Volume I: see https://www.gutenberg.org/ebooks/64187

Some characters might not display properly in this UTF-8 text file (e.g., empty squares). If so, the reader should consult the html version or the original page images noted above.

Transcriber’s note:

Text enclosed by underscores is in italics (_italics_).

Text enclosed by equal signs is in bbold face (=bold face=).

[Decoration]

MEDIÆVAL MILITARY ARCHITECTURE IN ENGLAND.

by

GEO. T. CLARK.

Vol. II.

With Illustrations.

... Time Has moulder’d into beauty many a tower, Which, when it frown’d with all its battlements, Was only terrible....—MASON.

London: Wyman & Sons, 74–76, Great Queen Street, Lincoln’S-Inn Fields, W.C.

1884.

CONTENTS OF VOL. II.

DOLFORWYN CASTLE, MONTGOMERY _Page_ 3

DOVER CASTLE, KENT 4

DUNSTER CASTLE, SOMERSET 24

DURHAM KEEP 32

EATON-SOCON CASTLE, BEDFORDSHIRE 36

EWIAS HAROLD CASTLE, HEREFORDSHIRE 39

EXETER CASTLE 44

FILLONGLEY CASTLE, WARWICKSHIRE 47

FONMON CASTLE, GLAMORGAN 49

FOTHERINGAY CASTLE, NORTHAMPTONSHIRE 50

GROSMONT CASTLE, MONMOUTHSHIRE 51

GUILDFORD CASTLE, SURREY 53

HARLECH CASTLE, MERIONETH 72

HASTINGS CASTLE, SUSSEX 82

HAWARDEN CASTLE, FLINTSHIRE 88

HELMSLEY CASTLE, YORKSHIRE 100

HEREFORD CASTLE 108

HERTFORD CASTLE 119

HOPTON CASTLE, SHROPSHIRE 123

HUNTINGDON CASTLE 126

HUNTINGTON CASTLE, HEREFORDSHIRE 129

KENILWORTH CASTLE, WARWICKSHIRE 130

KIDWELLY CASTLE, CAERMARTHENSHIRE 153

KILPECK CASTLE 162

KNARESBOROUGH CASTLE, YORKSHIRE 168

LEEDS, OR LEDES, CASTLE, KENT 176

LEICESTER CASTLE 182

LEYBOURNE CASTLE, KENT 188

LINCOLN CASTLE 189

LLANQUIAN TOWER, GLAMORGAN 201

LONDON, TOWER OF 203

LUDLOW CASTLE, SHROPSHIRE 273

ST. LEONARD’S TOWER, WEST MALLING 291

MIDDLEHAM CASTLE KEEP, YORKSHIRE 293

MITFORD CASTLE, NORTHUMBERLAND 300

MONTGOMERY CASTLE 303

MORLAIS CASTLE, GLAMORGAN 312

NORHAM CASTLE, DURHAM 322

NOTTINGHAM CASTLE 336

ODIHAM CASTLE, HANTS 336

OSWESTRY, SHROPSHIRE 346

PENMARK CASTLE 351

PENRICE CASTLE, IN GOWER 353

PENRITH CASTLE, CUMBERLAND 357

PEVENSEY CASTLE, SUSSEX 359

PICKERING CASTLE, YORKSHIRE 368

PONTEFRACT CASTLE, YORKSHIRE 375

PORCHESTER CASTLE, HANTS 388

RICHARD’S CASTLE, HEREFORDSHIRE 401

ROCHESTER CASTLE 405

ROCKINGHAM CASTLE, NORTHAMPTONSHIRE 423

OLD SARUM, WILTSHIRE 447

SCARBOROUGH CASTLE 458

SKENFRITH CASTLE 467

SOUTHAMPTON, THE ANCIENT DEFENCES OF 472

TAMWORTH CASTLE, WARWICKSHIRE 481

TAUNTON CASTLE 488

THURNHAM CASTLE, KENT 492

TICKHILL CASTLE, YORKSHIRE 494

TRETOWER, BLAEN-LLYFNI, AND CRICKHOWEL CASTLES 499

TUTBURY CASTLE, STAFFORDSHIRE 505

URQUHART CASTLE, INVERNESS-SHIRE 509

WAREHAM, DORSETSHIRE 513

WHITE CASTLE, MONMOUTHSHIRE 517

WHITTINGTON, SHROPSHIRE 521

WIGMORE, HEREFORDSHIRE 526

YORK, THE DEFENCES OF 534

ILLUSTRATIONS TO VOL. II.

DOLFORWYN CASTLE, Plan and Sections _face p._ 3

DOVER CASTLE, View „ 16

DOVER KEEP, Plans of First Floor and Basement „ 17

Ditto Plans of Second Floor and Second Floor Gallery „ 19

DUNSTER CASTLE, Ground Plan „ 25

EATON-SOCON CASTLE, Ground Plan „ 36

FILLONGLEY CASTLE, Ground Plan „ 48

FONMON CASTLE, Ground Plan and Elevation „ 49

GUILDFORD CASTLE „ 69

HARLECH CASTLE, Bird’s-eye View „ 72

Ditto General Plan „ 73

HASTINGS CASTLE, Ground Plan „ 83

HAWARDEN CASTLE, Ground Plan 92

Ditto Spur-work enclosing the Main Entrance 97

HELMSLEY CASTLE, Ground Plan 101

Ditto Plans and Section of Keep 102

HERTFORD CASTLE, Ground Plan 120

HOPTON CASTLE, Ground Plan 124

HUNTINGTON CASTLE, Ground Plan 129

KENILWORTH CASTLE, Plans of Mortimer’s Tower and of First and Ground Floors of Keep _face p._ 141

Ditto General Plan „ 145

KIDWELLY CASTLE, Bird’s-eye View „ 153

Ditto Ground Plan „ 154

Ditto General Plan „ 155

Ditto Elevation and Section „ 159

Ditto Gateway and Main Entrance „ 160

Ditto Interior of Chapel „ 156

KNARESBOROUGH KEEP, Plan of Dungeon and Elevation of Keep „ 171

Ditto Plans of Main Floor and Basement „ 172

LINCOLN CASTLE, Ground Plan „ 192

LONDON, TOWER OF, Ground Plan in 1866 206

Ditto The Keep, Third Stage 210

Ditto Fireplace in Keep 211

Ditto St. John’s Chapel—South Aisle 212

Ditto The Keep, Upper Stage 213

Ditto Ditto Vertical Section, East and West 214

Ditto Bloody and Wakefield Towers, Plan of Basement 221

Ditto Wakefield Tower, Palace Entrance and Oratory 223

Ditto Bell Tower. Plan of Basement 227

Ditto Ditto First Floor 228

Ditto The Curtain, Ground Plan 230

Ditto Devereux Tower, Plan of Basement 232

Ditto Salt Tower—Plan of Basement 236

Ditto Ditto First Floor 237

Ditto Well Tower—Plan 239

Ditto Cradle Tower, Plan of Basement 239

Ditto Ditto Window 240

Ditto St. Thomas’s Tower, Plan of Basement 241

Ditto Ditto Detail of Ring Stones 242

Ditto Ditto Piscina 243

Ditto Byward Tower and Postern, Plan of 245

Ditto Byward Tower, Plan of First Floor 246

Ditto Middle Tower, Ground Plan 248

Ditto Ditto First Floor 249

LUDLOW CASTLE, Ground Plan _face p._ 274

Ditto The Keep, Ground Floor „ 277

Ditto Ditto First Floor „ 278

MONTGOMERY CASTLE, Ground Plan „ 309

MORLAIS CASTLE, Ground Plan and Sections „ 313

NORHAM CASTLE, Ground Plan „ 324

Ditto Plan of Basement, Keep „ 326

NOTTINGHAM CASTLE, in the Sixteenth Century, Bird’s-eye View „ 336

ODIHAM CASTLE, Ground Plan 337

OLD SARUM, Ground Plan _face p._ 449

PENMARK CASTLE, Ground Plan 352

PENRICE CASTLE, Ground Plan _face p._ 353

PENRITH CASTLE, Ground Plan „ 358

PICKERING CASTLE, Ground Plan and Sections 373

PORCHESTER CASTLE, Plans _face p._ 389

RICHARD’S CASTLE, Ground Plan and Section „ 403

ROCHESTER CASTLE Keep, Plans of Basement and First Floor „ 410

Ditto Plans of Main Floor and Main Floor Gallery „ 412

Ditto Sections of Mouldings 413

ROCKINGHAM CASTLE, Ground Plan _face p._ 426

SCARBOROUGH CASTLE, Ground Plan „ 462

Ditto The Keep „ 464

Ditto Ditto South Face „ 465

Ditto Ditto Interior „ 466

SKENFRITH CASTLE, Ground Plan „ 470

TRETOWER CASTLE, Elevation and Ground Plan „ 500

WAREHAM, Ground Plan and Section „ 513

WHITE CASTLE, Ground Plan „ 517

WHITTINGTON CASTLE, Ground Plan „ 523

MEDIÆVAL MILITARY ARCHITECTURE.

DESCRIPTIONS.

“ . . . . . Time Has mouldered into beauty many a tower, Which, when it frowned with all its battlements, Was only terrible.” . . . . —MASON.

DOLFORWYN CASTLE, MONTGOMERY.

Dolforwyn, or “The Maiden’s Meadow,” is a name evidently transposed from the meads of the adjacent Severn to the ridge occupied by the castle, which rises 500 feet or 600 feet above, and half a mile west of, the river, from which it is separated by an intervening hill. The approach is by a steep road, which becomes still more so near the top of the ridge, and finally skirts along, and is commanded by, the works of the castle.

These works are very simple in plan, and of rude construction. A platform about 200 yards long by 100 yards broad occupies the centre of the ridge. Its rocky sides are scarped and revetted all round to a height of about 10 feet, and upon this wall was built a curtain of from 20 to 30 feet more, and about 5 feet thick. At each end a cross ditch was quarried in the rock, so as to isolate the castle from the equally high ground beyond. Probably there were no bridges across these ditches, and the entrance seems to have been by a plain doorway in the curtain upon the northern face of the works. The curtain appears to have been quite plain, without either buttress or pilaster or flanking tower, save at the eastern end of the area, near the centre, where are the remains of a circular tower about 30 feet in diameter with walls 5 feet thick. The curtain to the south, or most exposed, side is broken away; on the opposite side it is more perfect, and contains a doorway, broken, and now a mere hole in the wall. Within is a fragment of a building into which probably the gateway opened. The platform is very irregular, partly natural, chiefly from the heap of rubbish covering up the foundations of the domestic buildings. The building is not unlike Dinas Brân and Dinas Powis, and is probably of the age of Henry III. or Edward I., early in the reign. The material is the tile-stone of the country laid in courses. There is no sign of ashlar.

Dolforwyn has no history. All that is known is that it was granted by Edward I. [7 Edward I.] to Roger Mortimer of Wigmore as “the Castle of Dolvoron,” with the territories of Keddewy and Kery, to be held by the service of three knight’s-fees. In 14 Edward I. the castle was still held by a Mortimer, for Richard Labaunk was in prison at Wigmore, by reason of arrears in his account to Edmund de Mortimer whilst constable of his castle of Dolfnovan. He was liberated on bail. In 18 Edward I., Bogo de Knovill, being constable of Montgomery Castle, had a pardon for £90 due on his farm of lands of Kery and Kidgewenny. This, however, was from the king, who seems to have resumed possession. Dugdale says the castle was built by David ap Llewelyn, who flourished 1240–46, but the Welsh attribute it to Bleddyn ap Cynfin between 1065 and 1073. Bleddyn may have had some kind of stronghold here, as a place very convenient for a raid upon the flat country, then held by the English; but he certainly did not build the existing walls. These are not unlikely to have been the work of Roger Mortimer, and their destruction probably followed at the first convenient opportunity. The name of the castle has not been found in the Mortimer inquisitions, nor is it mentioned save as above, among their possessions, or those of any other landowner. After the settlement of Wales it would cease to possess any value.

DOVER CASTLE, KENT.

“Est ibi mons altus, strictum mare, litus opacum, Hinc hostes citius Anglica regna petunt, Sed castrum Doveræ, pendens a vertice montis, Hostes rejiciens, litora tuta facit.” —DE BELLO HASTINGENSI CARMEN, i. 603.

The tract of chalk which forms and gives character to the isle of Thanet and the south-eastern portion of the county of Kent, rises towards the sea to a line of cliffs upon which the promontories of the North and South Foreland and of Dover are the most conspicuous. The cliff line, however, is not continuous. It is broken at intervals by various valleys and gorges, down which the waters from the interior find their way to the sea, producing havens which in former days and for vessels of light burthen were much in request. Of these waters the chief is the Stour, which at no very remote period, near to the present Canterbury, fell into the head of a considerable estuary, the waters of which, guarded by the ancient fortresses Regulbium, Rutupiæ and Lemanis, maintained Thanet as an island, and gave to the trade of the period free access to the interior of the district. This estuary was not exempt from the general tendency to become silted up. Before the Norman Conquest the waters had receded from Canterbury to the parish thence named Stourmouth, and Thanet from an island had become a peninsula. The process of silting up has been since continued, and the mouth of the Stour carried many miles lower. The river, after a very winding course, falls into and in part forms what remains of the ancient English port of Sandwich, opposite to the anchorage known as the Small Downs.

About three miles to the west of the South Foreland, another and much smaller stream, fed from the lower chalk and greensand, flows down a deep valley, and, reaching the sea between two considerable heights, has given origin to the port, town, and castle of Dover, so called, without doubt, by derivation from a British name represented in the Roman times by Dubris. The town and port lie deep in the valley. Of the heights, a part of that to the west has been rendered famous by Shakespeare, and has long borne his name. That to the east is known as the Castle Hill, so called from the fortress by which, under some form or other, it has been crowned from a very remote period. A position so convenient and so capable of defence would, upon any shore, have attracted the notice of the very earliest inhabitants; for not only was the height strong and the port convenient, but these advantages were found at the point at which the island approached nearest to the Continent, and at which those who crossed the straits, whether as friends or foes, would first make the land, and, if not obstructed, would come ashore. The position, therefore, was of far more than local importance, and would be sure to receive the attention not only of the chiefs of the Cantii, but of the rulers, if such there were, of the whole of Britain. It is therefore probable that haven, town, and fortress date from very nearly the first settlement of the country.

Although the Western or Shakespeare’s Cliff is part of a larger range, the Castle Hill is better suited for defence. It is, in fact, an isolated knoll about 1,000 yards north and south, and 500 yards east and west, the summit being a steep and narrow ridge. Towards the south its boundary is the sea cliff, 320 feet high, and to the west the deep valley of the town. To the east and north are other valleys, less deep but by no means inconsiderable, and the sides of which are steep. Moreover, the whole hill is of chalk, that is to say, of a material easily scarped and capable of retaining any general outline to which it may be cut.

But though, on general grounds, a very remote antiquity may safely be attributed to both town and fortress, it is difficult to find any precise or special evidence on which to rest the claim. Here, as at Durovernum or Canterbury, the Roman form indicates a British origin, and, if the ancient name of the stream be indeed, as asserted, the Dour, may well be derived from it, and the Castle Hill is just the place upon which a British camp is likely to be found. The commerce of the Britons, known to have been carried on with activity through the Cornish ports and the Isle of Wight, has also been claimed for the route through Dover. The actual present traces of British occupation in this southern country are indeed very scanty, and confined to a few names of rivers and hills, a very few of towns or villages, and to occasional entrenchments upon high ground, and of an irregular outline. The great roads, whatever their remote origin, in their present form carry the stamp of Rome upon every mile of their course, and the oldest known works in masonry are due to the same people, while the general topography, all that relates to property and self-government, hundreds, lathes, rapes and tythings, parish and hamlet, grange and farm, and the crowd of bourns, dens, hams, hangers, hirsts, ings, tons, wolds and worths point with overwhelming force to the English settlers. Even the tenure in gavelkind, claimed as a British custom, and known in Wales by the expressive name of “Randyr” or “partible” land, is by most legal antiquaries regarded as Teutonic.

Although Cæsar does not mention Dover by name, there can be no doubt that his fleet lay before it in August, B.C. 55, the period of his first invasion of Britain. Dr. Guest has clearly demonstrated that the Portus Icius whence he sailed was a small and now silted-up haven between Cape Gris Nez, the Ician promontory, and Wissant, whence a ten hours’ course brought him in the morning abreast of Dover. Here he found the natives, in great numbers and armed, drawn up to oppose his landing. He therefore anchored in Dover Wick, the roadstead east of the town, to give time for his slower ships to arrive, and thence proceeded to Deal, where he probably landed with two legions, or from eight to ten thousand men. Dr. Guest has pointed out that the word Icius coincides closely with the Irish name for the English Channel, “Muir n’Icht,” “the Ician Sea,” “icht” being a form of “uch” or “ucha,” upper in height, which plays so important a part in the names of places in Wales and the north of Scotland. It was natural that the Channel should be named from its most remarkable feature, and to this day its name in Dutch is “De Hofden,” or “the heights.”

Cæsar stayed but three weeks in the country, and may not have visited Dover; but as, when he returned in the following year with a much larger force, he seems to have embarked and landed at the same points, he must have been familiar with the aspect of the port from the sea. As on this occasion he traversed Kent and crossed the Thames, he probably left no dangerous force behind him at Dover, which he does not mention, and which clearly was not then made a rallying point by the Britons. Had the heights been held in force, he would probably in the first instance have reduced them, or at any rate have mentioned the fact in his narrative. During the century that followed Cæsar’s appearance, Rome took no active part in British affairs, but it is probable that a considerable trade sprang up between the island and the Continent, and the Britons made great advances in commerce and civilisation. Towns were founded and coins struck. The next military invasion took place, A.D. 43, ninety-eight years after the first appearance of Cæsar, and under the reign of Claudian, and command of Aulus Plautius, who landed with four legions. The Cantii then held a tract nearly corresponding to the present county of Kent. Durovernum had been founded amidst its indigenous alders, and Camalodunum, beyond the Thames, was the chief city of the Trinobantes. Where Plautius landed is not precisely known. Probably at several points on the open beach between Rich borough and Laymen. Whether the Britons mustered north or south of the Thames is also unknown; but, when Claudian followed his lieutenant, the way lay open to that river, and he marched at once upon Camalodunum, where the Trinobantes were put to flight. Plautius probably subdued the country as far west as the Axe and the Tamar, and his progress may, it is thought, be traced by the remains of his rectangular camps opposed to those of larger area and irregular outline thrown up by the retiring Britons. As, in the year A.D. 40, Caligula had caused a lighthouse to be set up on the heights of Boulogne, it is not improbable that Plautius was the builder of the corresponding tower at Dover.

Ostorius Scapula is said to have occupied with a camp the Castle Hill. He, Suetonius Paulinus, and Agricola were busied mainly with the midland and northern parts of the island, and the southern province, Britannia Prima, seems to have been at peace. Roads were laid out, towns built, the metals were smelted, and agriculture prospered. Dubris (Dover), Durobrivis (Rochester), Rutupiæ (Richborough), Lemanis (Lymne), Regulbium (Reculver), and Anderida (Pevensey), came afterwards into notice as towns or havens. The Watling Street, which ran from Canterbury by Rochester northwards, seems to have been commenced at Dover. There were indeed roads from Canterbury to Lymne and to Richborough, but Dover would be the port reached by the production of the road in a straight line from Canterbury. In the reign of Valentinian, A.D. 364–7, a cohort of the Second Legion of 1,100 men was stationed at Dubris, where stamped tiles show them to have built a bath, and which is mentioned as a port in the Iter of Antoninus and as a town (civitas) by the geographer of Ravenna. In the time of Constantine, Dubris was one of the six ports south of Thanet, under the Count of the Saxon shore, the others being Rutupiæ, Regulbium, Lemanis, Anderida, and Adurnus (Portsmouth). It is, however, not included in the list of the twenty-eight towns existing when the Romans retired from Britain, of which Rutupiæ was one. Still, even if Dubris were one only of three heads of the Watling Street, its importance under the Roman sway was considerable.

To the Roman period is to be referred the burial-ground laid open near the edge of the cliff in 1797, and the bath discovered on the brook west of St. Mary’s Church. No mention is made of a Roman fortress, nor was it in accordance with the practice of that people to place a permanent camp, still less a military station, upon a height so inaccessible as the Castle Hill. The existing earthworks show no traces of Roman outline, nor, when they had possession of the whole district, was there any need to fortify the lighthouse. The lighthouse alone, of the works upon the hill, can with certainty be pronounced to be Roman, but this, of course, implies the existence and employment of the port. All that the topographer can affirm is that the earthworks do not now present, and, so far as description may be relied on, do not appear ever to have presented, anything of a Roman character.

Kent was probably the part of Britain first invaded by the Northmen, and certainly the first actually subdued and settled. It was the only independent state established and maintained by the followers of Hengist, the Jutes, a people who did not contribute largely to the conquest of Britain, neither did they occupy any considerable portion of the conquered country; but what they did retain became and still remains intensely Teutonic, and their early supremacy, during the conversion of the English to Christianity, is marked by the fact that their chief city became, as it has since remained, the ecclesiastical metropolis of the island. Dover was a considerable Jutish port, and before long was regarded as the key of England. Very probably the inner earthworks still to be seen, though too much altered to be recognisable, were the work of this people, and the collegiate church of St. Martin, founded in the town by Wihtræd, King of Kent (690–725), is said to have been removed by him from the Castle Hill. In the time of Alfred, Dover was placed in the bailiwick of Stouting, and the lathe of St. Augustine. Its history, however, properly so called, does not begin till the reign of the Confessor, whose charter to the Cinque Ports, which John is said to have inspected, was confirmed by many later kings.

In September, 1051, Eustace, Count of Boulogne, brother-in-law to King Edward, paid a visit to the English Court at Gloucester, and returning through Canterbury there rested his escort, and thence went to Dover on his way homeward. Before entering the town, he and his men put on their coats of mail, and attempted to take free quarters in the houses of the burghers. This led to a fight in which much blood was shed, and the Count finally, being expelled the town, returned to Gloucester with his complaint. The subsequent tale has often been told. Godwin, then Earl of Kent, took part with his injured countrymen, and withstood the strong Norman interest about the king, and was, in consequence, banished. Godwin proposed, says Malmesbury, “Ut magnates illius castelli blande in curia regis de seditione convenirentur.” Whether “castellum” can be taken for more than the fortified town is uncertain. It is not probable that Eustace would have ascended to the castle, since he sought quarters in the town. On the whole this passage can scarcely be taken to prove the existence at that time of a regular castle on the hill. Nevertheless, the existence of such a castle at that time is exceedingly probable, for in 1064–65 Harold, says William of Poitiers, swore to Duke William that, on the death of the Confessor, “se ... traditurum interim ipsius militum custodiæ castrum Doveram, studio atque sumptu suo communitum, item per diversa loca illius terræ alia castra.” Eadmer is more precise. In his account Duke William insists, “et insuper castellum Dofris cum puteo aquæ ad opus meum te facturum.” Malmesbury says, “Castellum Doroberniæ (Dubris) quod ad jus suum pertineret.” Here there can be no mistake as to what is meant. A well in the town could be of no special value, but a well on a chalk hill 290 feet above the water springs was an addition to the castle worthy of special notice. Such a well, moreover, was a very laborious work, and must have taken some time to complete. Harold’s oath is no doubt involved in a good deal of doubt and uncertainty, but the tale may at least be taken to show that there was, before the Norman conquest, a castle upon the hill, now crowned by the Norman keep, and that it had been strengthened by Harold. Malmesbury here, as in his account of the fray with Count Eustace, for Dover puts Canterbury, evidently in error.

Domesday Book opens with Dover. “Dovere tempore regis Edwardi,” &c. That king held two parts of half its rents, and Earl Godwin the other third, that is, the earl’s penny. The burgesses provided the king with twenty ships annually for fifteen days, in each twenty-one men. The king’s messengers also had certain valuable privileges, showing that the port lay in the usual route to the Continent. There was a mill at the entry to the port, much in the way of the shipping. The castle is not mentioned, but we read of the Gildhalla, or Guildhall, of the burgesses. The town, though thriving, had recently been burned, and the rent reduced in consequence.

The Normans landed, as their Teutonic ancestors had landed centuries before them, beneath the ancient walls of the Roman Anderida, which, under the name of Pevensey, had become an English, and was to become ere long a Norman fortress. Pevensey is a haven no longer, but within its circuit may still be seen huge fragments of Roman and Norman masonry, and the simpler, but, at least, equally durable mound of its English occupants, a grand and striking composition, and, to the instructed observer, eloquent of great events. William took possession of the ruins, and, on the following day, marched upon the battlefield, hastily fortifying with wood and earth the hill of Hastings in support of his position. After the battle Romney first felt the weight of the Conqueror’s hand, and he then turned to Dover. The castle was, says William of Poitiers, impregnable, and the town even then considerable; but the people, though assembled in vast numbers, had no leaders, and town and castle were at once surrendered, though in the transfer, either by design or accident, much of the town was burned. William paused here eight days, detained by sickness among his troops. He treated the people with great lenity, placed a Norman garrison in the castle, to the defences of which he added, and then proceeded towards the Thames, receiving the submission of Canterbury and Rochester on his way. Kent was placed under the command of Bishop Odo, who held Dover as its military centre. This continued to be its position, and in 1074, when the fierce Jutish blood broke out, and the men of Kent, headed by Eustace of Boulogne, rose against William, their first object was to gain Dover Castle, to which they laid siege. Its defences, whether English or Norman, were strong, but the attack was sharp. Bishop Odo and Hugh de Montfort, who had the castle in charge, were absent, but they exerted themselves in the county, and the castle was relieved. Towards the end of the reign the town is said to have been walled, and to have had ten gates. Its position, as the most important of the Cinque Ports, seems to have been established, to which its castle largely contributed. The five ports were Hastings, Romney, Hythe, Dover and Sandwich. Winchelsea and Rye were among the members. Their service was to provide fifty-seven ships annually, of which number Dover furnished twenty-one.

Important as the castle continued to be under the reigns of the Plantagenet and Tudor kings, its history does not possess any very particular interest. It was not the scene of any very remarkable event, and, though accounted the key of the kingdom, was not in any very intimate manner bound up with its history. It has always been held by the Crown and governed by a constable, usually a man of eminence, of which officials Lyon gives a list of 138. Three barons of the house of Fiennes held the office under the Conqueror, Rufus, and Henry I., and in their time seem to have been built the outer curtain and many of its towers. Walchelin Magminot, placed in office by Stephen, held the place against him in 1137, and in 1138 surrendered it to the queen, just before the battle of the Standard. Stephen died at Dover, probably in the castle, and Mr. Puckle has fought gallantly for the recent discovery of his remains. Henry II. is supposed to have built the keep and the wall of the inner ward in 1154, soon after his accession to the crown, and, if so, it was probably the actual work of one of the Barons Fiennes, who held the office of constable in that reign. Of these lords, the last, James Fiennes, was constable at the accession of Richard I., and in 1191 received, as a prisoner in the castle, Geoffrey, Henry II.’s natural son, on his way to take up the archbishopric of York. He was taken from the sanctuary of St. Martin’s church, and imprisoned by order of the Bishop of Ely, the chancellor. For this outrage Fiennes was suspended, and eventually the chancellor was excommunicated and banished. In 1198 the constable was one of the five officers appointed to inspect the treasures of the church of Canterbury. In the reign of John, the constable was Hubert de Burgh, to whom it fell to defend the castle against Prince Louis and the French invaders. Louis made great attempts to win over De Burgh to his party, but without success, and upon his final refusal he laid siege to the castle. _Trebuchets_ and _petraria_, and much siege artillery were brought over from France, and a covered way was run along the slope outside the castle ditch on the north-west quarter. Wooden towers (_Malvoisins_) also were erected on the edge of the ditch to quell the fire from the walls, which were actually shaken by strokes of the ram. Also a small colony was established in temporary huts, so as to give the siege the aspect of a blockade. Hubert, however, was a true and loyal subject of England. He regarded his charge as held under his sovereign, whoever that sovereign might be, and his castle as the “clavis Angliæ et repagulum.” He returned attack for attack, stone for stone, until the death of John, by removing much of the cause of dissatisfaction, led to the retirement of the French from the country. The great spur-work in advance of the northern gateway, and which still, though much altered, remains, is probably the work of De Burgh, no doubt suggested to him by the direction of the French attack. Henry III. was here in state in 1255, on his way from Spain, and in 1259 the castle was in the hands of the barons, who refused to allow his brother Richard, King of the Romans, to enter the place until he had sworn to adopt their cause. Henry recovered the castle in 1261 from Hugh le Bigot, but was unable to hold it, and was refused admission when, with his brother and the Earl Mareschal, he presented himself before the gate in 1263. In the great struggle, Henry held the Tower of London, and the barons the castle of Dover. After Lewes, the Princes Edward and Henry were for a time held in durance in Dover. In 1265, when the prince had the upper hand, and Kenilworth had fallen, fourteen nobles of the royal party were imprisoned in the keep of Dover. Here they defended themselves, turning the tables upon their captors, who, attacked from without by Prince Edward, and thus placed between two fires, surrendered the castle. The Countess of Leicester, who was within it, was allowed to retire to France.

On his return from Palestine, after his father’s death, in the summer of 1274, Edward I. landed at Dover. He again visited the castle in 1278, 1296, and 1299. Dover was his usual port when he visited or returned from the Continent, and in his time the castle was maintained and strengthened, and some of its most considerable parts, such as the Constable’s Gate and St. John’s Postern, are of the conclusion of the reign of Henry or early in that of Edward, and most probably the latter.

Edward II. was here in the January following his father’s death, and here he embarked for Boulogne on the occasion of his marriage in 1308, and here he received the queen. Edward III. was not infrequently at Dover, but is more likely to have lodged in the Maison Dieu than in the castle. He probably drew troops from hence when he sailed from the port on his celebrated secret expedition to Calais in 1348. In the reign of Henry V. the Emperor Sigismond was received at the castle as a visitor, and from hence the king embarked his army for France in 1421, as did Henry VII. in 1491.

Henry VIII. suppressed the Maison Dieu, the celebrated foundation of Hubert de Burgh, within which very many kings and princes had been entertained, but the castle was an object of his special care, and was repaired and garrisoned. Henry also built the blockhouses of Sandown, Deal, Walmer, and Sandgate, and placed them under the charge of the Constable of Dover. Three bulwarks also were constructed under the cliff and upon the pier of the harbour. Later on the castle fell into neglect, and in the reign of Charles I., being garrisoned by a small force, and but little cared for, it was taken by surprise and held for the Parliament. In 1648 the Kentish royalists made a vigorous effort to recover it, but were repulsed by Algernon Sydney, then its governor.

DESCRIPTION.

Dover Castle is called by Matthew Paris “the very front door of England,” and described by William of Poitou as “Situm est id castellum in rupe mari contigua quæ naturaliter acuta undique ad hoc ferramentis elaborate incisa, in speciem muri directissima altitudine, quantum sagittæ jactus permetiri potest consurgit, quo in latere unda marina alluitur.” It presents a good combination of the defences of several architectural periods, the general result being a concentric fortress, the growth of many centuries, and which, a century ago, presented much both of earthwork and masonry of great and unmixed antiquarian interest. More recently, however, the works have been delivered over to the military authorities of the country, and the result has been a series of alterations, additions, and removals, necessary, it is in all courtesy to be presumed, for the defence of the country, but very destructive of the ancient features of the fortress. The ancient earthworks have been scarped, extended, retrenched, and tunnelled, barracks and magazines have been built, the keep has been converted into storerooms and water-tanks, and in its basement are two powder magazines. In its present condition, and having regard to the strict regulations, prohibiting even the use of the pencil, under which the whole is placed, an accurate examination of what remains of the ancient works, whether in earth or masonry, is almost impracticable.

The British camp was oblong, following the figure of the hill. It was composed of a deep and broad ditch, the contents of which were in part thrown inwards and upwards so as to form a bank. The area thus enclosed measured within the bank about 875 yards by 350 yards, the latter being its diameter at the cliff, by which it is cut off, the northern end running to a point. This main ditch has probably been deepened and more or less altered during the Norman period, and it is now connected with various bastions, hornworks, and caponnières, but its general outline is sufficiently clear, and it may, from internal evidence, be presumed to be British. Within the area, rather nearer to its northern end, a second earthwork, also composed of bank and ditch, has been thrown up. This includes the lighthouse tower, and is therefore called the Roman ditch, but there is nothing about it of a Roman character, and it is far more probable that it was the work of the English, and formed the inner defence of the castle which Harold undertook to surrender. The space between this and the great earthwork forms the outer ward of the Norman castle; the inner work has been divided into two parts. That to the south, containing the lighthouse, forms the middle ward; that to the north contains the keep, and is the inner ward. The lighthouse has been employed as a belfry to the adjacent church of St. Mary, which is the subject of an excellent memoir by Mr. Puckle, though he can scarcely be admitted to establish its connexion with the British church. The two are undoubtedly the oldest buildings within the castle area.

The _Keep_ and _Inner Ward_ are Norman, of the reign of Henry II., but the curtain and most of its towers have been refaced or rebuilt, a great part recently, but more or less on the old lines. The plan is an irregular polygon, about 120 yards each way, with fourteen rectangular mural towers of no internal projection, and not rising above the curtain, which, however, is lofty. Two of those to the south-west take their name from Magminot and one from Gore. The walls were of flint rubble, quoined with ashlar, and battering outside at the base, the top of which is marked by a bold cordon of stone. Upon the sides of some of the towers are bold machicolated openings from garderobes. The keep stands detached in the centre of the ward, and within the area, built against the walls, are several buildings, as Arthur’s Hall, the guard chamber, and the officers’ quarters, some of early English character, others modern.

There are two entrances, the north, or King’s Gate, and the south, the Duke of Suffolk’s, or Palace Gate. The gateways are vaulted passages, with a flat segmental arch, opening externally between two square flanking towers. There is a groove for a portcullis, and the vaults have an early English import. Each of these gates is defended by an additional work. A sort of hornwork is thrown out in advance, enclosed within a wall with towers, and with a second gate placed obliquely to the first, to allow the approach to be commanded. The north, called the King’s Gate, is tolerably perfect. The southern outwork is nearly destroyed. It had two gates, one to the south, connected with which was the tower containing a well, and one to the east, called King Arthur’s Gate, close to which was the Armourer’s Tower, and near this was Earl Godwin’s Postern. The masonry of this inner ward was probably of the date of the keep. It was included within a broad and deep ditch, now incomplete, on the southern front.

The _Middle Ward_ includes the southern half of what was probably the English earthwork. Its southern two-thirds is rounded, but at its base are, or were, walls and towers connecting it right and left with the curtain of the outer ward, which it thus divided into two parts. In advance of this work were three towers: two, to the east and west, Clinton and Mortimer, were square. Both seem to have been of Norman date. The central tower to the south was circular, and was called Valence. It was, no doubt, later, probably of the reign of Henry III. The foundations of Clinton Tower were laid open and removed in 1794. A fragment of Mortimer’s Tower remains. Valence Tower was destroyed in the last century. The gatehouse of the ward is called Colton Gate. It is Norman, but an octagonal story has been added to the square base. The curtain connecting these towers is gone, and the lofty south bank has been scarped, and its outline changed. In this ward is a well 380 feet deep, once covered by a tower.

The _Outer Ward_ is contained within a curtain, much of which is reduced by an internal ramp to a parapet wall. Upon the three landward sides it is narrow, and chiefly occupied by the ditch of the middle and inner wards. Upon the remaining or southern side it expands and includes a large space between the middle ward and the cliff. Its circuit is only not complete because its walls rest at either end upon the cliff. The plan of the wall is irregular, with an occasional angle or shoulder for raking the ditch. Exteriorly it rises from the outer dry ditch, and upon it are twenty-seven towers of various dates, figures, and dimensions, square, circular, and multiangular. Most of those on the western face are simply hollow bastions, and have no internal projection. Those on the eastern face are mostly mere sentry-boxes, or bartizan turrets of small dimensions. Upon the seaward front, where the cliff is perpendicular, there is no wall. The ingenuity of the engineer has been exercised on the landward and weaker sides. Five of the towers are connected with gates. These towers, commencing at the south-west angle of the ward, near the cliff, are: 1. Canon’s or Monk’s Gate, now destroyed; in it was a well. 2. Rokesley’s Tower, semicircular. 3. Fulbert de Dover’s Tower, square; near which was long the office of the “bodar,” or sergeant-at-arms, to whom all civil warrants of arrest for debt or breach of the revenue laws were addressed. Fulbert’s Tower was also his prison. 4. Hirst’s Tower, semicircular; it commands a shoulder or re-entering angle in the wall. 5. Arsick’s Tower, semicircular. 6. Gatton Tower. 7. Peverill’s, Beauchamp’s, or Marshall’s Tower. This is also the gatehouse between the two divisions of the outer ward, which was strong, with a ditch on the south front, and a drawbridge. It was also the marshal’s prison. 8. Port, Gosling, or Queen Mary’s Tower, having been repaired by that sovereign. 9. Fiennes’s Tower, or the Constable’s Gate. This is one of the grandest gateways in England. It is in plan a triangle with its obtuse angle presented to the field. The angles at the base fall within the line of the curtain, and are capped by two large drum towers. The salient angle in like manner is capped by an oblong tower, rounded at each end and flat in the centre, through which the entrance passes. These three towers are large and lofty, and are connected by an embattled curtain. Within the triangle a central tower rises to a still greater height, and commands the whole. The entrance passage is broad and vaulted, and provided with gates and a portcullis. Within, it opens upon the level of the outer ward; without, it terminates abruptly upon the scarp of the ditch, there about 50 feet deep. From this gateway a bridge communicates with the opposite bank. A single lofty pier rises from the centre of the ditch, and from it an arch springs to the outer abutment, carrying a regular roadway and parapets. In the opposite direction the parapets alone spring, as two arches, from the pier to the gateway, and serve to steady the pier, but the roadway is omitted, and its place supplied by a drawbridge. This arrangement is not uncommon, but is here specially necessary, owing to the height and consequent weakness of the pier. Upon the counterscarp of the ditch is a _tête-de-pont_, from which a steep road descends by a traverse towards the town. This gate, though open, is but little used. The ordinary gate is modern, and near the site of Canon’s Gate. The Constable’s Gate is of the Decorated period. Its interior is said to be very curious, but is not shown.

Sixteen towers, including Clopton, Godsfoe, Magminot (4), and Crevequer’s Towers (2), protect the north-west face. Clopton is a hexagon; the name of Magminot is borne by four towers. The two towers bearing the name of Crevequer mark the position of the great postern, a very curious work. Passing from the north gate of the inner ward, a range of arches cross the ditch and the outer ward, and terminate abruptly in a large low pier with salient angles to the right and left. Opposite to the pier, and no doubt at one time connected with it by a drawbridge, rise a pair of circular towers (Crevequer), connected by a heavy curtain and flanked by lesser towers (Magminot) at short distances, all forming part of the _enceinte_ of the outer ward. Towards the field the curtain has a salient angle, and from its base a covered gallery descends into the outer ditch, and there reaches St. John’s, a drum tower built in the middle of it. The gallery passes through the first story of the tower, and terminates in the counterscarp, in a circular chamber cut in the chalk, and from this chamber three tunnels radiate to different parts of the glacis, of which one formerly led to a distant postern, and another still communicates with the old spur-work, attributed to Hubert de Burgh, and converted into a modern ravelin. Two other tunnels, apparently of Edwardian date, leave the main gallery under the castle wall, and the basement floor of St. John’s contains two sally ports, opening into the bottom of the ditch. The modern access to these galleries is by a shaft sunk in the pier of the old drawbridge, but the old entrance was nearer the curtain. The French siege of 1216 was directed upon this quarter. The approaches were made from the west below the Constable’s Gate, and under cover of a trench and breastwork. While the attack was impending Sir Stephen de Pencester brought a reinforcement into the castle by the postern under Godwin’s Tower. De Burgh, taught by experience, threw up the advanced work which still, under a changed form, covers the northern end of the castle, and it was to reach this in safety that the gallery from St. John’s Tower was executed.

Fitzwilliam’s Tower (18), placed about 80 yards east of the north gate, was connected with a second postern, not unlike the last, and now connected with a caponnière. Beyond this are 19 and 20, two watch turrets, and farther on 21, Albrinci’s or Avrenches’ Tower. This contains a third postern of peculiar arrangement. It is a low, polygonal structure, placed on a shoulder on the ditch, so as to rake its continuation southward. It was reached by a covered gallery from the south gate of the inner ward, which is continued through its basement so as to open on the counterscarp of the main ditch. Connected with this gallery was Veville or Pencester’s Tower, placed upon the curtain which on this side closed the connexion between the two divisions of the outer ward. Of the remaining towers, five in number, three are called Ashford’s, and near one of them was another well. This part of the defence has been completely remodelled. The names of the several towers are those of the knights by whom they were built, or whose duty it was to defend them, for to no castle in Britain, not even to Richmond, was the practice of tenure by castle guard so extensively applied as to Dover, and very numerous and valuable were the Kentish manors so held, amounting to 230½ knight-fees, of which 115¼ were attached to the office of Constable.

The _Keep_.—This is a very fine example of a late Norman keep. It is very nearly square, being, at the base above the plinth, 98 feet north and south by 96 feet east and west, with a forebuilding 15 feet broad by 115 feet long, which covers the east side and the south-east angle of the main structure. The angles are capped by pilasters 19 feet broad and of 5 feet projection, which meet to form a solid angle, and, rising to the summit, become the outer faces of four square turrets. On each of the three free faces is an intermediate pilaster, 15 feet broad by 5 feet projection, which rises to the same height with the parapet, and forms a bay in its line. There is a battering plinth, 6 feet high, from which the pilasters rise, and the total height of the wall is 83 feet, and of the turrets 12 feet more, or 95 feet. The base of the keep is 373 feet above high-water mark. The top of the plinth is marked, on the face of the pilasters, by a bold cordon or roll, and there are two sets-off of 6 inches common to both walls and pilasters, one at the first and the other at the second-floor level. The walls are of unusual thickness, even for a Norman keep. That to the west, between the pilasters, is 21 feet reduced to 19 feet at the first and to 18 feet at the second floor. The north wall is 17 feet, the south 19 feet, and the east 18 feet. The cross wall, which runs north and south, and divides the building nearly equally, is 11 feet at the base, and reduced to 7 feet and 6 feet at the top story.

The main entrance is in the east face, near its north end, at the second-floor level. The forebuilding which covers it is the finest in England. It is in fair preservation, all but its roofs and part of its east wall, which are modern. As at Rochester, it is of masonry inferior to the keep, at least outside, and there is no cordon at the base of its pilasters, but it contains within more ornamental work than the keep, with which it is so intimately connected that it cannot be an addition. In the north-east and south-west angles of the keep are well-staircases remarkably commodious and well lighted. They are 14 feet 6 inches in diameter, the stairs being 6 feet 6 inches and the newel 1 foot 6 inches. They rise from the basement to the roof by 114 steps, and communicate with each floor, the two lower by lobbies, the upper and upper gallery by branching passages. From the north-east lobbies doors open into the two tiers of vaults below the upper part of the forebuilding. The extraordinary thickness of the walls is intended to allow of the construction of a very unusual number of mural chambers, of which there are altogether twenty-seven. Besides the main entrance there seems to have been one at the first-floor level, also from the forebuilding. Others have since been made, one probably in the fourteenth century, direct into the basement, and another very recently into the base of the south-western staircase. Besides these, divers loopholes have been converted into doors, to give external entrance to the basement mural chambers which are used as water-tanks and powder-magazines.

The _basement_.—There is some doubt about the original level of this floor. At present it is 9 feet above the ground outside, but it contains two doors, the sills of which are 6 feet or 8 feet below the floor, and on opening the ground near the centre of the keep a pavement was found at the same level, so that the floor may have been filled in with earth and raised. If so, however, there must have been some kind of stair to supplement the well staircases, for the floor of their lobbies is certainly at its original level. The cross-wall between the two chambers is pierced by three plain round-headed arches of 11 feet opening.

The east chamber, 50 feet by 20 feet, has in its north end a deep recess, and a loop up to which the sill is stepped, while the arch overhead rises as the recess contracts. It is evident that the recess at the south end was similar, but has been converted into a doorway. This is the opening in general use, from the first landing of the exterior staircase. The wall here is 24 feet thick, and there are rebates for two doors, with bars within each. The arches are segmental and the angles rounded off. The alteration seems to have been made in the Decorated period, with a trace of later work. Close east of this door, and 6 feet below it, is another door, now partially concealed, which led into a vault 28 feet by 15 feet, occupying the south-east angle of the building. In its south wall a door leads into a second vault, 23 feet by 15 feet, below the lower vestibule and chapel. Each vault had a loop towards the east. They are now used as water-tanks, and reached by external openings. Returning to the east chamber, in its east wall are two doors, one at its north end and one near the middle, now blocked up. The northern door opens into the lobby of the north-east staircase, and from it a door led into a vault 28 feet by 12 feet, in the forebuilding, now used as a magazine, and entered through the outer wall. The middle door probably led into another vault, of which nothing is known. Thus at the ground level of the forebuilding there are three, if not four vaults, all originally entered from the keep. They are about 6 feet high to the arch-springing.

The west chamber, 50 feet by 16 feet, has a large recess and loop in each end, and in the west wall two doors, one opening into the lobby of the south-western staircase, where is the modern door, and the other, now blocked up, which led into a mural chamber 39 feet long by 12 feet broad, now a powder-magazine, and entered from the outside.

The first floor also contains two main rooms which communicate by a small door near the north end of the cross-wall.

The east chamber, 53 feet by 22 feet, has a recess 7 feet wide in its south end, which ended in a loop, now converted into a window. In its right or west jamb is a fireplace; in its left a door opening into a mural chamber 18 feet by 11 feet, with a loop to the south, and from the east end of which a door opens upon a staircase of the forebuilding. This door and the window above it are of Tudor date, but there are indications that there was an original door here, of which the Tudor frame is a replacement. In the exterior wall is an arch of relief in fine ashlar, which matches in size with the vault within, and looks as though intended to protect an original door. In the east wall of the great chamber is a door opening into the lobby, in which six steps descend to the north-east staircase. From this lobby a door leads into a vault 24 feet by 12 feet, with a recess in its north and south walls, and a loop in its east end. From this a short passage leads into a second vault, 17 feet by 2 feet, with a loop to the north. These two vaults are placed below the upper vestibule and great guard-room of the forebuilding. In the north wall of the great chamber, at the east end, a small door opens into a mural gallery, 16 feet by 4 feet, with a loop to the north, and which ends in a chamber 5 feet by 6 feet, also with a loop to the north. Above this gallery is a recess 7 feet wide, raised about 10 feet from the floor, to clear the gallery. Its loop is replaced by a modern window.

The west chamber, 52 feet by 20 feet, has also a 7-feet recess in its south end, with a modern window, and in its east jamb a door opening into a mural chamber, 13 feet by 9 feet, with a south loop, and a fireplace in the east wall. This, and the fireplace already mentioned, are placed back to back, and, though with Tudor fittings, may possibly be original. In the north wall of the great chamber is a high recess and window, similar to that in the east chamber. In the west wall are four openings; that at its north end opens into a chamber in the north-west angle, 23 feet by 10 feet, with loops to the west and south. From it branches a passage 18 feet by 4 feet, in the north wall, which leads under the high window recess to a chamber 7 feet by 6 feet. Both passage and chamber have loops to the north. Next follows a window recess of 6-feet opening at the floor level, rising by four steps to the modern window; then a door opening into a mural chamber, 20 feet by 9 feet, with a fireplace in the east wail and two loops to the west. Finally is a door opening upon the lobby which leads to the south-west staircase, rising three steps, and having a loop to the west. Besides the three fireplaces in mural chambers, there were two others under arches of 12 feet span, one in the centre of each face of the cross-wall. These, however, are closed up. In the same wall, at the north end, is a door between the two chambers.

The second floor is the main or state floor of the building, and that into which opens the great entrance. As in the keeps of London, Rochester, and Hedingham, it had two tiers of windows, the upper passing through a mural gallery. Early in the present century this floor was covered in by two large brick vaults, slightly pointed. The object was to convert the ramparts into a platform for cannon. This clumsy addition completely conceals the upper half of the walls and destroys the effect of two very fine chambers.

The east chamber, 55 feet by 24 feet, is entered on the east side, near the north end, by a large, full-centred doorway, flanked outside by nook shafts. From this a vaulted passage, 5 feet broad, traverses the wall, here 16 feet thick, and descends by eight steps into the main chamber. This inconvenient height was evidently given to secure headway for the vault below the vestibule. In the entrance passage, on the left or south side, a door opens into the well chamber, a vault, 16 feet by 8 feet, with a loop to the east upon the great staircase. At the south end of the vault, upon a step, is the well, 4 feet diameter and 289 feet deep, lined, as far as can be seen, with ashlar. In the east wall is a recess for a spare bucket. Near the entrance, at the north end of the east wall of the great chamber, a door leads into a curved passage which descends eleven steps into the north-east staircase. At the south end of the same wall a large arch, 5 feet above the floor, opens into a mural chamber 8 feet by 7 feet. This arch has been reduced by modern brickwork, to support the great vault. In the same chamber a loop opens upon the grand staircase, and in the south wall is a door and a descent by eight steps towards the chapel. It is possible that the loop was originally a small door leading into the upper floor of the middle tower of the forebuilding, and that thence, in Tudor days, a small wooden gallery led to the north door of the ante-chapel. In the south end of the great chamber is a window recess of 7 feet, opening with nearly flat sides. In its west jamb is a fireplace, in the east a door leading into a mural vault 18 feet by 12 feet, over that by which the first floor is entered. This vault has a loop to the east upon the great staircase, on the north the eight steps, already mentioned, and in the south wall a narrow door and mural passage leading to the robing-room and ante-chapel. In the north end of the great chamber, near the north-east corner, a door opens into a chamber 7 feet by 5 feet, which lies in the north wall below the window recess, here of 8 feet opening.

The west chamber, 55 feet by 21 feet, has in its south end a window recess 7 feet broad, in the east wall of which a door opens into a mural chamber 14 feet by 9 feet, with a loop to the south and a fireplace in the east wall. In the west wall of the great chamber is a window recess 5 feet wide, commencing at the floor level, and there are three doors. That at the north end opens into a chamber, 22 feet by 9 feet, in the west angle, with loops to the west and north. From this opens a passage in the north wall 17 feet long, which ends in a chamber 11 feet by 5 feet; chambers and passage have loops to the north. Near the centre of the west wall of the great chamber a door opens into another mural chamber, 22 feet by 14 feet, with two loops to the west and a fireplace in the east wall. Another door leads, by a curved passage with a loop to the west, into the south-west staircase. In the great cross-wall is a door near its north end, and in each of its faces two large fireplaces, now walled up. In the north wall is a high window or recess of 8 feet opening.

The main gallery, though it threads the wall nearly all round, lies at different levels, and at one point is stopped. It is entered from the two staircases. That at the north-east gives off two branches, of which one rises by fourteen steps, with a loop to the north, and enters a passage in the north wall 16 feet long, having a loop to the north, and to the south an opening, now blocked up, into the great chamber. The passage ends in a chamber 18 feet by 8 feet, in the north wall of which is a loop, and near it a deep recess for a garderobe; and in the south wall is what seems to be the mouth of a gallery threading the cross-wall, where, however, it would be stopped, or very much reduced, by the shafts of the fireplaces. The other branch from the north-east staircase rises by six steps, when it gives off a branch to the east of nine steps, which lead to the roof of the upper tower of the forebuilding. The main passage then curves and rises ten steps more, in all sixteen steps, when it enters the substance of the east wall, where it is 60 feet long and 4 feet 7 inches broad, and has a door to the east, opening on the middle tower of the forebuilding, and a loop, and opposite to these are two openings in the west wall, which formerly opened into the great chamber, and now are blocked by the brick vault. The passage then turns and lies for 67 feet in the south wall. In its east end is a loop; in its south wall, there 10 feet thick, a door opening upon the roof of the lower tower of the forebuilding. In the same wall are two, and probably three loops, opposite to two of which seem to have been openings into the great chamber. The third is much broken. In the inner wall is a recess, probably the mouth of the cross-wall gallery. At its west end the main gallery descends twenty steps to reach the south-east staircase. From this staircase, at twenty steps higher up, a door opens into the gallery, which is continued along the west wall 68 feet. It has three loops in the outer wall, opposite to two of which are apertures, now closed, which looked into the great chamber. This gallery is continued 15 feet, with a width of 6 feet, in the north wall.

The roof of the keep is now an artillery platform, pierced on the south by six and on the west by five embrasures, the top of the pilasters forming a bay in the centre of each face. The two other faces are solid, and protected by guns “en barbette.” Of the four turrets, which are 21 feet square, with two entrances on each face, the two to the south have entrance stairs, and doors from the stair-head open in the side of those at the north-east and south-west angles. The north-west turret seems to be entered by an opening in its east face. Though these turrets have been much pulled about, their substance seems original.

The forebuilding covers the whole of the east and about 45 feet of the adjacent south face of the keep. Its breadth ranges from 15 feet to 23 feet. It was strengthened by three towers, one over the north end or top of the staircase, one over the south-east angle or bottom, and one on the east front over the middle of the staircase. The middle tower, in which was the middle doorway, is unusual. These rose to about four-fifths of the height of the main building, and their battlements were reached, as has been shown, by doors from the upper gallery. The object of this forebuilding was to contain and protect the great staircase, and in it are three vestibules, a lower and upper chapel, and an ante-chapel, several mural chambers, and a well. The entrance at this time begins from the ground on the south front by an open staircase of ten steps, probably modern, as the original ascent seems to have been against and parallel to the keep wall. At the top of these steps is a landing, upon which opens the present entrance into the basement floor. The staircase then turns to the right and rises eleven steps, still open, having on the left the keep and on the right the flanking projection of the forebuilding, in which is the lodge, and above it the robing-room. In front is a lofty doorway, 7 feet wide, with a segmental arch, quite plain, and above it a loop which opens from the chapel passage. This was the lower entrance, and was closed by a barred door. Entering the doorway, eleven steps under cover lead to the lower vestibule, which is thirty-two steps, or 20 feet, above the ground. This vestibule is a handsome chamber, 15 feet by 12 feet, with an arcade of two arches in its south wall, each pierced for a loop. In the west wall, near the entrance, is the door of the lodge, a plain barrel-vaulted chamber, 13 feet by 6 feet, having a loop to the west. In the east wall of the vestibule is an arch of 7 feet opening, springing from coupled columns and flanked by two others. The head is moulded with the chevron pattern. This arch opens into the lower chapel, 14 feet by 13 feet, which occupies the south-eastern angle of the building, and is so placed that the altar could be seen if desired by each person who entered the keep. In its north and south walls are arcades of two arches, divided by a pier carrying two nook shafts and a third shaft in the centre. In the south-east arch is a loop, and opposite to it a cupboard. In the east end is also a loop, placed in a recess flanked by two shafts. The floor of this chapel is one step above the vestibule. The ceiling of both was flat, and of timber. Beneath chapel and vestibule is the vault already described. The walls of this part of the forebuilding are from 2 feet 6 inches to 5 feet thick.

From the vestibule a doorway, the second, opens in the north wall. This has a segmental head, and is original, but it has been reduced in breadth, probably when the basement door was opened, by the insertion of new jambs. It opens into a vaulted passage, 6 feet wide and 5 feet long, being the thickness of the north wall of the lower tower. It is occupied by three steps and a landing, from which four steps ascend into the middle vestibule. This is a chamber 25 feet long by 15 feet broad, having a modern roof and two modern windows in its east wall, which has been in part rebuilt. On the left, on entering, is a Tudor doorway, and above it a square-headed window of two lights, of the same date. This opens into a mural chamber already described, and thence into the first floor of the keep. A large arch of relief is seen above in the wall, and it is probable that there was always a door here.

Beyond this door the staircase rises by twelve steps to a broad landing at the end of the vestibule. Besides the Tudor doorway and window, on the same side, higher up, are loops from two mural chambers of the second floor of the keep, and in the south wall over the stairs a small door, not now used, which opens into the ante-chapel. In the north wall of this vestibule, a doorway of 5 feet opening, the third in order upon the stairs, opens into a passage 6 feet wide, which pierces the wall of the middle tower, here 6 feet thick, the whole tower being 14 feet. From this doorway a flight of twenty steps, 8 feet wide, ascends into the upper vestibule, to a large landing 19 feet by 14 feet. The vestibule itself is 25 feet by 14 feet, and lies between the middle and upper tower of the forebuilding. In the east wall, which has been rebuilt with the roof, are two modern windows. In the south wall a narrow passage, 10 feet long by 3 feet broad, leads to the outer well, the ashlar pipe of which is 4 feet diameter. It seems to have been used in modern times as a cesspit, and is choked up. The well is placed in the centre of the tower, which seems to have been a mass of masonry. At present the wall round the well has been broken away so as to form a rude chamber. It cannot now be ascertained whether the well stopped at this level, as is probable, or was carried up to the roof of the tower. In the north end of the vestibule, looking down upon the staircase, a door leads into a vaulted guard-chamber, 16 feet by 10 feet, with loops to the north and east, and a deep recess in the west wall. In the west wall of the vestibule is the great door of the keep and a loop from the well-chamber. There are traces on this wall of ashlar, as though it was originally intended to vault and groin this vestibule.

In the forebuilding, on the level of the second floor of the keep, remain to be mentioned the robing-room, ante-chapel, and chapel. They are placed in the lower tower of the forebuilding over the lodge, vestibule, and lower chapel, and were entered only from the keep, through the mural chamber already described. From this chamber a passage, only 2 feet 5 inches broad, lies in the wall over the outer doorway, and from it a loop opens over the staircase, outwards. This passage, 17 feet long, ends in two small doors, right and left, one entering the robing-room, the other the ante-chapel. Originally this was the only way into the chapel, and a very stout person could scarcely have reached it. The Plantagenet princes, though mostly big in the bones, were rarely corpulent. The robing-room is over the lodge. It is 10 feet by 7 feet, vaulted, groined, and ribbed; the ribs are of plain roll section, and there is no boss. They spring from nook shafts at the four angles. There are loops, or rather small windows, of a foot opening to the west and south, flanked by small columns. The ante-chapel, 16 feet by 13 feet, has an arcade of two arches in the north and south walls, and a loop in the north-east space, converted into a small door, so as to give a separate entrance. In the east wall an arch of 8 feet span opens into the chapel. It springs from coupled columns, flanked on each face by two others, and the arch has a chevron moulding. The floor was of timber, and rested on five beams. The roof is vaulted, groined, and ribbed, and the moulding a roll with a band of dog-tooth. The chapel is 14 feet by 13 feet. It is peculiar in its position, being placed over the base of the entrance stair, instead of, as at Middleham, at its head. It also has an arcade of two arches in the north and south walls, with nook shafts in pairs at the angles behind which, from a corbel capital, spring the ribs of the vaulted roof. These are of a roll section with a band of the dog-tooth ornament, and a central flowered boss. There is a small east window, and also south of it a piscina with a trefoiled head and projecting basin, now broken off. This seems a Decorated addition. There are two loops in the south wall, and one in the north near the east end, and in the other north bay a cupboard. The floor was of timber, resting upon four joists.

The material of this keep is chiefly the rag stone of the country worked as rubble, not very regularly coursed. Ashlar, mostly of Caen stone, is used freely for the door and window dressings and the quoins. The joints are close. The pilasters are of unusual breadth and projection. The chapels and the lower vestibule are highly ornate, with much of the chevron and roll mouldings, and occasionally of the dog-tooth. The arches are sometimes segmental, but more commonly full centred. There is no portcullis in the building; the entrances were closed with doors only, secured with wooden bars. None of the loops that open on the staircase could be used in its defence. Excepting about the main door there is no ornamentation in the keep itself. Doors, windows, and fireplaces seem to have been quite plain. This is the only known keep in which there is a second well, and it is difficult to understand why there should be two so near together, the expense of making which must have been so great. The upper gallery is of very rude masonry indeed; the lines of the passages do not coincide with the general direction of the walls, and the execution is very inferior. It is unfortunate, with ample buildings all round, that the authorities should pervert this very curious keep to vulgar and dangerous uses. The stores should be kept elsewhere, and the brick vaulting and additions be removed. The external breaches and doors in the walls should be closed, the second well cleared out, and the whole building as far as possible restored to its original condition. It might then be fitted up as a museum of arms, and every part, including the ward, made accessible. The reputed date of this keep is 1153, when the foundations are said to have been laid by Henry, grandson of Henry I., shortly before he succeeded to the throne. This coincides sufficiently well with the evidence of the building itself, which is late in the Norman style.

The general view here added is taken from the north, and shows the constable’s gate and bridge, and in the foreground the towers and curtains of the outer ward on this side. Immediately behind them are seen the square towers and the curtain of the inner, or Norman ward, and within all rises the keep, applied to the east side of which are seen the three towers of the forebuilding.

DUNSTER CASTLE, SOMERSET.

The Castle of Dunster is of high antiquity, and for many centuries was a place of great military consideration in the western counties. It was the _caput_ of an extensive Honour, and the chief seat of a line of very powerful barons. The hill upon which it stands is the north-eastern, or seaward extremity of a considerable ridge, from which it is cut off by a natural depression, and thus forms what is known in West-Saxon nomenclature as a tor. The tor covers above ten acres of ground, and is about 200 feet high, with a table top in area about a quarter of an acre. It stands upon the western edge of a deep and broad valley, which contains the park, and below the castle expands into a tract of meadow about a mile in breadth, skirting the sea from Minehead to below Carhampton. The park is traversed by a considerable stream, the Avill, one head of which springs from Croydon Hill, and the other, flowing past Wootton-Courtenay, rises about six miles distant under Dunkery Beacon. The home view, one of exceeding richness, is limited and sheltered on the south and west by the Brendon Hills and the high ground rising towards Exmoor. To the east it includes the vales of Cleeve and Williton, bounded by the Quantocks. Seaward on to the north the eye ranges over Bridgwater Bay to the headland of Brean Down and Worle, and commands the west or opposite coast from Penarth Point to Aberavan and Swansea.

West, and at the foot of the castle hill, and under the immediate control and protection of the old fortress, is the town of Dunster, a small and compact cluster of old-fashioned houses, many with timber fronts, in the midst of which is the parish church, once connected with the priory, the foundation of one of the early Norman lords. The eastern or monastic part of the building now forms the private chapel of the Luttrells, and contains several of their tombs.

The fancy cloths once known as “Dunsters” have long ceased to be fabricated, and of the fulling mills the ruins have well-nigh perished. The haven at which these manufactures found shipping is also silted up, and the privileges conceded to the townspeople, being now shared by the community at large, are no longer commemorated, and are known only because the charters granting them have been preserved. Of the neighbouring hills, “Gallocks” is thought to be so called because there the high judicial powers of the lords were exercised in the view of all men, and “Grabhurst,” the castle ridge, is said to be named from an entrenched wood, though this use of the word “graff” is unknown or unusual in English nomenclature, and “Hirst” or “Hurst” belongs rather to Sussex and Kent than to Somerset. The fact is that the hill in old deeds is spelt “Grobefast,” and is at this time colloquially “Grabbist.” Near it is a lofty detached hill known as “Conygaer,” crowned by a tower of the last century. At present it is thickly planted, but no camp has been discovered there, such as the name might indicate. The castle mill remains. It is placed on the verge of the park, upon the stream, and concealed and protected by the castle.

The castle is composed of two parts, due to the natural disposition of the ground; these are the tor or keep, and the lower ward. The tor is in form oval, and its summit, naturally flat, has been further levelled by art, as the slopes also have been trimmed, and rendered almost impracticable for direct ascent. The summit measures about 35 yards east and west, by 70 yards north and south. The keep, which stood here, has disappeared, and its existence, long a matter of tradition, may now be deduced from a sewer and some foundations in the south-west corner laid open a few years ago. The present surface was laid down as a bowling-green in the last century, and a summer-house constructed at the north-east corner, in which is a window in the Perpendicular style taken from some earlier structure.

The artificial scarping of the hill sides is confined to the upper 80 or 100 feet. At this level are two platforms or shelves, one, a small one, towards the south, the other much larger, also chiefly natural, towards the north, and which forms the lower ward. The lower ward is of a semilunar or semioval figure, the hollow side being formed and occupied by a portion of the skirts of the tor. It measures about 33 yards north and south by 126 yards east and west, and covers about half an acre of ground. The outer or convex edge, steep by nature, has been cut into a low cliff, supported by a retaining wall, which, with its flanking towers and superstructure of parapet, protected this ward. At the foot of this wall, part of which supports the present house, the slope recommences, and, though now terraced by roads and paths, formerly descended unbroken to the base of the hill.

The keep was probably either circular or polygonal, approached as at Lincoln by a direct flight of steps from the lower ward. Its gateway seems to have been defended by a portcullis, as one is mentioned in the records, which could not have been in the earlier gateway or the later gatehouse. The buildings and inhabited part of the old castle were in the lower ward or its north-eastern quarter, upon the _enceinte_ or curtain wall, and on the site, generally, of the present house. The wall was strengthened and flanked by half-round towers, of which the lower part of several remain incorporated into the later building and connected with fragments of wall, now a part of the house, and betrayed by their excessive thickness. One of these walls has a core of the natural red sandstone rock, enclosed in masonry, but traceable by an occasional exudation of dampness. The gateway of this ward remains between two of these flanking towers. It is 9 feet broad, with plain chamfered jambs, and a low stiff drop arch. There was no portcullis, and probably no drawbridge, the only defence being a door composed of bars of oak, 4 inches square and 4 inches apart, forming a grating, planked vertically outside with inch-and-a-half oak plank. Upon each oak bar was laid a bar of iron, and the whole fabric was spiked together with iron fastenings, having diamond-shaped heads. The meeting line of the two valves was guarded by an iron bar. In the right valve, on entering, is a wicket-gate 3 feet 8 inches high by 2 feet 6 inches broad, fastened with a huge iron lock in a wooden shell. This very curious specimen of carpenter’s and smith’s work, though of later date than the gateway, is old, not unlike that of Chepstow Castle, and probably of the time of Henry VIII. or Elizabeth. The gateway itself belongs to those of Henry III. or his son. In the last century the gates were permanently closed, and behind them was built a wall backed with earth. The gateway has recently been restored as far as possible to its original condition, and now gives access by steps to the lower ward.

The mural towers flanking the gateway are parts of circles, 16 feet 6 inches external diameter, and the lower 12 feet are original. One contains a curious vaulted basement with the usual three loops, and in the rear a doorway which opened into the ward. The other, or eastern flanker, had a basement chamber until recently filled with earth, and had also three loops, of which two are still visible outside. This tower was connected with a building in its rear, the foundations of which are original, and now form a part of the offices. Also upon the retaining wall, but about 20 yards beyond, and to the south-west of the old gateway, is another similar flanking tower, of which the upper story remains, and a part of a doorway. This tower is open in the rear. The towers, curtain, and entrance gateway are, in substance, all of one date, and what ashlar remains is of good quality and well jointed. The superstructure has been renewed recently.

The approach to the castle was steep, as it still is, from the town up to the old gateway, to enter which the road made a sharp turn. Just below the gateway, upon this approach, has been built a gatehouse, which projects from and is connected with the curtain, being incorporated into the tower, flanking the old curtain on the west side. This structure, the great gatehouse, still remains perfect, and, though evidently intended more for ornament than defence, makes a most appropriate approach to the castle, and gives to the whole structure much of a mediæval and something of a military character. This building is rectangular, 63 feet broad by 23 feet deep, and about 45 feet high, sixty-two steps leading to its battlements. It is pierced by a passage 10 feet 6 inches broad, having a plain pointed waggon vault, and at each end a not very highly pointed arch, with good moulded jambs continued round the head. There are no lodge doors opening into the archway, and neither portcullis nor drawbridge. The fronts are plain, save that the exterior has two flanking buttresses, and over the entrance is a rectangular panel containing nine coats of arms in three rows—one, four, and four. The interior front is flanked by two half-hexagonal turrets, of which that at the outer end contains a well-stair, entered from the outside by a small four-centred doorway, and communicating with each floor, and with the battlements. The corresponding turret is built upon one of the old mural towers which flanked the gateway of the lower ward. It contains a small chamber, probably a garderobe, on each floor. Against the outer end of the building are two buttresses, a large and small one, probably added to support the wall which then stood upon a steep slope and showed signs of settlement. The gatehouse is of three stages. The basement has a chamber on each side of the main passage, entered, one by the well-stair, the other by an exterior door. That next the well-stair is 14 feet 6 inches by 9 feet 8 inches, and has a window to the front. That on the other side of the archway is 21 feet 6 inches by 16 feet 3 inches, and is entered from the outside by a small doorway, probably an insertion. Opening from this chamber are two closets, and a well-stair ascends to the two floors above.

The first floor contains two rooms, 22 feet 10 inches by 16 feet 6 inches, and 21 feet 6 inches by 16 feet 6 inches, and 13 feet high. In its inner end are two closets.

The second, upper, or principal floor, was formerly of two rooms, but has recently been converted into a handsome hall, 47 feet by 16 feet 6 inches, with an open roof. It has five windows and a fireplace, and is entered on the level from the lower ward by a doorway, which seems an insertion of the date of Henry VIII., and which has the head of the well-stair on one side, and beyond, on each side, a closet. The windows of the gatehouse are mostly of two lights, divided by a transom into four, with the upper lights cinquefoiled and in the head quatrefoiled. The summit is embattled, and at the four angles are turrets, of which the two to the outer or front face are apparent only.

The shields on the exterior panel are, in the upper line, 1. Luttrell with crest and supporters. Below, in the next line, 2. Luttrell impaling Courtenay; 3. Luttrell impaling Beaumont of Sherwell; 4. Luttrell impaling Audley; 5. Luttrell impaling Courtenay of Powderham. In the lower row, 6. Luttrell impaling Hill; 7. Luttrell impaling blank; 8 and 9 blank. The Luttrell supporters were two swans chained and collared, derived from Bohun through Courtenay. The date of this gatehouse is uncertain. It has been thought to be the “novum ædificium castri de Dunster,” with the construction of which the accounts show Henry Stone to have been charged in the 9th of Henry V., but the lower part is of the style prevalent under Richard II. The door from the lower ward into the lobby is scarcely earlier than Henry VII. or VIII.

It is probable that the gatehouse was for some time used in combination with the gateway by its side, until the latter was closed. The approach and entrance, however inconvenient, were strong, and almost precluded any regular attacks by battering-machines, or even by escalade.

The history of Dunster commences with Domesday, in which it is recorded that William de Mohun holds Torre, and there is his castle. Aluric held it in the time of King Edward. “Ipse [Willielmus de Moion] tenet Torre. Ibi est castellum ejus. Aluric tenuit T.R.E.” These words are very appropriately inserted over the great chimney-piece in the hall. The Exeter Domesday also confirms the holding both of Mohun and Aluric. Who Aluric was is unknown. That he was a considerable Englishman none can doubt, but the name was common, occurring sixty-four times in Domesday, as does Alric, probably the same name, twenty-six times.

Mohun no doubt found the tor strongly fortified, after the English manner, for not only was it a frontier fortress against the western Celts, but it must have been exposed to the piratical invasion of the Northmen, who gave name to the opposite islands of the Holms, and the not very distant port of Swansea. The place was, in fact, a natural burh on a large scale, such as Æthelflæd and Eadward the Elder were wont to throw up artificially on a smaller scale in the early part of the tenth century. There was the conical hill with its flat top for the _aula_ or _domus defensabilis_, and the courtyard below for the huts and sheds of the dependents and cattle.

William de Mohun was no mere adventurer. He was a great baron of the Cotentin, having the castle of Moion in la Manche. He fought at Hastings with a knightly following, and received from the Conqueror from sixty to seventy manors in Somerset, Devon, Dorset, and Wilts. These manors were in his time, or in that of his successor, combined into an Honour, as was the case with those attached to the chief seats of Plympton, Totnes, and Barnstaple. Dunster became the _caput honoris_.

The Honour of Dunster was one of about eighty-six in England, though in what they differed from baronies is not precisely understood. The nucleus of either was almost always an estate held before the Conquest, added to largely by the Norman who conquered it. In all cases it extended into more than one county, and was held of the king _in capite_ by homage, fealty, and military service. By the laws of Henry I., every lord could summon his liegemen before the court, “et si residens est ad remotius manerium ejusdem honoris unde tenet, ibit ad placitum, si dominus suus summoneat eum.” The Honour is not a jurisdiction mentioned in Domesday, unless it be in a passage relating to Cornwall where it is recorded, “Hæ terræ pertinent ad honores chei;” chei being a place. The term is said to have been first used by the Conqueror in his charter to the Abbot of Ramsey. Most of the Honours seem to have fallen into disuse by the alienation of the manors composing them, as was the case with Dunster, although the records show that for many centuries the rights were maintained by the lord of the castle in full rigour.

To what extent the Mohuns were content with the earlier defences of the castle is unknown, but it is remarkable that no mouldings or fragments of Norman ornament have been dug up in or about the building, although there is original Norman work in the parish church. From the configuration of the ground the lines of the old fortress must have been where they still are, so that there would be no reason for pulling down the earlier works to enlarge the area; and yet it is difficult to suppose that works as durable as was the case with those of the Norman period could have fallen to decay by the reigns of Henry III. or Edward I., the date of the oldest extant parts. However this may be, it is certain that the castle of the Mohuns was one of the most important of the western fortresses; and in the lawless days of Stephen it was held for the empress against the king, during the great revolt of 1138, its lord being then William de Mohun, the second baron.

William, indeed, was not content with passive resistance. He is described as the “Scourge of the West,” ravaging and plundering the country up to the gates of Barnstaple, where he was held in check by Henry de Tracy. He is said to have been created earl either of Dorset or Somerset, or both, by the empress in 1140; but this creation rests on very uncertain authority, and has never been admitted as valid. The earldoms of that period were very irregular, and some were afterwards set aside. This lord founded the Augustine Priory of Brewton, in Somerset, and, according to the Black Book of the Exchequer, he held forty-four knights’ fees. It is not improbable that to him is due the circular or polygonal keep, which was common at that time where a castle possessed a mound, and which is known to have stood on the summit of the tor.

No mention of the castle occurs till the reign of John, who held the castle and Honour during the minority of Lord Reginald, when the fines, &c., for the Honour were levied by the king’s officers. In the Chancery roll of 1201–2, Nicholas Puinz accounts for 15s. 2½d., half a year’s pay allowed to the janitor of the castle, and the same to the watchman; and these payments are repeated by Reginald de Clifton, who, in 1204, was ordered to place Reginald de Moyon in possession of the castle of Dunster and the heritage then in his custody. A very little before this, the coming of age of Reginald, Hubert de Burgh was in charge, and had accounted “de finibus militum” of the Honour; and 25th February, 1202, John called upon the knights and free tenants to contribute through De Burgh for strengthening the castle. “Our castle,” the king calls it, probably not merely as holding it in wardship, but as asserting the general rights of the crown to all castles. A second Reginald seems to have founded a mass for the weal of his ancestors, to be said daily by a monk or a secular priest, to be provided by the prior, in the upper or St. Stephen’s chapel, in the castle, or during war in the chapel of St. Lawrence, within the priory. If the same was neglected, power was reserved to distrain upon the goods of the prior. Leland mentions St. Stephen’s chapel as connected with the keep. There seems also to have been a second chapel, as usual, in the lower ward. Upon the death of Lord Reginald, about 1213, the Honour again fell into the hands of the Crown during a long minority. Henry Fitz Count was placed in charge, and Alice the widow was allowed dower and “maritagium.” It is curious that John does not appear to have visited Dunster, although he was at Stoke-Courcy.

Henry III., in 1220, placed the forest of Dunster in charge of Peter de Maulay. He retained the castle in his own hands, and there occur several charges for the payment of Roger and William de Vilers, as “balistarii regis,” who dwelt in the king’s castle of Dunster. A specific order in 1222 places the Mohun lands in Carhampton in charge of William Briwer, probably next of kin to the widow, but reserves to the king’s hands the castle and vill of Dunster. Soon after, Watchet market, being unlicensed and injurious to Dunster, was put down (Close Roll I., 137, 418, 605). The above were not the only persons to whom from time to time this valuable wardship was committed.

Of the condition of the castle at the close of this wardship, nothing is on record; but the wealth of the family was much augmented by the match of Reginald, King John’s ward, with a Briwer co-heiress, and either his son Reginald, the founder, 1246, of Newenham Abbey, who died 41st Henry III., 1256, or his grandson John, who died 7th Edward I., 1278, the last baron by tenure, might have built the curtain and mural towers containing the lower ward, of which the bases remain. The keep was probably left unaltered, and indeed, from the great and inconvenient height at which it stood, could have been but little used. The purely defensive parts of castles, when not inhabited by the owner, were usually but little cared for, and the allowance for a porter and a single watchman shows that in this respect Dunster was no exception.

In 1376, John de Mohun, the eighth baron and the tenth in lineal descent from the founder of the family, died, leaving daughters only, and a sale of the castle and the rest of the property was effected by his widow Joan Burghersh. The purchaser was another widow, Elizabeth, daughter of Hugh Courtenay, Earl of Devon, and widow, first, of Sir John de Vere, and afterwards of Sir Andrew Luttrell, of Chilton, a cadet of the barons Luttrell of Irnham. Elizabeth was a lady of high rank, of kin, through the Bohuns, to Edward III., and with the command of great wealth. Her son, Sir Hugh Luttrell, became the new lord of the castle and Honour, and probably built the great gatehouse.

The Luttrells were steady Lancastrians, and their representative, Sir James, took knighthood on the field of Wakefield, and fell in the second battle of St. Alban’s, when his estates were confiscated, 1st Edward IV., though only to be restored, 1st Henry VII., in the person of his son, Sir Hugh. Sir Hugh Luttrell stood high in the favour of Henry VII., and seems to have lived at Dunster in great splendour. To a second Sir Hugh, Leland attributes the great gatehouse, and he may have completed or repaired it, and opened its south door leading upon the lower ward. Probably he also inserted the armorial panels over the entrance portal, the last of which, complete, bears his coat impaling that of Margaret Hill, his wife. He also repaired the chapel of St. Stephen. Leland describes the donjon, or keep, as having been “full of goodly buildings,” which, however, had disappeared even before his time. The inhabited part of the castle was then, as now, in the north-east angle of the lower ward. Sir Andrew, Sir Hugh’s son and successor, “built a new piece of the castle wall by the east.”

The next possessor who left his mark upon the castle was George Luttrell, Sheriff of Somerset in 1593. He built the market-house in the town and the older part of the present dwelling-house, which bears date 1589, incorporating with it much of the curtain wall, towers, and walls of the older and more distinctly military building. The entrance to the ward seems to have remained as before, through the gateway between the flanking towers.

During the wars between Charles and the Parliament, the Luttrells sided warmly with neither party, and were out of favour with both. Its owners at this time were Thomas Luttrell, who died 1644, and George, died 1655. In 1643 a Royalist garrison, under Colonel Wyndham, took possession, and the castle was visited by Prince Charles, whose chamber is still pointed out. In 1646 Blake laid siege to the castle for the Parliament, and battered it from the north-west, behind the Luttrell Arms. It was surrendered by Wyndham in April, 1646. A few iron cannon balls, memorials of this siege, have been found.

The government, although they apologised for the military occupation of the castle, levied a local rate for pulling it down. Probably this referred only to the upper part of the curtain wall on either side of the gatehouse. It is said that the gatehouse was injured, but its present condition shows that the injury could not have been of a very serious character.

A century later the accounts show that the Luttrells raised the surface of the lower ward, probably about fourteen feet, evidently with earth obtained by scraping the adjacent slope of the tor. This, which involved the closing up of the old gateway, was probably combined with the construction of a new approach, which passed below and outside of the gatehouse, wound round the castle and the tor, and entered the lower ward at the new level. Matters thus remained until the accession of Mr. George Luttrell in 1869, when, under the judicious advice of Mr. Salvin, a great addition was made to the Elizabethan house, a new tower was constructed on the west front, and the foundation and pavements of buildings along the north front, and connected with the entrance gate and the gatehouse were laid open, and the walls restored and rebuilt, and a terrace formed along a part of the curtain. The approach for carriages was also much improved, though, as before, at the cost of avoiding the great gatehouse.

The ancient walls incorporated with the later residence prove that there must have been very considerable buildings upon the ground now occupied by it, but there is some reason to suppose that both hall and chapel stood near the site of the later gatehouse, and, therefore, to the right or west of the original entry. If this be so, the extent of buildings in the lower ward must have been very considerable indeed, as in the other direction they certainly extended, as does the present house, to the foot of the tor, and were flanked by it. Nevertheless, considerable as the alterations have been, and handsome and convenient as are the rooms of the present mansion, it represents very fairly the original fortress, and, like it, is sheltered by the tor, and predominates over the park, the town, and the sea-coast, commanding a very extensive view, and, as becomes the representative of so important a military post, is itself visible from the tract of country of which it was sometimes the terror, but more frequently the protection.

THE KEEP OF DURHAM.

The Castle of Durham is now given over to the use of the Northern University, and is in great part occupied by the students. Parts of it only are open to visitors, and any description of its details, under the present restrictions, would be of little value.

In a recent volume of the publications of the Surtees Society, Mr. James Raine, the worthy son of a distinguished sire, has given to the archæological world a very curious poem, now first printed, entitled “Dialogi Laurentii Dunelmensis Monachi ac Prioris,” a work of the time, and which records the intrusion of William Cumin into the see of Durham. This was a period of extreme interest in that important see, once including the city of Carlisle and the territory of Teviotdale, and at the date of the poem still holding the Castles of Durham and Norham, fortresses of the first rank, even in a district which contained Bamborough.

The strife between Stephen and Maud, severe all over England, was nowhere conducted with greater severity than upon the Tyne, the Tees, and the Wear. David of Scotland, Maud’s uncle and active supporter, unsubdued by his defeat at Northallerton, claimed the earldom of Cumberland in his own right, and that of Northumberland in right of his wife. Durham alone stood in his path, and its bishop, Geoffrey Rufus, strong in his impregnable castle, steadfastly adhered to Stephen. His death in 1140–41 enabled a certain William Cumin, an adherent of David, to obtain by force and fraud possession of the castle and the temporalities of the see, although he failed to secure his election to the bishopric. The result was a severe contest between Cumin and the lawful bishop, William de St. Barbe, in the course of which the cathedral was occupied by soldiery, and its monks were ill-treated and slain. It was not till 1144 that Cumin was put down, and peace restored to the house and patrimony of St. Cuthbert.

Laurence, who was born at Waltham and brought up in its holy house, came to Durham during the episcopate of Flambard, who probably completed the castle, the masonry of which, at least, was begun during the reign of the Conqueror. As an ordinary monk, he was celebrated for his facility in metrical composition. He became first precentor, and then a chaplain, to the bishop. The episcopal seat and church of Durham has been described as

“Half church of God, half fortress ’gainst the Scot,”

and the bishops themselves partook largely of this double character. In the bishop’s household, Laurence saw much of secular life. He became a hunter of the wolf and boar, a fisherman, and a judge of horse-flesh; and, if not actually a warrior, he certainly understood the principles of military defences. At the death of Bishop Rufus his connexion with the episcopal household ended, and he took an active part against the intrusion of Cumin and in the election of St. Barbe. He was for some time expelled from the monastery; but after his return became prior in 1149. On St. Barbe’s death, in 1152, he led in the election of De Puiset, Stephen’s nephew, and supported him against the Archbishop of York, by whom he was excommunicated and sentenced to a penitential flagellation at the door of Beverley. Nevertheless, he stood firm to the election, and was one of those who accompanied De Puiset to Rome, and witnessed his consecration by the Pope. He did not, however, live to return to England, but died in France; and his bones only were laid at Durham.

The Dialogues are but one of several of his poems. They may be referred to the first half of the twelfth century, when their author was probably resident within the castle with Bishop Rufus, and must have been very familiar with that nearly completed structure.

The castle still retains many of the features and some of the buildings described in the poem. The ditch which cut off the fortress from the cathedral is, it is true, filled up, and the pasture ward to the east is built over and obscured, but the south gate, though rebuilt, stands on the old site, and is still the main entrance; and the wall on the right on entering still extends towards the keep. The keep itself is a late work; but the mound upon which it stands is a part of the original fortress, and the masonry is laid on the old lines, and in outline the tower no doubt represents pretty clearly the work of Flambard. A strong wall still connects the keep with the lodgings of the castle, and forms the front towards the river. The chapel also remains but little altered, and the walls and arches of the dormitory are original. The well is still seen in the open court, and is, or was recently, in use. Notwithstanding various repairs, rebuildings, and additions, there can be but little doubt that the Castle of Durham resembles in its general aspect the fortress of the Conqueror and of Flambard; nor is there in England any more remarkable example of a Norman castle of the shell-keep type. The publication of the description of it by Laurence possesses, therefore, a peculiar charm; and this must be the excuse for the following attempt at its translation. The poem is here and there very obscure, and occasionally scratches Priscian’s head; and it may be that I have misapprehended one or two lines in the original:—

DESCRIPTIO ARCIS DUNELMENSIS, “LAUR. DUNELM.,” I.L., 367.

Arx in eo regina sedens sublime minatur, Quodque videt totum judicat esse suum. Murus et a porta tumulo surgente severus Surgit, et exsurgens arcis amœna petit. Arx autem tenues condensa resurgit in auras, Intus sive foris fortis et apta satis. Intus enim cubitis tribus altius area surgit, Area de solido facta fidelis humo. Desuper hanc solidata domus sublimior arce Eminet insigni tota decore nitens. Postibus inniti bis cernitur ipsa duobus, Postem quippe potens angulus omnis habet. Cingitur et pulchra paries sibi quilibet ala, Omnis et in muro desinit ala fero. At pons emergens ad propugnacula promptos Et scandi faciles præbet ab æde gradus. Cumque venitur eo via lata cacumina muri Ambit, et arcis ita sæpe meatur apex. Arx vero formam prætendit amœna rotundam, Arte, nitore, statu, fortis, amœna, placens. Hinc in castellum pons despicit, atque recursus Huc et eo faciles pons adhibere solet: Largus enim gradibus spatiatur ubique minutis, Nec se præcipitat sed procul ima petit. At prope murus eum descendit ab arce reflectens In zephyrum faciem flumen ad usque suam. Cujus ab āēra largo sinuamine ripa Se referens arvum grande recurvus obit. Obditus et siccis aquilonis hiatibus arcem Exsurgens repetit fortis ubique feram. Nec sterilis vacat aede locus quem circinat alti Ambitus hic muri; tecta decora tenet. Consita porticibus duo magna palatia præfert In quibus artifices ars satis ipsa probat, Fulget et hic senis suffulta capella columnis, Non spatiosa nimis, sed speciosa satis. Hic thalami thalamis sociantur, et aedibus aedes, Et datur officio quælibet apta suo. Hic vestes, ibi vasa nitent, hic arma coruscant, Hic (_sic_) æra latent, hic caro, panis ibi. Hic fruges, ibi vina jacent, hic potus avenæ, Hic et habet propriam munda farina domum. Cumque sic hinc domus atque domus jungantur, et aedes Ædibus, inde tamen pars ibi nulla vacat. Castelli medium vacat æde, sed exhibet altum Ille locus puteum sufficientis aquæ.

Queen-like the castle sits sublime, and frowns O’er all she sees, and deems the whole her own. Straight from the gate the gloomy wall ascends The mound, and thus the stately keep attains. A close-built citadel, piercing the clear air, Outside and inside strong, well fitted to its use. Its base, of heaped-up earth three cubits raised, Solid and firm, the floor does thus support; On which firm base the supereminent keep Rises, unrivalled in its glittering sheen. On twice two timbers stayed, are seen to rest The buildings there, for each main angle one: While round each half circumference are wings, Each ending in a formidable wall. Springing from these a bridge, by easy steps, To the high battlements an access forms, Where the broad wall all round gives ample path, And thus the summit of the keep is gained. Stately that keep! a circle in its form, Splendid and strong by art, and by position fair. Thence, downward to the castle, leads the bridge, And offers easy access to and fro; For broad its path with many a shallow step, The base attaining by a gradual slope. Hard by, the wall, thrown backwards from the keep, Faces the west towards th’ encircling stream, On whose high bank continued, it enfolds With a bold sweep an ample pasture there; From parching northern blasts protected thus, And so curves round to the stern keep again. Nor does the space within the wall embraced Stand without buildings: such there are, and good. Two porches to two palaces belong, Of which the work to th’ artist brings no shame. Here, too, a chapel fair six columns boasts, Nor large, nor small, but fitted to its needs. Here beds lie near to beds, and halls to halls, Each for its province suitably disposed: Robes here, bright vessels there, here glittering arms, Here bread, there flesh, and tempting coin concealed, And corn and wine laid down, and barley beer, And the clear flour here finds its proper bin. Thus on one side house joins to house, and hall To hall. The other too is occupied. The court alone is free, and there is seen The well, full deep, with water well supplied.

EATON-SOCON CASTLE, BEDFORDSHIRE.

The Ouse, rising in the shires of Northampton and Bucks, and finally falling at King’s Lynn into the head of the Wash, flows deep and sluggish past Bedford, St. Neots, and Huntingdon, intersecting broad tracts of low and level land, now fertile meadow, but formerly almost impassable swamp, opposing great difficulties to the march of an invading force, especially if advancing from the eastern coast. At Eaton-Socon, between Bedford and Huntingdon, and a little above the town of St. Neots, the Ouse impinges upon the rising ground to the west, upon which stand Eaton Church and Village, and which afforded facilities for the construction of a large and lofty earthwork.

This earthwork, known as the Castle Hill, is placed upon the west or left bank of the river, about 30 yards from its present brink, and a furlong or so from the fine parish church. It is possible that when the earthwork was first formed, the course of the stream lay at the foot of the banks. At present there is a large mill upon the river, a few yards above the castle, the leat of which is reunited to its parent stream opposite to the south-eastern edge of the fortress.

The appended sketch shows the general plan of the place. It is, roughly speaking, a triangle, the east side, of 160 yards, resting upon the river, and the north side, of 140 yards, projecting from it at a right angle. The third side, or hypothenuse, is convex and irregular; measured upon the curve, it is in length about 220 yards. The area, complete, is about 3½ acres.

The work is composed of three parts—an Inner, Northern, and Outer ward. The inner and north wards lie side by side upon the river, separated by a cross ditch. The two are contained within another ditch, which communicated at each end with the river. Beyond this, covering the south-western front, is the outer ward, and beyond this again the outer ditch, which commences at the south-east corner in the mill leat, covers the south-western front, and at the north-western angle sweeps round to join the ditch already mentioned, and thus, through it, to communicate with the river at the north-east corner of the work.

The _Inner Ward_ is about 45 yards north and south, or along the river front, by 54 yards east and west. Its figure is a rectangle, with the angles so rounded off that its aspect is almost as much that of a circle as of a square. Towards the river is a steep slope of about 20 feet. On the other sides a similar slope falls towards the ditch. On the crest of the slope on these three inland sides is a bank of earth about 8 feet to 12 feet high, especially strong at the north-west corner. The entrance was at the south-east corner. The inner area or platform of the ward is about 15 feet above the level of the exterior soil. On this platform, about 8 yards from the north-east corner, is a low circular mound about 5 feet high and 40 feet diameter upon its table-top. It may have been somewhat higher. It has no ditch of its own. The ditch of this ward is from 40 feet to 50 feet broad, and still contains water.

North of this is the _North Ward_, above 35 yards north and south, by 80 yards east and west. This also has a ditch about 40 feet broad, on the west and north fronts. From the inner ward it is separated by the cross ditch common to the two, and towards the river is a steep slope about 20 feet high. Besides these defences, the slope on the west and north is crested by a steep bank of earth, and towards the south, or inner ward, is one somewhat slighter.

The _Outer Ward_ is in figure long and curved. Its breadth ranges from 29 yards at the north end down to 22 yards near the south-east end, beyond which it terminates upon the mill leat in a point. This ward is separated from the other two by a common ditch, which communicates with the cross ditch, and thus at three points, directly or indirectly, with the river. The south-eastern end opens at the junction of the leat with the river. The leat covers this end of the ward, and from the leat springs the outer ditch, from 40 feet to 50 feet broad, which covers the outer front of the ward, and at its northern end, sweeping round by a sharp angle, is continued till it joins the north ditch, of which it thus forms a part. The entrance was at the south-east corner, where a modern causeway crosses this outer ditch. The road thence skirts the edge of the outer ward, along the margin of the leat, and thence, by a second causeway, crosses the inner ditch, and, ascending the slope, gains the inner ward. No doubt these causeways represent drawbridges. It does not appear how the inner and north wards communicated. The bridge between the latter and the outer ward was probably at the south-west corner of the northern. The ditches are in parts reduced in depth, and evidently were originally fed from the river. There is not a trace of masonry, but depressions in the bank of the north ward seem to indicate towers at its north-east and north-west angles. It is convenient to use the term angle, but the lines are more or less curved and irregular, and are largely rounded where they meet. The plan has, in fact, nothing of the squareness of a Roman work, and the rounding of the angles is quite different. It is probable that the tendency to sharpness in its outline is due to the walls and towers of the work in masonry, which, though now gone, is reported to have been at one time present.

Eaton, or Eiton, appears in Domesday as held _in capite_ by the Bishop of Bayeux, but there is no mention of a castle either then or at any later time. The Beauchamps, its later lords, do not there appear as Bedfordshire landowners. In the “Liber Niger,” however, about 1165, Simon de Beauchamp held a barony, under which Hugh de Beauchamp held one knight’s fee.

This Hugh was of Eaton. He was the eldest son of Oliver, a cadet of Milo de Beauchamp, of Bedford. Oliver and Hugh were founders of Bismead Priory, in Eaton parish. Hugh was a considerable baron. He was Custos of Rhuddlan Castle 3 Henry II., and, 22 Henry II., one of those who conveyed Henry’s daughter to Palermo, on her marriage with the King of Sicily. He was slain in Palestine, 33 Henry II., and succeeded by Roger, his brother, and he by a third brother, William, whose son John inherited, and, 6 Henry III., was seized of the family manors of Eaton and Sandy. His son was William, living 42 Henry III., father of Ralph, summoned as Lord Beauchamp of Eaton, and who died 21 Edward I., holding one fee in Sandy and Eaton, no doubt that held by Hugh, his ancestor. He held Eaton _in capite_, “per baroniam.” His son, Roger de Beauchamp, of Eaton, was the last summoned. He was aged 21 at his father’s death, and his descendants have not been followed up.

44 Edward III., John de Goldington and Jocosa his wife, John Hemmyngford and others, were seized of a third part of Eaton juxta St. Neots, and half of Sandy. 1 Henry IV., Sandy was held by Katherine, wife of Sir Thomas Engayne; and 1 Edward IV., Lord Zouch had Eaton. Long after the extinction of the barons, the barony remained, and Wyboldeston Manor was held of it by Sir John de Greystock. In the reigns of Henry VIII. and Elizabeth, Eaton was held by the Lords Vaux of Harrowden, who sold it in 1624. They were cadets of Vaux of Gillsland, and are said to have inherited Eaton from the Beauchamps.

Although the manor of Eaton is not unfrequently mentioned in the records, we never hear of the Castle, nor is the addition of “Socon” ever used. Leland writes of “Eiton a good village in interiore ripa” [of the Ouse] ... “where be seen vestigia castelli between the church and the ripe, and almost hard on the ripe.... The ruins of Eiton Castle long to Lord Vaux.” In Leland’s loose language these terms may apply to the earthworks only. It is, however, probable, from present appearances, and from the place having for six descents been the chief, if not the only seat, of barons of the realm, that there was a castle, and Speed, writing a little later, names Woodhill, Tempsford, Eaton, and Ampthill, as the castles of Bedfordshire. It is possible that a very moderate excavation would lay bare the foundations, though no doubt, building materials being scarce in these parts, those who dismantled the castle would make a clean sweep of the masonry.

The position is a very strong one, and the earthworks much resemble those of Huntingdon, and are probably of the same age. The mill is no doubt a very old one, though this does not apply to the present structure.

THE CASTLE OF EWIAS HAROLD, HEREFORDSHIRE.

The “Castellaria Aluredi Ewias” of Domesday was a tract, the particulars of which are not known, but which no doubt lay among those lines of hill and valley which converge like the fingers of a hand upon the Worm and the Monnow, between the Golden Valley and the Black Mountain, and form the south-western portion of the county of Hereford. The actual castle, “Castellum Ewias,” stands about six miles within the border, and about three miles outside or west of the presumed line of Offa’s Dyke at this point. The country is hilly, but fertile, well worth the defence, for which it affords many natural advantages. The immediate position is chosen with great skill, though it required an immense application of human labour to make it an almost impregnable fortress against the fierce and active hordes of Welshmen whose alienated patrimony it was intended to grasp. While the Mound of Builth remains an evidence of English rule, that of Ewias can scarcely be regarded as the advanced post, the “Castle Dangerous” upon the British territory; but it must nevertheless at all times have been a post of very great danger, and have borne, with Kilpeck, the brunt of the ordinary and frequent attacks of the men of South and West Wales upon Hereford.

In selecting the position advantage was taken of a tongue of high land, broad towards the west and north, but which came rapidly to a narrow and almost abrupt termination in a point about 300 feet above and within the junction of the two adjacent streams. Of these the larger flows along the northern front of the position, and the smaller down a deep valley along its southern front. The two meet a few score yards below the high ground; and upon the left or further bank of the larger stream, and a short distance above the junction, is the church, and attached to it the village, to which the castle and its English lord have given the distinguishing name.

It was decided to convert the point or eastern end of the high ground into the proposed strong place, and to form thus, in the northern fashion, an isolated mound. With this intent a broad and deep ditch was cut across the ridge, curved so as to embrace about one-half of the future elevation. At its north end the ditch was carried straight down the hill-side towards the brook. At its south end it came to rather a sharp conclusion, running out upon a natural bank and slope. Here, however, it was in some sort resumed at a lower level, and ended in a shallow ditch at the southern or principal entrance to the castle. The part thus isolated by the ditch formed the circular base of a mound of about 120 yards diameter and about 30 feet high. This the addition of the soil from the ditch raised to about 70 feet, and thus gave it, in the military sense, a command over the adjacent part of the original ridge. On its opposite, or eastern side, the mound does not descend at once towards the junction of the waters, but at its foot is a broad semicircular platform, which covers its east, north-east, and south-east fronts, and from the outer or convex edge of which descends a steep slope towards the water, which slope is again succeeded by inclinations of a far more gentle character, not included in the military works.

A fair general idea of this stronghold may be given by supposing a circular platform of 200 yards diameter to be bordered on the east and adjacent sides by a steep natural slope falling from its edge, and on the west and adjacent sides by a steep artificial slope falling to its edge. Then on the western margin is placed a conical table mound, 60 feet or 70 feet high, and about 120 yards diameter at the base, which necessarily converts the western slope into the further side or counterscarp of a ditch, and reduces the eastern side to an open crescent-shaped platform. Such is the original plan of the Castle of Ewias, and such its present appearance after the complete removal of the masonry which for about 600 years adorned or encumbered its earthworks.

The top of the mound is oval, about 34 yards north and south by 40 yards east and west. Upon it has stood a shell-keep, either circular or many-sided, about 30 yards diameter. Although no masonry remains, the outline of the keep is plainly indicated by the trench which has been dug while the foundations were being grubbed up. The keep seems from this to have stood, not in the centre, but nearer the eastern margin of the mound, probably to allow room for a couple of exterior towers, or perhaps a gatehouse, which seems to have stood where now are some circular pits. Towers would be well placed on this the weakest side, so as to give a still greater command over the approach along the high ground. There is no trace of any regular ascent to the keep,—no mark of an original winding path up the mound, that now in use being evidently very modern. The side is so steep that no wheeled carriage could ascend it, and scarcely any heavily-laden horse. Probably the way up lay by a direct flight of steps, as at Hawarden and Carisbrook.

There is no trace of a well. The material of the keep was evidently a hard schistose bed of the old red sandstone, fragments of which are seen in the excavations.

The outer ward or crescent-shaped platform, below and west of the keep, runs out to a point towards the southern end, but to the north or north-west it is stopped at a breadth of about 42 yards by the prolongation of the keep ditch. The breadth of the ward at its greatest is about 60 yards. Along the north-west front it is strengthened by large earth-banks thrown up from the contiguous ditch, but elsewhere the natural slope of from 30 feet to 40 feet, steeply scarped, needed neither ditch nor bank. This ward had a curtain wall along its outer edge, of which the foundation diggings remain open. The north-west end was continued up the mound, and probably the circuit on the opposite side was completed in a similar way, so as to make the mound and keep, as at Tamworth and Durham, a part of the general _enceinte_. A group of excavations shows that this ward contained a considerable number of domestic buildings placed in its north-eastern and eastern part, near to the curtain wall. At the foot of the mound to the north is a sort of notch in the line of bank, possibly indicating a postern. The main approach evidently rose gradually from the village bridge, and skirted the foot of the eastern slope of the outer ward nearly to its south end, where it turned inwards and entered that ward by a roadway or slight cutting.

There is no trace of masonry to be seen within or about the castle _enceinte_; the material seems to have been in request as building-stone, and to have been everywhere collected and even grubbed up with most covetous care. There is a limekiln on the south side, near the line of the entrance, no doubt built of the materials of the castle, and a sort of house, now a shed, between it and the brook, but the material shows no mark of the tool and no old mortar.

There are some mounds between the castle and the brooks, possibly thrown up on the occasion of some attack by the enemy. On the other or high side there are no outworks nor any indications either of attack or defence.

There are no remains of the priory save what are included within the parish church. This is a good-sized building, recently repaired or restored, and in excellent order. It is composed of a tower, nave, south porch, and chancel. The nave has been so completely restored that little of old work is to be seen in its walls or roof. It is probably in substance of Decorated date, judging from the buttresses on the south side. The porch is new. The chancel has in the north wall a sepulchral recess, of Decorated pattern, covering the original recumbent figure of a female with her hands in prayer, holding what looks like a covered cup. In the south wall are two lancet windows of one light, under pointed recesses, and between them a late Decorated window of two lights, trefoiled, with a plain four-sided opening in the head. The whole is in a round-headed recess. The arch into the nave is new.

The tower is the best part of the church. It is of large size, square, and short for its size, probably having had another story. It rests upon a bold plinth, about 5 feet 6 inches high, at the top of which is a bold half-round cordon, with a band. The south-west angle is covered by the pilaster buttresses, of 8 feet 6 inches breadth, and a foot projection, which die into the tower, near the present summit. In this angle is a well-stair. In the south side is an unusually large door, of 8 feet opening, with high lancet arch. In the centre of the flat jamb on each side is a half-column, 2 feet diameter, with a water-bearing moulding, and a sort of bell-cap, with several bands of moulding above it. The arch is plainly chamfered, and the cordon of the tower is carried round it as a hood. Above this is a clumsy window of two lancet lights under a pointed head, very plain. Above this again is a small broad window, with a trefoiled head, and above all an early English window of three lights, with three-quarter shafts before each mullion, with bell-caps. In the nooks of each jamb are two similar shafts, seven in all. The head is a drop-pointed arch, plainly chamfered. There is a window similar to this in the north wall. The church contains nothing earlier than this mixture of the early English with the Decorated style. The masonry of the castle was probably, from its plan, of a late Norman, or transitional date. The earthworks are of the regular Herefordshire type; attributable to the English of the early part of the tenth century. They resemble generally, in the presence of a mound, those of Kilpeck and Builth, Caerleon and Cardiff, of Brecon, Abergavenny, and many places in this county or district. No doubt this and the similar works were thrown up when the early Saxon inroads were made into Wales, and were the strongholds of the invading chiefs.

Ewias Harold certainly does not bear the name of its founder, and that founder was probably as completely forgotten in the eleventh century as now.

There are two places called Ewias in Herefordshire, distinguished by the names of their eleventh-century owners, as Ewias Lacy and Ewias Harold. Both are mentioned in Domesday, and both as the seats of a castelry, a sort of Honour or superior lordship attached to the castle. Under the lands of the church of Hereford, we are told that “in the manors of Dodelegie and Stane are ten hydes, all waste save one in Dodelegie. Of the nine, one part is ‘in castellaria Aluredi Ewias,’ and the other in the King’s enclosed land.”

Another entry explains that Alured was Alured de Merleberge or of Marleborough, a great tenant in chief, especially in Wiltshire. We read, “Alured de M. holds the castle of Ewias of William the King. For that king conceded to him the lands which William the Earl [Fitzosbern of Hereford] had given to him. Who refortified [_refirmaverit_] this castle.” Of it held seven knights, whose Christian names are given, besides other persons. The castle was then valued at £10. Agnes, the daughter of Alured, married Turstan of Wigmore.

How or when Alured gave up the castle does not appear; but in 1100 it was held by a certain Harold, also a large tenant in Domesday, though not in Herefordshire. He is called “Heraldus filius comitis Radulphi,” and as such held Sudeley, in Gloucestershire. Earl Ralph, called the Timid, was the Earl of Hereford who was beaten by the Welsh and English forces in 1055, when his son was a mere child. Ralph was a considerable man by descent, being great-grandson of Æthelred and great-nephew to the Confessor. Harold probably obtained some of his father’s possessions when he came of age, and Ewias may have been part of them. He and his descendants were liberal donors to St. Peter’s, Gloucester, in its behalf founding the Priory near the Castle of Ewias.

The names and order of Harold’s sons are preserved in the Gloucester Cartulary, and they correct Dugdale and all other authorities. They were Robert, Roger, John (to whom his father gave Sudeley, and whose issue were barons), Alexander, and William. Robert de Ewias, the eldest, is described in the “Gesta Stephani” as “vir stemmatis ingenuissimi.” According to the “Liber Niger,” he held _in capite_ upwards of forty-seven fees, the mesne tenants of which were twenty knights. Dugdale mentions only twenty-two fees, and confounds him with a second Robert, his son, also Lord of Ewias. The elder Robert had by his wife Sybilla, Robert, and Richard de Ewias, who left a daughter and heiress, Sybilla, who married Philip Spenser, and left issue.

Robert de Ewias, the third owner of the castle, and the second baron, married Petronilla. He was living 1194–6. He also left a Sybilla, daughter and heiress of Ewias. She married, first, Robert de Tregoz; second, William de Newmarch, whom she married during her father’s lifetime, in the reign of Richard I. He was living 11 John. Third, Roger de Clifford, probably the second brother of William de Clifford. From this match spring the Earls of Cumberland. Newmarch had no children. Sybilla was dead 20 Henry III., and was followed by her son, Robert de Tregoz, slain at Evesham 1265. He was father of John and Henry, father of a line of barons who ended about 1405.

John de Tregoz died 1300, leaving two co-heirs, Clarice and Sybil. Clarice, who died 29 Edward I., married Roger la Warre, and had John, aged 23, in 1300; and Sybil married Sir William de Grandison, ancestor in the female line of the St. Johns, Viscounts Grandison. In the partition, John la Warre had the “body of the castle,” of which, 4 Edward III., he enfeoffed John de Cleydon. He died 21 Edward III. John, his eldest son, died before him, and as early as 12 Edward III. he had enfeoffed his grandson, Roger la Warre, and Elizabeth his wife, with Ewias Castle and Manor.

Roger la Warre died 44 Edward III., seized of Ewias Harold, and was succeeded by John, his son. 13 Richard II., Sir John de Montacute, sen., is seized of Ewias Harold, and three Wiltshire fees in the Honour of Ewias, and Teffont-Ewias, in Wiltshire, besides other Ewias lands in Herefordshire. 18 Richard II., these same lands were held by Margaret, wife of Sir John Montacute, Bart.; and 10 Henry IV., by Thomas de Montacute, Earl of Salisbury.

The nature of this alienation is obscure; for, in the midst of it, 22 Richard II., Sir John de la Warre and Elizabeth his wife are seized of the Castle of Ewias Harold. However, there seems to have been an actual and permanent alienation to the Montacutes; for, 7 Henry VI., Thomas, Earl of Salisbury, has Ewias Harold. Thence it passed to the Beauchamps, of whom Joan, widow of Sir William Beauchamp, of Bergavenny, had the castle, vill, and lordship in 14 Henry VI.; and, finally, the Beauchamp heir, Edward Nevile, Lord Abergavenny, died seized of the castle, &c., in Herefordshire, and of Teffont-Ewias, in Wiltshire.

THE CASTLE OF EXETER.

The Castle of Exeter is not only a fortress of high antiquity, but is in many respects peculiar. It occupies the northern angle of the city, forming a part of its _enceinte_, and it crowns the summit of a natural knoll formed by an upburst of Plutonic rock, of a red colour, whence it derived its Norman appellation of Rougemont. The knoll rises steeply on the north-east and north-west from a deep valley, but on the other two sides the slope, though still considerable, is more gradual. The sides of the knoll have been scarped, and at the foot of its upper part a deep and broad ditch has been excavated, beyond which, to the north, a second scarp descends to the bottom of the valley. Towards the south, where the ground allowed of and required it, there was a second and outer ditch. The contents of the inner ditch were carried upwards and inwards to form a high bank round the original summit of the knoll, the central part of which was thus converted into a pit, and became the inner, and indeed the only, ward of the castle. In figure this ward, taken at the level of the circumscribing bank, is something between a square and a circle. Probably its outline was governed by the natural figure of the ground, and such angles as it now has are due to later modifications of the works.

Originally then, the fortress was a hill camp, composed of a bank about 30 feet high, cresting the edge of the knoll, and outside scarped down about 60 to 80 feet deep to the bottom of a broad ditch, which again was reinforced on the less steep side by a second ditch. The main ditch towards the north-east and north-west has been filled up and converted into a broad public walk and garden, but the outer or second scarp still remains, and descends to the valley now occupied by the station of the London and South-Western Railway. Towards the south-east and south the ditch remains unaltered, and is a very fine example of an ancient earthwork. Towards the east it seems to have been filled up.

The camp thus described is probably older than the city, and was an ordinary earthwork, constructed in the usual fashion of the Britons, with one main ditch, reinforced with parts of others where needed. The main entrance was probably always on the south-eastern face, where the ground is less steep than elsewhere. Here, no doubt, a cut traversed the bank, and the ditches were crossed by narrow causeways, as at Old Sarum and elsewhere. These original works were probably British, and were no doubt occupied and slightly modified by the Romans. When the city, if such there was, of Caerwisc, was founded by the Britons, they probably made it an appendage to the south side of the camp, on the site of the present city, the spot being indicated by nature for such a purpose. The city occupies an oblong, elevated platform, contained between the Exe on the south-west, and its tributary streams, with their valleys, on the north-west and south-east, and connected with the higher and distant ground to the north-east by a long narrow isthmus, pierced recently by the tunnel of the London and South-Western Railway.

The Isca Damnoniorum of the Romans was certainly this enclosure, though, no doubt, they gave their _enceinte_ more of a rectangular figure than it afterwards maintained, and laid out the cruciform roads which are occupied by the two main streets of the present city. The camp was their citadel, and they would, of course, continue the defences of the city up its faces, so as to make it a part of the general _enceinte_. The Saxons, on their arrival, no doubt contented themselves with these previous arrangements, and made the best of them against the Danes in 876 and 894. Rather later Æthelstan walled in the city and the castle, and, amidst the varieties of ancient masonry still to be traced round the town, Mr. Freeman thinks it just possible that some of this great king’s work may be seen. These were the walls which enabled the citizens to hold at bay Swend of Denmark in 1001, when he threw up the earthworks at Penhow to the north of the city, and won a victory in the open field. When the Norman Conqueror appeared before Exeter in 1068, he approached from the north-east, and summoned the city at the east gate, just below the castle. Æthelstan’s walls were then in good order, and it was in them that the breach was effected. Probably, however, neither the city walls nor the defences of the castle were up to the Norman standard, for Baldwin of Okehampton was left in command with the usual instructions to build a castle, as the Normans understood that formidable structure. How long Baldwin contented himself with repairing the existing defences, and in what order he replaced them, is unknown, but enough Norman work remains to show the general plan upon which he or his immediate successor proceeded. A strong retaining wall was built against the face of the upper bank. This wall rested, and does still rest, upon the natural edge of the hill, and it supports, as a revetment, the made ground behind it, being about 30 feet high, and having carried a parapet of about 4 feet more. Probably this wall was carried on slowly, the old outer defences being tenable.

The earliest masonry now seen, earlier probably than the wall, is the gatehouse, which may safely be attributed to the latter part of the eleventh century. At the western angle, where the city wall joined the castle, was built a rectangular tower, the base of which still remains, and it is said at the north angle was a similar tower, the two thus flanking the north-west face. The wall had a high base or plinth, battering somewhat, and carrying the superstructure, which is vertical. There remain upon it two half-round solid bastions; one at the north end of the south-west face has three flat pilasters rising from the plinth, and is evidently pure Norman; the other, near the centre of the north-east face, is similar in pattern, but the pilasters are rather narrower and chamfered, and probably very late, or transition Norman. Most of the wall is rubble, but a portion of the north-eastern front, near the site of the castle chapel, is composed of good blocks of ashlar, possibly of the age of Richard II. The bank and wall have been removed in the centre of the north-west front, to make room for the Sessions House, an ungainly structure, ugly anywhere, but here especially out of place. The chapel stood in the court, near the western corner.

The gatehouse is decidedly original, and a good example of a rude Norman gatehouse. It is about 30 feet square, with walls 6 feet thick. At each end is a full-centred archway, of 12 feet opening, very plain, having a square rib 2 feet broad, with deep recesses or “nooks” of 2 feet on each side. The southern capital of the inner archway shows traces of Norman carving. There was no portcullis, each portal having doors; the space between the portals was covered with timber. On each of the two outer sides are two broad flat pilasters. The superstructure is lofty, and seems to have contained two stories. Above each portal are two windows, of 2 feet 6 inches opening, divided by a space of about 2 feet. The jambs are square, with a plain Norman cap or abacus. The present covering of each is formed of two inclined stones or lintels, which may be original, but are more probably late insertions. Above each pair is a larger single window. The inner portal opens at the level of the court. Outside, the ground is about 10 feet below that level. No doubt there was a drawbridge falling upon a detached pier, whence a causeway, probably with one or two bridges, crossed the ditches and carried the approach. The _enceinte_ wall abuts against the gateway flush with its inner face, so that it has a projection outward of about 24 feet, flanking the adjacent curtain. In later days, probably during the time of Richard II., two buttresses, or rather pilasters, 4 feet broad by 5 feet deep, have been built against the inner face, one on each side of the portal; and at the other end are a similar pair, but of 14 feet projection. These latter, at the battlement level, outside, are connected by a flat segmental arch; and the sort of barbican, or forebuilding thus formed contained the drawbridge, covered the gateway, and above had a flat roof, where archers could be posted to protect the approach. The old entrance is walled up, and pierced with two loops, which look early, but can scarcely be so. In the east side of the gatehouse a small doorway, in the Decorated style, has been pierced, possibly as a postern, for any lodge connected with it would have been outside the castle. The present entrance is, and for very many years has been, close west of, and outside the main gatehouse. This evidently was due to a wish to preserve the gatehouse, but to avoid the inconvenience of entering at so high a level. Probably when the new entrance was made the ditch at this point was filled up, all but a narrow gut, across which fell the drawbridge shown in the later drawings of the castle. When this was dispensed with the whole was made smooth, and Castle Street took its present aspect.

There is no evidence as to what buildings, save the chapel, were contained within the court of the castle. There must of necessity have been a hall, kitchen, lodgings, stabling, and barracks; and probably most of these buildings stood near or on the site of the Sessions House, where there seems to have been a postern gate. There is no evidence of a keep, nor, at so great a height, was any needed. Rectangular keeps, though found at Corfe, Sherborne, and Taunton, were not common in the west. A shell-keep, as at Trematon, Launceston, Dunster, Restormel, or Truro, would, in such a position, have been the usual structure; but the previous earthworks had converted the only site for a shell-keep into a pit so deep that it would have been commanded from the ramparts. Probably the Normans regarded the whole court as a shell-keep.

Whether the city walls were built concurrently with the castle is unknown. Probably they were, for the water-gate, removed in 1815, had certainly a Norman arch, as had, though later, and in the transition style, Broad Gate, of which also Lysons gives a view. These walls crossed the ditches, and abutted upon that of the castle. That from the east gate, seen in the Club Garden, has been rebuilt; but the north-west wall is very perfect, and though the buttresses on its outside are of Decorated date, as were most of the gates of the city, the substance of the wall is original, and very strong. In its base, where it crosses the ditch, it contains a hollow place, much enlarged, and said to have been a dungeon, which is absurd. It probably was a culvert or sluice-gate, to allow the ditch to be drained and cleared out, for though these ditches could scarcely have permanently contained water, a wet season would have converted them into a pond.

FILLONGLEY CASTLE, CO. WARWICK.

THE CASTLE YARD.

The scanty remains of this, the seat of the family of Hastings before they rose to their earldom of Pembroke, and now known as Castle Yard, are placed about a quarter of a mile south of Fillongley Church upon a small triangle of land formed by the beds and junction of two brooks which flow down from the south and the south-east to meet and form the point of the triangle to the north.

In the space thus protected towards the north, west, and east, was constructed a rudely oval shell of masonry about 50 feet east and west by 80 feet north and south, traces of which are seen in five masses of stone rubble-work, of which four are overthrown and a fifth remains in its proper place. A light ridge connects these detached masses, and indicates the line of the curtain of which they formed a part. The fragment standing in its place seems to have been connected with a curved recess, or, perhaps, a well-stair, but the others have been overthrown by gunpowder. The whole was on a small scale, and there are no visible traces of towers or gatehouse. The earth and rubbish heaped up in the interior of the area may cover foundations, and, no doubt, a light excavation here would disclose the plan of the whole structure and the place of the entrance, which, probably, was towards the east.

Outside the walled area the ground falls slightly, and there are traces at about 30 yards’ distance, towards the south, of a ditch and internal bank protecting this, by nature, the weakest side. The valley towards the east is occupied by one of the brooks which flows a few feet only from and below the wall. There is no trace of any second or outer circle of masonry, so that probably the castle was composed of a single ring wall, possibly set with turrets, but if so, of no great size.

The whole castle must have been a very inconsiderable place, though from the steepness of the little valleys between which it stands, and the marshy character of the ground watered by the brooks, it evidently possessed much passive strength, and if garrisoned by determined men, could have held out for some time.

Fillongley is unnoticed in any record older than the Conqueror’s Survey. We learn from Domesday that “Filungelei” contained two hides of land, of which one half-hide belonged to the Bishop of Coutances, and was held by Lewin; a second by the monks of Coventry; a third by a certain Alsi, who held it from before the Conquest; and a fourth by Robert le Despenser, which, though of less value, contained the church. Alsi may have resided upon the Castle hill. The half-hide held by the monks is identified with Old Fillongley, on the western side of the parish, and was held of them in the reign of Henry III. by Gerard de Alspath, whose name is preserved in Alspath Hall, in the south-west quarter of the parish. He held it as a fourth part of a knight’s fee. A part of this holding passed in some way from the ownership of the monks, who, in the time of Edward III., held in it one-eighth of a fee of Lord Hastings. Part of Le Despenser’s share passed to the Marmions, and thence, it would seem, to the Earls of Leicester, one of whom probably enfeoffed of it either Walter or Hugh de Hastings, one of the earlier members of that family, who long continued to hold it of the Marmions and Le Despensers. Thus the family of Hastings were of Fillongley in the reign of Henry I., and it speedily became their chief residence in Warwickshire, and the nucleus of a considerable estate. John de Hastings, 29 Edward I., had a license to crenellate “manerium suum et villam de ... Filungeleye,” though the castle is probably considerably older. With the church was the advowson which, in the reign of Edward III., Lawrence Hastings, Earl of Pembroke, sold to Clinton, Earl of Huntingdon. Fillongley, though it appears in the inquisitions of the Hastings family, is not mentioned as a castle, being, in fact, too inconsiderable to be claimed by the Crown, and so entered upon the public records, but it seems to have remained the chief, if not the only, residence of the family, until Henry de Hastings, in the reign of Henry III., married Joan de Cantelupe, who became a very great heiress, so that John, their son, the recipient of the above license, inheriting the castle of Abergavenny, resided there, as did occasionally his descendants. On the death of the last Hastings, Earl of Pembroke, Fillongley passed with much of his property to the Beauchamps, and so to the Nevilles, who held it when Dugdale wrote his very valuable history.

Fillongley Church contains no monuments of the Hastings family, who buried at Polesworth and with the Grey Friars at Coventry.

FONMON CASTLE, GLAMORGAN.

Fonmon Castle was, no doubt, built by Sir John de St. John soon after the conquest of Glamorgan; and part of the present building is original.

The castle rises from the western edge of a narrow and deep ravine, which conveys a streamlet from Fonmon village into the Kenson.

On its north front, but at some little distance from the castle, a similar steep bank slopes down direct to the Kenson, which there traverses a meadow, in earlier days probably an impassable morass. On its west and south sides the castle stands on table land, and was covered, no doubt, by a moat and outer wall. The keep, a rectangular building 45 feet high, and 25 feet north and south, by 43 feet east and west, including its walls, which are 5 feet thick, appears to be late Norman, and may be presumed to be Sir John de St. John’s work. Additions, probably of early English and early Decorated date, enclose it on the north, east, and partially on the south sides; on the latter forming a considerable wing, a part of which is a square tower which caps the south-east angle, and is a principal feature in the general view of the building. Two bow towers of the same date project from the east front. The principal additions on the north are of the seventeenth century, and were erected without reference to defence.

The outworks, with the exception of one tower which stands alone on the south-eastern front, about 140 yards from the castle, long since have given way to stabling, barns, and formal terraced gardens, most of which have in their turn disappeared. The remaining tower seems to have been the south-eastern termination of the defences of the outer court.

The St. Johns resided, more or less, at Fonmon until towards the fourteenth century, when, by intermarriage with the heiresses of Paveley, Pawlet, and, finally, of Beauchamp of Bletsoe, they became powerful English lords, and removed their head-quarters into Bedfordshire. Fonmon was probably left to the care of a bailiff, though some cadets of the family settled at Highlight, and there remained after the sale of the property to Colonel Jones, about 1655. Since that period it has been regularly inhabited by the Jones family, as it now is by their descendant and representative, Robert Oliver Jones, Esq.

The castle contains portraits of Cromwell and Ireton, and a fine one of Mr Robert Jones, grandfather of the present proprietor, by Reynolds.

FOTHERINGAY CASTLE, NORTHAMPTONSHIRE.

The masonry of this castle has been entirely removed, but the original moated mound and earth-banks remain, and the lines of the masonry may be traced by the trenches dug when the foundations were grubbed up. The mound stands on the left bank of the river Nene, and of the buildings the principal, disposed nearly east and west, was very evidently the great hall in which Mary of Scotland was tried and executed.

There is in “Ellis’s Letters” (First Series, Vol. II.) a draught plan, in Burghley’s hand, of the hall as it was ordered to be prepared for the trial.

The “great chamber” was 23 yards long with the window, and 7 yards broad. The length without the window was 21 yards. The window, therefore, was probably a bay or oriel, of 6 feet projection within. The draught plan was sent down with the figures blank to be filled in by special measurement.

At the upper, probably the eastern, end, and in front of the window, was a chair for the Queen of England. Against the wall on her left, a bench for thirteen barons; on her right, leaving space for passage between it and the wall, was a bench for twelve earls. Across the hall, 13 yards from the upper (east) end, and 8 yards from the lower end, was a bar, extending on the barons’ side nearly to the wall, but on the earls’ leaving space for passage. Within, close to the bar and opposite Elizabeth’s seat, was a chair for Mary.

The points of the compass are not marked upon the plan, but it is clear, from the disposition of the passage, that the door was on that or the north side, and most probably near to its west end. The door of a castle-hall would never have opened towards the defences of the place, but into the inner court.

The block for the execution was placed at the upper end of the hall; probably, therefore, where Elizabeth’s chair had stood during the trial.

The fireplace in which, after the execution, the block and bloody drapery were burned, probably was placed in the centre of the south wall.

Possibly, during a season of drought, some local antiquary may succeed in establishing the actual position of the door, the bay window, and the fireplace.

GROSMONT CASTLE, MONMOUTHSHIRE.

Grosmont is one of five strong places disposed along the right or south-west bank of the Munnow river, the others being, below it, Skenfrith, and above it Oldcastle, Longtown, and the fortified house of Perthir; Monmouth Castle, and the town beneath its protection, occupied the junction of the Munnow with the Wye. These are some of the fortified buildings scattered broadcast over the Welsh marshes, and especially abundant in the county of Monmouth, and the remains of which, always picturesque, are often tolerably perfect.

In the rear of these castles on the Munnow were those of Brecknock, Tretower, Crickhowell, and Abergavenny, upon the Upper Usk; and over the whole of that country there is scarce a hill-top or point of vantage which is not occupied by some defensive earthwork, showing the importance attached to it by each of the several races, Celt, Roman, Saxon, and Norman, who in turn either attacked or defended this devoted soil.

Grosmont, about four miles above Skenfrith and five below Oldcastle, is placed, like the former fortress, upon the high concavity of a sharp bend of the river, about a hundred yards from its margin. Very near to it is the fine old cross church, which, having shared in the prosperity of the castle, has escaped its decay, and still remains in tolerable repair, although requiring a few subtractions and restorations at the hand of a judicious architect.

The castle is composed of a court or ward of irregular plan, more or less rectangular, with projections upon the south side, the wall of which contains a space of 110 feet by 70 feet, strengthened on the south by a larger and a smaller three-quarter mural tower, having a gateway upon the east face, and on the west traces of a building exterior to the curtain wall. The north side is occupied by a hall, also exterior to, or rather replacing the line of, the curtain, three of its four walls forming a part of the exterior defences of the building.

The whole is placed within a ditch of great depth, and, indeed, the earthworks generally are of so formidable a character as to make it probable that they are earlier than the present building, or than any other work in masonry. The actual platform occupied by the walls, and contained within the crest of the ditch, is about 150 feet in diameter.

Outside the ditch, to the east and south, and covering the entrance of the castle, is a large demi-lune, or platform of earth, scarped towards the field, and upon which are traces of walls and a defence of the nature of a barbican. The main ditch, now traversed by a modern embankment, was evidently at one time crossed by the usual bridge, of which a part lifted. The gatehouse, if such it can be called, presents two lateral cheeks of wall, projecting on either side of the bridge, and thus forming a covered way, from each side of which a cruciform loop is directed along the ditch. The pointed vault of the entrance is broken, but there remain the ragged grooves for the portcullis, and the two holes which received the large wooden bar fastening the gate.

Entering, on the right is the shell of the hall, 80 feet long by 27 feet broad, out of all proportion to the area of the defences. The floor, of timber, was laid 6 feet above the level of the court, so as to give height to a spacious basement store-room or cellar, but which, however, has a large fireplace in its east wall. The hall has windows at each end, and four in each side; but probably only the six to the east belonged to the hall, the other two lighting a withdrawing-room. The position of the fireplace on the north side seems to mark the centre of the hall.

On the left of the entrance the curtain extends to the south-east or smaller drum tower, and probably supported a spacious lean-to roof marked by the corbels or bearers for the upper wall-plate. This south-east tower seems to have been massive, but low, and to have been altered and enlarged at the gorge, on the side towards the court, which now projects inwards in a rectangular form. When this addition was made the tower seems to have been raised to three or perhaps four stories, and near its summit is a bold cordon.

A strong curtain extends from this to the south-west drum-tower, of larger dimensions, and broken down towards the court. The floors of these two towers were of timber. Between them, and parallel to the curtain, seem to have been some buildings, probably barracks.

The buildings outside of, and built against, the west curtain projected boldly into the moat. They are in great decay. Here was the fireplace, the flue from which, wrought out in the substance of the curtain, rises above it as an elegant octagonal chimney shaft, the summit of which is crowned by the elegant lanthorn or spiracle which has so often been drawn and is so well known.

Grosmont, as it now appears, is of moderate size and much mutilated; but its towers and walls, though stripped of their ashlar, are still standing, and the earthworks are large, bold, and well-defined.

Whatever may be its ancient history, the present building presents nothing earlier than the reign of Henry III. The additions seem to have been in the Decorated style, and, probably, are of one date, that of the reign of Edward I. After the South Welsh conquest, Grosmont was one of the numerous De Braose castles, and passed by inheritance to the Cantelupes. It then fell into the possession of Henry III., who granted it to Hubert de Burgh. In the well-known war waged by the Welsh and Richard Mareschal, Earl of Pembroke, against Henry III., it was besieged by Llewelyn and relieved by the king, who occupied it as head-quarters during the latter part of the campaign. After De Burgh’s fall, Henry regranted the castle to the Earl of Lancaster, and it has since, with the somewhat earlier castles of Skenfrith and Whitecastle, remained attached to the duchy. Henry, Crouchback’s grandson, was here born, and hence styled Henry of Grosmont. Probably he was the author of the principal additions.

The adjacent church contains a late Norman font, with cylindrical base and octagonal bowl; and the pier arches of the central tower are also Pointed Norman. Most of the remainder of the church is early English, and probably of the date of the castle; but there is a Decorated north porch, and also some other parts in the same style, which may have been the work of the artist who completed the additions to the castle and designed its elegant chimney-shaft and finial.

GUILDFORD CASTLE, SURREY.

Guildford Castle, of which the keep was always the most prominent, and is now the chief remaining, feature, is in position, age, structure, and dimensions, a very remarkable fortress.

It is true, indeed, that though of great age, neither the town nor the castle have played any great part in English history. The town was never walled; the castle never stood a siege. No considerable battle was ever witnessed from its towers; no parliament or great council was ever held within its hall. Though always a royal manor, and long maintained as a royal residence, it was used also as a prison, and is but rarely mentioned, either in the records or by the chroniclers. The castle was not garrisoned in the great civil war, and so escaped being dismantled and blown up by either king or parliament. Its state of decay is due to the effects of time, powerfully aided by the local greed for building materials. Nevertheless, though wanting in many of the points of interest often attaching to English military buildings, Guildford Castle has certain peculiarities of its own not unworthy of notice, and which it is the object of this notice to set forth and explain.

The great chalk range which forms the bulwark of London and the southern limit of the vale of the Thames, from its mouth to the border of Hampshire, is contracted towards the west into a narrow but elevated ridge, which extends from Reigate nearly to Farnham and, resting upon the firestone and gault of the Weald of Kent and Surrey, supports the clay and gravels of the London basin.

This ridge, generally unbroken, is traversed by two well-known gorges about twelve miles apart, of which the eastern is occupied by the Mole at Dorking and Mickleham, and the western by

“The chalkey Wey, that rolls a milky wave.”

The Wey, the tributaries of which rise widely over much of Surrey and the eastern part of Hampshire, is, where it cleaves the chalk, a considerable stream, much less milky than in the days, or rather in the verse, of Pope; and twelve miles below the pass it falls into the Thames at Weybridge.

The town of Guildford is placed upon the right or eastern bank of the river, well within, that is, north-east of the gorge, and within and a little above the town is the castle. As the ridge is here steep and lofty and the gorge deep and moderately narrow, it might be supposed that this pass would at all times have been important to whoever wished to defend London and the Thames from invaders from the south. In position it is to the south of London what Berkhampstead Castle is on the north; both are placed in gorges of chalk, both upon tributaries of the Thames; both are late Norman castles, founded upon earlier English earthworks; and both were for centuries held direct by the Crown. Guildford, however, unlike Berkhampstead, though the nucleus of a large town, has no military history. Although Surrey and Sussex are by no means deficient in traces of early occupation, the immediate neighbourhood of Guildford is in this respect almost a blank. An early trackway has been talked of as taking the ridge of the Downs, and traces of an irregular, and therefore British, camp are said to have been formerly observed on St. Martha’s hill; but the long chalk crest and slopes of the Downs, so tenacious of the slightest works ever executed on their surface, are not known to exhibit any traces of the encampments or pits or other works usually attributed to the British, which, considering the dense forest that certainly extended to the foot of the high ground on the south side, and probably on the north, is very singular. Antiquaries, indeed, have placed the capital of the British Regni at Guildford and the city of Vindomis at Farnham, but nothing beyond general probabilities have been brought forward in favour of either supposition. It is curious, also, to observe how completely Celtic names have disappeared from the neighbourhood. Even the rivers, the first to receive and the last to change their names—the Wey, the Wandle, and the Mole—are Saxon, as are the names of Guildford, the chief town of the county and district, Farnham, the ancient episcopal seat, and the villages about them.

Neither are there any very decided marks of Roman occupation in Guildford. The Castle Hill at Hescomb, Hilbury in Puttenham, and Holmbury in Ockley, are said to be rectangular, and therefore Roman earthworks; but neither of the two great Roman roads from the south passes through Guildford. The Watling Street leads from Canterbury, by Rochester, to Southwark; and the Icknild Street, from Chichester, takes the pass of Mickleham in its way to the same destination. It is very curious that a town so remarkable in position, so strongly posted, and so directly in the way from the south-west to London, and withal so sheltered, and placed close to pastures so fertile, should exhibit no marks of occupation by the Romans or the earlier or later Britons.

The early history of Guildford, like its name, is Saxon, and, like its name, savours wholly of the arts of peace. Of the “guild,” or mercantile community, which in early times must have been established on the “Ford” of the Wey, nothing is recorded; but from the lingering presence of such names as Burgh Road, Burgh Field, and the Bury, it has been supposed that the earliest Saxon municipality was seated on the west, and not, as now, on the east bank of the river. It has been said that the cause of this was the establishment of the fortress on the east bank, and the consequent want of space for private dwellings. But the fact is probably just the reverse. A fortress, whether Saxon or Norman, would, as a rule, attract inhabitants to place themselves under its protection; and, however spacious may have been the area enclosed—and a little under six acres is the very utmost that has ever been assigned to it—there must always have been ample room between the walls and the river to the north, where the present town is located. If ever the town stood upon the west bank, the balance of probability is in favour of its having been transferred across the stream as soon as the earliest stronghold was established there.

There is reason to believe that the principal thoroughfare of the present town—the High Street—existed in the thirteenth century, and probably some centuries earlier. Guildford is a borough by prescription, and therefore may be of any English date, however early. It has paid the castle the fitting compliment of placing it on the borough shield, which bears “on a mount vert, a castle.” The town stands in three parishes: St. Mary’s, which includes the castle; Trinity; and on the west bank of the river, St. Nicholas.

The recorded history of Guildford has no ignoble beginning. It was the property of Alfred, and is first mentioned in his will, between 872 and 885. “To Ethelwald, my brother’s son,” says the great king, “I bequeath the manor at Godalming and at Gyldeford, and at Steyning.” On the death of Ethelwald, childless, Guildford reverted to the West Saxon crown. In the following century, in 1036, Guildford was the scene of the capture of Alfred, the elder brother of the Confessor, and of the massacre of his Norman attendants. As to the particulars of the event, and as to the parts played in it by Godwin, Queen Emma, and Harold Harefoot, testimonies differ, but all agree in the mention of Guildford as the place to which the Atheling was conveyed.

When the Conqueror marched northward from Canterbury, he went by the Watling Street, through Rochester, to Southwark, and thence ascending to Wallingford, turned the position of Guildford, and placed himself between it and the Thames. Its name even does not occur till late in the reign, and then only in the General Survey. From that survey, it appears that it had remained Crown property. No castle is there mentioned, but that it contained a residence is more than probable, both because it had been so long a royal demesne and from what is stated as to the Atheling’s reception there.

In Domesday Book, as now, Guildford was in the Hundred of Woking. The chief of the royal tenants was Ranulf Flambard, afterwards so celebrated both for his rapacity and his magnificence. He was rector of Godalming, and as such held lands in Guildford, which were afterwards appended to his canonry at Salisbury, to be eventually resumed by Henry II., and attached, with the castle, to the Crown. In 2 Henry II. the king gave Godalming hundred and manor to the church of Sarum, in exchange for the castle of Devizes and Rueles, or Erlestoke, then held by the bishop of that see.

The Conqueror granted a large plot of ground, upon which much of the modern town north of the castle and south of High-street now stands, to a family of the name of Testard, who held it for several generations by a singular tenure recorded in Blount, and are reputed to have built the two churches of St. Mary and Trinity for the use of their tenants—a fact which would go to show that the town was already standing within convenient reach of these churches, of which one is still mainly Norman, and of large area; and, further, makes it improbable that the castle _enceinte_ ever extended far to the north, as the Conqueror was not likely to have granted away any part of the older area. The historians of Surrey estimate the population of Guildford recorded in Domesday at 700 persons.

The internal evidence of the buildings of the castle makes it most probable that the whole of it, keep, hall, and domestic buildings, with its _enceinte_ wall enclosing above five acres, was constructed by Henry II. very early in the reign; but the castle is not mentioned in his reign, nor in that of Richard I. In the Pipe Rolls, the town appears from time to time as contributing to tallage and other imposts, and in 1 Richard I. the park is named in connexion with the canons of Sarum. It also appears from the Rot. Curiæ Regis, 6 Richard I., that an assize was held there. Henry II., probably when he built the castle, seems to have formed a royal park on the opposite side of the river, north of the Hog’s Back, the site of which is still indicated by such names as Guildford Park, Wilderness, Stag Hill, and the Manor Farm, the latter being probably the site of the royal lodge.

Captain James, who conducted the Ordnance Survey of the district and paid great attention to the ancient boundaries, and to whose researches much that concerns the castle is due, is of opinion that the area of the park was on the north, west, and east, conterminous with the parish of St. Nicholas, and that on the south it was bounded by the crest of the Hog’s Back. This tract is said anciently to have contained four manors, but at this time it is composed of three very ancient farms, all within one manor.

King John, whose suspicious nature and feverish activity led him to be always in motion, was at Guildford nineteen times in eleven different years. In 1200 he kept Christmas here and equipped his household in new liveries, which, to the king’s great but dissembled disgust, the Archbishop of Canterbury proceeded to surpass in splendour. In 1202 he was not here; but there is a charge for £6. 5s. 8d. for work done upon the king’s houses, and £1. 6s. 6d. for the transport of wine, and 4s. for the repair of the gaol in the castle. This is the first mention of the castle, and it is curious that it is connected with its use as a prison.

In 1204, John was here 9th October and 7th, 8th, 9th November; and in this year £10 was paid for the repair of the king’s houses, and £40 for the expenses of his chamber. John Fitz-Hugh, sheriff of Surrey and Sussex, 1208, 1210–11–12–13 and 1214, was then made keeper of the park. In 1205, the king was here 9th, 10th, 11th of April; 1st August; and 30th, 31st of October. On the 7th February, two tuns of wine, the king’s prisage, were sent here, and 15th May, two hundred porkers went from hence to Southampton, a supply of flesh to London, and a net to Southwark.

King John was here 28th, 29th, 30th December, 1206, and in 1207, 27th, 28th December. In this year, 28th August, the sheriff of Hants was to take certain prisoners from Sarum Castle and deliver them to the constable of Guildford Castle. This is the second mention of the fortress, and also as a gaol. In 1208, John was here 25th, 26th, 27th January; and 5th, 6th, 7th April; and in 1210, on the 1st, 2nd, 3rd January; and 8th March. Also in 1212, on the 12th, 13th, and 14th May.

In 1213, on the 14th, 15th, 16th, 17th December, King John was here. 3rd January the custody of the county of Surrey, with the castle of Guildford, was committed to Reginald, son of Reginald de Cornhill, to be held during pleasure, and John Fitz-Hugh was ordered to give it up to him. The Cornhills were a family of farmers-general of the revenues of the counties south of London, and between 1164 and 1215 Gervase, Henry, Ralph, and Reginald de Cornhill appear as sheriffs of Surrey and Sussex then combined.

In this year, 1213, also one hundred deer, “damos et damas,” were given from Guildford to the Archbishop of Canterbury to replenish his park. In 1214, 24th August, 53s. was allowed for the entertainment of the Papal legate, then on his way to revoke the Interdict. Rochester and Guildford Castles are mentioned together in this year as undergoing some repairs.

In 1215, King John was here for a whole week from the 15th to the 21st of January. His forces had been beaten at Bovines in the preceding July, and he had come to Guildford from London, after receiving the demands of the confederate barons agreed and sworn to at Bury. He was probably at that time actively employed in obtaining support from the clergy, in the hope of evading the great charter, which, nevertheless, he was forced to agree to in the following June. On the 18th November, John Fitz-Hugh, who again was sheriff, was ordered to give up the castle to whomsoever Peter Bishop of Winchester should name to receive it. It was probably made over to Reginald de Cornhill.

In 1216, John paid his last visit to Guildford, and remained there during the 20th, 21st, and 22nd of April. In the June following, the Dauphin Louis was here on his way from Sandwich. With Guildford, he held Reigate, then a castle of the Warrens, and Farnham.

Guildford, both castle and park, are mentioned not unfrequently in the reign of Henry III. In 1222, 19th November, £200 was paid for the royal expenses going thither from London. In 1223, 19th January, King Henry was at Guildford. 18th April, allowance was made for building a house of alms in the king’s court there. This was probably an office for the receipt of deodands, fines, forfeitures, escheats of felons’ goods, and other moneys accruing from incidents of feudal tenures, and, it has been supposed, appropriated to charitable uses. 14th May, works were in progress on the king’s houses, and 27th May, Richard Dale had ten marcs for repairs in the park; and again, in October, money was paid for fencing it. In 1224, repairs were done to the king’s houses, and half a marc paid for making a door. The fencing of the park was proceeded with. In 10 Henry III., William de Coniers was governor for the king, as were in 30 and 53 Henry III., Elias Mansel and William d’Aguillon.

In 24 Henry III., 4th April, the sheriff of Surrey was ordered to repair the glass windows of the king’s houses and chapel at Guildford, broken by the storm, and the houses unroofed thereby were to be restored. In 29 Henry III., the vill of Guildford is mentioned as vested in the king; the sheriff was to enclose the area by the kitchen which the king had purchased, with a wall conveniently answering to the other wall by which the said court is enclosed; and he is to repair the two piers of the king’s hall, which need repair because they are out of the perpendicular. In 30 Henry III., 3rd February, the sheriff of Surrey and Sussex is to make “a certain chamber at Guildford, for the use of Edward, the king’s son, with proper windows well barred, which is to be 50 feet long and 26 feet wide ... with a privy chamber ... so that the chamber of the same Edward be above, and the chamber of the king’s noble valets underneath, with fitting windows, and a privy chamber, and a chimney in each chamber. And he is to make under the wall towards the east, opposite the east part of the king’s hall, a certain pent-house, which, although narrow, shall be competently long, with a chimney and private chamber, for the queen’s wardrobe; and to make in the queen’s chamber a certain window equal in width to the two windows which are now there, and as much wider as may be, between the two walls, and as high as becomingly may be, with two marble pillars; and to wainscote that window above, and close it with glass windows between the pillars, with panels which may be opened and shut, and large wooden shutters internally to close over the glass windows; and to cause the upper window in the king’s hall towards the west, nigh the dais, to be fitted up with white glass lights, so that in one half of that glass window there be made a certain king sitting on a throne, and in the other half a certain queen likewise sitting on a throne.”

In 40 Henry III., these decorations and alterations were still continued, for on the 3rd January, the king being at Guildford, orders the sheriff of Sussex to deliver £100 to the wardens of the king’s works at Guildford, “to pay off certain arrears due for the same works, and for wainscoting the king’s chapel, the queen’s chapel, the king’s chamber, and the other chambers newly built there; and for making the great windows in the king’s chapel; for barring the windows of the king’s new chamber with iron; making the porch to the hall, of stone; for painting in the hall there, opposite the king’s seat, the story of Dives and Lazarus; making a certain figure with beasts on the same seat; and lengthening the chamber of the king’s chaplain there.”

Also on the 5th May following, the sheriff of Surrey and Sussex is ordered to whitewash the king’s hall at Guildford within and without. On the 17th June, 45 Henry III., 1261, the king visited Guildford, and doubtless examined and took pleasure in the various improvements and decorations he had ordered. All this luxury was probably confined to the hall and royal apartments in the middle ward, for an entry on the Hundred rolls at the commencement of the reign shows that the Sussex county prisoners were kept at Guildford, and no doubt in the keep.

In 50 Henry III., Prince Edward was at Guildford engaged in putting down Sir Adam Gordon, a soldier who, having been outlawed after Evesham fight, had turned freebooter, and made the Surrey woodlands very insecure. Edward came up with and attacked him between Farnham and Alton, took him prisoner in single combat, got him a pardon, and presented him to the queen then at Guildford. At the end of Henry’s reign, 52 Henry III., the “King’s Mills” were removed farther down the stream, probably to the site now occupied by their modern successors.

Edward I. became possessed, in due course, of Guildford, and in 27 Edward I., 1299, the park, castle, and farm of the town were assigned as part of the dower of Margaret, the king’s second wife, and on her death, 10 Edward II., they reverted to the Crown. Edward was here, 20th January, 31 Edward I., 1303, resting on his way from Odiham to Windsor.

In 35 Edward I., 1306, Henry de Say, keeper of the prisoners indicted at the Sussex Assizes, and lodged in Guildford Castle, petitioned that an officer might be sent to receive their fines and chattels, according to their offences, and that a stronger prison may be provided, the castle being insecure for so many prisoners. The answer, recorded on the rolls of Parliament, is terse:—“Si carcer sit nimis debilis, faciat, Custos, emendari; si nimis strictus, faciat elargari; quia Rex non est avisatus mutare locum prisonarum suarum: vel saltem teneat eos in vinculis fortioribus.” “Double iron the prisoners” was at one time a usual, and certainly an economical, way of securing their safety. It is probable that it was under these circumstances of great pressure that the mural oratory in the keep was employed as a prison.

In the king’s circular to the sheriff, 1 Edward II., 1307, 15th December, which was followed by the edict confiscating the goods of the Templars, the sheriff of Surrey and Sussex was ordered to repair to Guildford. In 15 Edward II., Oliver de Burdegala, governor, had a writ of privy seal directing the castle to be victualled and garrisoned.

Guildford, 2 Edward III., 1328, was the head-quarters of the sheriff of Surrey, who was ordered to go there to prevent tournaments from being held. On the 8th March, 1329, the king was at Guildford; also 28th February and 26th December, 1330; 18th–20th November, 1331; 2nd September, 1334; and 18th–24th April, 1336; so that the castle was in not infrequent use as a royal residence. In this last year the king granted the town in fee farm to the corporation, reserving only the park and castle. On the 23rd April, 1337, 11 Edward III., the king ordered that Robert d’Artoys should have a right to be hospitably received should he visit the royal castles of Guildford, Wallingford, or Somerton, and he might sport in the park at Guildford.

In the same year, 24th December, Edward was himself at Guildford, as he was in 1340, and again 27th–28th December, 1347, in which year the commonalty of Sussex petitioned that Chichester in place of Guildford might be the county gaol. The petition was set aside, and in 41 Edward III. the sheriff still held his official residence in the castle, which was the prison, as before, for the two counties. 42 Edward III., 1368, 28th July, the king was here. 43 Edward III., the custody of the castle and park was given to Helmyng Legatte for life. Edward was again here, 45 Edward III., 1371, on the 12th of May, probably for the last time.

In 1 Richard II. Sir Simon Burley was constable of the castle, and afterwards Sir Hugh Waterton, on whose death, 10 Henry IV., Sir John Stanley had the office also for life, and his appointment was confirmed by Henry V. By that time the custody of the park was evidently an office more coveted than that of the castle.

What occurred in the castle during the wars of York and Lancaster is not known, save that it was the scene of no event of importance, and it certainly continued still to play the ignoble part of a common prison, for in 3 Henry VII., 1487, the old complaint is revived, and the county of Sussex again petitions for a gaol of its own, and under its own sheriff, suggesting Lewes as a proper place. This time the prayer was granted, and probably the Surrey prisoners either then or soon afterwards were bestowed elsewhere, though the two counties continued long after this to be placed under one sheriff. As late as 1620 a Sussex gentleman, Nicholas Eversfield, was sheriff of the two counties, and the jurisdiction does not appear to have been finally divided till 1637.

Finally, in 1611–12, after having been attached to the Crown at least from the days of Alfred, or 700 years, the castle and its _enceinte_ were granted by James I. to Francis Carter, of Guildford, who died in 1617, and whose son, John Carter, is described as dwelling in the castle in 1623. His eldest son Francis died in 1668, leaving a daughter only, and it is his brother, the second son, John Carter, whose initials, “J. C., 1699,” stand upon a tablet within and above the great gateway in Quarry Street. The castle has since remained in private hands, and is now the property of Lord Grantley.

The above extracts, mainly taken from those given by Mr. Parker in his valuable volumes on Domestic Architecture in the Middle Ages, will have shown that the fittings and adornments of the castle were chiefly due to Henry III. That prince, who was a great patron of the arts, and especially of architecture and painting, paid great attention to the royal residences. Unfortunately, his decorations were for the most part confined to the hall and principal domestic apartments, but few of which anywhere have survived. At Guildford, the destruction has been peculiarly sweeping, and the only remaining structure, the keep, does not seem to have participated in the royal care. The keeps of Norman castles, inhabited but rarely, and only during a siege, even by those who built them, seem very soon to have been altogether deserted for more convenient lodgings in the lower and more spacious wards. The keep was then used as a storehouse or barrack, or, as at Guildford, as a prison, and very little was spent upon its repairs, and nothing upon its decoration. It is, however, in consequence of this neglect, that the Norman keeps, where they have not been pulled down, remain pretty much as they were originally built, or with only such additions as may easily be detected, or such diminutions as may readily be supplied. This is particularly the case with the keep at Guildford, the additions to, or alterations in, which are of the rudest character, and may readily be detected, while of the masonry of the original structure little is wanting save the parapets and angle turrets and some details connected with the approach and entrance.

Guildford Castle occupied a natural platform of nearly six acres upon the slope of the chalk hill, far below its summit, and from 40 to 80 feet above the river. The platform, inclining gently towards the stream, terminates at about 80 yards from its bank in a low cliff of from 10 to 12 feet high, in parts replaced by, and in parts resting upon, a steep, natural slope or talus, which dies away 40 feet lower down into the meads traversed by the river. The crest of this cliff or talus is occupied by Quarry Street, and forms the west front of the castle. Towards the north, the river is more distant, and the slope of the platform far more gradual. On this side, the High Street and the present town of Guildford intervene between the castle and the river.

The keep stands on the eastern and highest part of the platform, and commands the rest of the castle, as the castle commands the town; and here are what appear to be the remains of the English residence. At the foot of the steep, a mound, wholly artificial, but resting upon an inclined natural base, has been thrown up, composed of chalk, in form conical, truncated, and with a level summit, no doubt originally circular, and still nearly so, and about 90 feet diameter. The base is about 200 feet. Between the mound and the adjacent steep hill-side is the main, and perhaps a trace of a second and outer, ditch. This inner ditch, about 60 feet broad and 12 to 20 feet deep, sweeps round the foot of the mound on the east, north, and south sides, the ends dying out on reaching the platform on the west below. The ditch, always dry, has long been cultivated as a garden, and was no doubt once considerably deeper. Its north limb is partially built upon by the houses in Castle Street, and is, in consequence, nearly obliterated. It is traversed at the north-east quarter by a narrow causeway of earth, which, no doubt, represents an older causeway of stone, provided with a drawbridge, and forming a direct entrance for foot passengers and perhaps horses, to the keep. Beyond this, the east and south-east, in the extra-parochial plot called the “Bowling Green,” are very slight traces of what may have been a second and outer ditch, a not unlikely security to have been provided by the inhabitants of the mound against an attack on this the weakest, because the commanded, side.

The mound on the eastern face, measured from the scarp of the ditch, is about 30 feet high, but on the western side, where it rises from a lower level, it is about 50 feet, or 92 feet above the river. The mound and the ditch evidently supported and protected the dwelling of the English lord, and it is probable that upon the platform below, where the Norman king afterwards placed his hall and offices, were lodged the serfs and dependents of the earlier household. Judging from the close analogy of Leicester, Tamworth, Tutbury, and other earthworks of known date, the earthworks of Guildford may, with great probability, be referred to the earlier part of the tenth century.

The keep, a rectangular structure, covers the eastern slope of the mound, but is placed a little to the south of its central line, so as to allow of a gateway (now gone) at its north-east angle, and a passage up the mound outside the north wall. The east or lower wall rests on the undisturbed ground, a little above the level of the scarp of the ditch, and the west or upper wall upon the edge of the level summit of the mound, nearly the whole of which thus extends undisturbed to the west and north-west of the building. The difference in level of the base of the two faces of the keep is about 15 feet. It is exceedingly rare to find a rectangular keep placed upon an artificial mound. Guildford, Christchurch, and Clun are the only recorded examples.

The keep stands nearly by the points of the compass, measuring 46 feet north and south, by 52 feet east and west. The wall is perfect to the base of the parapet, a height on the west front of about 63 feet. The masonry of the lower side contains more ashlar, and is of better quality than the rest, to prevent the structure from slipping. Of the depth of the foundations nothing is known, but the thickness of the wall—at least 11 feet at the visible base—would serve to distribute the load, and chalk, even when made ground, does not make a bad foundation. There was, no doubt, a risk in placing so heavy a building upon an artificial hill, even though a couple of centuries old, but the result has justified the means employed, for there is not a crack or mark of settlement in the whole edifice. Grose represents some half-buried arches on the south side, not now visible, but which, if they ever existed, which is more than doubtful, might indicate that parts of the building rested on piers, carried down to the solid ground. However, enough of the wall is bared to show that this is not the case. What Grose took for an arch was probably a low course of inclined or half-herring-bone masonry. Others have described an opening on this side supposed to lead to a sub-basement vault which there is no reason for supposing to exist. The machicolations cited in evidence as defending this fabulous doorway are the vents of a garderobe in the upper story.

The four faces of the keep are generally alike. Each is flanked by two pilasters of 4 feet 6 inches wide, by 9 feet projection, so placed as not to cap the angle, but to convert it into a hollow or re-entering one. This hollow was left open, not filled up, as at Scarborough and elsewhere, by a bold bead or engaged column. In the centre of each face is a third and similar pilaster, but 5 feet wide. Probably these rested below upon a plinth common to the whole building; but if so, this is gone. Each pilaster is of equal breadth and projection throughout, having no sets-off. The central pilasters run up to the base of the parapet, now gone. Those at the angles were continued to form the usual square turrets, of which some slight though clear remains still rise above the curtain.

The material employed for the exterior is chiefly Bargate stone, from the bed representing the chalk marl immediately beneath the chalk. This is worked up as rubble, interspersed irregularly with courses of the same stone laid herring-bone fashion, for which the larger and flatter stones have been selected. The work is very rough. The herring-bone courses are laid at all heights and distances; some broken, some mere single inclined stones, and here and there, especially near the top, are occasional courses of flints, some of which look like insertions. The angles, salient and re-entering, of the pilasters, are of the same stone, cut as ashlar and well jointed; but between these quoins the pilasters are usually of rubble, sometimes herring-boned. Above the parapets the angle turrets seem to have been of ashlar. There is no string-course, shelf, or set-off upon the face of the wall. The west-central pilaster, being pierced by the entrance, is mainly of ashlar, as are the pilaster and adjacent wall, about the north-east angle of the building, where the gate of the ward seems, from traces in the masonry, to have abutted. Here, too, the joints being very wide are made good with single or double rows of thin ordinary bright-red roofing tiles. The base of the east face was repaired about forty years ago, and now has a modern ashlar plinth of about 15 feet high. The ashlar within reach on the other faces has been pillaged, and the base of the wall generally is very hollow and ragged. The hearting of the walls throughout seems composed of chalk and Bargate stone, very roughly laid and grouted.

The walls are everywhere pierced with putlog holes, about 4 inches square, indications of the method of construction, and probably originally but loosely stopped to allow the work to dry and for the convenience of future repairs. There are no large holes above, and no signs of a bretasche having been employed. Four double windows on the upper floor and one on the east face of the middle floor, though original, have been fitted up with cut brick mullions and arches of perhaps two hundred years ago, the work, no doubt, of the first purchaser. All earlier alterations seem to have been effected in stone.

Having thus disposed of the general exterior of the keep, the next step is to describe its interior details. Allowing for the removed plinth or casing, three of the faces are about 11 feet thick and the fourth or east about 14, so that the interior dimensions are 24 feet north and south, by 27 feet east and west. The building is composed of a basement and two upper stories, and the floors and roof were of timber. There is no evidence of any subterranean chamber, and no reason for supposing one.

The basement on the level of the top of the mound is about 12 feet high. The walls are pierced in the centre of the north and south faces with a round-headed recess 5 feet wide and about the same height to the springing. The sides and roof converge to an exterior loop, and the base is stepped up to it. The work is good plain rubble. The east and west walls were originally solid, and the only entrance to this floor must have been by a ladder and trap from the floor above. It was of course a store or cellar, as usual.

At a later date, a doorway, 4 feet 6 inches broad and 8 feet to the springing, has been cut through the west wall near the north end. This has a slightly pointed arch. Its masonry is of small weak rubble without any dressings; and this, and the absence of bond with the older work, show it to be an insertion. In the north-east corner, the wall has been rudely cut away to some depth to form a fireplace and an oven. The bricks composing these have been removed, and a recent pier of masonry supports the wall above. The chimney shaft of this and a fireplace in the floor above have been formed by cutting away the inner face of the wall, which has been rudely restored. No doubt all this is the work of the purchaser, who seems to have lived in the keep and converted the basement into a kitchen. In the south-west corner is a small platform of stone, said to have carried a wooden stair communicating with the floor above, and of the date of the kitchen. This is probable enough. One of the stones is a late Norman capital, brought from some other part of the castle. The whole interior of this basement is rubble. It contained neither fireplace nor garderobe. The two loops, its only light, are about 18 inches high, and were probably 4 inches broad, though now increased by weather to 6 inches.

The first or state floor was about 30 feet high, fairly lighted, and contained various mural chambers. In the centre of the west side was the entrance from without, and in each of the other three sides a window. These were of two lights, or rather composed of two tall, narrow, round-headed windows, coupled under one round head outside and a similar recess inside. These recesses commenced about a foot above the floor level, and are 4 feet 4 inches wide and to the springing about 12 feet high. Their sides are parallel, not convergent, and each contained four steps ascending towards the window. There are no mouldings or decorations, but the quoins are ashlar. The window, arches, imposts, and jambs are plain and good. The central window piers or mullions are gone, but in two cases the small head arches remain. In the third case, that in the east face, the window has been removed and replaced by one in brick, but the recess is untouched.

The entrance is 3 feet 4 inches broad, 9 feet high, and about 14 feet from the ground. It is lined with good ashlar; but with a barrel-vault, round-headed, in rubble. The outer portal occupies the whole breadth of the central pilaster, being about 5 feet wide. It is very slightly, but decidedly, pointed. There is no portcullis groove, and but one, an outer, door, well strengthened by bar holes. Below the springing are two small holes, now stopped, for an iron bar, rather low for a centring, and possibly connected with a light drawbridge. The door is in the centre of the west face, as is the opposite window of the east face; but the north window is at the west end of its face, about 3 feet from the corner, and the south window is placed diagonal to it at a similar distance from the south-east corner. The three windows were all of one pattern.

Besides these openings, there are at the same level three mural chambers and a staircase. The principal chamber occupies the south-west angle, and is in plan a right angle with two limbs, like the capital