ABSTRACT
The general relationship of the mediaeval city wall to its Roman predecessor has long been known. The late wall follows the line of the first wall every where on the landward side, departing from it only at the south-west, where below Ludgate the line was moved outwards to practically the east bank of the River Fleet. These events have been discussed in connexion with the siting of the Roman wall in this area (pp. 52-6). But while the two walls run together with the mediaeval wall overlying
the other and using its remains as a base the actual manner in which they were joined does not appear to have been as straightforward as the wellknown reconstruction in the Roman London volume of the Royal Com mission would suggest.2 The evidence has of course only come to light since that volume was published and it has shown that at any rate in the north-western sector the later work does not always rest tidily upon the Roman. At St. Alphage (17) the outer part of the Roman wall had evidently com
pletely collapsed. It was refaced at an uncertain date on an irregular line which lay behind that of the original wall (Fig. 19). At both the east and the west ends of this sector the straight joint between fort and city ‘thicken ing’ survives only to a height of about 6 feet above the original plinth-level. Above this the double wall is replaced by solid post-Roman masonry and the junction between the two is irregular and not easy to define precisely. At Falcon Square (5) on the other hand, an entirely new wall had been
built on the outer wall of the south turret of the fort gateway. Its face was set back behind the Roman face which remained to a height of only 2
feet above the foundation. The rebuild was a good deal wider than the Roman wall and projected behind it for at least 7 feet, its rear part resting upon a mixed clay filling in the lower part of the turret. It seems fairly certain that this wall is of mediaeval date, though no significant mediaeval material was associated with it. The relationship of the mediaeval to the Roman wall in the north-western
part of the city is even clearer at the Barber Surgeons’ Hall. Within the hall the rear face of the wall is largely masked by later walls which have some times destroyed the internal thickening: it would appear, nevertheless, that on the inside the fort wall survives to a height of at least 4 feet. Externally, on the other hand, the Roman wall has been completely refaced from plinth-level to modern surface. Only in the angle of wall and Bastion 13, where the structure has been protected by the projecting bastion, does any Roman face-work survive at all; and there it remains to a height of only four or five courses over a horizontal distance of about 5 feet. This refacing does not resemble the random rubble of uncertain but possibly post-Roman date which has already been described both at St. Alphage and behind Bastion 14 (pp. 65, 64): the rubble is squared, but entirely without regular coursing, and most of the material has the look of being reused (Plate 20). In the Cripplegate region therefore little of the Roman work on the wall
can have been visible during the Middle Ages. Not only was the damaged front of the wall renewed, perhaps at more than one period, but its height also was made up, no doubt to replace the collapsed upper part of the original wall. These additions are thinner and slighter than the original wall on which they are carried. Their distinctive feature is a series of tile or brick courses at intervals of three or four courses of Kentish ragstone which might at first glance be mistaken for Roman work. In fact, however, the courses are too closely set and they are not continuous, quite apart from the un-Roman character of the bricks themselves. A small fragment of this type of construction still survives at the north end of Noble Street (6) where also recent work has revealed that the mediaeval masonry, which is un coursed in the core of the wall, is closely bonded to the Roman masonry at irregular heights. Until 1958 there was another to the north which had an interest for the final phase in the wall’s history (p. 90). The actual date of the brick-coursed fragments remains uncertain, but
the work bears a sufficient resemblance to datable remains presently to be
described to suggest that they may be of the mid-fourteenth century. It should be noted that the courses are present only on the face: the core of the wall is made up of compacted rubble, often containing much chalk and showing none of the structural method of the Roman wall. From the documentary evidence it has long been clear that the wall was
at various times repaired or reconstructed. One such rebuilding in Cripple gate also has long been well known: it is the brick wall of which a portion now survives in the churchyard of St. Alphage, though until into the last century it had been more extensively preserved. The distinctive feature of this wall, which survives to almost its full height, is its diaper-pattern of dark brick, now made more obvious as a result of recent cleaning. Its outer face had been covered by modern buildings on the north side, and with the destruction of these buildings, the City Corporation have taken the oppor tunity of preserving the wall so that it is now visible on both faces.1 Its total height is about 25 feet above the Roman foundation and most of this height must have been exposed in mediaeval times. Only the top 6 feet of the wall are taken up with the later brick-work, the core of which appears to merge with the earlier structure. This work is generally ascribed to Mayor Joceline who in 1477 was
responsible for a large-scale repair of the wall between Aldgate and Alders gate.2 The exposure of the external face of the St. Alphage wall revealed that Joceline’s work abutted on the west a rebuilding which seems to be nearly a hundred years earlier (Fig. 19). The distinctive facing of this wall survives, with breaks, over a distance of about 50 feet. The masonry con sists mainly of small close-jointed ashlar in which there are at narrow vertical intervals courses of knapped flints or of tiles set in the face (Plate 25). Masonry of identical character is to be seen at Westminster Abbey in the transept-like additions to St. Katherine’s Chapel (the Infirmary) built in the mid-fourteenth century (Plate 26).® The resemblance is so close, even to the presence of occasional deeper courses, that the two structures must have
been built by the same hands. London Wall in Cripplegate is thus brought into close relationship with another important London building. This type of work does not appear to have occurred throughout the full length of the Cripplegate salient, for there is evidence that the wall confronting Cripple gate churchyard (la) was part of Joceline’s reconstruction.1 The tile-coursed wall in Noble Street (6) presumably was part of the work done in Joceline’s time under the aegis of the Goldsmiths’ Company, who accepted responsi bility for repair between Cripplegate and Aldersgate. But the repair was no doubt piecemeal, and (as has appeared in the case of St. Alphage) not all the work in the same length need be of the same date. On the whole it would seem that the Roman wall outside the area of the
Cripplegate fort is better preserved than in the parts just described, so that away from Cripplegate there would have been more Roman work visible in the later wall. At Coleman Street, for instance (23 above, p. 51), the faced wall survived in places to a height of only one or two courses above the plinth; but here, no doubt, it had suffered from being on the frontage of the old London Wall. Further east, at All-Hallows-on-the-Wall and elsewhere, the wall-face had survived to as much as 10 feet. This sort of variation may be due to the differences in construction between the fort wall and the city wall proper, the latter being much more massive and provided with levelling courses. On the other hand the poor state of the city wall at the Newgate angle when the bastion was built has also been noted (p. 64); so that the conditions were evidently not uniform. No mediaeval masonry capping the Roman work had survived at Cole
man Street, but this length of the wall was notable for the presence behind it of a series of massive brick arches (Fig 20; Plate 27), backed against its inner face and evidently intended to reinforce the wall and carry some considerable superstructure. A first reaction was to regard this work as being of the Civil War period, for as will be seen (p. 88) some refurbishing
interpreted as part of the Roman wall. His interpretation was accepted by the Royal Commission. His drawing shows, however, that the arches were on the road-frontage behind, the wall, the broken stump of which, with bonding courses, appears on the right of the picture (Fig. 20). The bonding courses are in the same relative positions in the Roach Smith illustration as in the sector exposed in 1957. No additional information has been forthcoming about the surviving
stretches of the wall on the western side of the city, and the conditions prevailing on the south side of Ludgate Hill as this is being written make it impossible to investigate the one area in which the mediaeval wall departed from the Roman line. The problems presented by the wall there have already been discussed (pp. 52-6). On the eastern side consolidation by the Ministry of Works of the postern-gate in Trinity Place (53) revealed also that the Roman wall is preserved in mutilated state to a height of about 14 feet and again carries the mediaeval wall above street-level. The mediaeval wall retains its wall-walk and is at least 25 feet high above the Roman plinth, but has none of the structural refinements of the western sector.1 It has already been seen (pp. 7 iff.) that some of the bastions have now
to be accepted as mediaeval in origin and this might indicate that the mediaeval authorities felt obliged to extend the unequally distributed series that they inherited from the Romans. Bastions 12 and 14 have usually been quoted as examples of adaptation to mediaeval purposes, on the assumption that they were added to or rebuilt in their upper parts, which as with the wall were thought to use the surviving Roman structure as foundation or base. The original date of Bastion 12 is uncertain and that part of it which has long been visible above ground has been tidied up in relatively recent times. Bastion 14 can be seen to be mediaeval above street-level. Its typically varied stonework contains arrow-loops which are contemporary with the rest, as well as a single-light window, at present blocked, which looks late mediaeval (Plate 15). But the upper work in its present condition looks very different from that exposed at the lower level when the bastion was excavated
in 1 947 (Plates 1 6, 1 7). The probability is that when the masonry is stripped of the modern brick and other obstructions now masking it it will be seen to be a structure of two building ‘periods’, both of them mediaeval.1 Opportunities of examining the city ditch have also been restricted. In
Castle Street, just north of the fort gateway (5), the inner slope of this ditch was seen to cut across the shallow ditch, here flat-bottomed, which was associated with the Roman city wall. The mediaeval ditch went to a depth of about 10 feet below basement-level (19 feet below the street) but owing to obstructions could not be followed for more than a few feet (Fig. 3). At Cripplegate Buildings (18: where now stands Roman House) the evidence of a complicated section is more difficult to interpret. The post-Roman ditchfilling had been cut through by later disturbance and there were the remains of a massive timber framework supported on large square-sectioned piles. These constructions were of late mediaeval date and had the appearance of some kind of wharf or landing stage. Stow records that in 1569 a ‘warf of tymber’ was made ‘from the head of the Posterne [i.e. Moorgate] into the towne ditch’.2 It would appear that the same arrangement was provided at Cripplegate (Plate 28). In 1965-6 changes in St. Giles Cripplegate Churchyard in connexion
with the Barbican Redevelopment Scheme exposed partial profiles of the mediaeval ditch. It has not so far been possible to study a complete section and it seems likely in any case that here as elsewhere its outer part will have been destroyed by the fairly massive foundations of modern buildings which run with it along the inner side of the lip (see below). The general character of the ditch appeared however to be fairly consistent. It had a flattened-V section, with a width of 30-35 feet from a berm about 9 feet wide; its depth from the offset of the Roman wall (which seems to mark approximately also the base of the mediaeval wall hereabouts) varied between about 6 and about 9 feet. To a depth of 1 1 feet or more from the graveyard surface the ground
was completely disturbed by burials; below this the filling of the ditch was a dark clayey silt with some dumped building rubbish and other debris. There was practically nothing in the way of small finds. The only significant piece was a sherd of ‘bellarmine’ ware which came from the black silt within 1 8 inches of the bottom of the ditch: since this was below the burial level it must be taken to indicate that the ditch was open to almost its full depth until (presumably) well into the sixteenth century. Also set in the bottom of the ditch, but towards its outer side were several upright stakes about 4 feet long, at intervals of about 6 feet, all leaning slightly westwards. These were squared timbers of about 5 inches scantling. Their purpose was not obvious and they showed no sign of having formed part of a single structure. The observations at St. Giles have proved valuable for the interpretation
of features observed in a ioo-foot cutting that was made near Aldermanbury Postern (St. Alphage: 17) in 1949-50. Unfortunately-sited later obstructions here had destroyed the junctions of
the surviving features in this cutting, so that not only was it impossible to see exactly how they were related to one another but also their original dimen sions remain vague (Fig. 21). In spite of this, the profile of the Roman city ditch, its width 12-13 feet, was nearly complete. Beyond it were a broad U-sectioned ditch, apparently of two periods; and to the north of that a flatbottomed expanse the full width of which could not be determined because it continued under Fore Street: its visible width was about 4 5 feet. The Roman ditch had a depth of about 6 feet below the probable level of the Roman berm, which would have been at or near the present cellar floor; the deeper U-sectioned ditch a depth of 9 J feet; and the shallow flat bottomed area beyond a depth of 6 feet. (The last two may have been rather more if their contemporary surface was at all raised above what seems probable.) From the evidence of the Cripplegate sections the conclusion is justified
that the U-sectioned ditch here is at any rate one version of the mediaeval city ditch though its measurements appear to differ somewhat. The filling was in two parts: the lower dense and clayey and wet, particularly towards the base, but producing very little in the way of finds; the upper, defined at the base by a light sandy layer, somewhat similar but containing streaks of sand and yielding much seventeenth-century pottery (Plate 29). The upper part of the ditch had the appearance of a late re-cutting. Its scale did not make it a very effective obstacle, though its surviving depth of only 6\ feet
-- -'
may have been a good deal less than its original depth if by the seventeenth century the general surface of the ground in the area had risen as would be expected. A feature having a further bearing on its date was the brick-built sewer which appeared in this section as well as further west in Cripplegate churchyard to run more or less parallel with the wall: the sewer has been seen in the course of the recent work in the churchyard and was recorded outside Bastion 12 in 190 1.1 The contact between the sewer and the upper ditchfilling is not a very good one; it nevertheless suggests that the ditch was already in existence (and partly filled) when the sewer was constructed in 1648. A ditch open into the early part of the seventeenth century suggests a connexion with the Civil War; and while to the west there is no evidence of re-cutting the ditch south of Bastion 12 also seems to have been open at this time. Although it is well established that in 1642-3 London was encircled
by defences (with forts at intervals) to protect it from the threat of attack by the king’s forces, the situation with regard to the original ditch front ing the city wall would seem to be still unresolved. Contemporary maps show that while in the mid-sixteenth century the ditch was preserved and open throughout the whole of its length from Aldersgate to the postern on Tower Hill,2 by the following century buildings and gardens had en croached upon it in the Minories and Houndsditch sectors and the ditch is not shown at all. Returning now to the St. Alphage section, it is this encroachment of
buildings upon the ditch that makes for difficulty both in determining the
size of the ditch itself-as in Cripplegate churchyard-and in relating it to other features in the section. The modern foundations are likely to be the more massive successors of the earlier walls which no doubt reluctantly respected the ditch while moving as far across its outer lip as their builders dared. At St. Alphage the modern foundation cuts the section at the point where the mediaeval ditch would have made contact with the flat expanse to the north: it is impossible therefore to say what their relationship was; to decide (for instance) whether the mediaeval ditch had been dug through the filling of the broad hollow or vice versa. Until the evidence of Cripplegate churchyard became available it was
thought likely that the broad hollow represented the original mediaeval ditch, which was ‘begun to be made’ according to Stow, \ . . in the yere 1 2 1 1 & was finished in the yeare 1 2 13 ’.1 Stow goes on to say that the ditch was then made ‘of 200. foot broad’ ; and while this dimension cannot have been constant there was the possibility that the St. Alphage feature might be part of the exceptionally broad stretch that according to the early maps fronted the wall in the Moorfields area. This possibility still remains and if it could ever be demonstrated as a fact would involve a rapid narrowing of the ditch on the west before Cripplegate is reached. On the other hand, the hollow may have been part of the waterlogged area of Moorfields, frequently referred to by Stow and other early writers from fitz Stephen (ob. 1 19 1) downwards. The conditions are generally thought to have been due to the impeding of the natural drainage by the city wall, which acted as a dam, in spite of culverts left for the Walbrook and its tributaries. The sandy floor of the hollow was overlaid by nearly two feet of dense clay silt, dark grey in colour, which was completely barren of finds. The site was covered by mixed materials which looked as if they had been deliberately dumped in order to raise the level. Over the moor in general this activity went on into the sixteenth century; of it Stow said sceptically' . . . if it be made leuell with the Battlements of the Cittie Wall, yet will it bee little the dryer, such is the Moorish nature of that ground’.2 The St. Alphage section suggests natural or semi-natural marsh conditions alleviated by the artificial raising of the ground level. The encroachment of buildings upon the city ditch already mentioned,
which took place at varying rates in different parts of the defences, marks 1 Survey of London (1908 ed.), I, p. 19. 2 Op. cit., I, pp. 32-3.
