the town walls, but near the river, from which its trenches were fed.' It was without doubt a motte-andbailey castle; the motte still remains, and the accompanying bird's-eye view by David Loggan, 1675, shows that the later stone walls of the bailey stood on the earthen banks of D'Oilly's castle. The site is now occupied by a gaol. On the line of the walls rises the ancient tower of St George's Church, which so much resembles an early Norman keep that we might think it was intended for one, if the Osney chronicler had not expressly told us that the church was founded two years after the castle. It is evident that the design was to make the church tower work as a mural tower, a combination of piety and worldly wisdom quite in accord Iwith what the chronicler tells us of the character of Roger d'Oilly.

Henry II. spent some £260 on this castle between the years 1165 and 1173, the houses in the keep, and the well being specially mentioned. We may presume that he built with stone the decagonal [shell?] keep on the motte, whose foundations were discovered at the end of the 18th century. There is still in the heart of the motte a well in a very remarkable well chamber, the masonry of which may be of his time. The area of the bailey appears to have been 3 acres.

The value of the city of Oxford had trebled at the time of the Domesday Survey.*

In the treaty between Stephen and Henry in 1153 the whole castle of Oxford is spoken of as the "Mota" of Oxford.5

1 Osney Chronicle, 1071.

1 See Ingram's Memorials of Oxford for an account of the very interesting crypt of this church, p. 8. The battlement storey of the tower is comparatively late. 3 Mackenzie, Castles of England, i., 160.

4

♦ D. B., p. 154.

Rymer's Fœdera, vol. i.

PEAK CASTLE, Derbyshire.-The Survey simply calls this castle the Castle of William Peverel, but tells us that two Saxons had formerly held the land. There is no motte here, but the strong position, defended on two sides by frightful precipices, rendered very little fortification necessary. It is possible that the wall on the N. and W. sides of the area may be, in part at least, the work of William Peverel; the W. wall contains a great deal of herring-bone work, and the tower at the N.W. angle does not flank at all, while the other one in the N. wall only projects a few feet; the poor remains of the gatehouse also appear to be Norman. It would probably be easier to build a wall than to raise an earthbank in this stony country; nevertheless, behind the modern wall which runs up from the gatehouse to the keep, something like an earthbank may be observed on the edge of the precipice, which ought to be examined before any conclusions are determined as to the first fortifications of this castle. The keep, which is of different stone to the other towers and the walls, stands on the highest ground in the area, apparently on the natural rock, which crops up in the basement. It is undoubtedly the work of Henry II., as the accounts for it remain in the Pipe Rolis, and the slight indications of style which it displays, such as the nook-shafts at the angles, correspond to the Transition Norman period. The shape of the bailey is a quadrant; its area scarcely exceeds 1 acre.

1 "Terram castelli Pechefers tenuerunt Gerneburn et Hunding." D. B i., 276a, 2.

2 There are similar nook-shafts to Henry II.'s keep at Scarborough, and to Castle Rising. Mr Hartshorne (Arch. Journ., v., 207) thought that there had been an earlier stone keep at Peak Castle, because some moulded stones are used in the walls, and because there is some herring-bone work in the basement. But this herring-bone work only occurs in a revetment wall to the rock in the cellar; and the moulded stones may be quite modern

PEAK-PENWORTHAM

183

The value of the manor had risen since the Conquest, and William Peverel had doubled the number of ploughs in the demesne. The castle only remained in the hands of the Peverels for two generations, and was then forfeited to the crown. The manor was only a small one; and the site of the castle was probably chosen for its natural advantages and for the facility of hunting in the Peak Forest.

1

It is

PENWORTHAM, Lancashire (Fig. 24).—“ King Edward held Peneverdant. There are two carucates of land there, and they used to pay ten pence. Now there is a castle there, and there are two ploughs in the demesne, and six burghers, and three radmen, and eight villeins, and four cowherds. Amongst them all they have four ploughs. There is half a fishery there. There is wood and hawk's eyries, as in King Edward's time. worth £3." The very great rise in value in this manor shows that some great change had taken place since the Norman Conquest. This change was the building of a castle. The modo of Domesday always expresses a contrast with King Edward's time, and clearly tells us here that Penwortham Castle was new. It lay in the extensive lands between the Ribble and the Mersey, which were part of the Conqueror's enfeoffment of Roger the Poitevin, third son of Earl Roger de Montgomeri. Since Penwortham is mentioned as demesne, and no insertions for repairs, and may have come from the oratory in the N.E. angle, or from some of the ruined windows and doorways. The sums entered to this castle between the years 1172 and 1176 are less than half the cost of Scarborough keep, and do not appear adequate, though the keep was a small one. But there is some reason to think that the cost of castles was occasionally defrayed in part from sources not entered in the Pipe Rolls. 1 Rex E. tenuit Peneverdant. Ibi 2 carucatæ terræ et reddebant 10 denarios. Modo est ibi castellum. . . . Valent 3 libras. D. B., i., 270.

2

2 We need not resort to any fanciful British origins of the name Peneverdant, as it is clearly the effort of a Norman scribe to write down the unpronounceable English name Penwortham, 3 See ante, under Clitheroe.

under-tenant is spoken of, we may perhaps assume that this castle, which was the head of a barony, was built by Roger himself. He did not hold it long, as he forfeited all his estates in 1102. At a later period, though we have not been able to trace when, the manor of Penwortham passed into the hands of the monks of Evesham, to whom the church had already been granted, at the end of the Conqueror's reign.1 Probably it is because the castle thus passed into the hands of the church that it never developed into a stone castle, like Clitheroe. The seat of the barony was transferred elsewhere, and probably the timbers of the castle were used in the monastic buildings of Penwortham Priory.

The excavations which were made here in 1856 proved conclusively that there were no stone foundations on the Castle Hill at Penwortham.2 These excavations revealed the singular fact that the Norman had thrown up his motte on the site of a British or Romano-British hut, without even being aware of it, since the ruins of the hut were buried 5 feet deep and covered by a grass-grown surface, on which the Norman had laid a rude pavement of boulders before piling his motte.3

1 Mr Halton's book (Documents relating to the Priory of Penwortham) throws no light on this point.

2 Transactions of the Historic Society of Lancashire and Cheshire, vol. ix., 1856-1857, paper on "The Castle Hill of Penwortham," by the Rev. W. Thornber; Hardwick's History of Preston, pp. 103-11.

3 In a paper published in the Trans. Soc. Ant. Scot. for 1900, on “AngloSaxon Burhs and Early Norman Castles," the present writer was misled into the statement that this hut was the remains of the cellar of the Norman bretasche. A subsequent study of Mr Hardwick's more lucid account of the excavations showed that this was an error. There were two pavements of boulders, one on the natural surface of the hill, on which the hut had been built, the other 5 feet above it, and 12 feet below the present surface. The hut appeared to have been circular, with wattled walls and a thatched roof. Several objects were found in its remains, and were pronounced to be Roman or Romano-British. The upper pavement would probably be the flooring of a Norman keep.

PENWORTHAM-PETERBOROUGH

185

Among the objects found in the excavations was a Norman prick spur, a conclusive proof of the Norman origin of the motte. No remains appear to have been found of the Norman wooden keep; but this would be accounted for by the theory suggested above.

Penwortham is a double motte, the artificial hill rising on the back of a natural hill, which has been isolated from its continuing ridge by an artificial ditch cut through it. The double hill rises out of a bailey court which is rudely square, but whose shape is determined by the ground, which forms a headland running out into the Ribble. The whole area cannot certainly be ascertained. There was a ferry at this point in Norman times. The castle defends the mouth of the Ribble and overlooks the town of Preston.

2

Penwortham was certainly not the caput of a large soke in Saxon times, as it was only a berewick of Blackburn, in which hundred it lay. It was the Norman who first made it the seat of a barony.

Peterborough. The chronicler, Hugh Candidus, tells us that Abbot Thorold, the Norman abbot whom William I. appointed to the ancient minster of Peterborough, built a castle close to the church, "which in these days is called Mount Torold." This mount is

1 Mr Roach Smith pronounced this spur to be Norman. As its evidence is so important, it is to be regretted that its position was not more accurately observed. It was found in the lowest stratum of the remains, but Mr Hardwick says: "As it was not observed until thrown to the surface, a possibility remained that it might have fallen from the level of the upper boulder pavement, 5 feet higher." We may regard this possibility as a certainty, if the lower hut was really British.

2 Mr Willoughby Gardner says the castle commands a ford, to which the ancient sunk road leads. Victoria Hist. of Lancashire, vol. ii.

3 Hugh Candidus, Canob. Burg. Historia, in Sparke's Scriptores, p. 63. This passage was kindly pointed out to me by Mr Round. Hugh lived in Henry III.'s reign, but he must have had the more ancient records of the monastery at his disposal.