CITY WALLS, GATES, AND POSTERNS

The first town wall was probably built in the late 9th or early 10th century when Oxford formed part of the West Saxon system of burbs, built for defence against the Danes. The text of the Burghal Hidage referring to Oxford is slightly corrupt, but between 1,300 and 1,500 hides seem to have been assigned to the town, implying a wall a little over a mile long, considerably shorter than the later medieval wall of c. 1½ mile. (fn. 1) The late Saxon wall has been identified at only two points, both on the north side of the town, at North Gate and further east in Exeter College; it was built of turves with a stone revetment. (fn. 2) The Saxon north wall was on the line of the medieval wall at least from North Gate to Smith Gate. The line of the south wall is uncertain, but there is some evidence for the positions of the east and west walls.

The later street plan suggests that the first east wall ran east of the medieval Schools Street and Oriel Street and west of Catte Street and Magpie Lane, and when in 1899 part of the medieval north wall between the Bodleian Library and the Clarendon Building was excavated a wall containing herring-bone masonry was found, apparently bonded into the town wall and branching from it. It appeared to be the revetment for a bank rather than a free-standing wall, but the excavators seem to have been in some doubt as to whether it was turning away from the wall to run south-east or east, or whether it was the inner wall of a double north wall such as later existed between Smith Gate and East Gate. (fn. 3) No trace of such an east wall was found in 1909 when the underground book stack for the Bodleian Library was made, (fn. 4) but the remains of a turf bank might easily have been unrecognized at that date. A ditch just west of Catte Street might explain the list to the east developed by St. Mary's tower as it was being built in the late 13th and early 14th century. (fn. 5) The line of the west wall is more problematical, but an east-west ditch running just west of the line of St. Ebbe's Street, below the 11th-century Church Street, may have formed part of the 10th-century defences; (fn. 6) if a rampart had run just west of New Inn Hall Street and St. Ebbe's Street it would have taken advantage of a steep natural fall in the ground. In the Middle Ages there was a straight line of property boundaries running north from Queen Street, west of, and parallel to, New Inn Hall Street. (fn. 7)

An undated ditch running from north-east to south-west below the quadrangle of Corpus Christi College may have formed part of the south-eastern defences, (fn. 8) but no trace of a wall or rampart has been found there, or on the line of the medieval wall between Littlegate and the castle, where the 13thcentury wall had been built on a domestic site. (fn. 9) Trill mill stream might have made a major southern defence superfluous.

The wall was probably extended both eastwards and westwards in the early 11th century, perhaps after the sack of the town in 1009. (fn. 10) Church Street was laid out across the suggested line of the 10th-century west wall in the early 11th century, and the parish of St. Peter-in-the-East was presumably within the wall by 1086 when houses there were among those responsible for the maintenance of the wall. (fn. 11) The construction of the castle in 1071 probably led to the further modification of the western wall, but the new wall, like the old, was probably of earth with a stone revetment. The position of the rampart is marked by the strips of city property at the re-entrant curve of the modern Bulwarks Lane, which were left waste, and hence acquired by the city, when the rampart was replaced by a wall. (fn. 12) Henry I gave St. Frideswide's priory permission to inclose part of the intra-mural road and of the wall itself, (fn. 13) and between 1136 and 1140 Stephen granted to the priory the gate in the wall within its inclosure, and permission to build on the wall, provided that the priory maintained the wall. (fn. 14) In 1142 the town was said to be very strongly defended with deep water on all sides. On one side were the castle and a tower, presumably St. George's; another side was described as having a second line of defence, possibly outer ditches, although at a later date one section of the wall was double. (fn. 15) The late-12th-century town seal shows a town surrounded by a stone-built and crenellated wall. (fn. 16)

Between 1226 and 1240 the walls were thoroughly overhauled and the remaining sections of rampart replaced by stone. The king gave firewood for lime kilns and hurdle platforms between 1226 and 1231, and oak joists and planks for turrets and bastions in 1233. (fn. 17) The surviving sections of the wall, notably that in New College garden, seem to date from the early13th-century rebuilding; the wall is of local rubble with ashlar dressings, the bastions are hollow, and a rampart walk survives in New College. (fn. 18) Between Smith Gate and East Gate the wall was double, a feature unique in England. The outer wall, which may possibly have been built in the late 13th century, apparently rose straight from the southern side of the town ditch; its bastions corresponded to those on the inner wall which stood c. 33 ft. south. Much of the space between the walls was filled by ponds in the later 14th century. (fn. 19)

In 1244 the Grey friars were allowed to pull down part of the south-western town wall where it ran through their enlarged precinct, provided that they built a strong, crenellated, precinct wall to replace it; they were unable to do so, and in 1248 were allowed instead to build the north wall of their church in the gap in the town wall. (fn. 20) Despite the documentary evidence no footings of a stone wall predating the friary have been found. (fn. 21) Further work was done on the wall in the 1250s and 1260s, (fn. 22) and in 1257 Henry III gave permission for the building of turrets, presumably additional to those built in 1233. (fn. 23) Oak bars, presumably for gates and posterns, were made in 1264, and in 1266 Osbert Giffard, keeper of the town, was ordered to repair the bars, crenellations, and rampart walks. (fn. 24) Repairs were intended in 1285, 1301, 1321, 1326, and 1347, (fn. 25) but little appears to have been done, and in 1371 the walls were reported to be undermined and cracked. (fn. 26) In 1378 Richard II ordered the repair of weak and ruinous walls. (fn. 27) A turret was repaired in 1423-4 (fn. 28) and the city repaired the wall in 1555, 1557, and 1558; (fn. 29) but much of it seems to have fallen into decay, and by 1583 the outer wall at Smith Gate had completely disappeared. (fn. 30) In 1612, when the wall no longer served any defensive function, a tenant was allowed to pull down 7 yd. of it near East Gate and replace it with a timber structure, (fn. 31) and in 1616 William, Lord Say and Sele, another tenant, was allowed to replace a near-by section with a footpath 2 ft. wide. (fn. 32) By 1675 much of the wall had disappeared. (fn. 33)

The Burghal Hidage system, whereby between 1,300 and 1,500 hides of land in the surrounding countryside contributed to the defence and presumably to the maintenance of the wall, was probably short-lived; it would have been pointless once Oxford ceased to be a refuge burh for the surrounding countryside. In the mid 11th century the burden of repair was borne by 210 houses in the town, known as 'mural mansions', of which all but five were free of all custom except military service and the repair of the wall; one house became a mural mansion during the Confessor's reign which suggests that the system was still developing. (fn. 34) Eight of the 210 houses belonged in 1086 to rural manors, and it has been assumed that all mural mansions had such an attachment and that the system had somehow developed from the Burghal Hidage system. (fn. 35) Places such as Bloxham, Shipton-underWychwood, and Princes Risborough (Bucks.), however, seem too far from Oxford to have been included in the original 1,300 or 1,500 hides, and there is nothing to suggest that in other towns houses belonging to rural manors were responsible for the wall. On the other hand the system of mural mansions did not simply lay the responsibility for the wall on the houses adjoining it, for one of those identified stood near Carfax. (fn. 36) It may be that the mural mansions were the larger houses of the town; in any case their freedom from custom presumably made them attractive to country thegns who wanted a town house, perhaps mainly for market purposes. The ownership of town houses by such thegns could account for the connexion with the country properties in 1086.

The system of mural mansions seems to have continued until the time of the building of the medieval stone wall after 1226. (fn. 37) In 1227 34 landowners, including the prior of St. Frideswide's and the master of St. John's hospital, were distrained to carry out the service of walling the town, (fn. 38) and several houses were later taken into the king's hand for default of the service. (fn. 39) As late as 1251 some houses in the town were alleged to owe murage. (fn. 40) Payments from mural mansions were inadequate for the work undertaken in the 1220s and 1230s, and extra money was raised by another form of murage, a toll on goods coming into the town, which was taken almost every year between 1226 and 1239. (fn. 41) In 1228 and 1235 sums of 52 marks and £35 owed by the town to the king were given back to the town for the repair of the wall. (fn. 42) Later-13thcentury and early-14th-century repairs were financed by further murage grants, (fn. 43) but there were complaints that the money so raised was not being spent on the wall, and commissions were set up to inquire into the spending of the money in 1325, 1329, and 1330. (fn. 44) In 1371 the king ordered all those living, staying, or trading in the town or its suburbs to contribute to the repair of the walls. (fn. 45)

From the later 14th century leases or grants of land close to the wall included the condition that the tenant should repair the wall adjoining his holding. (fn. 46) The earliest example was a grant of 1378 to Adam River of the town ditch and land between the wall and ditch from East Gate to Smith Gate; (fn. 47) the largest such grant was that to New College in 1379 of the wall around its site. (fn. 48) Already in 1375 the town was leasing two bastions near South Gate to a burgess, perhaps as storerooms, and one at Smith Gate to the vicar of St. Peter-in-the-East, apparently as a chapel, (fn. 49) but it is not clear who was responsible for the maintenance of the bastions. By 1387 four other turrets and land below the wall at North Gate, between Greyfriars and West Gate, and from Smith Gate to East Gate were being leased, and some of the tenants were maintaining the wall. (fn. 50)

The 10th-century town presumably had gates on each of the roads leading to Carfax, and the North, South, East, and West Gates remained the principal gates throughout the Middle Ages. Littlegate, earlier Water Gate, at the bottom of St. Ebbe's Street, and Smith Gate at the north end of Catte Street, were first recorded in the earlier 13th century. (fn. 51) North Gate stood across Cornmarket Street, beside St. Michael's church. In the 18th century it comprised a vaulted tunnel 70 ft. long and c. 12 ft. wide, having been extended as the prison above it was enlarged. On the north the gate was flanked by round towers and in the Middle Ages was approached by a bridge or causeway over the town ditch. There was a portcullis at the gate by 1325. (fn. 52) The medieval East Gate, across High Street just east of the junction with Merton Street, was probably similar to the North Gate, though not so long; the room above it was Holy Trinity chapel. (fn. 53) What is probably an east view of the gate appears on a misericord in New College chapel, (fn. 54) which depicts a round-headed arch in a square gate tower flanked by two smaller square towers. The same gate seems to be shown, from the west, in a drawing of 1625, without the flanking towers, (fn. 55) and although Anthony Wood implied that parts of the towers were still standing in 1661 he thought that they had been round. (fn. 56) In 1675 the gate was shown as having two square towers, apparently with only a wall between them, (fn. 57) but the round-headed archway between two square towers shown in 18th-century drawings (fn. 58) seems to have dated from 1711 when the city council agreed to rebuild the gate. (fn. 59) Both North and East Gates were demolished by the Paving Commissioners in 1771. (fn. 60)

The South and West Gates, which disappeared in the earlier 17th century, (fn. 61) appear also to have been vaulted tunnels with flanking towers. The parishioners of St. Michael at the South Gate leased a turret at the east side of the gate in 1445. (fn. 62) The church may have been related to the defences at an early date, as was St. Michael's at the Northgate. Part of the South Gate was presumably demolished by Cardinal Wolsey at the building of Cardinal College; (fn. 63) the remainder fell down in 1617, and although the city council resolved to rebuild it, the work was probably never begun. (fn. 64) The West Gate was repaired by John Claymond, president of Corpus Christi College, in the early 16th century; (fn. 65) it fell down or was demolished in the mid 17th century. (fn. 66) A drawing which may be of the West Gate (fn. 67) shows a 13th-century archway flanked by circular towers, but a map of c. 1605 shows a round archway. (fn. 68)

Littlegate consisted of a small, arched gateway for pedestrians and a larger one for carts. Above it was a large room, usually, in the Middle Ages, leased to scholars; two adjoining rooms were perhaps in a flanking tower. (fn. 69) The gate was apparently falling down in the early 17th century, but the pedestrian arch survived until 1798. (fn. 70) Smith Gate seems to have been little more than a postern, a simple archway in the wall next to the octagonal tower known as the Lady chapel, until between 1633 and 1643 it was enlarged to allow the passage of carts. (fn. 71) It appears to have been removed between 1661 and 1675. (fn. 72) There were perhaps as many as a dozen other posterns, some of them little more than holes in the wall and comparatively easily made or blocked up. The earliest recorded was that confirmed to St. Frideswide's between 1136 and 1140; (fn. 73) Windsor's postern at the north-east corner of the wall was recorded in 1378, when it may have been newly made to give John Windsor access to his plot of land between the two walls. There was a postern in the outer wall just east of Smith Gate. (fn. 74) The town made a postern next to the castle ditch, probably at the north end of Bulwark's Lane, in 1460-1. (fn. 75) 'The hole in the wall', recorded in 1550-1, was called c. 1590 the Turl, after the 'twirl' or turnstile there which kept cattle out of the town; (fn. 76) by the later 17th century it was known as Turl Gate. (fn. 77) An archway was built there in 1614-15, and the gate was demolished in 1722. (fn. 78)

A gate known as the New Gate was built, probably in the late 13th century, on Folly Bridge. Timber, iron bars, and locks were bought for it in 1310-11, and repairs were made regularly throughout the later Middle Ages. (fn. 79) As late as 1565 the city, in leasing the gatehouse for the Berkshire archdeaconry court, reserved the right of entry at all times for the city's defence. (fn. 80) Shortly after 1611 the gatehouse, known as Bachelor's Tower, or Friar Bacon's Study (in the belief that it had been used as an observatory by Roger Bacon), (fn. 81) was heightened by Thomas Waltham alias Welcome, and became known as Welcome's Folly. (fn. 82) The tower, demolished by the Hinksey turnpike trustees in 1779, appears to have been hexagonal, the road passing under it through a round arch. (fn. 83)

The walls were surrounded by a ditch or moat which on the north side was wide enough to be a fairly formidable obstacle, but was elsewhere little more than a small ditch in the later Middle Ages. On the west and south-west, however, the Castle mill stream and the Trill mill stream flowed round the walls. Part of the moat at the North Gate and between Smith Gate and East Gate was made into fish ponds in the 14th and 15th centuries. (fn. 84) In 1371 the ditch was blocked with filth and rubbish, and in 1378 it was ordered to be cleaned. (fn. 85) In 1381 the town obtained licence to enlarge it to a width of 200 ft., but the grant was cancelled after protests from Merton College, whose manor of Holywell extended up to the north wall. (fn. 86) Merton's tenants had earlier hampered clearing operations by throwing earth and sand back into the ditch because it was obstructing their road. (fn. 87) During the 16th century the ditch became filled with rubbish, and a few cottages were built on it; in the course of the 17th century the whole northern moat was built over. (fn. 88)

The medieval defences were decayed and inadequate by the time of the Civil War. (fn. 89) A preliminary line of earth-works built north of the city was destroyed in September or October 1642 when the town was in parliamentary hands. (fn. 90) Between November 1642 and May 1646 the royalists embarked on an ambitious new scheme of defences much of which seems to have been completed. An area from St. Giles's church in the north to Friar Bacon's Study on Folly Bridge in the south and from Magdalen Bridge in the east to St. Thomas's church in the west was surrounded by a ditch, rampart, and palisade, with gates defended by drawbridges. (fn. 91) On the north the line was double. Cannon were positioned at Magdalen Bridge, the Botanical Gardens, St. Clement's, Holywell, Folly Bridge, and St. Giles's. (fn. 92) Strongpoints were made outside the main defences at Dover's Spere in Addisons' Walk, around St. Clement's church, at Eastwyke Farm by Grandpont, at Hart's Sconce on the river Thames, at Oseney mill, at Rewley, and north of St. Giles's church. Inside the defences there were 'guards' at the back of Christ Church, at the Hollybush in St. Thomas's parish, at Holywell and Holywell mill, and at the magazine in New College; there was a 'horse guard' at Wadham College. (fn. 93) In addition, barriers were placed in the Thames just below the town, and both the Thames and the Cherwell were made to flood the meadows around the town.

Richard Rallingson of Queen's College apparently designed the defences put up in 1643; in June 1643 Charles Lloyd was engineer in charge of the work. In March 1644 the king's engineer in ordinary Dietrich Boekman seems to have been in charge at Oxford, but the final plans for the defences were made by another of the king's engineers, Bernard de Gomme in November 1645. (fn. 94) Strong works around New College were finished by March 1643, and in April and May work was in progress at the entrances to the city, in St. Clement's parish, and around Magdalen College. (fn. 95) By early June the outwork at St. Clement's was finished. (fn. 96) In June and July Charles I ordered all men between the ages of 16 and 60 to work one day a week on the defences or pay 1s. (fn. 97) In June work began at Holywell, and in August new bulwarks were being made to the north of the city. In December work outside the west gate was hindered by parliamentarian sympathizers who pulled down at night defences built during the day. (fn. 98) In January 1644 the city at first refused to pay £200 a week towards making a quarter of the fortifications but some money was paid at the end of the month. (fn. 99) Labour was not given willingly: in April the governor of the town ordered recalcitrant citizens to be distrained by officers with musketeers, and in May the king ordered those who would not pay to be expelled from the city. (fn. 1) In March 1644 bulwarks 14 ft. wide were being built outside the north gate, and by that summer the whole north side of the city was palisaded. (fn. 2) In July batteries were being made in Magdalen College walks. (fn. 3) Some guards were apparently being manned by the autumn of 1644, (fn. 4) but work was still in progress, at the city's expense, in October and November that year; (fn. 5) as late as August 1645 the order for all men to work on the defences was repeated. (fn. 6) In May 1646 work on de Gomme's defences, probably the outer line on the north, was 'newly finished'. (fn. 7) Some defences were slighted in 1647 after the city's capture by the parliamentary forces, (fn. 8) and others north of the city, in 1651. (fn. 9)

Footnotes