Western Approaches: the Original Entrance Front of Caerphilly Castle?
Total Page:16
File Type:pdf, Size:1020Kb
Western Approaches: The original entrance front of Caerphilly Castle? Derek Renn Fig. 1. Caerphilly castle. The Western approach - (or Western Island, the hornwork) leading to the Central Island through the west outer and inner gates. Image © Paul R Davis THE CASTLE STUDIES G UP THEU NAL CASTLE N 29 210 STUDIES 2015-16 G UP U NAL N 31 2017-18 Western Approaches: The original entrance front of Caerphilly Castle ? Western Approaches: The original entrance mention in 1307 [IPM of Gilbert de Clare's wid- front of Caerphilly Castle ? ow] of both the pasture of the road before the 4 Derek Renn gate of the castle, and….’. ABSTRACT In 2000, Rees’ ‘possibility’ became a probability when the Royal Commission on the Ancient and That Caerphilly Castle was originally planned to Historical Monuments of Wales’ report on the be approached from the west is not a new castle stated: ‘At first, it appears that the main theory, but is here examined in depth. There approach to the castle was from the W. through may have been a pre-castle settlement on the the unfinished outwork [ie. the Western Island], site. It is proposed that the Inner West gate- which may explain the apparent priority given house (and a possible hall-and-chamber) were to the fine Inner West Gatehouse. Both inner adapted when the Inner East gatehouse was gatehouses are impressive, but that to the west planned. The probable date of the similar Ton- is less developed and presumably the earlier’.5 bridge gatehouse is reviewed in the light of the John Owen, a local resident, argued in 2000 that career of Roger de Leyburn. the first castle consisted of the Western and INTRODUCTION Central Islands with little or no water defences, Various names have been used for each of the citing G. T. Clark’s statement that Gilbert de 6 three main islands which now make up the site of Clare had leave to enditch the castle in 1271. Caerphilly Castle. Here I shall use the latest guide- But Clark had misread his source (actually a book term ‘Western Island’ as being both descrip- Letter Close of 22 February 1272) which was an tive and neutral instead of ‘hornwork’ used order that the castle should remain unaltered in 7 previously.1 Today, the Western Island appears walls, ditches and brattices. Paul Davis, in his like a poor relation, out-housed beyond the back 2013 biography of the de Clare family, wrote door and neglected. Everywhere else there is that the first castle at Caerphilly had ‘a simple powerful display in stone, with the main entrance twin-towered gatehouse approached from the 8 from the east protected by three successive twin- west across an outer bailey.’ towered gatehouses and other twin-towered So there is muted support for the premise that gatehouses in the middle and at each end of the the original approach to Caerphilly Castle was masonry dams retaining the great lakes. from the west, but was subsequently changed But was it always thus? In 1937 Professor Wil- to be from the east. There were ‘back to front’ liam Rees wrote: ‘We cannot exclude the possibil- changes at two of the ‘Three Castles’, about 30 ity that the castle, as originally designed, miles away to the north-east (in royal hands consisted only of the citadel [ie. the Central Is- from 1239). Grosmont lost its entrance from the land], the main gateway of which was on the north in favour of one from the south about 9 west side and was covered by the hornwork’ [ie. 1256-7. At the same time, White Castle’s ‘horn- the Western Island]. In his expanded edition, work’ to the south was relegated, the bailey on ‘perhaps’ was inserted after ‘cannot’.2 The anon- the far side walled and given a plain gate-pas- ymous official pamphlet guide of 1958, based on sage, the moat doubly-dammed. and a very sim- 10 Rees’ 1937 book, changed this to: ‘It appears that ple twin-towered gatehouse inserted. The the hornwork may have come at a comparatively third castle (Skenfrith) never seems to have had late stage in the development of the defensive a ‘proper’ gatehouse, just a round-headed arch- scheme’.3 Cedric Johns wrote in his detailed 1978 way through the west curtain wall with a win- 11 guidebook: ‘The western outwork or bailey out- dow above it. side the middle west gate served to protect it Was there a pre-castle settlement in the Caer- when it was the usual entrance, as inferred from philly basin? Does the Welsh name for the THE CASTLE STUDIES G UP THEU NAL CASTLE N 29 211 STUDIES 2015-16 G UP U NAL N 31 2017-18 Western Approaches: The original entrance front of Caerphilly Castle ? Fig. 2. Caerphilly castle - Central Island, Hornwork, Eastern dam and barbican - ground-floor plan. Inset: Showing the extent of the surrounding water. The plan is taken from A J Taylor’s annotated copy of William Rees’s ‘Caerphilly Castle - A history and description’, 1937, written following the bulk of the reconstruction work commissioned by the Marquis of Bute. THE CASTLE STUDIES G UP THEU NAL CASTLE N 29 212 STUDIES 2015-16 G UP U NAL N 31 2017-18 Western Approaches: The original entrance front of Caerphilly Castle ? Western Island, Y Weringaer/Caer y Werin assemblage faces north-west rather than west, [‘people's fort’] hint at more than a castle bai- and aligns with an ‘old paved road’ here, per- ley providing a refuge for dependent townsfolk haps the ‘road before the gate of the Castle’ in case of attack?12 Perhaps the site of an earli- mentioned in 1307, whose alignment inter- er llys, the name contrasting a civilian settle- sects (at right angles) with a terrace-way ment with that of the adjacent Roman military which, extended, bisects the site of the Roman 18 fort?13 auxiliary fort and the Civil War earthwork. The line of the Cardiff-to-Brecon Roman road DESCRIPTION OF THE CASTLE here is unknown. Probably begun in 1268 the castle was sited in The Western Island changed the orientation of a natural basin and built on a gravel spit be- the approach from the ‘old paved road’ to that tween streams flowing east, divided by three of the Central Island’s axis. It was a place which wide ditches, the spoil used to level up stone- might be used to check visitors, to store sup- revetted platforms and to create outer plies, to provide temporary accommodation, embankments.14 Johns thought that the Cen- assemble troops or hunting parties, to act as a tral Island was the first to be ditched, serving tournament ground (like The Brays at Kenil- worth Castle of 1266,with a somewhat similar as a protected camp during construction,15 but ‘theatrical’ screen wall) and to provide a garden it looks too small to provide enough space for or a field for cultivation from time to time.19 It the workforce and guards (particularly at was a defensive cushion; attackers would have night), as well as to store simultaneously all the to storm it first, and not be distracted by the building materials and equipment. An aerial opportunities to create damage, destruction, 16 photograph published in 1949 appears to in- plunder and ransom there, then find a way dicate the line of the former south bank of the through its buildings and finally regroup in front Central Island. At least one of the other areas of the entrance to the castle proper of the Cen- would have been occupied as well as the Cen- tral Island. tral Island while work was going on. The West- ern Island, being nearer to the hills to the north Why does the Western Island look unfinished? Was it never completed, destroyed, or aban- from which came the heavy Pennant sand- doned after its use as a construction site ended stone from which the castle was built (perhaps or when work moved round to the east, with its from a roadside quarry near Chapel-Martin) grand show-front and foreworks? The medieval 17 seems the better choice. Some four miles settlement is usually assumed to have been to north-west of the castle is the isolated early the south, on the evidence of the South Barbican church of Eglwysilan, in whose large parish the entrance and a single medieval building outside castle was founded. the castle: the altered ‘Court House’. However, THE WESTERN ISLAND (FIGS. 1, 2 Inset) the castle’s Outer West Gatehouse was remod- elled as a court-house in the late 16th century, The Western Island has a very odd shape: and the first edition of the Ordnance Survey map stone-revetted in eight straight varying shows the village on the east side of the castle.20 lengths, two of which form a right angle at the north-east corner. One straight length sepa- THE CENTRAL ISLAND (FIG. 2) rates two unequal semi-circular salients. too far apart to have formed an orthodox gate- Caerphilly is rightly regarded as a concentric house, despite a bridge-pit between them. The castle, but it has many asymmetric features, ditch here on the west side is steep-sided, even on the Central Island. It has not the geo- curving and narrower than the moats further metric near-perfection of royal Harlech or east, which may have been recut later. This Beaumaris castles, begun in 1283 and 1295 THE CASTLE STUDIES G UP THEU NAL CASTLE N 29 213 STUDIES 2015-16 G UP U NAL N 31 2017-18 Western Approaches: The original entrance front of Caerphilly Castle ? respectively.21 The plan of the inner ward is a • has spiral stairs reached from barred lobbies parallelogram, not a rectangle: each gate- in the rear wall, not in half-round turrets house is aligned exactly at right angles to its reached through barred doors from the gate- curtain wall, but is sited about a metre south of passage into the flanking rooms.