The use of ceramic building materials in the medieval period was subject to even more regional variation than was the use of pottery vessels. To take one example, the areas in which flat ceramic roof tiles were used seem to have been dicontinuous and quite sharply defined. Therefore, the scraps of evidence for medieval ceramic building materials cannot always be combined from different areas to produce a generalisation of the true state of affairs in the study region and there is even less likelihood that the sequence of use revealed in the study region can be transferred to other regions of the country.
Another reason for the lack of certainty about the development of the industry is that the raw data for this study has been very unevenly sampled. Only half a dozen sites have been excavated in which ceramic building materials have been adequately sampled. The reasons for this are not difficult to find. At 143-5 Bartholomew Street, Newbury, the author attempted to keep every flat tile with more than one dimension and every fragment of curved or otherwise featured tile. Even the collection of this partial sample placed a great strain on the processing of finds from the excavation. It is therefore not surprising that the collection of flat roof tile has been uneven. It is also not generally realised by excavators that the use of brick and tile varies significantly both between areas and also with the status of the structure under excavation. There has been, therefore, no forceful argument in favour of collection to justify the amount of effort required to collect, process and store the samples.
Some types of ceramic building material have suffered more than others from this differential collection. Decorated floor tiles have been collected with the same rigour as potsherds but undecorated floor tiles, especially unglazed examples, may well be under-represented while the collection of bricks, flat roof tiles and daub is exceptional and their absence from the collection of finds from a site cannot be taken to imply that they were not used there.
With these warnings in mind it is possible to reconstruct the outlines of the history of the ceramic building material industry, which is discussed below first in chronological outline and secondly type by type.
Until the late 12th century there is no evidence from the study region for the production for trade of ceramic building materials. Clay was utilised for walling, hearths and the production of loom weights and it is unlikely that all this clay could have been obtained from within the property of those using it. However, petrological analysis has not revealed any long-distance trade in unfired clay and it may be assumed that carting of clay over distances greater than a couple of miles would have been rare.
Preparation of clay by tempering is probable at Gloucester in the 10th century, where the same fabric was used for the daub walls of a pottery kiln and for the manufacture of loom weights. This fabric was not used for the pottery made in the kiln, which had a limestone temper.
The evidence to date therefore shows an awareness of the properties of clay at all levels of society thoughout the late Saxon and early medieval periods but no exploitation of clay pits on more than a domestic scale.
During the late 12th to mid-13th centuries there is evidence for widespread but intermittent development of a ceramic building material industry. Flat roof tiles were used, albeit only occasionally, at Gloucester and Hereford, and were in general use in the Kennet valley (figs.9.8, 9.9). The fabric of tiles from Gloucester indicates that they were not made by potters experimenting with tile production but were rather the products of a specialist tiler. The only early 13th century fragment of flat roof tile from Hereford is, however, in a pottery fabric, Hereford A2.
Where flat roof tiles were made of fired clay ridge tiles would certainly also have been ceramic. In south Worcestershire, ridge tiles were produced in Worcester-type ware and are found at sites in the Severn valley from Gloucester to Droitwich (fig.9.1). Both the characteristically laminated fabric and the distinctive tall, hand formed knobs of these tiles show that these ridge tiles originated in the same centre.
The production of floor tiles in England also began in the late 12th century but no tiles of such an early date are known from the study region.
The pattern of use and production of ceramic building materials seems to have stabilised at the end of the 13th century so that regional patterns of tile use developed (figs.9.2, 9.9). Flat roof tiles were used in the Kennet valley, the south-east of Wiltshire and possibly in and around Worcester (fig.9.10). This production was normally carried out by specialist tilers and the only evidence for any combination of tile and pottery manufacture comes from Laverstock. Even there, documentary evidence seems to indicate that specialist tilers were also in operation in the area although working in separate centres.
In the Severn valley, south Wales and north Wessex, excluding the Kennet valley, ridge tiles, finials and louvers were manufactured by potters and were traded over similar distances as hollow ware vessels (fig.9.2). In the Kennet valley this roof furniture seems to have been made by tilers. The general division between the manufacture of flat roof tiles and possibly undecorated ridge tiles and the manufacture of decorated ridge tiles, louvers and finials seems to be general over the whole of southern England and is illustrated by documentary evidence from Banstead, in Surrey, where in 1372/3 flat roof tiles were obtained from two sources, Ashtead and Reigate, whereas the decorated ridge tiles were obtained from the pottery at Cheam (Moorhouse, 1981, 108). This division seems to make sense from a technical standpoint, since finials are often wheelthrown and all types of roof furniture except the simplest undecorated ridge tiles require dexterity and skills alien to tile production yet commonly used in the production of pottery vessels.
There seems to have been no major change in the mode of production of floor tiles between the late 13th and early 16th centuries. Two main types of production have been demonstrated. The first is 'factory' production in which the tiles were produced by a large stable industry, often encompasing several generations of tilers. Examples of this type of production range from the earliest tiles used in the Severn valley, the St. Bartholomew's Hospital, Gloucester, - Cleeve Abbey type, to some of the latest, produced at Malvern Chase. In three cases floor tiles were produced in the same areas as pottery vessels, at Malvern Chase, Nash Hill and in the Salisbury area. There is, however, no evidence for the production of floor tiles by potters, nor for the firing of floor tiles in the same kilns as pots. In fact, whereever evidence has come to light it seems to show the opposite, even when, as at Nash Hill, pottery and floor tiles were produced in sucessive kilns. There is no evidence from the study region for the production of floor tiles and flat roof tiles in the same centres, although it would not be surprising to find such evidence in the Newbury area, where both pegtiles and floor tiles were being produced in the late 14th to early 15th century (Eames, 1980, 215-8). Floor tiles and bricks were produced together at Wigmore Abbey at the begining of the 16th century and there may be a connection between the production of Canynges-type floor tiles and that of brick and tiles at Worcester, although there is as yet no archaeological proof for this suggestion.
The distances over which ceramic building materials were travelling during the later medieval and early Tudor periods are comparable to those travelled by contemporary pottery. Both pottery and ceramic building materials produced in south Worcestershire show an increase in the distance travelled in the 15th century, especially, but not only, down river to site in the Bristol Avon and beyond. In the late 13th to 14th centuries sites such as Chepstow and Bristol were recieving few if any ceramic goods from south Worcestershire, although manufacture of pottery, ridge tiles and floor tiles was already on a large scale there. However, in the early 15th century Droitwich-type floortiles were exported to sites in the Lower Severn, as were the later Great Malvern, Canynges-type and Malvern Chase tiles. From the late 15th century onwards there is plentiful evidence for the export down river of Malvern Chase pottery.
With the exception of the Worcester industry, production of bricks and flat roof tiles began in the Severn valley in the late 16th century (fig.9.11). Only three source areas have been examined in detail, Hereford, Gloucester and Malvern Chase, and of these only the Malvern Chase products have a distinctive petrology. Brick and flat roof tile from Malvern Chase has been found at Hereford, Breinton and Gloucester and there is documentary evidence for the use of Malvern Chase brick in the construction of St. Katharines Chapel, Ledbury.
The distances over which Malvern Chase brick and flat roof tile have been demonstrated to travel are much lower than those of the late 16th century pottery from the same source. However, as noted in the introduction to this chapter, there are great dangers in using negative evidence in the study of ceramic building materials and although likely it should not yet be assumed that late 16th century bricks and flat roof tiles were not being transported over similar distances to pottery vessels.
In the early 17th century there is a notable change in the location and organisation of the pottery industry, with many small potteries taking the place of the large industry based at Malvern Chase. It is possible that a similar process took place in the brick and tile industry with large industries such as those at Worcester and Malvern Chase itself being replaced by small centres. Evidence from parish registers, mainly of a slightly later period in the late 17th century, shows that brickmaking was a common occupation in most areas while the first editions of the Ordnance Survey maps of the region show that brick kilns were a regular feature of the landscape in the 19th century.
Details of the development of individual ceramic building materials are given below.
Ceramic ridge tiles can be used on any type of pitched roof, whether it is made of turf, thatch, wooden shingles, stone tiles or ceramic tiles.
The first archaeological evidence for the use of ceramic ridge tiles in the study region is from Gloucester in the 12th century, where they are undecorated and probably made with flat roof tiles in Gloucester TF89 (see below). Decorated ridge tiles start in the mid- 13th century. The earliest recognised are in Worcester-type ware and are probably the same date as the associated pottery (mid- to late 13th century).
In Hereford the earliest tiles were thought to be knife- cut coxcomb tiles in Hereford A7b of a type with short, stabbed crests. This type was found only at Hereford Castle. If they are the same date as the pottery in this fabric they must be mid-13th century or later. From sites in the town of Hereford itself, the earliest ridge tiles derive from late 13th to 14th century deposits and are also made in Hereford A7b fabric. A similar date is likely for the introduction of decorated ridge tiles over much of the study region. With one exception wherever a detailed petrological analysis has been made the ridge tiles are made in the same fabric as a contemporary hollow ware. This leaves little doubt that in the Severn valley and the west of England decorated ridge tiles were made by potters rather than tilers. The exception to this rule is found in Berkshire, where decorated ridge tiles are found in Newbury tile fabric. Significantly, Newbury Group C ware is the only late 13th to 14th century ware known in the study region which did not also produce ridge tiles. It is likely that in regions where flat tiles are current in the medieval period ridge tiles were made in these centres rather than with pottery.
There are several regional variations in the decoration of ridge tiles. Knife-cut coxcombs occur virtually everywhere, although there are minor variations between different centres. Fretwork ridge tiles occur in Berkshire and Hampshire (Platt and Coleman-Smith, 1975, nos.1402-3, 1406, 1411, 1415 & 1417) while knobbed ridge tiles occur in the industries of Herefordshire and Worcestershire only, although the products have a much wider distribution (figs.9.1, 9.2, 9.3).
Minety, Wiltshire, is the only centre in the region to produce handformed coxcomb tiles in the medieval period but similar tiles were made in the 17th century in North Devon and at Stroat, Gloucestershire, which also looks to the south-west in the style of its bowls and jars (see Ch.11). It is noteworthy that despite the overlap in distribution found in the products of different centres there is little borrowing of decorative styles. Malvern Chase produced a few knife-cut coxcomb tiles and Hereford A7b produced a few knobbed tiles but in general once a style was established in an area it continued to be produced there.
Finials and louvers, although rare, were produced in the same fabrics as decorated ridge tiles, so that those in Berkshire were made in tile fabrics and those in the rest of the region were made in pottery fabrics. In London, however, ridge tiles were made in tile fabrics but louvers are made in pottery fabrics, perhaps because the forming techniques used are the same as those used in making hollow wares. No louvers are known from Berkshire, but possibly in their place one finds vessels which may be chimney pots made in a pottery fabric (Newbury Group B). The distribution of decorated ridge tiles is shown in figs. 9.1 to 9.3 and 9.7. For the late 14th century onwards much of the dating is based on knowledge of the history of the contemporary pottery industries rather than on stratigraphic evidence. It is very difficult to date the end of production of a ridge tile type since the tiles could have such a long life. One mid-18th century pit group from Gloucester contains Gloucester TF88 flat tiles and an almost complete two-knobbed Malvern Chase ridge tile which must have been made in the early 17th century at the very latest.
There is little difference between the distribution of ridge tiles and that of the associated pottery with a few minor exceptions. Minety ridge tiles are not found in Glouester to any great extent although Malvern Chase and Hereford A7b tiles are. There are also no known Gloucester TF99 ridge tiles found in the study area.
This general similarity in distribution areas is despite a difference in the method of marketing. Ridge tiles are often itemised on building accounts and sometimes the potter or tiler involved is mentioned by name. Elsewhere it is recorded that the mason in charge of the building went to the pottery to bargain for tiles. In only two cases do we have any documentary evidence for the source of tiles; firstly at Newport Castle in 1448, where a Cardiff potter provided them (Pugh, 1963). "et solutum Johanni Croker de Kaerdyffe pro ij duodenis de crestis de eodem emptis pro coopertura summitatis camera predicte iij s." (Item, paid to John Croker of Kaerdyffe for two dozen ridge tiles, purchased for covering the roof of the said chamber, 3s. 0d). Secondly at Ledbury in the late 16th century ridge tiles were obtained from Hanley Castle, the main manor in Malvern Chase (Morgan, 1955).
Salzman (1952) quotes the price of ridge tiles at 1.5d each at Moor End, in 1366 and 36 crests at 6s 2d in 1432 at Kings Hall. An exceptionally expensive purchase was 2s for 2 crests made in the fashion of mounted knights at Banstead Manor, Surrey, in 1373. The fact that only two were bought may indicate that they were replacements for an earlier roof or that they were used as finials at either end of the ridge. Salzman also notes references to 'Holtill' (hollow tile) and 'fyneax' or 'festeaux' which may be finials and louvers. Ridge tiles were on average 30 or 40 times as expensive as contemporary flat tiles, while the Banstead Manor tiles were between 200 and 250 times as expensive (see below). The building accounts for St. Katharine's Hospital, Ledbury between 1584 and 1595 show us the sort of demand for Malvern Chase tiles that may have been typical.
|
1587 |
Dec. 8th item for vij crests |
vijd |
|
1591 |
May 21st item twoe crests |
ijd |
|
1591 |
August itm iij crests |
iijd |
|
1593 |
July 28th item vi crests |
vjd |
|
1593 |
Sept. item two dussen of crests for the same house [a mill house] |
ijs |
|
1595 |
July item for xvj crests |
xvjd |
Thus over a period of 8 years only 58 ridge tiles were purchased at a cost of 4s 10d compared with over 13s 4d spent on c.1400 Malvern Chase bricks (the buildings were timber- framed and stone tiles were used for the roofs).
The decline in production of decorated ridge tiles began in the early 17th century with the production of plain ridge tiles by the Herefordshire kilns and Ashton Keynes, Wilts. However, Stroat, Gloucestershire, was still producing decorated ridge tiles and in the late 17th to early 18th century North Devon decorated ridge tiles were imported to Gloucester. The longevity of this tradition is such that there are still medieval or early post-medieval decorated ridge tiles to be seen on roofs in Worcestershire and Gloucestershire to the present day.
There are also at least two Malvern Chase finials still in use on houses in Hanley Castle but it is likely that their production ceased when brick chimneystacks were introduced during the late 16th century. Ceramic chimney pots are not found until the late 18th century in the region, with the exception of the two Newbury Group B examples from Berkshire (see ch.2).
In the distribution maps (figs.9.1-3, 9.7) only Malvern Chase tiles can be securely split into chronological groups, and these of only the broadest nature: In the late 16th to 17th centuries these tiles have a clear glaze and are higher fired than in the previous periods (fig.9.7). The remaining maps are based on the earliest occurrences and analogy from pottery made in the same industry.
The earliest documentary reference to tile manufacture in the study region is in a Saxon Land Charter for Wooton-under- Edge (Grundy, 1935-6, 282). "These are the bounds of Wooton: First from the ash tree along the top of the ridge slope to the lea where the tiles are made..."
Grundy identifies this locality as 'Tyley' but nothing is known of the type of tiles produced. They were most likely stone tiles and could as easily have been used for walling or flooring as for roofing. The term 'brick' although used on the continent from the 13th century is first found in this country in the 15th century (Salzman, 1952).
The earliest evidence for the use of ceramic flat roof tiles comes from Eastern England, including Berkshire. At Newbury and Reading Abbey there is evidence for the use of pegtiles from the second half of the 12th century onwards, while at Reading Abbey flanged and curved tiles were also found in the late 12th century, but in a different fabric from the peg tiles. Other evidence for tile production in Berkshire comes from place-names. Both 'the Tilery' in Windsor and 'Tilehurst', to the west of Reading are first recorded as place-names in the late 12th century. Archaeological evidence for 12th century pegtile use comes from London (where they are found in waterfront deposits dated pre-c.1170 by dendrochronology) and Essex, where pegtiles are visible in the fabric of several mid-12th century churches examined by Messrs. P. Drury and W. Rodwell (Drury, 1980). Shouldered pegtiles are known at present only from London and a single example from Gloucester (fig.9.8). Although other fragments of Gloucester TF89 tile may be of shouldered pegtiles most are from curved tiles, presumably ridge tiles. Flanged and curved tiles are known from Reading Abbey, London, Southampton and Bayham Abbey but a single fragment from Bewell House, Hereford may be either from a flanged tile or just possibly a ridge tile with a flange at the base. In either case it is the earliest glazed roof tile from the town and was found in an early 13th century context. The fabric, Hereford A2, is first found in the late 12th century.
Occurrences of these three types of flat roof tile are extensive and it would not be surprising to find further instances of their use at this time, probably on buildings of high quality employing non-local craftsmen (for example abbeys and castles). It appears that the shouldered pegtiles and flanged and curved tiles, were unsuccessful experiments and are not found after the early 13th century.
Pegtiles were not generally adopted in any part of the study region in the medieval period except the Kennet valley (fig.9.9). Pegtiles alone are found in the late 12th century at Newbury. They were used in large quantities on the Bartholomew Street sites from the late 13th century onwards, both for walling and set on edge as hearths. During the late 14th century, and possibly later, floor tiles were also produced in Newbury tile fabric. There is a documentary reference to the use of pegtiles at Marlborough in the mid-13th century (see below).
Worcester is the only site in the study region outside Berkshire which might have a medieval flat roof tile industry. There are references to tilers in Worcester in the 13th, 14th and 15th centuries (1299AD Northwich and Whitestones, Johannes tegulator quoted in Hollings, 1950, 7) and although expressly forbidden to form a guild by the Kalendar of Ordinances of 1467 (Toulmin Smith, 1870) which were re-enforced in 1497 (V.C.H. Worcs. II, 275) there was a building outside St. Martin's Gate known as the Tilehouse (Habington, ed. Amphlett, 1899, 45). Study of the floor tile industry suggests that only two possible floor tile groups could have been made in the town; the Droitwich-type tile group and the Canynges-type tile group. The first is known to have been produced at Droitwich, but could also have been made in Worcester (the towns are only c.6 miles apart and would probably have access to petrologically identical sands and clays) and the second is first produced in the late 15th century and is thus too late to be responsible for all of the documentary references. It is therefore most likely that some of the nibbed and pegged tiles found in Worcester are of late medieval date. By the 16th century bricks were being made by Worcester tilers (Dyer, 1973) and it is also possible that some of the late medieval tilers were making 'wall-tile' or bricks. Evidence for the Worcester tile industry is summarised by Carver (1980).
A flat roof tile industry is known from documentary evidence at Alderbury, near Salisbury. This industry was supplying Clarendon Palace with large quantities of tile between 1354 and 1388. These tiles included crest tiles (Musty et al., 1969, 83).
Salzman quotes several prices for roof tiles taken from contemporary documents (Table 9.5). Although it is sometimes not clear whether bricks or flat roof tiles were being referred to there are several which from their location must refer to flat roof tiles.
| Site | Date | Price per 1000 |
| Marlborough | 1237 | 3s |
| London | 1258 | 3s |
| London | 1278 | 3s |
| Guildford | 1291 | 2s ** |
| London | 1350 | 5s |
| Clarendon Palace | 1354 | 4s @@ |
| Battle Abbey | 1355 | 2s 6d |
| General | L14th C | 4s to 5s 6d |
** plus 3d for carriage
All data from Salzman (1952) except @@ (Clarendon Works Accts. PRO E 101/459/29).
An Act passed in 1477 states that tiles were then lasting only 4 or 5 years instead of the expected 40 to 50 years and regulates the manufacturing processes and the sizes of tiles. Peg tiles were to be 10.5" by 6.25" by at least 5/8" thick, Ridge tiles were to be 13.5" by 6.25" and gutter tiles were to be 10.5" long (Celoria & West, 1967)
Drury and Pratt (1975) have published the excavation of a tile kiln from Danbury in Essex which, from its range of products, is probably comparable with the Newbury industry. Floor tiles, glazed pegtiles and ridge tiles were produced at Danbury in a rectangular, tile-built kiln.
From Wigmore Abbey comes evidence for the manufacture of brick in the late 15th or early 16th centuries alongside floor tiles, a combination not yet found elsewhere in the study region but which might be expected at Worcester. Although the kiln was not found it is likely that the Wigmore tiles were made on the site of the Abbey for use there, perhaps because of the remoteness of the site.
At Reading and Newbury too brick was a 16th century introduction, predating the Dissolution at Reading Abbey, and at both sites was made in a sandier fabric than the peg tiles.
Elsewhere in the study region there is no certain evidence for the use of brick before the Dissolution of the Monasteries. At Malvern Chase, Hereford and Gloucester both bricks and nibbed tiles were introduced together in the late 16th centuries. Since both Hereford and Gloucester received some Malvern Chase bricks and flat tiles it is possible that production in Malvern Chase began earlier than at Hereford and Gloucester but there is no stratigraphic evidence for this. The building accounts of St. Katharine's Hospital, Ledbury show the quantities of brick used in the late 16th century at a time when most buildings were timber-framed (Morgan, 1955,88-131).
|
1587 |
Sept 17th |
item to Grundie [John Grundy, Mason] for going to Hanley to bargayne for brick and for half a days work |
Vijd |
|
1588 |
Sept 9th |
item for brick |
Iijs viijd |
|
1588 |
Nov 25th |
item for 100 of brick |
Xvjd |
|
1593 |
July 17th |
it. a thousand of brick for the same [2 new houses and renewing a furnace] |
Viij iiijd |
The latest evidence for brick production in the Chase is from 1637 when Edmond Thold and Edmond Sacrament, brickmakers, of Castlemorton are recorded (Q.S.R. part ii, 625). By the late 17th century brickmaking had moved to the side of the River Severn, exploiting the silty alluvial clays rather than the keuper marls of the Malvern Chase. Hanley Castle church was rebuilt in silty brick in c.1674 and an agreement of 1684 records the setting up of a brickworks at Upton-on-Severn by Edward Addis. A member of the Addis family was recorded as a potter in Hanley Castle in the late 16th century.
Although not studied in detail, as it is outside the scope of this study, it appears that references to late 17th century brickmakers are so numerous as to suggest a very local distribution of their products. There may in fact be a change in the organisation of the industry from the late 16th to 17th century, when the demand was for brick as a minor element in timber-framed structures to the late 17th century when brick was increasingly being used on its own.