Volume 2 — HEREFORD CASTLE.

THE borderers of England, whether on the north or on the west, were bold, hardy, and aggressive races of men, having for their patrimony the moor and the morass, the wild bleak upland which occurs on the Scottish frontier, or the steep crags and rugged defiles which gave strength and, in ordinary cases, impunity to the Welsh. The Scots, indeed, came of the same stock with the English, and had recently formed a part of the same kingdom, under the same lords; nor was the land immediately south of the Tyne or the Tweed materially richer than that to its immediate north. But the Welsh had other more potent and better causes for their continual and ferocious transgressions. They were the earlier possessors of the whole country, a possession still attested by the names of hills and rivers, and, indeed, of many of the chief cities of England. They were of an entirely different blood and language from either Saxon or Norman: by the one they had been gradually driven into the western and less fertile tracts of the island, and the other held them cooped up and at bay within their mountain fastnesses, and responded to their continual partisan warfare by occasional invasions on a large, and, for the time, irresistible scale.

Full in view of the Welsh mountains were spread out some of the most fertile lands in Britain—cultivated, inhabited, and fortified by the invaders, who, not content with the slopes of the Cotteswold and the velvet meads on the left bank of the Severn, had pushed their conquests far beyond the right bank, holding the stately ridge of Malvern and the rich pastures of the Teme, the Lug, the Frome, the Worm, and the Wye, even up to the Dyke of Offa. That dyke, still to be traced along much of its length, between the estuaries of the Severn and the Dee, was felt to be a perpetual mark of inferiority, an abiding affront to the patriotic feelings of every Welshman. No marvel, then, that the Welsh were ever insurgent, ever breaking forth with fire and sword into the English territory, and especially into the most rich and, by nature, least-protected tract of it, the fair county of Hereford.

Herefordshire was first subjugated by the Romans. Three miles north of the city are the remains of a camp, probably British, covering about thirty acres of ground, and now known as Sutton Walls, a little south of which runs the line of the road which led to the Roman town of Magna Castra, four miles to the west, and was continued onward in the same direction.

At what precise period Herefordshire passed under the Saxon yoke, and became annexed to Mercia, is unknown—probably under Creoda, early in the seventh century. The see was founded in 680, when a synod was held here under Putta, the first bishop. Sutton, or Southton, so called from its position close south of the Roman camp to which it has given its name, was at a very early period a residence of the Mercian kings. Here, in the middle of the eighth century, resided the celebrated Offa, probably during the construction of his dyke, which passes about six miles to the west; and here in 794 was committed the murder of Ethelbert, his intended son-in-law, a deed which led to the aggrandisement of the cathedral church, to which Offa’s penitential donations were largely paid. A charter of King Cœnulf, A.D. 799, mentions Hereford, as do others by Deneberht, in 802 and 803, the last recording certain “monasteria quæ olim in antiquis dictis ad Herefordensem ecclesiam præstita fuerunt.” Sutton continued a Mercian palace until the union of the seven kingdoms, under Egbert of Wessex, in 827.

Edward the Elder, a great constructor of strong places, who repaired the citadel and walls of Chester in 909, is said by Grafton also to have fortified Hereford, erecting a strong castle there. His sister Ethelfleda, a still greater castle-builder than he, and to whom are due the mounds of Tutbury, Tamworth, Leicester, and some others, was the widow of Ethelred, Duke of Mercia, in 915, when the Danes had advanced by the Wye, and taken captive the Bishop of Archenfield. She attacked and defeated them before Hereford, no doubt making use of the new defences. She died shortly afterwards.

Early in the eleventh century, the Saxon sway had been so far extended as to have left the Dyke behind, and Wales was regarded as a part of England, and an attempt made to force it to contribute to the common tax for the defence of the island against the Danes. For this purpose Edwin, the Ealderman of the Mercians, led an army as far as St. David’s, punishing the people for their refusal. At that period, and during the reign of the Confessor, the Welsh were led by Griffith-ap-Llewelyn, the ablest of their princes, and, indeed, the only one who ever held the Welsh together, or gave the English cause for anything like serious apprehension. Aided sometimes by the Danes from Ireland, sometimes by traitorous Saxon chiefs, and sometimes even by the discontented Normans, he is found during the first half of the eleventh century waging on the whole a successful war against Herefordshire. It is certain, from the number, magnitude, and character of the existing earthworks, that Irchen or Archenfield, called from its beech-trees “Trefawrth” by the Welsh, and sometimes by the English “Fernley,” had from a very early period been inhabited by the English; but either distrusting these, or because he wished to quarter afar off the visitors whom he attracted, but of whom he was afraid, the Confessor made large grants of land along this district to his Norman courtiers, with whom, therefore, Griffith had not unfrequently to deal.

The earliest of these grants seems to have been made to Richard Fitz-Scrob, whose fortress, built after a fashion till then unknown in England, gave great and general offence. His original castle has been replaced by later structures, now also in ruin, but the earthworks are probably original; and the name of Richard’s Castle shows how deeply the fear of its builders was impressed upon the people; and it is, moreover, a very rare example of a parish bearing a purely Norman name.

These grants were opposed by Earl Godwin and his sons, and it was to enforce his remonstrances against them that the English Thane led a force from Beverstone, and challenged the Confessor to give up his stranger favourites—a struggle which finally ended in the temporary banishment of that truly English noble.

In 1052, during Godwin’s exile, Griffith invaded Herefordshire, and advanced as far as Leominster before Fitz-Scrob and his Normans were in the field to meet him. They were beaten in a pitched battle, and upon open ground.

In 1055, the earldom of Hereford was in the hands of Ralph, surnamed “the Timid.” Griffith, uniting with Ælfgar, the Saxon lord of the East Angles, who was accompanied by an Irish force, burst into Archenfield, and again laid waste the border. Two miles from Hereford, Griffith and Ælfgar were met by Earl Ralph, who seems, with the Norman contempt for infantry, to have placed undue weight upon his cavalry, the result of which was a complete defeat. Griffith entered Hereford, which was undefended, sacked and burnt the city, treated the cathedral and the clergy with excessive severity, and destroyed what the “Brut” calls the gaer, that is, the castrum or fort. The account seems to imply that the city was not then fortified. Mr. Freeman thinks the gaer was a work of masonry. However that may be, there is little doubt but that it was a work on the site of the later castle, for by the riverside, for defence, it would certainly be placed; and as the position of the cathedral has doubtless always been the same, there would scarcely be room for a fortress between its precincts and the western marsh. The position of the bridge, too, is probably ancient, and this would lead into the city, not into the gaer. The appearances of the earthworks, as they existed before the removal of the mound and the filling up of its proper ditch, much resembled those of Tamworth, and other works attributed to Edward and Ethelfleda, and the gaer may well have been of that date, so far as the earthworks were concerned. This inroad of the Welsh in 1055 was the most severe and the most lasting in its effects of any on record. All Archenfield suffered, and traces of the spoiler are recorded long afterwards in Domesday.

Although Godwin’s return from banishment in 1052 had been followed by the putting forth of most of the Norman intruders, Richard Fitz-Scrob, one of the most offensive, seems to have remained, and long afterwards to have put down Edric the Wild; and his son Osborn, after a short exile in Scotland, came back to Herefordshire, and held office and dignities both before and after the Conquest.

In 1055 Godwin was dead, and to Harold, as Earl of the West Saxons, it belonged to redeem the disgrace incurred by Ralph the Timid. He lost no time in preparation. In the course of the same year he mustered his forces at Gloucester, and by his mere presence cleared Hereford of the Welsh and Ælfgar. He at once fortified Hereford. Whether he restored the castle is unknown, but he surrounded the city with a wall, no doubt along the line of the later structure. Mr. Freeman supposes Harold’s work to have been a mere “dyke of earth and loose stones,” Florence of Worcester describes it as “Vallum latum et altum.” Domesday, however, records a “murus” at Hereford as having stood in the time of the Confessor, so that Harold, when Hereford came under his immediate government, may, as Mr. Freeman suggests, have replaced his vallum with a wall of masonry. Harold’s defences probably did not include the suburb, which even then must have existed, since we read in Domesday of burghers within and burghers without the walls; though the latter would derive a not incomplete protection from the broad belt of marsh which then surrounded the city.

Griffith had sought and received terms of peace. Nevertheless it did not suit him to allow Hereford to become a strong post. Early in 1056 he again crossed the border. He was opposed by Leofgar, the new bishop, who, however, was slain in the first combat. His successor, Ealdred—a man of equal determination with better fortune,—held the Welsh in check, and negotiated a peace, and the fortifications of Hereford were completed.

In 1062 Griffith again appears upon the scene. Probably he traversed Herefordshire, for he crossed the Severn in the diocese of Worcester. On this occasion Harold—appearing, not as the defender of this or that province, but of the whole kingdom—executed a counter movement, and invaded North Wales. This was followed in 1063 by Harold’s great invasion, in which Griffith was murdered by his own countrymen, and Wales submitted, having deprived herself of her greatest son. It was at the conclusion of this war that Harold employed himself in constructing a sort of hunting-lodge for his sovereign in the low lands of Gwent, at Port-skewet, where earthworks are still seen. The lodge was attacked and destroyed while in progress by Caradoc-ap-Griffith-ap-Rhydderch, of South Wales.

Hereford played no part in the Conquest; but the city and shire occupy a respectable place in the Domesday Survey, where the customs are related in great detail. The king had six moneyers there, and the bishop one. Of the city burgesses 103 held of the king, and twenty-seven had held of Earl Harold. All the tenants of the burgh were liable to military service against the Welsh.

The customs called of the Welsh in King Edward’s land in Arcenefeld, or Irchenfield, also there recorded, are curious, and show not only the early existence of a local militia to resist the Welsh, but that the people were mostly of Welsh blood, and were employed against their countrymen. The king held three churches there, and the priests of them were to be the king’s legates or ambassadors in Wales, and when the army marched against an enemy, to the men of Archenfield was committed the post of trust and danger. During the advance they were to form the “Auantwarde,” and in retreat the “Redrewarde.”

Such, then, having been the antecedents of Herefordshire, it is not to be wondered at that it bristled with strong places, nearly all of which show indications of early dates, and in many may be traced the mount or motte, which in England there is strong reason to regard as in favour in the tenth century. Domesday, usually so silent as to fortresses, and enumerating only forty-nine in all England, mentions in this county eight, and two strong houses. There were, however, many more, and at this day there remain traces more or less considerable of twenty-eight, of which many preserve the mount, and others earthworks of an early character. Similar works are found at Brecon and Builth, places known to have been held by the English at an early period. No doubt these strongholds, originally strengthened with timber, were burned again and again by the insurgent Welsh; but the positions of most of them were well chosen, and each had its surrounding estate, so that when, after the Conquest, the great Norman barons marched into Wales, they constructed upon these sites castles of masonry after their fashion, of which a few remain, though many, having been destroyed, have been rebuilt in the reign of Henry III., or later.

William’s arrival no doubt confirmed and extended any local power that may have been allowed to Richard Fitz-Scrob and his son, under Harold. Osborn, as sheriff, held Hereford, and either he or the first Norman earl probably rebuilt the castle in the Norman manner.

William FitzOsborn, the great Norman chief, “Magister militum bellicosus,” Earl of Hereford from 1067 to 1071, was a fearful scourge of the Welsh, whom he drove back and vanquished on the banks of the distant Rhymny. To him are attributed the castles of Strigwil, Clifford, Wigmore, and Ewias—that is, the Norman part of them, for some at least preserve older earthworks.

Roger de Bretuil, William’s third son, succeeded him in the earldom. He plays no part in the history of shire or castle. Failing in rebellion in the eastern counties, he ended his days in prison.

When Bretuil died is unknown, but in 1138 Hereford, commanded by William Talbot, held out successfully against Stephen during a long siege. The defence was much aided by Milo, the Constable of England, who received the earldom of Hereford from Maud, in 1140, and it remained in his family till the end of the century. Milo’s patent gave him the moat, probably the “motte,” and all the castle. As earl, however, he was unfortunate. Stephen returned and took the castle, and wore his crown in state in the cathedral.

Roger, son of Milo, was in opposition to Henry II., but escaped by the wise counsels of Foliot, the bishop. In the reign of John, a less discreet or bolder prelate—Giles, Baron de Braose—united with Llewelyn; but failing to bring over the men of Hereford to support him, died in exile. The castle seems then to have fallen into the possession of the Crown, and so to have remained. Prince John gave the custody of it to Roger Bigod, and in the sixth of his reign as king, and in the seventh of Henry III., William, brother to Thomas, Lord Cantelupe, and sheriff, was also governor. The turbulent reign of Henry III. gave value to the castle of Hereford. 15 Henry III., John Fitz-Terrick and William de Stowe, surveyors of works at Hereford Castle, were allowed £20 for repairing the walls. 16 Henry III., Terrick and Roger Carlton were surveyors of mangonels and petards within the castle. 26 Henry III., John and Roger le Werrur were surveyors, and had spent £7. 8s. 6½d. on artillery, and £12. 1s. 4d. in making a trebuchet called “Blythe,” and £40. 14s. 4d. in building a tower in the castle. 42 Henry III., the castle was in the hands of the sheriff, Henry de Pembridge, until the barons forced the king to appoint John de Grey.

44 Henry III., Roger Mortimer was governor, and penetrated into Wales as far as Builth. He was, however, in turn attacked, and held the castle, though for a few weeks only, against Simon de Montford and Prince Llewelyn. Bishop Acqua Blanca was taken and imprisoned at Eardesley. Peter de Montford, the Earl’s son, had charge of the county, and employed its issues in the repairs of the castle. After the battle of Lewes, Prince Edward was a prisoner here.

Edward’s escape hence has often been related. Widemarsh, whence he galloped off, still bears that name, and lies about a mile north of the city.

During the whole of this time, from 1199, the title of Earl of Hereford had been borne by Henry de Bohun, whose mother was a co-heir of a previous earl, but neither he nor his celebrated descendants, Earls of Hereford and Essex, although they held Huntington Castle and the lordship of Brecknock, and built the Castle of Caldecot, ever seem to have been seized of that of Hereford, which remained in the Crown.

Soon after the time of Edward’s flight, John le Werrur and William Valet were surveyors of works for the king, at Hereford, and charged £45. 6s. 10½d. for repairs of the walls and of the king’s houses. William Capon also held lands at Marden, by the serjeantry of keeping the door of the castle, and this, under Henry III., was commuted into military service in the field.

Possibly the castle was employed as a local treasury; for, 1 Edward I., Henry Pigot held 46 acres in capite by the tenure of transporting the king’s treasure from Hereford Castle to London. In the same year Richard Porter and Richard le Panner, surveyors, charged £21. 1s. 6d. for expenses and repairs of the king’s house in the castle, and in the next year they had fifteen oaks from the king’s forest of Haywood, a mile south-west of the city, allowed for repairs.

2 Edward I., Giles Berkeley, sheriff, had spent £4. 15s. in repairs, and in 7–8 Edward I., Roger Burghull, his successor, had spent £2. 5s. 4d. Hugh Turberville, one of a very unruly race, had, it appeared, when sheriff, burnt and destroyed in the castle the king’s house, and certain engines of war and military stores, to the value of £100. For this he had a pardon, November 5, 48 Henry III.; but the debt remained, and, 9 Edward I., the sheriff was ordered to distrain upon his goods.

12 Edward I., a new chapel was erected in the castle, and the barons of the exchequer, by the king’s precept, allowed Roger de Burghull £10. 6s. 8d. on that account. 15 Edward I., Henry de Solers, a Herefordshire man, had charge of the castle, with arms and stores. 32 Edward I., Miles Pychard charged 40 marks for repairs, which were disallowed because not certified by the surveyor. John de Acton was governor, 33 Edward I.

1 Edward II., Alan Plukenet, keeper of the king’s forest of Haywood, was to allow twelve oaks and stone for the repairs of the castle walls and towers. Ralph Freeman held lands in capite, in Fromyngton, by the service of carrying the cord round the castle walls when they were measured. 5 Edward II., he paid half a mark for his relief, and commuted future payments for 7s. 7d. per week. 10 Edward II., Hugh de Hacluyt charged £5 for repairs.

13 Edward II., Sheriff Richard Walwyn charged £6. 13s. 6d. for repairs. At the close of the reign the queen held a great council at Hereford, and Hugh le Despenser, the younger, was hanged upon a tall gibbet, outside Friar’s Gate. During the reign of Edward II., Wales was loyal, and Hereford, therefore, neglected. John of Gaunt was its governor, 1 Richard II., but even his great love of building was not exercised here. As late as 8 Henry VI., the city had a grant of timber to replace the wall where wanting.

The castle lay unnoticed, and more or less of a ruin for some centuries, until it became the scene of one of the struggles between Charles and his people. It was first seized by Sir William Waller for the Parliament in 1643, with the city, then also walled, and the position of which is naturally strong. He retired from it before Prince Maurice, but shortly afterwards, without stroke of sword, recovered possession, again to retire, so that in 1644 it was still held by the king. In 1645, however, its troubles began in good earnest. Leslie and the Scots laid regular siege to it, and from the south of the Wye opened a destructive fire upon the castle, cathedral, and bishop’s palace. Sir Barnabas Scudamore, with eleven guns, held out stoutly, beat off an attempt at a storm, and forced the enemy to be content with a blockade. As the mills without the walls were destroyed, others were extemporised within them. The Scots then encamped on Burtonshaw Meads, close under the castle, between it and the river. Scudamore, whose defences were out of repair, and his garrison weak, sent out for country folks to assist as workmen. The enemy found this out, and, entering under the guise of workmen, succeeded in taking the place. Much injury was then done to the public buildings, and the castle, as belonging to the Crown, and a ruin, was sold about 1652, for the value of the materials.

DESCRIPTION OF THE DEFENCES.

The castle of Hereford was one of the strongest, most advanced, and most important fortresses upon the Welsh March, and one which, being posted in a very fertile and open district, was peculiarly offensive to, and very liable to the attacks of, the Welsh people. The present remains are not inconsiderable, but, as is often the case with fortresses of pre-Norman date, they are confined chiefly to earthworks, and include but slight traces of the later defences of masonry.

The castle was placed upon the left or northern bank of the Wye, east of and below the city bridge, the cathedral, and in some measure the city, of which it occupied the south-eastern corner. Though excluded from the city liberties, it is, in common with the city, covered at some distance by a steep slope, at the foot of which lies the Yazor brook, which, rising far off to the west beyond Kentchester, and the remains of Magna Castra, supplied the broad tract of lowland still known as Forster’s Moor, Moorfield, Canon’s Moor, Prior’s Moor, Eastern Moor, Moor Barn, and Widemarsh, to the north-west of the city, and which, after skirting its north front below Baron Court, turns Monkmoor and Scult Mills, and finally falls into the Wye, at Eigne Mill, some traces of which remain about half a mile in advance of, and below the castle. The ground thus included is to the north-east, a broad and dry platform of gravel, very fertile, and probably employed as a safe pasture in wild times. The lower tract to the north-west, now drained, and either cultivated or built upon, must, in former days, have been an almost impracticable morass.

The city of Hereford, within its walls, was, in plan, about three-fourths of an irregular circle, placed upon the Wye, which forms its concave chord. Its dimensions were, north and south, 600 yards; east and west, 770 yards; and along the river bank, about 600 yards; including the cathedral and the castle. The total girth was about 2,350 yards. The walls—no doubt Norman—upon the older lines, were confined to the landward sides, excepting about 50 yards above the city bridge, and were in length about a mile and a quarter. They were pierced by six gates—Wye Bridge, removed in 1782; Friar’s Gate and Eigne, on the west, removed in 1787; Widemarsh Gate and Bye or Bishop’s Gate, towards the north, removed in 1798; and St. Andrew’s or St. Owen’s Gate, on the east, removed in 1786.

The walls were reinforced by fifteen hollow mural towers, placed from 75 yards to 125 yards apart, or at a “flyte shoot,”—explained to be within 400 yards, and here much less. They were rather more than half round—that is to say, their side walls were a little prolonged or stilted in plan, and they were 34 feet high, having a cruciform loop in front, and probably two others laterally. Of these, two at least remain and are accessible, though their upper half has been removed. The intervening curtain was 16 feet high on the interior, but it was built against the old English bank, which thus covered the lower 6 feet of the interior face, and served as a ramp. Besides the two towers already mentioned, which stood on the north-west face, the wall may be traced along the western side from Bishop’s Gate—now “the Commercial Road”—to the river bank. Where it has been pulled down the step remains, and the difference between the level within and without is brought to view. Of this wall the part from the Friar’s Gate to the river is open to view from the exterior paddock. The ditch has been partly filled up, but the wall remains about 12 feet high, unaltered, save by the removal of its upper 4 feet or 6 feet. At regular intervals along it of about 50 feet are broad pilaster buttresses, of slight projection, and without sets-off. These die into the wall below its present top. There do not appear to have been any mural towers on this part of the wall. Some kind of manufactory conceals the termination of this wall upon the river, where there was probably a tower; and from this point to the bridge a line of modern houses effectually conceals any foundations that may have escaped. The tracing out the town wall, or what remains of it, though aided by parallel streets, representing the rampart within and the ditch without, is a delicate operation; for the lowest houses of the city are here found, and the inhabitants seem to drive the trade of those who, at Jericho, dwelt in a similar locality. That part of the wall showing the pilasters is, no doubt, the original Norman wall. The part beyond, or, at least, the towers upon it, is probably of the time of Henry III. The wide modern street, called Commercial Road, which takes the place of the old Bishop’s Gate, has, of course, caused the removal of all traces of the wall near the site of the gate; but the line may still be traced round the eastern quarter of the city as far as the angle of the castle. Here, as before, are two roads—one inside and one outside the line of the wall—the actual place of which is shown, sometimes by a few stones built into the walls of later houses, and sometimes by a dip of nearly 6 feet in level. Beyond St. Owen’s Gate the actual wall remains for some yards, and may be seen from the exterior road. It extends to the counterscarp of the castle ditch, where it ends abruptly, having been in part pulled down.

The city bridge crosses the Wye by six arches, round headed, and all apparently old. One is stiffened with three plain square ribs, and another shows traces of a similar addition. The bridge has bold piers, and advantage has been taken of this to widen it by turning a subsidiary arch, about 5 feet wide, on each face, so that the old work is partially concealed. It does not appear where the fortified gateway was placed; probably upon the last narrow arch on the city side. This does not, indeed, show any trace of having been perforated for a drawbridge, and tradition places the gate at the outer end.

Next to and a little below the bridge, near the river bank, and a little south of the centre of the city, stands the cathedral, and below and next to it the castle.

Besides the cathedral and its appendage of St. John’s, there stood within the walls four churches—St. Nicholas’s, All Saints’, St. Peter’s, and St. Owen’s. There was also St. Martin’s, beyond the Wye, destroyed during the civil wars. The disposition of the streets is irregular, and in no way cruciform or indicative of any Roman arrangement. Thus situated and defended, Hereford was a very strong place, as it had need to be, for it was exposed to the fierce and repeated attacks of the Welsh, who especially resented, as was natural, the conquest of their most fertile provinces by the invaders.

The castle, as has been said, occupied the south-eastern quarter of the city. It lay in the parish of St. John. Leland describes it as one of the fairest, largest, and strongest castles in England.

It was composed of two wards placed side by side along the bank of the Wye, not actually on the stream, which, when the defences were of earth, might have undermined them; but about 7 yards distant from and on the edge of a steep sloping bank, about 8 feet above the water.

In general plan the eastern ward was an oblong, the sides being nearly straight, and the east end narrower than the west. The eastern end runs at right angles to the river, and measures along the old line of its wall 100 yards. The south or river front measured 175 yards, as did the north front. The west front, connecting the two, measured 196 yards. This inclosure formed the eastern or lower ward, and contained about 26,000 square yards.

The upper or western ward was applied to the end of the eastern, and like it rested on the river. In form it was rounded, or rather irregularly polygonal, and composed of a large conical mound, wholly artificial, with a circular circumscribing ditch. Within the ditch it measured, on the east and south sides, each 100 yards, and, on the three remaining sides, 60 yards each. It was, therefore, in girth about 380 yards, and in area about 14,000 yards. Thus the area of the whole castle, ditches included, might be about 8¼ acres. The eastern ward is stated to have contained about 5½ acres.

The earthworks of the lower ward are tolerably perfect. They are composed of a steep bank along the north and east sides, from 15 feet to 30 feet high, the highest and broadest part being at the north-east angle, where there was probably a tower. This bank has evidently been thrown up out of a broad and deep ditch, which remains perfect and full of water, on the north front; but has recently been filled up along the east as far as the river; nor do there remain any traces of the castle mill, which stood at its junction with the river. The ground is raised along the river front about 4 feet, and may have been higher. Along the west or side towards the upper ward, the bank has been thrown into the ditch, and there is only a trace of either. Leland says there was “a great bridge of stone arches, and a drawbridge in the middle of it to enter into the castle;” but even in Leland’s time this was gone.

The entrance to this ward was evidently at its north-west angle, where the moat is crossed by a causeway, and the public now enter. There was a gatehouse in Speed’s time, which he places, probably by an error of drawing, near the middle of this face. But the bank is there far too high, and the ditch far too deep and broad, to allow of the entrance being placed there.

The modern museum, built on the river at the south-west angle of the ward, or rather between the two wards, and over the line of the ditch, contains some parts of an older building, and a doorway of the time of Henry III. or Edward I. This building probably guarded the opening of the middle ditch into the river, and was also the gatehouse between the two wards. Speed shows a sort of water-gate here, which is probable enough. The surface of the lower ward is level. In it stood the chapel of St. Cuthbert, with a semicircular apse, and in Speed’s time also two small dwelling-houses. Leland says, “There is a fayre and plentifull spring of water in the castell, and that, and the piece of the brook coming out of the ditch, did drive a mill within the castle.” This was in addition to the mill outside and north of the castle, and probably was a part of the present museum-house.

The earthworks of the upper ward have unfortunately been destroyed, the mound and banks thrown into the ditches, and all made level and much built over. The site of the mound is occupied by an enclosed three-cornered kitchen garden. The well or spring spoken of as St. Ethelred’s remains. It opens behind the museum building about 50 feet from the river, and 6 feet or 8 feet above it. As it is described as being further north, it is probable that when the ditch was filled up a pipe was laid to bring the water out at its original level. This ward contained the mound known as the Castle Hill, and which seems to have been removed early in the present century. It was girded at the base by a polygonal curtain wall, outside of which was the ditch. It is difficult clearly to understand how the mound was occupied. Leland says “there was one great tower in the inner warde.” Sir Henry Slingsby, in his diary, in 1645, describes Hereford city as not much unlike York, “for it hath a round tower mounted upon a hill, like to Clifford’s Tower, and the mills near it, with some little works about, having the river Wye running close by; but the walls, though they be high, yet are not mounted upon a ramp, as York walls are.” This is intelligible enough, the walls spoken of having been at the base of the hill; but Leland speaks also of a donjon or keep, of what plan is unknown; but upon its wall ten half-round towers, and within, what appears to have been a square tower of considerable height, in the base of which was a dungeon. We may safely conclude from Sir Henry Slingsby’s very clear account that the mound carried, like Cardiff and Kilpeck, a shell keep; but this could scarcely have been furnished with ten half-round towers. These probably belonged to the enceinte wall below. The keep was entered, it seems, from the south-east side by a flight of steep steps up the mound. In the mound was a well, lined with stone, as at York.

The castle ditches were wide, deep, and filled with water, not from the river, or but partially so, but from a brook, which seems to have fed the city ditch on this side, and on reaching the castle at the north-east angle to have divided, a part running direct along the east front of the lower ward, to the river, and the other part supplying the north ditch, and the ditch which divided the upper from the lower ward, and the ditch which passed round the east side of the ward, and divided it from the cathedral precinct. This latter ditch also received some little contribution from St. Ethelred’s Well, a spring on the north side of the upper ward, already mentioned. The castle mill stood at the junction of the eastern ditch with the Wye. The water in the castle ditches was, of course, penned back for the use of the mill, and to strengthen the defences, and it seems to have flowed back upon the city ditch as far as St. Owen’s Gate, thus strengthening and connecting the city and the castle.

We have, then, to recapitulate, as the constituent parts of Hereford Castle, an oblong space, with the river on one side, and high banks and ditches on the other three. One end was cut off from the rest, and had its proper ditch, more or less circular, and within it a conical mound, with a table summit, and upon it a shell keep, with some kind of central tower, probably an addition. At the base of the mound, within its ditch, was a second wall, many sided in plan.

The lower ward, no doubt, had its walls along the summit of the earthworks, and these probably were low, by reason of the height of the bank and the depth of the ditch beyond. There was a gatehouse at the land entrance at the north-west angle, and a water-gate at the south-western, where also was the passage between the two wards. Very likely there was a postern at the south-eastern angle, where is now a footway.

There is a general resemblance, as regards the mound and bank, between these arrangements and those at Wallingford, Wareham, and Tamworth, and with what is related of the castle of Worcester.