by Mr Pilbrow while constructing the deep-drainage system of the city in 1868, the number of Roman tesselated pavements, coins, and other relics found at various periods, and the traces of Roman cemeteries, abundantly prove that Durovernum developed at length into a large and populous place.

Among various discoveries may be enumerated Samian ware, coffins, conduit pipes, rings, bottles, urns, Upchurch pottery, spoons, arrowheads, and skeletons, as well as indications of a large iron foundry; and a long list of gold ornaments includes portions of châtelaines, fibulæ, studs, purses, combs; and (what is especially germane to this history) a purple enamelled Roman brooch of circular shape, and a looped Roman intaglio, found near St. Martin's Church. All these appear to show that the Roman occupation of Canterbury was at once complete and continuous.

Of Roman secular buildings above ground there are indeed no remains, and the ancient city must be traced some eight feet below the present level. But in St. Margaret's and in Sun Street there are undoubted evidences of Roman walls. It is not impossible that, when first occupied, the town of Durovernum was very small, consisting of a citadel surrounded by earth mounds, and that it gradually extended itself afterwards beyond its original limits.

The elegance of some of the enamelled brooches and rings, together with other discoveries, point to a considerable degree of luxury and civilisation. One writer fancied that he detected the remains of raised seats for spectators at a circus or amphitheatre in the so-called Martyr's Field, near the London, Chatham, and Dover Railway Station.

The exact dimensions and extent of the city are open to some doubt. Mr T. Godfrey Faussett fixed the site of the four gates as follows:—(1) Worth Gate, at the end of Castle Street; (2) Riding Gate, on the old road to Dover; (3) North Gate, near the present south-west tower of the Cathedral; and (4) a gate at the Ford, in Beer Cart Lane. Tracing the walls that lie between them, he concluded that the shape of the Roman town was an irregular oval, different from the usual square or rectangle, but accounted for by the low swampy ground that surrounded it, and not unlike the shape of Verulam and Anderida. The city's length, according to his plan, must have been nearly exactly double its breadth—namely 800 yards by 400.

FROM A PLAN DRAWN BY T. GODFREY FAUSSETT.

For actual existing buildings that may possibly have been connected with the Roman occupation, we must have recourse to the churches, which supply us with traces of early Christianity more rich and numerous than that of any other town in England. These are to be found in St. Martin's, St. Pancras, and a church on the site of the present Cathedral. Detailed investigation of them would bring us to some controversial points, for

the discussion of which one must be thoroughly conversant with all the recent discoveries and explorations that have been made. But we may, at any rate, state the documentary evidence.

With regard to St. Martin's Church, we have already quoted the statement made by the Venerable Bede.

The same historian also informs us that Augustine, "when the Episcopal See was granted to him in the royal city, recovered therein, supported by the king's assistance, a church which, he was informed, had been built by the ancient work of Roman believers; and consecrated it in the name of our Holy Saviour, God and Lord, Jesus Christ."

He does not mention St. Pancras, but we are indebted for an account of it (evidently based on older traditions) to Thorn, a Benedictine monk of St. Augustine's, in the fourteenth century. "There was not far from the city towards the east, as it were midway between the Church of St. Martin and the walls of the city, a temple or idol-house, where King Ethelbert, according to the rites of his tribe, was wont to pray, and with his nobles to sacrifice to his demons, and not to God—which temple Augustine purged from the pollutions and filth of the Gentiles; and having broken the image which was in it, changed it into a church, and dedicated it in the name of the martyr St. Pancras; and this was the first church dedicated by St. Augustine." St. Pancras, a Roman boy of noble family, was martyred under Diocletian at the age of fourteen, and was regarded as the patron saint of children. Dean Stanley reminds us that the monastery of St. Andrew on the Cœlian Hill, from which St. Augustine came, was built on the very property which had belonged to the family of St. Pancras, so that the name would have been quite familiar to the Roman missionary.

Now, these are the written traditions with regard to the early churches of Canterbury. How far, then, are they confirmed by actual discoveries? A great deal of light has been thrown upon the point within the last few years. In the course of explorations conducted in the Cathedral crypt by Canon Scott Robertson, Dr Sheppard, and myself, there was found at the base of the western wall some masonry of Kentish ragstone covered by a smooth facing of hard plaster, manifestly older than the columns of Prior Ernulf's vaulting shafts, and than Lanfranc's masonry in the upper portion of the wall. We may, therefore, consider it as more than probable that a portion of this wall (which was laid bare to the length of twenty-seven feet) formed part of the original building granted to St. Augustine by King Ethelbert.

The ruins of St. Pancras have also been carefully and minutely investigated, and traces have been found there of both an undoubtedly Roman, and a somewhat later, building. Though Mr J. T. Micklethwaite has satisfied himself that the present foundations can only be assigned to an Early Saxon period, asserting, indeed, that "we have evidence that it was used by St. Augustine himself," his arguments can not yet be accepted as conclusive, and much may be said on the other side.

We may observe an apparent difference in the shapes of these three churches. Of St. Martin's we shall speak at length hereafter, but we may note that, besides the different width of the nave and chancel, there is no sign of an apse at the west end, while indications of an eastern apse are more or less conjectural. In the plan of the original Cathedral, conjecturally drawn by Professor Willis from Edmer's description, and which he supposes was the old Christian church preserved by St. Augustine, the building was a plain parallelogram, with apses at both the east and west ends. The choir was extended into the nave, enclosed by a high breast-wall, and about the middle of the church (on the north and south) were two towers, the tower on the south side containing an altar, and also serving as a porch of entrance. This church was built, according to Edmer, "Romanorum opere," and in imitation of the Church of St. Peter, chief of the apostles, meaning the Vatican Basilica.

In St. Pancras there is a tower, or square porch, at the west end, and two transepts of the same size branching off from the centre of the nave, while the foundations of the chancel walls start farther in than those of the nave wall; and, at the distance of twelve or thirteen feet from the point of junction, can be detected the commencement of an apse. In this church we have discovered no doorways, except the one at the west end through the tower, and the possible indications of one leading into the southern transept, where we may yet see remains of an interesting altar