| Description: | The early Norman gatehouse which was the main entrance into the inner bailey of the castle was little studied until the 1970s and 1980s: two events have led to a greatly improved understanding of the building: R.A.Higham’s doctoral thesis on The Castles of Medieval Devon (Higham 1979), and a fabric survey undertaken in the context of the repair and repointing of the building under the aegis of the Property Services Agency, responsible for the maintenance of Crown Property at the time (Blaylock 1985). The medieval documentary history of the castle in general has been well covered by The History of the King’s Works (Brown et al. 1963); key documentary sources relating specifically to the gatehouse are the account of Orderic Vitalis, which says that William I personally chose the site of the castle in 1068; the account of the siege of 1136 from the Gesta Stephani, which implies that the inner earthwork had received stone walls (replacing a timber palisade, see below) by this date; and the expenditure of some £21 in 1250-51 on repairs to the King’s Tower (Brown et al. 1963, 647-8).
From the combination of typical features of late-Saxon masonry (long-and-short quoins; triangular-headed windows; etc.) with other features more typical of early Norman building (cushion capitals to the attached shafts of the entrance archway; billet ornament in the imposts of the top-floor doors and windows) it has been suggested that the building dates to the early 1070s, and was built by local masons in the process of adapting to new architectural details and features. This is a plausible thesis; supported by the early-looking selection of building materials (uniform use of dark purple trap rubble; vesicular trap and white Triassic sandstone freestone for quoins; and a very distinctive dark brown sandy mortar [on which cf. the descriptions of late-Saxon work on the city wall, Monument No. 11000]). The relationship of the gatehouse tower to the bank also supports a very early date, since it is presumed that the bank and ditch of the inner bailey must have been amongst the earliest features of the castle to be constructed. The bank clearly abuts the west elevation of the gatehouse. Evidence that the gatehouse was set within the bank was observed in the fabric of the base of the deep buttresses on the front elevation, where two grades of facework masonry were recorded: good quality facing above the line of the bank (subsequently robbed out); and lower-quality rubble facework below, ending in a diagonal scar representing the line of the bank (Blaylock 1987, 4-5, and Fig. 4). When a small excavation was carried out on top of the bank in 1990, the stone curtain wall was found to have been cut into bank layers (Recognition Event No. 99), implying that the first instance the rampart was crowned with a temporary (timber) defence. The top of the bank aligns (approximately) with the top of surviving early fabric in the buttresses in the rear elevation of the gatehouse, this level represents the position of a platform reached from the wall walk/top of the rampart, from which the upper floor of the interior of the gatehouse was reached by a door in the rear north) wall.
As originally built the gatehouse was intended to form the main entrance to the inner bailey, through an arch 3m wide and 3.5m high in the front elevation. The observer now stands within the position of the ditch; the archway above (now blocked) was originally approached by a timber bridge (although the precise details of how such a bridge worked, and on what gradient it lay, are obscure). Inside, the ground floor was wholly occupied by the entrance passage, with a second arch of similar dimensions to the rear. The ceiling of this passage was of timber (beam sockets were recorded in 1985); careful examination of the masonry showed that there had never been a stone vault in this position. A further tier of beam sockets appeared to suggest that there was a very low, unlit room immediately above the entrance passage only c.1.5m high; the significance and interpretation of this remain obscure (Blaylock 1987, 7). Above there was evidence for two further floors: an upper room lit by a pair of triangular-headed windows in the south wall, and a top floor with doorways in north and south walls, accessed from the northern wall walk (as described above), and carried forward in a sort of primitive machioulis by a large arch connecting the deep buttresses of the front elevation. There may have been another storey, with a parapet walk at roof level, although no original fabric survived at roof level (cf. the suggested reconstructed section in Blaylock 1987, Fig. 2).
Described as ‘the best preserved early Norman gatehouse keep in the country’ by Cherry and Pevsner (1989, 399), the building is a remarkable structure without many early parallels. There are similar early Norman stone gatehouse towers at Richmond Castle, North Yorkshire (rebuilt above) and Ludlow Castle, Shropshire. A useful reconstruction drawing of a ‘generic’ early stone gatehouse, based on the surviving fabric of the Exeter gatehouse, is given by Davison (1979, 131). Although under-scaled, this gives a good impression of how the gatehouse would have functioned, when set within the earthwork bank, and at the upper end of a timber bridge across the ditch.
Addition: One of Richard Parker’s conjectural reconstruction drawing of the castle commissioned by Exeter City Council Department of Leisure and Tourism in 2004 shows an artist's impression of the stone gatehouse tower of the 1070s within the Norman earthwork bank (crowned by a timber fortification), and with a timber bridge and causeway forming the approach to the original gateway within the gatehouse building. It is particularly useful from a topographical and architectural point of view in showing the approaches to the gatehouse, especially for attempting to address the problems of structure, precise location and difference in height presented by the topography of the site at this point.
A second drawing shows a similar view in the later medieval period, once the gatehouse had been replaced by a smaller gateway to the east, the timber defence replaced by a stone wall on top of the rampart (plus towers), and the timber drawbridge had been replaced by a stone bridge and barbican (the latter based on views in early 17th century maps, inter alia, see Filmer-Sankey et al. 2004, 174-5). See also notes on another, more general, drawing given under monument description EUAD 11209 (general description of the castle in the later medieval period).
[Original monument description by SRB, 23.v.00; addition SRB, 8.viii.06]
Extant: Yes |
|---|