We use cookies to improve your experience, some are essential for the operation of this site.

St Mary, Romsey, Hampshire

(50°59′20″N, 1°30′14″W)
SU 349 212
pre-1974 traditional (England and Wales) Hampshire
now Hampshire
  • Ron Baxter
13-14 April 2005, 04 April 2006

Please use this link to cite this page - https://www.crsbi.ac.uk/view-item?i=14518.

Find out how to cite the CRSBI website here.


Romsey is a town on the E bank of the river Test in the SW of the county, some 7 miles NW of the centre of Southampton. The abbey is on the western edge of the town, near the river.

The present church has an aisled nave and chancel and a crossing between with a crossing tower, and unaisled transepts with single eastern chapels. The eastern arm is unusual in that it is square-ended but has an ambulatory. The central vessel of the chancel is three bays long and terminates at the east with a two-bay arcade. The aisles are of four bays, the eastern bays ending in apses, and the straight ambulatory thus consists of four bays: the east bay of each aisle and between them the two bays east of the eastern arcade. There was once an axial chapel, of course, and the two deep arches leading into it from the ambulatory survive, now fitted with altars dedicated to St Mary and St Ethelfleda. It has usually been suggested that this chapel was originally an open square, two bays by two, but Fernie raises the possibility that there were two parallel chapels separated by a wall, each housing the relics of one of the founding abbesses (Elfleda and Merwinna). The 12thc chapel was replaced c.1270-80 by a Lady Chapel, itself demolished after 1544.

The original chancel elevation remains to north and south, and consists of a three-bay round-headed arcade with a tall gallery whose openings have central shafts and arches dividing them in two, but no tympana. There are no gallery windows. The clerestory has a passage and each bay has a tall central arch around the window, flanked by lower ones. Shafts run up the wall face ending at the straight horizontal wall-head, indicating that there was no stone vault. The arcade has compound piers and the arches of the arcade and gallery are decorated with chevron and other motifs. The aisles have transverse arches and quadripartite rib vaults between them. The elaborate figural and foliage capitals, including the ROBERTVS inscriptions, for which Romsey is justly celebrated are found on the aisle and ambulatory vault responds. The chapels at the ends of the aisles (St George’s on the north; St Anne’s on the south) are decorated with internal wall arcading with scallop or volute capitals and chip-carved decoration.

The crossing has round-headed arches carried on compound crossing piers. On the interior walls, above the crossing arches, a passage runs all the way around the tower, opening into the crossing space through three bays of paired round-headed arches on each face. What remains of the tower outside the church is short, plain and largely rebuilt.

The east walls of the transepts have similar elevations to the chancel; three storeys with two ground level arches framing the entrances to the chancel aisles and transept chapels (the S now the clergy vestry and the N the choir vestry), and gallery and clerestory arcades above. The west walls of the transepts, however, have significant changes in elevation. On the exterior the S transept has a triple window at the middle level of the outer bay, with simple round headed clerestorey windows above, and large areas of blank wall above and below the lower window. On the inside the S bay is articulated with blind arcading below the triple window, and the blank wall above it is largely occupied by the splays of the windows and their elaborate arches. The clerestory of the inner bay is again similar, but below it the gallery and the arch to the nave aisle have an enclosing arch, or giant order. This scheme continues in the nave arcade, and also appears on the external face of the south transept façade in a less plastic form. It is also notable that the gallery opening on the south, but not the north, has a heavy cylindrical central shaft, rather than the slender shafts used elsewhere in the transept and chancel galleries. On the N transept thel exterior W wall has a more conventional elevation with windows at 3 levels in the outer bay. The clerestorey is flanked by blind arches, reflecting the internal arrangement, the middle level has a simple wide arched window within and without, and at the lower level a triple opening similar to that at mid-level on the S transept has been blocked and a small three-light Perpendicular window inserted, although the original arrangement is clearly seen on the interior,

In the nave, the first bay is double on both north and south elevations, with compound piers framing the double bay and a heavy cylindrical central shaft, suggesting that an alternating system of giant-order supports may have been envisaged for the entire nave. Fernie (2000), 174-75 presents a good deal of convincing evidence that this was not, in fact, the original scheme, and that the cylindrical pier was intended to mark a feature such as the nave altar. Further west, the elevation reverts to single bays with compound piers; the giant order expressed by the enclosing arches of the gallery which are carried on nook-shafts descending right down to the arcade pier bases. The design of the gallery openings is similar to that found in the chancel and transepts. The nave is seven bays long (counting the double bay as two), but only bays 1-4 of the arcade and gallery are 12thc. The three western bays and the entire clerestory on either side are early 13thc work, and are not included in this report. The south nave aisle has quadripartite rib vaults with wall responds. Bay 1 contains the so-called Abbess’s doorway from the NE angle of the cloister, and bay 2 has intersecting wall arcading but the remaining 12thc bays are undecorated. The north nave aisle of the abbey church was formerly used as the parishchurchofRomsey. In a dispute in 1372, the parishioners asserted that the aisle was too narrow to accommodate them on Sundays and festivals, and a commission of inquiry was appointed by Bishop Wykeham to look into the issue. In 1403 the vicar and his parishioners were granted a faculty to pull down the north wall of the aisle and enlarge it, making it clear that the parish was responsible both for the work and for the maintenance of the enlarged aisle. After the Dissolution the parish bought the entire church, and the north aisle was returned to its original width. In the course of these changes the entire aisle wall was replaced except for the two eastern bays, which appear to have been spared, and which retain their original masonry but not their windows, which were replaced in the 19thc. The north porch is of 1908, by W. D. Caroe.

Romsey contains two important Anglo-Saxon carved stone roods. The smaller, now set over the altar in St Anne’s Chapel shows Christ elevated high on a tall cross with angels above the arms. To either side are the figures ofSt Johnand the Virgin and Longinus and Stephaton amid a sparse tangle ofWinchesterfoliage. The surface is eroded and retains little detail, but Talbot-Rice suggests a Byzantine prototype and offers a date in the early 11thc. Stone takes a similar view. The larger rood, now set in the west wall of the south transept, is generally considered to have been set originally above a chancel arch. Stone dates this to the first half of the 11thc, and Talbot-Rice to 1000-20 and this dating is usually accepted, although Fernie (1983) associated it with the building erected in 967 (see section VII).

The Quarr stone apse of an earlier church was discovered under the floor and reported by Peers (1901). Peers’s important article suggests a building sequence in which an aisleless, cruciform pre-Conquest church received its apse after the Conquest, c.1090-1100. From c.1120 the church was rebuilt starting at the east end. The Anglo-Saxon nave and transepts remained in place initially, and the big new transepts were built immediately to the east of the old ones. When it came to rebuilding the south nave aisle, work began west of the old transept leaving its façade wall in place, and this eastern section was not replaced until the rest of the aisle was complete. This explains the odd alignment of the first two bays of the south nave aisle, and the change in masonry. It also explains why the sculpture of the Abbess’s doorway in bay 1 of this aisle is more elaborate and advanced than that elsewhere in the Romanesque work. Later workers such as the Taylors, Pevsner, Hearn and Fernie have accepted Peers with only slight modifications. Hearn (1975) goes into most detail, identifying five distinct campaigns between c.1120 and c.1230. There are no documents that allow an accurate dating of the Romanesque work at Romsey, and estimates have therefore been based on sculptural forms and mouldings. Unfortunately, the dating of the figural and foliage Romsey chancel capitals was based on their similarities to some in the crypt of Canterbury Cathedral, and the comparison was made at a time when it was believed that the Canterbury capitals were carved in-situ c.1120. This view has now been discarded, and the Canterbury capitals redated to c.1100, but the old 1120 date has stuck at Romsey. McAleer was aware of the problem in 1983 but retained the later date at that time. Fernie also recognised that the redating at Canterbury constituted a challenge to the traditional dating of Romsey, but accepted an 1120 date on the basis of the chevron ornament which is more complex than that carved at Durham between c.1110 and 1120. This issue will be discussed further in section VIII, but for the present it should be stated that the present author prefers a starting date around 1110.

Romanesque sculpture is found in the chancel, the crossing and the western part of the nave at Romsey. In addition to the doorways, windows, arches, arcades, capitals and blind arcading mentioned above, there are internal carved stringcourses in the chancel main vessel, in the transepts and crossing and in the nave, and external ones on the nave aisle and transept walls, and corbel tables survive on the nave, chancel and transepts, although many corbels have been replaced. All of this work belongs to a single campaign, which must have extended over several decades, except the heavily restored Abbess’s doorway and the window above it. Both are set in a projecting frontispiece with a drip-course between them, protecting the doorway. The doorway has elaborate twisted shafts and symmetrical foliage motifs typical of the 1140s. The window, less tall than the other aisle windows, has decoration connecting it with the doorway.


According to John of Worcester, Romsey abbey was founded by King Edward the Elder in 907 for his daughter Elfleda, already a nun at Wilton, who became first Abbess of Romsey. William of Malmesbury, however, has the abbey founded by King Edgar under Bishop Aethelwold’s influence in 967, a refoundation according to VCH, when it was dedicated to SS Mary and Elfleda. The first abbess of the refoundation was Merwinna, who was succeeded in the post by Elwina around 993. Hampshire was overrun by the Danes at this time, and it is supposed (VCH) that the community fled toWinchester. By the time of Domesday the nunnery was valued at £136 8s, the twentieth richest religious house inEngland, and the third richest nunnery (behind Shaftesbury and Barking) according to Knowles’s calculations. The abbey’s holdings comprised the vill of Romsey itself, 14 burgesses in Winchester, Itchenstoke, Sidmonton, one hide in Totton and one hide in Sway (all Hants) and Edington and Steeple Ashton (both Wilts).

The nunnery retained its royal connections throughout the 11th-12thc. In 1086 Edgar Atheling’s sister Christine entered the house and became its abbess. Her niece Maud (daughter of Queen Margaret ofScotland) also came, but did not take vows and married Henry I in 1100. King Stephen’s daughter Mary became abbess here, but left the house in 1160 to marry Matthew, son of Theodoric, Earl of Flanders, by whom she had two children. She was later separated from her husband and may have returned to the abbey in penitence.

The abbey was dissolved c.1539 (although there is no extant formal surrender document), and in 1544 the parishioners bought it back from the crown for £100.


Exterior Features



Exterior Decoration


Interior Features


Chancel arch/Apse arches
Tower/Transept arches



Wall passages/Gallery arcades


Vaulting/Roof Supports


Interior Decoration

Blind arcades
String courses

Zarnecki (1950) is still the most complete and useful analysis of the sculpture of Romsey Abbey. He noted that the elaborately carved capitals were concentrated at the E end; eastwards from the E side of the transepts, and that there was no logic to the distribution of these capitals beyond observing that most of them were to be found in the E chapels, the south chancel aisles and the transept chapels. He thought at this time that their unity of style suggested that they were carved within a short period of time, and that this meant that they were carved in-situ, an opinion which he was later to reject, along with the majority of recent writers on the capitals. At this time he dated the sculptural work to c.1140, but again he was later to move towards an earlier estimate of the date.

For Zarnecki, Romsey belonged to the ‘Southern School’, a group of major sites linked by similarities of sculptural forms rather than an exchange of artists or workshop links, beginning at Canterbury and developing at Westminster Abbey, Reading, Hyde Abbey Winchester, Romsey, Christchurch and Rochester. The approach might seem old-fashioned and imprecise now, but it set a framework for English Romanesque sculpture that was useful to generations of students. The sheer number of Romanesque capitals at Romsey is unprecedented among surviving buildings, and most are variations of the scallop and cushion form with little or no ornament. A few, however, have been singled out for special attention. A sculptor named Robertus, called Master Robert by Zarnecki, was responsible for the first group. Two vault respond capitals in the N and S chancel aisles are in his distinctive style, and that in the S aisle bears his name on two scrolls. Their subject matter is uncertain: the unsigned capital in the N aisle shows a battle between two kings and their armies at a stage when only the kings remain alive to fight; their armies are reduced to body parts and discarded weapons. The signed capital in the S aisle is more impenetrable, although as Zarnecki pointed out, some of the same figures appear on both capitals and the subjects could be related. The best guess, that the first capital shows the battle of Edington, suggests that both could illustrate scenes of the abbey’s history. Zarnecki also attributed three other capitals to his hand: a scene of two of the Labours of the Months on a capital of the N ambulatory chapel arch, and two similar capitals showing smiling addorsed lions on the chancel E arcade and the ambulatory vault. The second master to catch Zarnecki’s eye he called the Second Master of the Acanthus Leaves (the first of this name worked at Canterbury). His work includes a large number of cushion or scallop capitals, mainly in the S chancel aisle, with figural or foliage motifs in the shields and acanthus foliage on the cones of the bell.


R. Baxter, The Royal Abbey of Reading, Woodbridge 2016, 210, 213-14.

E. C. Fernie, The Architecture of the Anglo-Saxons, London 1983, 150.

E. C. Fernie, The Architecture of Norman England, Oxford, 2000, 172-76.

R. Halsey, “Tewkesbury Abbey: some recent observations”, T. A. Heslop and V. Sekules (ed.), Medieval Art an Architecture at Gloucester and Tewkesbury (British Archaeological Association Conference Transactions for 1981), Leeds 1985, 16-35 (esp. 23-29).

M. F. Hearn, “A Note on the Chronology of Romsey Abbey”, Journal of the British Archaeological Association, 32, 1969, 30-37.

M. F. Hearn, “Romsey Abbey: a Progenitor of the English National Tradition in Architecture”, Gesta 14 (1975), 27-40.

D. Knowles, The Monastic Order in England: a history of its development from the times of St Dunstan to the Fourth Lateran Council, 940-1216, Cambridge 1940, 2nd ed. 1963, 137-39, 702.

J. P. McAleer, “The Romanesque Choir of Tewkesbury Abbey and the Problem of a “Colossal Order””, Art Bulletin, LXV (1983), 535-58.

C. R. Peers, “Recent Discoveries in Romsey Abbey Church”, Archaeologia 57 (1901), 317-20.

I. R. Scott, P. Budd and K. V. Nichols, Romsey Abbey: Report on the Excavations, 1973-1991. Hampshire Field Club and Archaeological Society / Test Valley Archaeological Trust, 1996.

L. Stone, Sculpture in Britain: The Middle Ages. Pelican History of Art, Harmondsworth 1955.

D. Talbot Rice, English Art 871-1100. Oxford 1952, 98, 108-10.

H. M. and J. Taylor, Anglo-Saxon Architecture. Cambridge, II, 1965, 520-22.

J. and H. M. Taylor, “Architectural Sculpture in Pre-Norman England”, Journal of the British Archaeological Association, XXIX, 1966, 3-51 (esp. 12-13).

Victoria County History: Hampshire. II (1903), 126-32; IV (1911), 460-69.

G. Zarnecki, 'Regional Schools of English Sculpture in the Twelfth Century: the Southern School and the Herefordshire School', Unpublished PhD thesis, University of London, 1950, 137-52.