Green Hill (which lies opposite the Fort Royal or Park), there was, till lately, a considerable mound of earth, most probably the site of the fort erected by King Stephen on the London Road, when he laid siege to Worcester Castle. It may, however, have been much more ancient. The other fort which he built was on the Bath Road: the mound on which it stood has also been removed.

In an old trench at the top of the ridge, between the supposed hypocaust and the mound, an ancient British coin was dug up by Mr. Holland's workmen, and also Roman coins of Alexander Severus, Gallienus, Victorinus, and Tetricus the younger. The ancient British coin is of common type, and I am informed that it cannot be appropriated to any particular chief, nor as yet to any particular district. The obverse of it probably represents a head, and its reverse exhibits a horse galloping towards the left. (See the woodcut.) This is the only ancient British coin which has come to my knowledge as having been found at Worcester.

AR

Some of the tiles found in the supposed hypocaust have a groove or channel across them. Several of these tiles I exhibited at a meeting of the Archæological Institute. Some of the members considered that they were Roman roofing and paving tiles, and that the grooves or channels might possibly have been made to receive either the recurve of Roman flanged tiles, or to carry off the water, or they might have been to enable workmen to break them in half when needed. The mortar or cement in which they were set still adhered to them, containing much pounded brick, and this was considered a further proof of the workmanship being Roman. In the "Journal of the Archæological Institutet," these tiles are described as Anglo-Roman; and

* In this latter case, they might either have been paving tiles or wall tiles. + Vol. vii., pp. 302, 303.

it is further added,-" The fragments exhibited presented some unusual peculiarities of fabrication, some of these tiles having been deeply grooved, in a manner differing from the scoring of common occurrence, serving to retain the mortar firmly: another tile, apparently for roofing, was formed with a knob at top, as a means of attachment. Lyon, in the History of Dover Castle,' speaks of wall-tiles in the Roman pharos, formed with hemispherical knobs at the angles; but this contrivance is unusual."

The fragments of the tiles in question are so imperfect, that it is impossible to say decidedly whether the channels ran along or across them; the former was most probably the case, as they are in the middle of the lengthwise centre of the tiles, but would not be quite in the middle crosswise*.

As Sidbury, or Southbury, lies on the south side of Worcester, it is probable that it was so called from its position in regard to the City. There is a Saxon charter, dated A.D. 963, in MS., Cotton, Tiberius A. XIII., which seems to establish this view. It is a grant from Bishop Oswald to Cynethegn, of two and a half manses or hides of land, at Oddingley, in Worcestershire. The charter goes on to say, Thonne is ealles thæs landes the oswold bisceop bocath cynethegne, thrinde healf hid and VI. æceras at haranlea and XL. æcera be eastan Lawern, and se haga be suthan byrig se is XII. gerda lang and IX. gerda brand," &c.

That is, in English, "This is the whole of the land which Bishop Oswald gave to Cynethegn, two and a half hides, and six acres at Harley, and forty acres east of Lawern, and the enclosure by the south beorh, [or borough,] which is twelve yards long and nine yards broad."

* See the "Archæologia," Vol. xxx., Appendix, p. 537, relative to channelled bricks found in Roman foundations at Thornham, near Maidstone, in Kent. + Leland, Habingdon, and others, wrote it "Sudbury," and it is so spelled in Saxton and Speeds' Map of 1610, and also in the map in "Boscobel."

Printed in "Cod. Diplom.," No. 507, which work also mentions Suthbyrig (Sudbury), in Suffolk, Nos. 685, 699.

In a survey of the Forest of Feckenham, 28th Edward I.*, the name is spelled Southburi. There are frequent instances of towns similarly designated on account of their position. Sidbury, or Chidbury Hill, in Wiltshire-a vast oval fortification, encompassed with two deep ditches-lies south of Everley. (Gough's "Camden," Vol. i., p. 158.) There are also Sidbury in Devon and Salop, Sudbury in Derbyshire and Suffolk, and Southbury (Chapelry) in Kent. In the "Worcester Miscellany" for 1829, it is contended that Sidbury, like Silbury Hill, in Wiltshire, is of ancient British origin, and derived from the Keltic word Sul, the Sun; and that the adjoining heights (now called the Fort Royal, or Park) were dedicated to the worship of Sul, or the Sun, the Keltic Apollo. This etymology, however, appears to be invalidated, for in early times it was designated as above. Still the " bury," or 'burrow," most probably was of ancient British or Roman origin; for the Saxons thus distinguished the fortified places of the Britons and Romans.

66

In September 1844, several Roman and other relics were found at Diglis, near Worcester, the particulars of which I communicated to the Worcester journals in that month nearly as follows:-At the south part of the cutting, across the meadow at Diglis, for the Severn Navigation Lock, at the depth of about twenty feet in the alluvial soil, were portions of small trees, bushes, and hazel nuts, intermingled with fragments of stags' horns and bones; a little nearer to the river, southward, at the depth of about twenty-five feet, portions of an oak tree; and still nearer the river, at the depth of about thirty feet, a great number of bones of the deer kind, and of short-horned cattle f and other animals, together with fragments of Roman urns and pans of red earth, and a piece of Samian ware; a little nearer to the river, at the same depth, the horns and part of the skull

* See Nash, Vol. i., Introduction, p. 65.

+ A small extinct ox, the Bos longifrons of Mr. Owen; fragments of the bones of which I sent to him. See the "Journal of the Archæological Institute," Vol. vi., pp. 34, 35, and 127.

of a stag or red deer (Cervus elaphus), weighing twenty-one pounds*. Alongside of this latter relic, was part of the underjaw of a horse, and a smaller antler; also the greater part of a

fine Roman urn, of slate-coloured

pottery, eight inches high, and twentysix in circumference-(see the woodcut here represented). It seems probable that there were Roman or Roman British pottery works near to the spot in question, like those discovered on the border of the Severn at Bow Farm, in the parish of Ripplet; and it is worthy of remark that the Diglis pottery, both red

and slate-coloured, exactly corresponds in character with that discovered in the Roman burial-ground at Kempsey. A coin of Marcus Aurelius was also found at the cutting. It appears to me that there was an ancient dyke at the spot, and that the rill of water which ran into the Severn having, in ages past, been diverted into another channel, the dyke became gradually filled up by the alluvium occasionally deposited upon the plains by the floods of the river, and thereby all the relics were buried at the great depth at which they lay; in proof of this, it may be remarked that the stratum on which they rested was muddy grit,

The antlers of the stag, or red deer, I presented to the British Museum, affixed upon a block of the oak tree.

+ Antiquarians have been in much doubt how such pottery was coloured. Perhaps the following extract from the "Archæological Journal" (Vol. i. p. 280), relative to a communication from Mr. Edmund Tyrell Artis, as to a Roman pottery-kiln discovered in the vicinity of Castor, in Northamptonshire, will throw some light upon the subject. The kiln 66 appears to have been used for making the bluish-black or slate-coloured kind of pottery, so frequently met with wherever Roman remains are found in England. This colour, Mr. Artis has ascertained, was imparted to the pottery by suffocating the fire of the kiln, at the time when its contents had reached the proper state of heat to ensure a uniform colour." Also see "Remains of Roman Art in Cirencester," pp. 78, 79, relative to how the colour was produced by chemical action.

such as we find at the bottom of water courses; and my opinion is, that in the Roman time the Blockhouse stream, and other rills from the adjacent heights, ran into the Severn at the point in question. In some proof of this, the black seam pointed that way through the whole width of the south part of the cutting. It would have taken an immense time for these relics to have been buried upon the surface of a level plain by the alluvium, at the depth they were; for it will appear in the accounts of Pitchcroft and Ripple, that the alluvium upon the level plains on the borders of the Severn has only accumulated about four feet since the Roman time.

Several of the fragments of the oak tree, before mentioned, still retain the bark. Fragments of bark also appear upon the oak coffin of a supposed ancient British chieftain, preserved in the Scarborough Museum, the particulars of which I communicated to the Society of Antiquaries*.

In the same year (1844), about a mile and a half below Worcester, and half a mile below the Diglis Lock, a bronze spear-head of very unusual shape was dredged up by some workmen employed in the improvement of the navigation of the Severn. It is ten inches and a half long, two inches and three quarters broad, and weighs eight ounces. A woodcut of it, as here represented, was given in the "Archæological Journal," Vol. ii., p. 187. It is there stated to be of " remarkable form and singular fashion, the blade being flat, and of greater breadth than usual; terminating at the lower extremity in a shape more resembling the barbed head of an arrow, than the head of a long-handled weapon." It is figured in the "Proceedings of the Archæological Institute at York, 1846," p. 39, plate v., fig. 4, and noticed in p. 34 of that work. It was also exhibited at a meeting of the Society of Antiquaries of London, 29th May, 1851, when a * Vide "Archæologia," Vol. xxx., pp. 458 to 462.