In the course of this study a number of questions and tasks have been presented, the pursuit of which would further enhance the state of knowledge. Earlier in this chapter, the desirability of excavation on the Watkinson site at Market Rasen was touched upon. The factory was demolished in the early 1970s and the site is free from any encumbrance. The cobbled yard surface visible in Hallgarth's photograph (Hallgarth 1969, 33), is unchanged today. The four naive sketches, reproduced in Appendix 2, show clearly the positions of internal furnishings. This presents a unique opportunity to relate excavated ground plans to documented usage with the added bonus of identifying the "footprint" of a tipping furnace. Collection of glazed stem tips from the site would enable chemical comparison with the materials recorded for this process. It is possible that artefactual evidence for the 'confined space' might be recovered allowing comparison with that from the Bristol sites.
Appendix 7 describes trial excavations on the site of a late seventeenth century pipe works at Pipe Aston in North Herefordshire. This is an example of cottage industry in a remote rural setting, possibly in conjunction with farming. Preliminary findings indicate termination of pipe production in the late seventeenth century with occupation on a small platform in the slope above the workshop continuing into the late eighteenth century. Although some pipes from the mid seventeenth century have been recovered from the site it is not certain that these represent an earlier phase of production. The site, which reverted to pasture once abandoned, offers the opportunity to examine a complete working complex of domestic buildings, workshops and kilns. Although stone from the buildings is likely to have been carried off for use in a nearby cottage of apparently early nineteenth century construction, the sloping site offers the possibility of kilns cut into the bank which might survive to a substantial height. A kiln site nearby at Birtly Farm yielded trapeziform bats which raise questions as to their function. Muffle material from the contemporary sites at 11 Benthall Lane and Gloucester, Quay Street, suggest possible variations in kiln design in this region. The opportunity to recover a more complete assemblage could lead not only to a fuller understanding of the process on this particular site but also aid interpretation of existing groups of material from elsewhere.
The muffle assemblage from Bristol, Temple Back, discussed in Chapter 5, appears to be substantially complete. It dates from the early part of the nineteenth century and represents the pinnacle of development. Although the condition of the pieces is generally stable, many were removed from a wall and are coated with lime mortar. Permanent storage must be found for this unique object; cleaning and some conservation is required; sorting together with spatial organisation would result in a complete understanding of the total form; the end product of over two hundred years of development. Once this has been done, drawings and photographs taken and a model made, the muffle, which weighs over 700kg, should be permanently accommodated in a secure store.
The surviving remains of the kiln in Pipe Passage, Lewes, should be secured for the future and are of sufficient note to warrant permanent display. The upper part of the arch is exposed to the elements, its body is filled with rubbish whilst the lower parts and the stoking pit are below ground level. The only known photograph of the kiln when it was excavated in 1956, reproduced in Chapter 9, shows the complete niche with ash pit and some internal structure. At the least this structure should be scheduled to protect it. Re-excavation, housing in situ and permanent display should be the ultimate goal.
Some questions posed by the artefactual evidence can only be settled by practical tests. To this end it would be useful to build a replica of a small muffle kiln using the same materials and techniques as those employed in the past. Repeated firings would provide data relating to the life span of the muffles and highlight any problems intrinsic to the design. Measurements of flue temperature around the muffle could either disprove or add weight to the interpretation of stem slag laminate as muffle covering material. The use of different fuels could be assessed and firing times could be established. These aspects have not been considered in this study because they are not considered to be central to the technological development of the kilns. Kiln management is an art rather than a science, often different fuels with differing properties were and are used at different stages of the firing. Although coal is known to have been used from the early seventeenth century and coal has been recovered from a number of sites, this does not rule out the use of other fuels such as brushwood or other fast burning long flamed material at the later stages of the firing. At Pipe Aston, where coal has been recorded amongst the waste deposits, there are also fragments of stone with pendulous 'fingers' of 'celadon' coloured slag which suggest wood burning. Referring to down-draught intermittent kilns in the present century, Rosenthal writes:
'At the commencement of the firing, short-flame coal, or even coke, may be used to advantage. At the next stage in the firing, it may be mixed with a long-flame (fat) coal. During the last period of the firing, however, the use of long-flame coal is essential. The long flames produced carry the heat to the middle of the oven which would be extremely difficult to heat by radiation or convection' (Rosenthal 1949, 105).
Reference to methods still used in non-industrialised regions illustrate the types of practice that might be expected from the pipe makers of earlier periods. Olsen describes mixed fuels used at different stages of the firing, by Spanish country potters. Of a small 'umbrella' kiln, five feet in diameter, at Moveros, he writes:
'The kiln is fired with heather and any available brush. The firing is started at the firebox mouth, then after 2 hours, brush is pushed farther into the firebox and stoked faster for about 6 hours. When flames come out of the top of the kiln, the firing is done' (Olsen 1983, 139).
Of a medieval type kiln at Mirovet:
'Long pieces of wood - three to four armfuls at a time - are stoked at about 4 to 5 minute intervals, until the sub-chamber glazed target pot turns darker than the pots next to it. Then the fuel is changed to wood chips. Six shovelfuls are stoked at a time, slowly building up to 30 shovelfuls at the end of the firing. Supplementary fuels are bundles of wood and grass, slash, slats and dehydrated olive pits. Sufficient time between stokings must be allowed for oxidation, determined by the time it takes for the fuel to burn down and clear the atmosphere. The firing lasts until the target pots in the dome flues are "done". Firing times with this type of kiln vary from potter to potter' (ibid, 146-7).
Of a bottle kiln at Breda (Spain):
'The firing is done with chestnut slats 3 feet long, and used about 1800 kilos of wood. The firing begins with a slow preheating period of about 3 to 4 hours. Stoking is then increased, building to a 4-minute-interval stoking rate at the end of 7 hours. The stoking interval is determined by flames coming out of the chimney cone holes; when they suck you stoke. The total earthenware firing takes about 10 hours' (ibid, 153).
Of a small traditional Muslim kiln at New Delhi, India, he writes:
'The firebox requires about 5 hours to warm up; then the actual firing takes about 6 to 7 hours. To warm the kiln, small amounts of wood are burned in the firebox and the coals shoved down into the pit bottom. Slow build up of coals in the pit allows the chamber temperature to slowly increase to 260oC, thus ensuring the safety of the wares. Large hardwood logs by stoking standards (from 2" to 6" diameter) are used. These give a relatively short flame and long coal life. The type of wood (hard or soft, branch or log or kindling) has an important role in controlling the temperature rise, the actual flames in the chamber and the chambers atmosphere. Hardwoods, for example, give short flames, intense local heat and good coals; softwoods explode quickly, giving an extremely long flame and moving heat quickly into the chamber' (ibid, 140-2).
Firing times generally cover the period from commencement of stoking until the desired temperature is reached and fires are extinguished. This often includes a soaking period at the end of the cycle, when fires are maintained but not increased, to ensure that all parts of the chamber reach the desired temperature. In the literature are several reports concerning firing duration. These range from 5 hours (Benham 1906), to 16 hours (Anon 1887, 191). Watkinson records that the quality of the coal determined the duration of the firing (Appendix 2). To this must be added the size and design of the kiln; a small kiln could obviously be raised to the required temperature more quickly than a large one. The physical limitations involved in firing raw clay are first the speed with which moisture can be safely released and secondly passage through quartz inversion. This takes place at 573oC when alpha quartz changes to beta quartz with a sudden increase in volume. Passing this point too quickly can result in cracking in a clay with a high proportion of free silica of small particle size (Cardew 1969, 38). Obviously, stoving the pipes to drive off most of the moisture held in mechanical combination is beneficial in reducing the time required for the firing. Leach states that, 'in kilns of 12 to 24 inches internal diameter three to five hours should be sufficient for biscuit firing small light ware'. The temperature required would vary slightly from one clay to another. The pipemaker's aim would have been the quality of his pipes rather than a specific temperature, when engaged in the process of firing his kiln. Although the pipe should remain porous for the best smoking qualities it should be hard enough for durability. Tests made on stem fragments by re-firing them in an electric kiln show that their porosity generally decreases markedly above 1000oC. The exact point of vitrification is dependant upon a number of factors; grain size, presence of oxides such as the alkalies which form relatively fusible silicates or alumino-silicates, and duration of firing (Scott 1929, 56-7). 'It has long been recognised that the duration of firing has a pronounced effect on the vitrification of a clay. Prolonged heating at one temperature may give the same result as shorter heating at a higher temperature. From the porosity determinations detailed below, there is evidence that vitrification in the case of certain ball clays begins at 940oC, and is complete at approximately 1100oC' (ibid, 57).
Clearly, if the optimum temperature were attained for every kiln filled with pipes, a considerable variation might be expected. In summary, firing times might vary between five and sixteen hours depending predominantly on kiln size, and temperatures between 850oC and 1000oC dependent on clay composition. On the continent where additional different clay types were used even higher temperatures were attained albeit with the sacrifice of porosity.
Although the greater part of mainland European evidence is for open flame kilns utilising saggars for firing tobacco pipes, there is evidence for muffle structures built on the same principles as the British norm. Duhamel du Monceau describes a small kiln from Rouen, in which the muffle is constructed from tiles. The recent discovery at Breda, in the Netherlands (Appendix 10), of a mid seventeenth century pipe kiln, yielded muffle fragments reinforced with pipe stems just as in British examples. Clearly, emigrating British pipemakers took with them the methods and technology with which they were familiar. Records of the pipemakers' guild of Gouda, in the Netherlands, dated 1760, detail a complaint by the potters' guild against the pipemakers. This highlights control by the potters of the firing process, pipemakers packed their pipes in saggars to be fired by the potters. The potters were complaining of the presumably new practice of packing the pipes in ground clay making the saggars too heavy to be handled by two men (Walker 1977, 101). This practice of filling the saggars with 'gruys' by pouring through the 'trompette' filling the space around the long stemmed pipes to prevent warping, is described by Duhamel du Monceau in his treatise of 1771. Possibly this development of long stemmed pipes, at an earlier date than those of the British industry, brought about a change of direction in kiln design, abandoning the muffle kiln in favour of open flame kilns with saggars specifically designed for pipe production. Paradoxically, in Britain the muffle kiln was developed to cope with longer stems, only to be superseded by open flame types with saggars when stem lengths were drastically reduced. The continental clay pipe industry would appear to have been less inward looking than in Britain adopting pottery type kilns at a much earlier period and in the case of Gouda, at least, co-operating with potters to have their pipes fired. Clearly further work on continental practice is required. This might begin with a synthesis of published material from contemporary accounts to retrospective studies, followed by a review of artefactual evidence in both public and private collections.
© Internet Archaeology URL:
Internet Archaeology
Last updated: Wed Oct 9 1996