Durham Castle was founded c.1072, but evidence for the foundation of the chapel built against its N curtain wall is uncertain (see Comments). It is a rectangular building, groin vaulted with stilted transverse arches between the bays, and has 4 bays from E to W and 3 bays from N to S. The vault is carried on 6 freestanding, coursed, cylindrical piers cut from sandstone from the Durham coal measures, distinctive for its swirling patterns of iron impurities. The capitals may be of a different stone altogether, perhaps an oolitic limestone, although geological analysis has not been carried out (see Bernstein, 278). They are all Corinthianesque but the Corinthian forms are inventively interpreted with human and animal forms, mythical creatures and foliage and geometric decoration. They all have separate impost blocks above them, which have a quirked quadrant roll below a quirked vertical face, except where otherwise indicated in the descriptions. Bases of columns and half-column responds have a roll below a quirked hollow chamfer. The transverse arches of the vault descend onto half-column responds with capitals on the E wall and short corbels with capitals on the W wall. On the N and S walls they descend onto rectangular pilasters with chamfered imposts but no capitals. The present entrance is through an archway in the S bay of the W wall, but this arrangement only dates from 1840. The original entrance was in the W bay of the S wall at the foot of a staircase descending from the SE of the North Hall, but in 1840 this newel stair was diverted and a tunnel cut directly from the lower hall.

In the following descriptions the capitals that form the bulk of this entry are treated as individual features within two separate arcades, so that the images appear close to the descriptions on the webpage.

Durham Castle was begun by Earl Waltheof of Northumberland about 1072, the work continuing under his successor in the earldom, Bishop Walcher (1071-80). No other bishop had a castle, and according to the Historia Regum which records that the Conqueror set a castle here 'to keep the bishop and his household safe from the attacks of assailants'. This is taken to refer to the burning in 1069 of Bishop Ethelwin's house where Robert de Comines, an earlier appointee to the earldom, was staying. This protest against his appointment by the townspeople of Durham was to be a factor in the Harrying of the North, William the Conqueror's response to opposition to his rule in 1069-70.

Following the account of Leyland (1994.2), Bishop Walcher was responsible for the curtain wall surrounding the castle, including the motte at its E end. The chapel is built against the inner N wall of this, and it has been a matter of some dispute whether this was built by Walcher or his successor William of St Calais (1080-96).

In view of the quality and inventiveness of the sculpture, the date of the chapel is of major importance. While it is generally agreed that the curtain wall against which it stands dates from the beginning of the campaign in 1072, the chapel is freestanding and could be later. Initially Zarnecki (1951, 25-26) accepted the earliest possible date of c.1072 for the capitals, and this date was used by Baylé as late as 1979 (Baylé (1979), 106), while Zarnecki had moved to a c.1080 dating (see Zarnecki (1978), 186). The patronage of Bishop William of St Calais (1080-96), is now generally accepted for the capitals (see e.g. Cambridge (1994), Wood (2010) and Bernstein (2018)), while Cambridge has argued convincingly for a date between 1080 and 1088.

A second issue to be addressed is the possibility that what remains today is the lower storey of a two-storey chapel, as at Aachen and Hereford Bishop's Chapel. This possibility may have arisen from a misreading of Laurence of Durham's description of the chapel written c.1141-43. His phrase: Haec nimis alta domus solidis suffulta columnis (This lofty building supported by strong columns), does imply a tall building in which the sturdy columns of the present chapel would act as supports for an upper storey, and the idea cannot be dismissed entirely despite the absence of any evidence in the building itself of a second storey. Nevertheless Wood and Leyland among modern authors both reject the idea out of hand, only Thompson (1994, 427ff) arguing for it.

It is generally agreed that the sculptors came from Normandy, either for this project or more generally to take advantage of the building boom that followed the Conquest. All the capitals are of the volute type, a simplification of the Roman Corinthian capital in which the angle volutes are all that remain of that complex acanthus ornament, and the bell of the capital is either left plain or decorated with various forms. As in Normandy too, the Durham capitals have examples in which the volutes are replaced by other forms, such as heads (capitals SE, S2) or standing figures (capital S1). This variation of the form goes back to earlier work in the Loire valley where, at Saint-Benoit-sur-Loire in the 1040s, experimentation with turning the volutes into animals and other forms was already under way. Specific parallels between the Durham capitals and those at sites in Normandy will be noted below, but the general point has already been made, especially in Zarnecki (1966).

The use of chip-carved backgrounds of saltires and stars on some of the capitals is not unusual in late-11thc sculpture thoughout the country, although it never occurs before the Conquest. This indicates that Normandy was also the source, and it was a common motif there too, especially in the Caen churches of La Trinité and Saint Paix, and at Graville-Saint-Honorine. Wood (2010) offers an iconographic interpretation of this motif, associating it with heaven (see also Wood 2001), and she also associates the foliage motifs widely used in the capitals as heavenly signifiers, but this is not universally accepted. The mask is another form that appears regularly on capital faces: either against a flat or a chip-carved background. These faces appear seven times, and in no case is there any feature that would serve to distinguish or identify them even, say, as a monk, a bishop or a king. This is another feature of 11thc. Norman capitals, seen, for examples, at Mont Saint-Michel, Rouen and La Trinité de Caen but rarely in such an isolated position on the capital face (see e.g. capital S3).

Turning to individual capitals, apart from capital SW with Carter's rather unconvincing interpretation, the only narrative capital is N3 with (following Zarnecki (1951) St Eustace's encounter with the stag shown on three of its faces (and a horrific head on the fourth). Zarnecki's suggestion that St Eustace was the hunter at Durham is based on the presence of a halo, indicating that this is a hunter who is also a saint. Apart from Eustace, the other well-known hunter saint was St Hubert, but although Hubert was venerated throughout the Middle Ages the stag story did not become part of his legend until the 15thc, when it was appropriated from the Eustace legend. The problem with the identification of the Durham capital with St Eustace is that the stag on the capital lacks the central feature of the legend: the figure of Christ on the cross between the stag's antlers (for Aelfric's 10thc version of the story of the conversion of Eustace, see Skeat in the bibliography). A formally similar hunting scene (without the halo and with the hunter shown mounted and blowing a horn) can be seen on a nave capital at St Gervais, Falaise. Another capital in the nave at Falaise depicts a pair of confronted lions on the angle which are very similar to those on capital S2 here. On the S face of this capital is a mermaid, or siren, and the normal interpretation of this is well known. It appears in all versions of the English Latin Bestiary including the First Family, which was circulating in English monasteries from the mid-11thc onwards according to the author's calculations (see Baxter (1998, 227-28). There were two Bestiaries listed in the late-14thc. library catalogue of Durham Cathedral (Baxter (1999), 162-64). Sirens lure sailors by their sweet music, lulling their unwilling ears and senses into sleep before tearing their bodies to pieces. They act as a warning against the delights of theatrical performance and licentious musical tunes, and their presence among the chapels capitals would certainly have been reinforced by sermons in which they provided exemplars. The pose of the siren on capital is the mirror image of the drawing in the earliest iillustrated English Bestiary, Oxford, Bodleian Library MS Laud. Misc. 247. The juxtaposition with the lions on the N and W faces, reminders of Christ's divinity, would therefore seem to offer a contrasting dipole that could be emphasized in sermons.

Capital N2 is something of an enigma. Viewed from the SW it appears to show a snake attacking a lion, which falls backwards as a result. The author can find no convincing explanation for this imagery, and neither does Wood (2010) or Bernstein (2018). One remote possibility is that it references Christ treading on the beasts, from Psalm 91 (Vulgate 90), 13, dealing with Christ trampling the lion and the asp; imagery commonly found on episcopal tombs and crosiers and thus relevant to the decoration of the bishop's chapel here. Finally mention should be made of the capital with Atlas figures on the angles (S1). Wood prefers to see them as orant figures without genitals and surrounded by foliage which, in her analysis, makes them temperate men rewarded for their restraint with a place in heaven. The present author cannot readily discard the notion that they are supporting the imposts and hence the vault and the complete chapel; a kind of visual pun.

Wood spends some time analysing the placement of the imagery within the chapel and especially the question of what was visible to whom when the chapel was in use: a fascinating and rewarding exercise. Bernstein, on the other hand, is concerned with the issues of hybridity and Normanitas: of William of St Calais' patronage and the fact that his private chapel was very different from the cathedral in terms of the extent to which it could be described as a Norman building.