the middle of the south front, and rooms were built at the east and west corners, that on the west being built over and destroying the apse or bay-window of the previous house. Few relics have been found to throw light on the dates of these buildings, but the coins range from Septimius Severus to minims possibly of the end of the fifth century.
It will be convenient to describe the various rooms as they appear on the plan (fig. 4). Room i, at the south-east angle of the block, is a large rectangular chamber with a doorway 6 ft. wide in its northern wall opening into the courtyard, no. 2. There was probably also a doorway into the long room or corridor, Room 3, but no trace of this was left. A curious leaden object, 64 in. long, of unknown use, was found in Room 1 and a couple of coins of Valentinian I. Also in this room was found a portion of a large sandstone column 1 ft. 6 in. in diameter, a coin of Valens, and a small sandstone finial measuring 7 ft. by 6 ft. by 5 ft. In the courtyard outside the north wall of the room some coins of Constantine and his sons were found, and some of the circular stone 'pot-covers ' which are so numerous at Caerwent.
There were doorways in both the north and south walls of Room 3, the former 4 ft. 6 in. wide, leading into the courtyard; the latter into an open space to the south, between Houses IX and X s. The south wall of Rooms 3, 4, and 5 is built against and partially upon the south wall of the earlier house. There is no sign of a doorway leading from the corridor, no. 3, to Room 4, the ancient doorway here, belonging to the second date of the house, having been blocked up. No. 4 was evidently the south porch of the house and its principal entrance, leading through doorways into Rooms 5 and 9. Room 5 probably had doorways into 6 and 8, but no traces remain. Room 6-10 was built partially over the apsidal chamber of the earlier building, its south wall being continued beyond the corners of the apse and forming a rectangular block to match the similar projecting building at the south-east corner, Room 1.
With its projecting porch and two wings this building must have had an imposing fagade, measuring no ft. from east to west. Room 7 was built entirely on the walls of the earlier house, and like it had a door in its north wall leading into the street. As previously stated, the north wall of the house could not be excavated, but it is probable that it had a wide gateway about the centre of the yard leading into the street.
The floor of most of the rooms of this later house had perished, or been destroyed, but in the north-east angle of the courtyard were remains of a floor of concrete or rammed gravel. North and south of the door leading from Room 1 to the courtyard were flat slabs of sandstone, one 2 ft. 6 in., the other 3 ft. long. There were traces of concrete floors in Rooms 7, 8, and 11, two feet below the turf level, on the same level as the street to the north, into which doubtless one or more doors oave access.
The remains of the earliest building on this site have been sufficiently described above, but some further account of the second house may be given. Although the boundary wall of this second building was almost as extensive as that of the later one, the house itself consisted only of six rooms instead of nine, the whole of the space east of Rooms 4, 9, and 11 having been occupied by a large courtyard or garden. There appears to have been no direct entrance into this second house from the south, but a doorway 6 ft. wide opened into the courtyard, to the east, and another led from the courtyard into Rooms 4, 9, and 11. The doorstep, with a central socket and rounded corners, remained on the south side of the doorway from Room 4 to Room 9. During the second stage of the house the room at the south-west corner was considerably enlarged by the addition of a large bay or apse on the south side. This differs from other apsidal additions to Roman houses previously found at Caerwent in that the corners of the apse are solid, that is, the spaces between the circular and rectangular walls are filled in with stonework, as in the temple found near the forum.1 The remainder of the rooms of the second house continued in use in the later building.
Note 1. When first uncovered this apse looked so like that of a temple that some unknown visitor to the excavations sent a note to the local papers, which was copied into various London and other newspapers under the heading 'Discovery of a Temple of Diana at Caerwent'.
Two curious features remain to be described. Running north and south under the south walls of both the second and third buildings, 2 ft. west of Room 1, is a wide wall of which some 18 ft. were excavated. It continued both to the north and south, but could not be further explored. Probably it is older than any of the above-described buildings.
A still more peculiar feature was found 12 ft. to the west of the house, the base of a semi-circular wall, consisting of a single course only of large stones, possibly the remains of a round building, of which the internal diameter would have been about 12 ft. It is probably not of Roman date.2
In filling in the excavations of House No. XXVI, Room 7, a curious bronze object was found (fig. 5). The face may be intended for the Gorgon Medusa. It has been suggested by Mr. Reginald Smith that this may have been an escutcheon for the bronze handle of a large terra-cotta vase, as shown in the sketches (figs. 6 and 7). There were traces of tin (or silver) on the bronze.
Note 2. It is quite possible that at one period the whole of the buildings which we have called Houses X, XI, and XXV s may have been one large house of the 'Caerwent type' — with rooms on all four sides of a central courtyard, the rooms formerly on the south-west having perished.
Books, Prehistory, Archaeologia Volume 64 1913 Section XI House No XXVIs
Like House no. XXV s this is situated entirely within the churchyard. It is to the north of House no. X s, its south wall adjoining the north side of the street, which is here from 20 ft. to 23 ft. wide. There are traces of buildings of two dates, but most of the walls excavated belonged to a range of rooms running north and south, with a courtyard or garden to the west, all apparently of the same period. Only three of these rooms and a part of the courtyard could be excavated, as the remainder of the house is under the vestry and chancel of the church. The house was bounded on the east by a paved area, which looks like a street running north and south, and of which only a small part has been uncovered. It is not in line with the street to the east of the forum, which can be traced southward through a modern lane to the street leading to the south gate of the city. This may have been only a private road leading to an eastern entrance to the house. Room 1 was a large, rectangular apartment, 26 ft. 6 in. by 18 ft., with a doorway 4 ft. wide leading into Room 2, and with a break in its western wall where no doubt was a doorway into the courtyard. The south wall of Room 1 and that of the courtyard are very well preserved, standing 1 ft. 9 in. above the level of the street and only a few inches under the turf. Some small late coins, a block of tufa, oyster shells, Samian and other pottery were found from one to three feet deep, but there was no trace of the floor. Room 2 is not rectangular, the east wall being a foot longer than the west ; a door in the east wall led into the courtyard. Room 3 could only be partially excavated as much of it is under the church, but there are considerable traces here of an earlier building. An old wall remains under the south wall of Room 2, a return wall running north and south under the north wall of this room into Room 3, and there turning east under the east wall of Room 3. Our excavation of this last-named wall ended a few feet south of the chancel wall, terminating in some large squared blocks of oolite, probably from the Mendip hills, similar to those built up in the tower and south aisle. These seemed to be in situ , and probably formed part of some important building on the site of the parish church, traditionally said to have been a ' Temple of Diana '. The blocks of freestone previously found in this vicinity have since been used in the rebuilding of the south aisle of the nave, but the block carved with birds hiding under foliage, no doubt part of a cornice, and some other ornamented pieces, have been built for preservation into the wall inside the church.1
Note 1. Close to the south-west corner of Room 3 remains of a rough grave were found, in which were a few bones, a pewter sepulchral chalice, and a square iron buckle, which apparently belonged to a priest of about the beginning of the thirteenth century.
Books, Prehistory, Archaeologia Volume 64 1913 Section XI Round Temple
The most interesting recent discovery at Caerwent was made quite accidentally early in September, 1912. The field outside the city wall on the north-east, between the east crate and the north-east corner of the wall, showed no sio-ns of Roman remains underneath the turf. Recently the ground lias changed hands, the larger part having been purchased by the County Council, and the corner near the gate by the Trustees of the late Col. Burton, for the erection of almshouses. On the last-named ground a handsome block of almshouses has been erected, and it was a little to the north of these buildings that some workmen, digging for stone to build a garden wall, hit on the remains of a solid Roman wall, the top of which was only two feet beneath the grass. They then proceeded to make use of this wall as a stone quarry, and it was only after some 50 ft. of the wall had been dug out and made use of that the Vicar of Caerwent heard of the discovery and wrote to inform me of it. Accordingly on the next morning I went down to Caerwent and found that some 50 ft. of a circular wall had been exposed and carried away, leaving a section of the wall 5 ft. deep and 2 ft. wide visible at either end of the trench. On examination I found that the remains, which were certainly Roman, extended under the ground east and west of the portion that had been excavated, and were unfortunately on three different properties. Viscount Tredegar having been communicated with and having generously undertaken to pay for the exploration, application was made to the owners for permission to dig.
The Burton Trustees were unable to grant permission, but fortunately we were more successful with the other owners. The County Council and the Highway Board both courteously gave their consent to the necessary excavations being made in their properties, and work was at once begun in the field east of 'the Homes', the circular wall being found and excavated to a length of about 110 ft. It was a well-built wall, averaging about 5 ft. high and 2 ft. thick, with a set-off on the outside. At first we were under the impression that the building we were excavating was a second amphitheatre and that the wall was probably that of the arena, its presumed circumference of about 390 ft. outside being however much less than that of most Romano-British amphitheatres. We therefore cut a couple of trenches radiating from the centre towards the east and south-east but could find no trace of outer walls. Continuing the excavation of the circular wall southwards, at 23 ft. from the hedge we found an opening in the wall and at a depth of 4 ft. a sandstone doorstep, with two socket holes for the bolts.
From this east door to the south the wall continued for 53 ft. to the hedge, which it passed under into the forbidden land of 'the Homes'. A couple of trenches from the wall towards the centre, as far as our boundary allowed, exposed the sloping clay floor, of which two sections were taken (figs. 9 and 10), but there was no trace of an inner wall. We then burrowed under the hedge so far as we dared, but still without finding anything. Being convinced that a few more feet, perhaps even a few more inches, might expose some inner building, I came to a rather rash conclusion. Having written to the Charity Commission in London, who informed me that they had no objection to the excavation being made on the property of 'the Burton Homes', and one of the two Trustees having already expressed his readiness to consent if his colleague would do so, I decided to run the risk and to 'take French leave' to dig a trench on the other side of the hedge, notwithstanding that one of the Trustees had refused his consent. After a few hours' excavation we were rewarded by the discovery of the remains of a portion of the inner wall. This, however, was not circular, but appeared to be quite straight, possibly a side of an octagon. While some ol the men were at work on this wall others were continuing the circular trench which had been left open by the Trustees' men. Fortunately two hours' work here exposed the north door (fig. 11), 8 ft. wide, with its doorstep containing square or oblong holes on either side for the gate-posts. At this point came a sudden end to all our hopes of further discoveries in the form of a telegram from the Cardiff lawyers, followed by a letter in which they ordered the trenches already opened to be immediately closed and the damage repaired. Under these circumstances I went down to Caerwent, carefully noted all that had been found, ordered the trenches to be at once filled in, and wrote a letter to the Trustees apologizing for my trespass.
Shortly after this the workmen employed by the Trustees cut a trench for drainage purposes a few feet west of the small piece of wall we had discovered. This was a most fortunate circumstance, as they cut through two more fragments of walling which we were able to add to our ground-plan, and which support the idea that the inner wall of the temple was octagonal. What was inside the octagon we do not know ; possibly another octagon, as has been suggested. Several Roman buildings with octagonal chambers inside a circular outer wall have been found in Britain, but I believe this to be by far the largest building of the kind yet known, if we may assume that this is a case.1
Note 1. The diameter inside the circular wall at Caerwent is 130 ft.; the diameter of the supposed octagon about 60 ft. This is about the same as the outer octagon of the building at Weycock, Berkshire, described by Mr. Neville in the Archaeological Journal , vol. vi, which 'consists of an octagonal wall 63 ft. 7 in. exterior diameter, enclosing a smaller octagon 35 ft. n in.' Possibly the Caerwent building also had a second octagon, but at Weycock there was no sign of an enclosing circular wall, and there were no traces of the entrances.
The excavation in the lane to the west, undertaken by kind permission of the Highway Board, led to the discovery of the west door, 32 ft. south of the point where the circular wall passes under the modern wall of the lane, the door being 4 ft. wide, 5 ft. below the surface of the road. The sandstone doorstep, 4 ft. wide, had a socket hole on the north side. This door is 270 ft. north-north- east of the south pier of the east gate of the city, and 50 ft. outside of the east city wall. The south door of the temple no doubt remains, but it is under the new buildings and cannot be excavated.
It has been suggested from the fact of the doorways being at the cardinal points that this may have been a solar temple. Nothing came to hand to throw any light on the subject, the only find being a number of round stones (the so-called potato-stones) found just inside the north gate. It is possible that these were sling-stones for the use of the unfortunate natives who had sought refuge in the temple during an attack of marauders, and whose bones still remain where they fell, several skeletons having apparently been thrown in carelessly one on the other, inside of and upon the ruins of the circular wall. This fact is interesting, as it shows us that the temple was in ruin when these interments took place.
Some of these human remains have been sent to Professor Macalister, of Cambridge, as mentioned in our last Report, where also his opinion that they are of post-Roman and pre-Saxon date is mentioned. He writes further, ' I am satisfied from their character that they are of the type commonest among post- Roman crania and are identical with others that I have obtained whose pre- Saxon age was confirmed by the fragments of pottery and other utensils found with them. I believe that they are most probably of late fifth or early sixth century age.' This would fit in with the raids of the Saxon or other marauders of the early sixth century1, with the early Welsh traditions of extensive migrations from Gwent to Armorica at that period, and the late unfinished buildings and numerous very late and much worn coins may be considered as evidence, even if it cannot be regarded as proof, that Caerwent continued to be inhabited by a considerable population for something like a century after the time of Honorius.
Note 1. In the new edition of his pamphlet on The Romanization of Roman Britain , pp. 63 and 64, Prof. Haverfield, in describing these raids, says: 'As the Romano-Britons retired from the south and east, as Silchester was evacuated in despair and Bath and Wroxeter were stormed and left desolate, the very centres of Romanized life were extinguished . . sites lay empty and untenanted for many years. Only in the far west, at Exeter or at Caerwent, does our evidence allow us to guess at a continuing Romano-British life.'
The building is on very low ground, and during our excavations was at times flooded to a depth of 4 ft. or 5 ft. At the north-east corner of the city, 300 ft. from the temple, there is generally a pond in wet weather (on the site of the moat), and it is difficult to say how the temple could have been kept free from floods, there being no sign of or room for a moat or ditch between it and the east wall of the city.