11.1.3 Furniture supplements.

The presentation of evidence for furniture supplements set out in Chapter 6 includes an alphabetical list of sites and types of object recovered from them (Table 4). This same data is presented below in chronological order to highlight the sequence of development (Table 8). It is immediately apparent that a high percentage of the fabrics used for furniture supplements are either clean white inclusion free clay or similar material with organic filler. Of the 133 fabrics used for this purpose, 81 conform to the former and 33 to the latter, together representing 85% of the total number. Of those fabrics with visible mineral inclusions only one, from Leicester 2, has sufficient quantity to imply either conscious selection of a clay with high grit content or addition of such material to the clay.

The following analysis is based on the number of sites using a particular type of fabric for a particular type of object (rolls, straps, wads etc). Quantities of material and typologies are ignored. Where there is an entry in any column against a site fabric this is counted as one site incidence. Using this method it is clear that the addition of organic material to the clay was predominantly used for the manufacture of thin sheets. Organic material occurs in 50 places in the table, 35 of which are thin sheet entries. The number of clean white clay incidences are 182 of which 50 are thin sheet. Clearly then thin sheets were intentionally made from both fabric types. Chronology does not seem to be a significant factor. At Trowbridge 2, Portsmouth 1, Chelmsford 2, Southwark 2, Wickwar 1, Chatham 2, Liverpool 2, Hull 1, Chatham 1, Northampton 1, Lincoln 1, Winchester 3, Sheffield 1, Burford 1, Canterbury 2, Athelney 1, Ipswich 3, Newark 2, Boston 1 and Nantgarw thin sheets occur on both types of fabric. It seems reasonable to suggest therefore that the fabric adjustment served some intentional purpose; either to save clay or to reduce weight. As both the incidence and quantity of material are greater for the unadjusted fabric weight saving seems the more likely. This is born out by the slag laminate evidence.

Within the table, of the 52 instances of roll occurrence, 45 are in clean white clay similar to that used for pipe manufacture; of the 18 instances of strap, 17 are in this fabric; of the 54 instances of wad, 42 are in this fabric; of the 17 instances of applied strip, 16 are in this fabric; of the 12 instances of rack, 9 are in this fabric. In the remaining fabrics the visible mineral inclusions are, with the odd exception, sufficiently small and sparsely distributed to be regarded as accidental. Organic material features in 3 roll incidences; 1 strap incidence; 5 wad incidences; 1 applied strip incidence and 2 rack incidences. Clean fine textured red clay has been used for rolls in 2 cases and for wads in two cases. Interpretation of the various uses to which these supplements were put is dependent in part upon their contact impressions and part upon gleanings from contemporary literature and photographs.

Rolls, Straps, Wads and Applied Strips

Simple rolls of clay were used to separate and stabilise ranges of pipes in the kiln where any shift in the load during the firing could prove disastrous. These soft clay rolls were able to adjust to any profile often displaying the impressions of stems or bowls with which they were in contact. Figure 31 illustrates a sample of the types of contact impressions encountered on rolls. Vestigial impressions as those of a, d, e and f are consistent with the support of raw pipes. Deep, more widely spaced impressions as in b and c suggest a framework of fired pipe stems which could be pressed with more force into the soft clay roll. The majority of fragments are too short for extrapolation of complete form. In a few cases such as a and d longer sections suggest a sub circular form. The positioning of the stem impressions in these two cases are indicative of different applications. In the photographs referred to above, similar rolls of clay are used to encircle the stem ends protruding above the saggar rim. It is conceivable that 31a might have been used in this way although the stem impressions might be expected to occur further toward the inner face. Serpentine rolls with impressions on the concave parts of opposing surfaces were clearly used between layers of pipes; in the case of 31e, between bowls. Rolls with stem end impressions are more scarce and most likely indicate accidental rather than functional contact. The best group is from Leicester, Causeway Lane, where of 144 roll fragments 9 have stem end impressions and 2 have transverse stem impressions. Although the purpose of this material is unclear, the stem end impressions suggest a crowning framework resting upon the top of the charge. This is the one group in which the fabric is densely braced with fine quartz sand. Tapered roll fragments are also scarce and probably come from the ends of hand formed rolls which tend naturally to taper off. Figure 31j is a perfect example of this. Type 5 rolls, pressed out to form struts, are known only from Leicester, Causeway Lane, where they appear to have been used to prop some unknown object (Figure 32a-f). It is difficult to conceive how such items could fit the known model. Together with the unusual roll assemblage discussed above, these items suggest some additional or divergent aspect to the emerging picture of eighteenth century practice.

From surface impressions it can be seen that straps were formed from rolls by pressing; often in the palm of the hand and sometimes on the bench or other convenient surface. They vary considerably in both size and in the delicacy or otherwise of their formation. This variation is more likely to reflect personal idiosyncrasies of makers than relate to differences of function. Straps appear to have been used in much the same way as serpentine rolls; laid over a layer of pipes with a subsequent layer placed above. Examples of straps with both stem and bowl impressions are illustrated (Figure 32h-j). No supporting evidence from either contemporary literature or photographs is known to the writer.

Wads were formed from rolls or balls of clay placed between other objects to form a support tailored to their specific surface requirements; acting in the same way as a mortar joint in brickwork. In order to build a stable column, of prefired props and buns, some form of bedding was required to negate the effect of irregularities, which would otherwise cause the column to rock. A wad of soft clay was the simplest available solution to the problem. Wads used for this purpose tend to be either in the form of rings, made from looped rolls (Figure 33o & p), or discs (Figure 33k & l), made from flattened balls of clay. It can be seen from the raised circular intrusion that 33l was used to bed a hollow prop. Wads were also used as bedding for bricks (Figure 33m & n) and between saggars. Strips of soft clay were applied to surfaces within the kiln chamber to provide an interface between these and the charge of pipes (Figures 34 & 35). It would appear that these soft strips served to locate the charge in a similar manner to the indented props and buns described above (Figure 108a-f). Applied strips usually bear parallel impressions of stems lightly impressed to form a restricting corrugation. In some cases strips have more deeply formed indentations suggesting that the strip housed a framework of fired rather than raw pipes (Figure 35a). A framework so formed could act either as a base for further ranges of pipes to be fired or as a support for the muffle or saggar cover. The radii of contact surfaces can provide clues to the likely sites that these strips occupied. Tight curves (Figure 35a) would be formed by the application of a strip to the the shaft of a prop whilst curves of greater radii (Figure 34q, r, s & w) might reflect the surface of an inverted dish or the outer surface of a saggar wall. All of the items described above were probably first formed as rolls. The straps were flattened prior to use, the applied strips deformed by the act of pressing them home. These latter range in section from variously flattened 'D's to triangular forms. A group, of triangular section, from Rainford, Rookery Farm , display their roll origins in the tapered horn terminals (Figure 35e & f).

Thin sheets

A second group of supplementary furniture is based upon the thin clay sheet (Figures 36 & 37). These sheets are characterised by one soft textured face with scrape or drag marks on the reverse. The soft face results from the sheet having been formed by spreading clay slip over a sheet of paper (or in some cases cloth). The paper remained attached to the surface of the very thin sheet of clay facilitating handling. By this means very thin light weight sheets could be formed and handled. In the kiln the paper or cloth was burned leaving only its soft textured impression on one side of the sheet. The artefactual evidence is supported by documentary evidence for this practice.

Duhamel du Monceau writing in 1771 of a kiln at Rouen, which bears startling similarities to English kilns, describes the use of paper covered with a thin layer of clay and horse manure to create a dome above the charge of pipes in the firing chamber.

Rees in his Manufacturing Industry of 1819-20 describing a damper in the dome of a pipe kiln says

'This is a mixture of horse-dung, sand, and pipe-clay, well worked together, and spread in thin layers upon coarse brown paper; a sheet of this being laid over the hole in the dome, so as to cover more or less of it, will give the means of increasing or diminishing the draught, and, consequently, the heat of the furnace'.

From 1881 there is the account in Tobacco Trade Review of McLardy's Manchester works, which describes the pipes,

'ranged round covered with very thin sheets of clay backed with paper'.

The practice survived into the present century, is noted in accounts of the industry at Leicester and Bristol and is recalled by Gordon Pollock who still makes pipes in Manchester. William Flanagan, a native of Broseley, moved to Leicester around 1885 and made pipes there until 1919. In 1959 his son T H Flanagan described the working practices employed by his father to J A Daniell who published an account from which the following extract is taken.

'The kiln was bee-hive shaped, built of brick and completely enclosed, apart from the draught holes. It was necessary to demolish one side of the kiln after each firing in order to remove the pipes and build up the brickwork again when the next batch was in position ready for firing. A batch consisted of about 40 gross of pipes.

It was customary to build up a central column of fire-clay inside the kiln and to stack the pipes against this in order to form a pyramid with the bowls pointing outwards. When the pipes had been stacked in the kiln it was essential that they should remain in position and that the heat should be retained at a very high temperature. In order to achieve this the pipes were first covered with several thicknesses of old play-bills, specially obtained for this purpose from the Pavilion and Palace Theatres. The paper pyramid thus formed was covered completely by a thick paste made from a mixture of clay and horse manure. The fire was lit and the kiln was bricked up. Firing normally took from eight to ten hours. When the fire had been extinguished the pipes were left to cool for a further five or six hours before being removed from the kiln. So far, in Mr. Flanagan's case, all the heavy work was done by himself but his wife was always at hand to assist him. No interruptions were allowed, even for meals, until the kiln wall had been bricked up' (Daniell 1964, 61).

Whilst this description is far from comprehensive, the reference to a central column, the covering for the pipes made from paper, clay and horse manure, and the absence of saggars all suggest a muffle kiln. A similarly anecdotal account by Mrs E Canning (née Offer), of her father's Bristol pipe works probably also describes the covering of a muffle kiln. Although her observations regarding the kiln appear garbled, the detailed description of the way in which clay was spread onto paper suggests first hand experience.

'This central dome was formed by rubbing the clay into paper by the base of the palm of the hand. This method adopted by Offer's was probably unique as sheets of newspaper were laid on a table and the stiff paste of clay rubbed in. When the kiln was lit the pipes had to be protected from the smoke. Larger firms used saggars as they do in the china industry. With many such sheets a huge dome was constructed over the pipes and the fire then lit and baked hard. The dome was then like a monstrous inverted basin'. (Jackson 1990, 27)

Such vivid recollection of the dome suggests that the entire top of this kiln was demolished between firings as in the description of the Rouen Kiln. Gordon Pollock, who is still making pipes at Manchester, relates how his brother-in-law used paper in the solid fuel kiln, at the old works. The kiln had a spy hole through which the temperature of the kiln could be judged by the colour within. The spy hole was cut through the centre of a larger plug, which, when the correct temperature was judged to have been attained, was withdrawn to give access to a test piece. The test piece was then withdrawn to check that it was properly fired. This test piece was a pipe or group of pipes set up in a saggar with an opening in the side adjacent to the inspection plug. In order to protect the pipes in this saggar from impurities carried by the fire a piece of newspaper coated with a thin layer of clay was damped and stuck over the opening in the side of the saggar. Being very thin it was easily broken with the iron hook used to withdraw the test piece. By this time with the kiln at the required temperature the protection was no longer required as the fire mouth and chimney were closed up so that the passage of gasses and other material through the kiln ceased (Interview with author 1991).

In April 1969 Walker visited two pipe factories in the Westerwald area of Germany, those of Richard Simonis of Hilgert and Wilhelm Klauer Sohne of Baumbach (Walker 1977, 122). Of the former he records that his topmost saggar was covered either with an inverted square saggar with rounded corners or layers of paper covered with wet clay. This covering was simply called a 'deckel'. Of the latter;

'A number of Klauer's saggars were bound together with wire which was plastered with wet clay before being put in the kiln, and the capping here was layers of paper smeared with wet clay'.

Work at the tobacco pipe factory of Lothar Hein at Hilgert in the German Westerwald has been well documented in print and on film (Freckmann 1987; Kõgler 1987; Simons 1974). Saggars were covered in the same way with sheets of paper covered in a thin layer of clay. These paper and clay 'deckels' can be seen clearly in photographs published by Freckmann (Freckmann 1987, 97). Thin sheet fragments are recorded from numerous sites within the compass of this study. By the advent of the nineteenth century the use of this material had become universal. With the adoption of open flame kilns in the latter part of the century the material continued to be used, adapted to the changed requirements. There is little reliable information from the eighteenth century. The best assemblage is that from Waterford, Arundel Square, which includes eight fragments of white clay sheet formed over paper and several hundred fragments, in red clay with voiding from organic tempering, formed over a woven cloth. Many of these fragments have the impressions of a supporting framework of pipe stems. One substantial piece is composed of a number of layers of white clay over a woven fabric base forming a thicker mass one surface of which is dimpled with bowl mouth impressions (Figure 45c). Two other sites which produced pipes into the early years of the eighteenth century also provide evidence for thin sheet material formed over woven cloth. These are Chelmsford, Moulsham Street, 1660-1710 and St Albans, Holywell Hill, 1680-1730. In both cases a white clay with added organic matter has been used. Similar material is also recorded from Portsmouth, Oyster Street, 1660-1700. Two large fragments from this site are dimpled with the impressions of many pipe bowl mouths. Thin sheet material is also recorded from Rainford, Church Field, 1630-50 and Trowbridge, 10 Church Street, 1650-1700. In both cases large numbers of fragments were recovered by excavation. The two groups are similar in that the fragments display the typical wiped or scraped surface opposed to an enigmatic contact cast. Other excavated assemblages of similar period (Gloucester, Quay Street, 1670-1700; Southwark, Arcadia Buildings, 1675-1700 and Benthall Lane, 1660-90) do not include any material of this sort.

Because of the fragility of these sheets and their expendability after firing, conclusive evidence of size and shape are unlikely to be recovered. The majority of fragments are from the interior of a sheet, defined only by broken edges. The presence or otherwise of stem or bowl impressions is largely accidental (Figure 36k-n). The absence of these is not necessarily indicative. There are many instances of fragments with one surviving, simply trimmed, edge. From Gibraltar Cottage, Chatham, there are 19 fragments with two parallel edges 58-60mm apart. These strips were probably used in a similar manner to the straps described above. It is possible to distinguish the fine cast of letter-press on one surface of these fragments. The cause of this phenomenon which has been observed elsewhere, notably from Rainford Graveyard, may either be reduced absorbency of the inked characters or a mechanical deformation of the paper by the print face.

Thin sheet edges, treated more elaborately, have been recovered from Barnstaple, Alexandra Road, 1857-65; Gloucester, Black Dog Yard, 1849-70 and Bristol, Temple Way, early 19th century. From the first two sites a strip along the edge of the clayed paper, about 20mm wide, has been folded over so that clay meets clay to form an edge strip of double thickness (Figure 37s & t). At Bristol, Temple Way, a roll of clay was added to the edge of the sheet and smoothed in towards the interior with fingers or thumbs. Into this thickened edge pipe stems have been embedded as a supporting framework (Figure 37u & v; Figure 38a & b). It is not clear exactly how this was achieved as there appears to be no pressure deformation on the obverse, only displacement of clay by the stem ends. The paper would of course mask this to a certain extent.

Thin sheets are also known that have been folded or rolled. The site at Athelney, Somerset, 1831-73, has yielded a number of these thin sheet rolls several of which have transverse impressions of pipe stems (Figure 36o & p). There is also a rolled thin sheet from Boston, Rosegarth Street, mid 19th century.

Clearly thin sheets evolved early on. It is not clear on what base the seventeenth century material was formed. In the eighteenth century cloth was used and in the nineteenth predominantly paper. The documented uses of this material are as covering either to stabilise or protect the pipes. At Rouen in the eighteenth century the material was used to cover the top of the muffle; at Manchester in the eighteen eighties each loaded stage of the tiered muffle was covered in this way; also at Manchester and at three sites in the German Westerwald in the present century this material was used to cover saggars.

Slag, stem and thin sheet laminates

There is clearly an overlap in usage between this group of composites and the material discussed above. All of the slag laminates probably served the same function. Some of the thin sheet material above has probably become detached from a slag laminate. The clean sides of the laminates were clearly protected from the fire path. Those fragments which display pendulous slag runs indicate the orientation of the material in the kiln. This invariably shows that the slagged surface was uppermost. The bulk of the material is positively associated with peripheral shelf muffle kilns. Of the 28 sites to produce slag laminates, 10 also include muffle fragments, 2 include muffle type furniture and a further 4 fall between 1820 and 1850, a period from which no other type of pipe kiln is known. The French description from the late eighteenth century of a dome over the top of the muffle formed of thick paper covered with a layer of moistened pipe earth over a framework of wastered pipe stems suggests an acceptable explanation for this material.

'The chamber or the pot having thus been filled with pipes, a dome over is made 12 or 15 inches high, with sheets of thick paper covered in a layer of earth 4 to 6 'lignes' thick, this we call 'dorure', (gilding).

This 'dorure' is powdered pipe earth, moistened with sufficient quantity of water, to make it workable so that the artisan can apply and spread it by hand on the sheets of paper which are placed on a ring of pipes already baked but waste, which are born by one end on the column of pipes to be baked, and by the other on the sides of the octagon which forms the chamber. Thus these baked pipes are like a sort of framework which supports the gilded paper.' (Duhamel du Monceau 1771)

There are accounts of similar mixtures of materials from later English sources. Daniell writing of Flanagan's kiln at Leicester in the early years of the present century records.

'When the pipes had been stacked in the kiln it was essential that they should remain in position and that the heat should be retained at a very high temperature. In order to achieve this the pipes were first covered with several thicknesses of old play bills, specially obtained for this purpose from the Pavilion and Palace theatres. The paper pyramid thus formed was covered completely by a thick paste made from a mixture of clay and horse manure' (Daniell 1964, 61).

Perhaps when Rees writes of the damper plate used over an aperture in the crown of the fire brick lining, which he describes as 'a mixture of horse-dung, sand, and pipe-clay, well worked together, and spread in thin layers upon coarse brown paper', he is confusing this with material used for the muffle cover. A similar mix of materials is described by Jack Tennant in his recollections of the Tweedmouth factory. He is not specific about the use of this mixture which might equally have been used as clamming for the wicket.

'Clay to seal the kiln was made from mud and dust swept off the street from between the cobbles complete with manure to give it bind' (Roberts 1988, 94).

The use of paper covered with a thin layer of pipe clay to form a dome over the pipes in the Offer kiln at Bristol has already been mentioned. This technology is not confined to the production of tobacco pipes. Walker records the use of similar material used in eighteenth century metallurgy.

'A highly unusual use of the term (muffle), though effectively the item described acted as a saggar, is noted in an account of gold and silver refining in the C18 (G. Smith 1750 ed: 7 and n.). A mixture of one part clay, one part sand, and two parts horse-manure - though sometimes pipeclay alone was used - was made into a square sheet "to the thickness of a crown piece", then bent into a U-shaped arch and dried. This was then placed over a type of crucible, called by Smith a coppel, and the whole was then covered with live coals to which more coal was added until cupel and muffle were red hot. The cupel was then used for assaying gold or silver with lead'. (Walker 1977, 997)

The antiquity of such mixtures of simply available materials is proven by the Piccolpasso description relating to maiolica production in northern Italy at the end of the sixteenth century.

'First the kiln is well swept out, the ashes remaining from the first fire being removed from underneath, and the kiln being cleaned of potsherds and other dirt; then lute should be taken made in the following way. "Sciabione" should be taken and this should be very well softened, then some ashes put into it, and asses' dung, and iron scales or the dust that gathers on the stocks of anvils; these materials well mixed together are put into a shallow bowl or pan; then it is brought under the kiln and thus roughly spread with the hand on the arches in such a way that it lies there to a depth of a finger, then you come out from underneath and in the name of Jesus Christ the setting of the kiln is begun'. (Lightbown & Caiger-Smith 1980, II, 106)

Of these six references all include an element of clay (the dust around Tennant's factory would consist mainly of clay and Piccolpasso's sciabione is a type of clay). In three of them paper is mentioned; in five manure; in two sand and in one ashes and iron scales. Pipe clay will vitrify between 940oC and 1100oC but it does not form a slag. A flux is required to bring about a melt at these temperatures. Of the materials listed manure, ashes and iron all act in this way to variable degrees. Fresh horse manure comprises approximately 70% water, 25% organic matter, with traces of nitrogen, phosphate and potash (Information supplied by Elm Farm Research Centre, Newbury, Berkshire). The effect this has when added to clay is simply to bulk it out, resulting in a voided fabric as the moisture is driven off and the organic matter is burned away. Small traces of glaze are sometimes formed in the voids by the flux action of ash on silica from the clay. Fresh horse manure could also be used to bind non-plastic materials as inferred by the account from Tweedmouth. This could only be of practical use if the non-plastic materials fused before the manure lost its binding capability. None of the references quoted above, specific to clay pipe production, include sufficient fusible material for this to have been a practical proposition. It is more likely that if the binding property of horse manure was in any way important it would have been in addition to a quantity of clay to give the mixture lasting stability.

When these mixtures were reproduced in the workshop it was found that without the ash element, listed only by Piccolpasso, there was no slag formation (Appendix 9). The simple paper and pipe clay mix quoted by Duhamel du Monceau produced material similar to the thin sheets described above. Paper, clay and horse manure as described from Leicester produced less dense sheet voided where the organic matter burnt out. Sand added to the mixture remained unchanged after firing, being still apparent as mineral inclusion in the fabric. Wood or coal ash added to any of the mixtures in varying proportions invariably resulted in a glassy or slaggy product. Coal ash, which contains varying quantities of earthy impurities, will form a slag, whilst wood ash produces a more vitreous material.

Although coal ash alone will form a slag similar to that encountered on pipe kiln sites, without some binding agent it could not be used to form sheets in the manner described above. Manure will bind at low temperatures but it fails in this capacity before the melt can take over. Clay added to the mix binds throughout the temperature rise finally combining with the ash to form a slag. The chemistry of slag formation is extremely complicated being not only dependant upon the relationship between the main constituents of the ash (silica and alumina), but also lesser flux elements such as lime, iron oxide and magnesium, as well as the balance of oxidising to reducing gasses in the kiln atmosphere (Lowry 1945, 485-551). The ashes of most British coals have melting points between 1050oC and 1350oC (Spiers 1962, 304).
Both the widespread occurrence and the consistency of these materials show clearly that slag formation over a bed of pipe stems, or thin sheets, represents some quite deliberate practice associated with the firing of tobacco pipe kilns in the nineteenth century. Table 9 shows the incidence of slag laminates in chronological sequence.

Type 1
slag over a layer of pipe stems, has been recorded from 19 sites covering the period 1820-1919.
Type 2
slag over a layer of pipe stems over a layer of thin clay sheet, has been recorded from 6 sites covering the period 1860-1900. In four of these the thin clay sheet has voids indicating organic inclusions.
Type 3
slag over a thin clay sheet, has been recorded from 12 sites covering the period 1820-1925.
In four of these the thin clay sheet has voids indicating organic inclusions. Types 1 and 3 occur together at Liverpool 2, Southampton 1 and Chatham 1. Types 1 and 2 occur together at Sunderland 1, Newark 2 and Leicester 2. Types 1, 2 and 3 occur together at Newark 1. Types 2 and 3 occur together at Rainford 10. It seems likely that these figures are effected by an element of chance; that in some parts of a cover composed of sheet, stem and slag, the slag was in contact with the stems whilst in other parts it was in contact with the clay sheets. A lack of adhesion between stems and the clay sheet would allow them to separate. In consequence, such a dome, once removed, could yield slag laminates of Types 1-3 together with thin clay sheets free from slag incursion.


The purpose of this group of objects is not known. Prior to this survey few had been recovered and none described. It was not until the scattered data was collected that they were seen as part of some widely accepted practice. Evolutionary forms occur from Aylesbury prior to 1710 and St Albans by 1730. Rack ends from the latter part of the nineteenth and early twentieth centuries from Gloucester, Ipswich, Belfast and Boston are of remarkably similar design.


© Internet Archaeology URL: Internet Archaeology
Last updated: Wed May 22 1996