Figure 5: Sketch reconstruction of a Rainford muffle.
A dump of pipes, pottery, and kiln material, excavated in the winter of 1978-9 at Rainford, Merseyside, includes parts of four vessels which are probably the earliest muffles recorded to date. This assemblage of pipes, pottery and kiln waste is dated by the excavator to the period 1630 to 1650 (Davey et al 1982, 192), arrived at by adjusting the typological date, based on Oswald, by reference to local excavated groups.
Parts of four muffle bases were recovered from this excavation. Of these one is c. 360mm diameter, two are c. 420mm diameter and the fourth is insufficiently represented for any calculation to be valid.
The first base is formed from a lightly iron-stained clay, firing pink to orange, with a high density of quartz sand included. The base encapsulates large sherds from a number of different pieces of kiln furniture, including one flat piece measuring 140 x 170mm, heavily vitrified and pierced with 23 holes, also four pieces from a cylindrical form c. 230mm in diameter. The base or perhaps base lining has no surviving wall but in the base wall angle are the remains of eight layers of white clay, bearing the impression of a woven cloth. This base is c. 360mm in diameter and 50mm thick.
The second set of remains may in fact represent more than one base. All of the fragments are formed from a similar clay to the previous item with the added inclusion of up to 25% volume of organic tempering. This is in the form of chopped grass and/or straw. The largest fragment, representing c. 25% of the circumference, displays on its underside the heat-masked outline of an 80mm wide radially aligned support. Some fragments have vitrified or flash-glazed outer surfaces. This base is c. 420mm in diameter and 45-65mm thick.
The third base is formed from the same clay as the first and, like the second, has evidence of a support outlined on the lower vitrified surface. A concentric scar on the outer edge of the upper surface indicates the position of a wall 31mm thick at its junction with the base. This base is c. 420mm in diameter and 38mm thick.
The fourth base is formed from a dense fabric fired to vitrification. There is no evidence of organic inclusions. Its limited resistance to thermal shock has resulted in spalling, into a series of flakes. A wall seating 30mm wide is evident on the upper surface. Insufficient material was recovered for a diameter estimate - incomplete thickness.
A number of fragments from cylindrical muffle walls were recovered. Those substantial enough for a diameter calculation yield internal dimensions in the region of 380mm. All are tempered with organic matter. None of the fragments display any evidence of external buttressing. A number of associated fragments, from a rim or collar, have been formed over the rims or edges of some other vessels or objects. There is great variation in the reflected rim or edge profiles.
These vessels are reported by King ( Davey et al 1982, 210-233) who makes a case, on the strength of rather inconclusive evidence, for a muffle with a permanent opening traversing the flue space. This concept relies entirely on the acceptance that the group of extended rim or collar fragments formed a bridge between the muffle and an outer shell of brickwork. Such an arrangement is unlikely to be a practical proposition. Differential rates of expansion between muffle and outer shell certainly rule out any possibility of a gas-tight seal at the rim to rim contact postulated. The group of rim extension fragments is more convincingly seen as the raising of a muffle wall, possibly to accommodate an increase in stem length.
The small size of these vessels, together with the complete absence of any internal or external features, makes the distinction between muffle and saggar extremely difficult. The evidence which tips the balance in favour of the former is the presence of radial support 'shadows' on two of the bases and the signs of intense heat in the vitrification of the fourth. These factors are both consistent with a fixed position directly above the fire.
Further support can be found by comparison with contemporary saggars from a nearby pottery kiln dump. If the Rainford pipemakers were using saggars to contain their pipes in open flame kilns, either for their own exclusive use or together with the already established local potters, the saggars of the two endeavours are likely to show marked similarities.
The contemporary pottery kiln dump, only 100 metres from the pipe kiln waste, was excavated in the winter of 1979 (Davey 1991b, 127-8). The material recovered provides the opportunity to compare the techniques employed in the construction of pottery saggars with the vessels used for clay pipe production. The four vessels used for the clay pipes are all hand formed, with roughly circular bases made from rolled cut or beaten slabs. The walls were formed separately and joined to the base periphery. In all cases the two elements have parted company along this joint, leaving a scar on base fragments. The fabrics used are predominantly light coloured, firing to a pink or buff colour. One of the four bases and much of the wall material has a high degree of organic tempering. Many saggar fragments were recovered from the pottery kiln dump. By contrast, these are all wheel-thrown with bases and walls of one piece. The fabric used fires to a deep cherry colour with inclusions of buff firing clay. There is no evidence of organic tempering in the fabric of any of these saggars. There is, however, a quantity of straw-tempered material from the site which is more massive than either saggar or pot and probably derives from the kiln structure.
In summary, it would appear that there is little or no common ground between the two endeavours. Both assemblages are dumps, with no associated structures, admitting the possibility that their ultimate proximity does not reflect their original relationship. There is no evidence in support of kiln sharing or joint production. In this context, and in view of the masking and intense fire damage to their bases, the vessels used to protect the pipes are more convincing as muffles than as saggars. They are exactly what might be expected in a prototype muffle: form similar to a common saggar, cylindrical without either internal or external projections; size on the upper limit for a saggar; base and walls thicker than usual in a saggar. There is no evidence for the hearth and firebox arrangement of the kiln or kilns to which these muffles belong.
Figure 6: Muffle reconstruction, 11 Benthall Lane.
Between March and June 1984 an excavation at 11 Benthall Lane, in what might be termed the Broseley production area, produced substantial remains of a pipe kiln muffle. Pipes associated with the kiln material, bearing the mark of Henry Bradley, are attributed by Higgins to the period 1660-90 ( Jones et al 1987, 5). In his report on the muffle material, Higgins includes a suggested reconstruction which is reproduced here (Figure 6). Examination of the material by this author confirms the accuracy of this figure. In conclusive support of the base wall junction as illustrated is a single fragment in which the two elements are fused together. Others fit loosely as illustrated. A single prop type buttress survives but amongst the unconnected rim fragments, which together make about half of the rim, there are four pieces with a horizontal pipe stem peg through the wall thickness, on which a buttress is likely to have been located. The probability is that there were a minimum of four buttresses projecting from the muffle wall just below the rim. The single surviving prop buttress is large, of sub-ovoid section, measuring c. 100mm and c. 60mm along its cardinal axes.
The external surfaces of the muffle wall exhibit a range of vitrification slagging and discoloration, intensifying from the base towards the rim. The underside of the base itself is merely discoloured except where it has been masked by the radially placed brick supports. There is a hole in the centre of the 'base'. The hole is c. 60mm in diameter, cut vertically through the base with a rounded upper arris and sharp right-angled lower arris. Three distinct layers of white clay lute cover the inside of the base and follow the rounded arris down into the centre hole. There is slight discoloration in the region of the hole but no slagging.
Figure 7: Sketch reconstruction of the Benthall muffle.
The material from which the Benthall muffle is constructed is difficult to evaluate. The base and lower wall fragments are formed from a clay which is low in iron, generally firing to a cream or grey colour. Both base and lower wall have a modicum of mineral inclusions and both are grass tempered. The wall fragments are reinforced with tobacco pipe stems arranged vertically at c. 50mm centres. There are no stems in the base. The upper wall, prop and rim appear to be made from a mixture of red and white clay of streaked or marbled appearance. Much of this has been subjected to a high temperature, displaying varying degrees of melt. In extreme circumstances it has the appearance of a charcoal grey pumice. It is not possible to see if this fabric had any organic content. Tobacco pipe stems are used vertically in the same manner as in the lower wall to reinforce the fabric.
Figure 8: Plan and Section of Muffle from Quay Street, Gloucester.
The Gloucester muffle has straight sides with right-angle corners. In Figure 8 the disconnected fragments are illustrated, arranged in the most simple possible regular form, that of a square. The base is 100mm in thickness and there is evidence of one hole through this which is shown in the centre. It is of course possible that the base was rectangular and that there was more than one hole passing through it. It is formed from an ill-prepared, predominantly white firing clay, with a rufous speckling of iron-rich particles. Only the upper 25mm of the base has any pipe stem content and this is randomly aligned, often breaking through the upper surface. The upturned outer lip is integrally formed. Into this lip pipe stems have been inserted vertically at c. 25mm centres and left projecting to key into the later formed wall. The outside of the base is punctuated by voids of sub-circular section c. 35mm diameter, reducing in size towards the interior. As none is complete, it is not possible to discover whether these voids terminate or form continuous airways. Slagging on the outside of the base is in patches, other areas being merely discoloured. The underside survives in minimal quantity, none of which is slagged.
The largest reassembled section of muffle wall measures 310mm in height and between 220mm and 350mm in width. The thickness measures 60mm at the junction with the base and 30mm at the rim. The external surface is variably discoloured and slagged, with the heaviest slagging towards the top. There are no external projections. The inner surface is covered with a number of layers of white slip wash. The wall has vertically aligned pipe stems included within it and has been formed in a number of stages of differing heights. Each stage has a rounded rim with pipe stems passing through connecting the stages. It is not possible to determine the original height of the wall, only that it was not less than 450mm from the lower surface of the base. Only one prop type buttress survives, of sub-circular section 90mm in diameter, formed to bridge a flue space of 100mm width between two flat surfaces. Included within the matrix of this buttress are pipe stems and bowls of Gloucester Type 4 dated to the period 1670-1700 ( Peacey 1979, 46).
Figure 9: Sketch representation based on the Gloucester evidence.
Figure 10: Plan and section of Muffle from Arcadia Buildings, Southwark.
The excavation of this site took place in 1977-9 by Martin Dean for Southwark and Lambeth Archaeological Committee (Peacey 1982, 3-12). The muffle is formed from a light-coloured clay with mixed mineral inclusions and voiding from burnt-out organic matter. The base is circular, 620mm diameter and 88mm thick, incorporating two horizontal layers of pipes - the second laid at right-angles to the first (Figure 10). In the lower layer the pipe bowls occur around the outer edge. No bowls are evident in the upper layer. The muffle wall is 35mm thick at the base and 15mm at the rim. A single layer of pipe stems is included within the thickness of the wall. These stems are placed diagonally in alternating horizontal bands, having the same effect as herringbone walling (ibid, 8, 3c).
Round the outside of the muffle are a number of projecting prop type buttresses to locate the muffle within the outer shell of the kiln. These buttresses are pegged into the muffle walls with a single radial pipe stem and are formed from the same material as the muffle body. The buttresses at the rim measure 22mm from the outside of the muffle wall to the inside of the outer shell. At its base the measurement is 70mm. This would have the effect of producing a tapering flue space all round the outside of the muffle. The assemblage contains 34 prop type buttresses. On the outside of the base are three ridge-shaped deflectors, also integral with the muffle body. These deflectors are clearly intended to smooth the gas flow over those parts of the muffle supports projecting beyond the outer edge of the muffle base. There is no evidence of any internal projecting features. In the centre of the base is a hole 18mm in diameter passing vertically through it. The hole is plugged at the upper end and covered by lute layers. Clearly this was carried out before the final firing. If plugged before the muffle was used then it must relate to some aspect of the construction, if later then to some abandoned practice. On the inner surfaces of the muffle are the remains of several layers of white slip wash. The outer surfaces are variously slagged and discoloured from contact with the fire. Figure 11 illustrates the basic form and features of this muffle.
Figure 11: Details of muffle from Arcadia Buildings, Southwark.
A second kiln at Arcadia Buildings, sharing the same stoke pit, and broadly contemporary with the first, yielded less muffle fabric. Notably differing from the muffle described above, this assemblage included a number of bar type buttress variations. There are fifteen fragments, two of which join to form a tapering bar 230mm long with a break at both ends. It has a bell-shaped horizontal section with axial pipe stems protruding from the rounded edge to peg into the muffle wall. The splayed edge is flat with irregularities caused by contact with brickwork.
There are seven short bars varying from 70mm to 105mm long with their ends either rounded or groined. The flue space bridged by these bars varied from 35mm to 50mm. The upper end and sides of these buttresses are splayed where they had contact with the outer shell. The lower ends are not splayed indicating inaccessibility during construction. It seems likely that the building of the outer shell progressed with the construction of the muffle and that buttresses were formed between the two, pegged by insertion of stems from the inside of the muffle and smoothed or splayed from above. Of the remaining six fragments, five have one rounded or groined and one broken end whilst the final piece has both ends broken. The flue space bridged by these varied from 25mm to 30mm.
Figure 12: Sketch based on the Southwark evidence [Detail Portsmouth].
There are thirteen bar type buttress fragments from the dumped assemblage PO1, and five found in association with a kiln PO2. Of these, only one has any pipe stem included, this being aligned vertically; the remaining seventeen are without either reinforcement or pegging. The stems in wall sections are aligned horizontally. The buttresses from PO1 are rounded at their upper end which coincides with the rim of the muffle. They exhibit stepped outer surfaces indicative of a corbelled outer brick built shell to the kiln. Two of these buttresses adjoined an opening in the muffle wall. This evidence is drawn together and illustrated in Figure 12.
The Chard group represents another demolition dump with no associated structure. It includes four bar type buttresses and four wall sections with scars from which such buttresses have been broken. Pipe stems included in the buttresses are vertically aligned. In the muffle wall, stems follow both vertical and horizontal alignments. There is no evidence of pegging between the wall and the buttresses.
Castle Street, Aylesbury, excavated in 1979, produced three prop type buttresses in association with various short bar types. There are ten short bars with rounded upper ends, nine with groined upper ends and ten where the upper end is coincidental with the rim. All of these have rounded sub-hemispherical ('tear drop') lower ends. The sides and upper ends of the rounded and groined examples are splayed into the muffle wall and outwards to the contact surface. Three of these have pipe bowls included within the matrix and four have voids from which bowls have become dislodged. These bowls conform to Oswald's Type G9, dating 1680-1710. There is no evidence of pegging into the muffle wall. The upper ends of the rim buttresses are formed square with the rim and the sides are flared into the muffle wall. All are pegged into the wall with pipe stems and bridge a flue space of 25-30mm. Stems in the muffle wall are at an angle to the rim in alternating 'herringbone' bands.
The kiln excavated at Holywell Hill, St. Albans, in 1970 has a slightly later date, producing an assemblage of pipe forms with a typological span of 1680-1730. The forty-one buttresses recovered from this kiln are all of prop type. All are roughly formed by hand around a cluster of pipe stems. In some cases the stems protrude from the outer contact surface and appear to have been located in the joints of the outer shell brickwork. Red brick fragments remain fused to some of the prop ends. None remain attached to the muffle wall but it appears from the protruding stems that they were pegged. The flue space bridged by these buttresses varied from 35mm to 100mm. Pipe stems included in the muffle wall are at an angle of c. 45 degrees to the rim.
© Internet Archaeology URL:
Last updated: Wed Oct 9 1996