The problem of the highly absorbent fired ball clay sticking to the smoker's lips must have been apparent from the very beginning of the industry. A common solution used well into the present century was to dip the mouth piece into beer or other liquid before putting the pipe to the lips (Harley 1963, 16). Various wax applications were used and are well documented (Walker 1977, 148-9). In Great Britain, the application of low temperature ceramic glazes to the tips of tobacco pipe stems appears, from the archaeological record, to have been introduced early in the nineteenth century. The colours of these glazed mouthpieces commonly range from green to yellow; blue has been noted from Norway (Walker 1977, 148). Walker also notes glazed mouthpieces from a rubbish-dump at Rosewell, Virginia, datable to between 1763 and 1772 probably the latter date (ibid, 148)
It is unlikely that any maker glazed the stems of his entire output, rather that the practice was reserved for particular, possibly higher status products. Because little attention has been paid to recording details of mouthpiece treatment there is insufficient evidence for statistical study. Kiln sites provide some insight to the problem though even here no study has been made of the relative frequency of glazed to unglazed mouthpieces. Although in the course of this study glazed and unglazed mouthpieces have been noted when encountered, no attempt was made to seek them out; the priority being to record structural material and waste. Kiln products were only examined with a view to providing dates for the assemblages. Many of the groups studied include only a sample of bowl types with stem fragments poorly represented if at all. The catalogue does not attempt an exhaustive record of pipes or stems.
Glazed and unglazed mouthpieces are recorded from eleven sites.
Barnstaple, BA2, Alexandra Rd G & un-G 1857-65 Boston, BO1, Rosegarth Street G & un-G 1861 Bristol, BRI5, Waverly Street Red Wax & un-G 1800-15 Bristol, BRI8, Mead Street G 1850-65 Bristol, BRI10, Temple Way G & un-G 1800-20 Durham, DU1, Back Silver Street G 1800-50 Gloucester, GL3, Black Dog Yard G & un-G 1849-70 Hull, HU2, Lowgate G 1825-50 Lincoln, LIN2, Cornhill G 1840-60 Manchester, MA1, Hurst Court G 1800-50 Truro, TRU1, Pydar Street G 1800-50
The earliest of these groups is that from Bristol, Temple Way, which is not later than 1820. Given the American data it seems likely that the practice of glazing stem tips was introduced in the later half of the eighteenth century. If this point could be reliably established it would provide a useful terminus post quem for products of a period which are often difficult to isolate on typological criteria.
There are few accounts of the pipemaking process which make any reference to this aspect of the technology. The following account of McLardy's Manchester Works is from Tobacco Trade Revue of 1887.'In the same building with these furnaces is a peculiar stove arrangement for vitrifying the coloured enamel on the ends of certain of the pipes. Those we saw being operated upon had the ends coated with what appeared to be red paint, but which, when burnt, came out a beautiful green enamelling'. (Anon 1887, July 1, 191)
Although published in 1936 the following description from The Bolsolver Robinsons purports to date from c. 1777. 'The final operation, necessary to prevent the pipe sticking to the smoker's lips, was the forming of the mouthpiece. This was done by pouring a solution of red lead over the ends of the stems and burning them in a clear red fire. A bright green glaze resulted and the completed pipes were then packed in boxes ready for sale.' (Robinson & Spence 1937, 196) A version of this same account has recently been discovered in the Robinson company archives. It dates from the late nineteenth century and is probably the source of the published account which it closely resembles. The document is reproduced with a discussion of its date and relevance in Appendix 3.
Details of the tipping process as practiced in Lincolnshire, published in 1969 by Hallgarth, are paraphrased from an unidentified source (Hallgarth 1969, 33-4). The original hand-written account has fortunately been traced in the course of this study. This document is in the form of auto-biographical anecdotes written by George Spencer Watkinson junior in 1918, 1928 and 1944. He describes working practices at his father's pipe-factory where he began working in 1864. The factory appears to have closed around 1895. A transcript of the entire document, with discussion is included here (see Appendix 3).
After some hours cooling, contents of kiln withdrawn, there but remaining the glazing of tip stems, that the users lip would not be injured - this was called tipping by liquid red lead Potash [permanganate of] or verdigris [;for different colour of glaze, then by] inserting the tips, over a coke fire in a confined space at white [heat] withdrawn immediately when clearly fused. Then placed with the tip on a raised wire [to cool].Two of these accounts describe the colour of the fire; Robinson, a clear red fire; Watkinson, white heat. Prior to the invention of the pyrometric cone, by Herman Seger in the late nineteenth century, kiln temperature could only be gauged by observing colour within the chamber or withdrawal of a test piece. Seger first published his results in 1886 (Rhodes 1969, 173). Throughout the greater part of the period under consideration colour observation and test withdrawal were the only means available to judge kiln temperature. Continual exposure of the eyes to high temperatures is not without risk. It is to this risk that John Frederick Bryant, the Bristol pipemaker, alludes in his verse entitled 'On a Piece of Unwrought Pipe-Clay', published in 1787. 'And in the furnace thy last change I speed: Ah! then how eager do I urge the flame, How anxious watch thee mid that glowing fire, That threats my eye-balls with extinction dire!' A dull red glow first appears in the kiln at a temperature of 600o-650o C (Cardew 1969, 209). Leach describes kiln colour as dull cherry to light cherry, 920o-1060o C; dark orange to pale orange, 1060o-1180o C; yellowish white, 1180o-1280o C (Leach 1940, 260). Although the terminology is too subjective to be certain of common ground, Robinson's clear red fire is more compatible with a soft lead based glaze than Watkinson's white heat.
There is a paucity of detail regarding these tipping furnaces. The Manchester account refers simply to a peculiar stove arrangement. Robinson offers no description at all while Watkinson records a confined space over a coke fire. In isolation these accounts do not offer sufficient detail for a visualisation or reconstruction. Until there is a body of archaeological evidence which can be linked with these accounts it is not certain that a standardised form was used. It is possible that different solutions were adopted in different localities. Artefactual remains of tipping furnaces have been recovered from five sites . Four of these are from Bristol, the other from Barnstaple.
The Bristol evidence is consistent with a uniform type of tipping muffle in use from as early as 1820 to at least the middle of the century. The single small fragment from Barnstaple could have come from a similar object. All have been used to melt green glaze. The type example illustrated is from Bristol, Temple Way (Jackson et al 1991). This tipping muffle is hand built from white clay with self coloured grog and some organic matter. It takes the form of an open ended vessel tapering to a spigot base reminiscent of an amphora. The spigot is reinforced with pipe stems. The open end has a thickened rim to a depth of c. 50mm. The outer surface is considerably damaged by spalling and for the most part covered with slaggy glaze. That it was aligned horizontally in the furnace is clearly shown by the crystalised slag drips formed on one more heavily slagged side. This "lower" edge is further attested on the inner surface by green glaze streaked from the depths of the interior to the outer rim. This streak occupies approximately one sixth of the circumference above which the interior is glaze free excepting its inner extremity. The rim is slag free, indicating the masking effect of a support. The spigot is partially slag free and generally less slagged than the body, indicating a further support. Slagging on the body decreases towards the "upper" surface. The muffle has been breached through the "lower" wall apparently by glaze erosion of a crack. Figure 97a illustrates in section the type of placing and support indicated by the slagging evidence. Figure 97b illustrates how a number of such muffles might appear set above a flue similar to those excavated at Boston (Wells 1970, 22; see Figure 98).
The site at Temple Way also yielded two body fragments from another tipping muffle. From Mead Street there are two spigot ends, twelve body fragments and four slag masked rim shards. This assemblage comes from three or more muffles. The site at Bath Road produced one slag masked rim fragment. The fragment from Barnstaple is from the body, displaying green glazed inner and slagged outer surfaces. It would appear from the testimony of slag deposits that these tipping muffles were supported at both ends bridging a space occupied by the fire. The spalling damage is indicative of a rapid temperature rise to the point where work could begin. This would limit the life of the muffle necessitating periodic replacement. The form of these tipping muffles is consistent with this scenario. It would be a simple task to withdraw such a muffle from the supporting structure and insert a new one sealing any surrounding gap with raw clay wadding. No demolition or rebuilding would be required.
A fourth Bristol site has produced possible evidence for the structure of the furnace into which tipping muffles were set. From Temple Back there are a number of fragments from slabs or bats variously coated with slag and green glaze. Four joining fragments form one corner of a rectangular slab, 38mm thick, with green glaze on one surface and slag on the other. A 50mm wide strip of the glazed surface and 100mm of the slagged surface, along both edges, is slag and glaze free where it has been masked by a covering structure. A further non joining fragment with similar coatings and masked areas probably comes from the same slab. The fabric, similar to that described above for the tipping muffle, suggests manufacture at the pipe works. From the same site a commercial fire brick slab 50mm thick, tongued and grooved to interlock, has green glaze on one surface and mortar adhering to the other. The glazed surface has a masked strip 60mm wide on one edge. A second fragment, in commercial fire brick fabric, has green glaze on one side whilst the obverse remains clean. The glazed side has a masked strip 80mm wide. If these slabs and the tipping muffles described above represent two features from similar structures then the slabs probably formed a horizontal bed in front of the open ends of the muffles. Glaze drips from stems as they were withdrawn would have accumulated on the upper surface of the slabs whilst the lower surface, where in direct contact with the fire, would have suffered slag build up. It is of course possible that the slabs and muffles are from dissimilar structures in which case the slabs might represent parts of box shaped muffles.
On the site excavated at Rosegarth Street, Boston, the base of a small furnace occupied a position adjacent to the kiln (Wells 1970, Figure 1). This consisted of two parallel flues, one metre apart, the space between being filled with compressed ash. The flues were built of brick with iron fire bars. This facility was tentatively identified as a drying chamber (ibid, 22). Amongst the recently discovered Watkinson papers are four small pencil and crayon sketches, each measuring 135mm by 85mm, depicting rooms in the Market Rasen pipe factory. These are entitled, 'Clay soaking trough and block', 'Rolling stems, wiring - moulding', 'Trimming - drying' and 'Kiln - tipping'. The last of these sketches shows the kiln with the tipping furnace built against one side in front of which a worker is dipping stem ends in the glaze. The sketch is not objectively accurate, the kiln being shown in section. The depiction of the tipping furnace is ambiguous. What appears to be freshly glazed pipes cooling on the raised wire may equally represent pipes being fused with their tips inserted through a horizontal slot into the muffle or fire. The size and position of this structure is remarkably similar to that from Boston. In the light of this new evidence it is possible that the plan originally published by Wells (Figure 98) records the only tipping furnace excavated to date (ibid, Figure 1). Eleven stem end fragments with pale yellow transparent glaze were recovered from the site. One fragment, from a layer postdating the construction of the furnace, and three fragments, from a pit sealed by the furnace, consist of two glazed stems fused together. These are clear evidence that glaze tipping was carried on at this site and that if the furnace is in fact a tipping furnace then it replaced an earlier tipping furnace on the site.
The three accounts of tipping quoted above provide some details of the materials used in preparation of the glaze. The Manchester account is a clear observation by an uninformed reporter. The material had the appearance of red paint which, when burnt, produced a beautiful green enamelling. The Robinson extract records the use of a solution of red lead which again produced a green glaze after burning in a clear red fire. The third reference, from the hand of a known pipemaker, George Watkinson, might be expected to be the most accurate. Here the materials listed are liquid red lead, potash (permanganate of) and verdigris, the latter two ingredients appearing as alternatives to produce different colours. Common to all three descriptions is red lead; lead peroxide, Pb3O4. A simple suspension of red lead in water applied to the stem tips and fired to 750oC under oxidising conditions results in a bright yellow transparent glaze. The Watkinson papers record two additional materials, potash (permanganate of) and verdigris. Potash is an inexact term originally applied to Potassium Carbonate, K2CO3, now generally applied to all Potassium salts particularly Potassium Oxide, K2O. The carbonate, when heated, releases Carbon Dioxide, CO2 leaving Potassium Oxide K2O. This material can also be used in the composition of a low temperature glaze. Potassium Permanganate, KMnO4, is an ionic compound of Potassium and Manganese Oxide. Manganese Oxide, added in quantities of two or three percent, gives a purple brown colour to a lead glaze. This writer has not seen any glazed mouth pieces which suggest use of this oxide. It would be interesting to excavate on the site of the Watkinson factory to obtain examples which could corroborate the account. Excavation could also yield a ground plan of the tipping furnace, shown in one of the sketches adjacent to the kiln.
Verdigris, basic copper acetate, Cu(CH3.COO)2+Cu(OH)2+H2O releases its volatile elements in the kiln leaving Copper Oxide which when added to Red Lead produces a lively green transparent glaze. Copper Oxide, CuO, is effective in quantities as small as one percent of the glaze mix, above five percent it produces a dark metallic black. This oxide also has the effect of making the glaze more fluid.
An understanding of the basics of glaze chemistry is essential to evaluate these accounts. A present day pottery glaze has three distinct ingredients. A glass former, Silica (SiO2), which has a melting point in the region of 1730o C. A flux, of which there are several, which, in combination with silica reduces the melting point of the mixture to a practical level. The third ingredient, Aluminium Oxide (Al2O3) acts as a stiffener reducing the speed of the melt and causing the glaze to remain viscous for one or two hundred degrees of temperature rise. A simple mix of silica with a flux tends to melt quickly and to flow rapidly to the point where much of the glaze will run off, spoiling both work and kiln furniture. The most powerful of the fluxes is Lead Oxide (PbO or Pb3O4). The simple application of the flux to the clay surface forms a glaze by use of free silica in the clay body. Free silica is in mechanical combination as opposed to that in chemical combination as part of the clay mineral Kaolin, Al2.2SiO2.2H2O. In practice all clays include quantities of free silica. Writing specifically on ball clays, with which pipemakers were predominantly involved, Scott states:
Ball clays never have the theoretical composition of kaolinite and invariably contain such oxides as silica (in excess of the amount required for clay substance), iron oxide, lime, magnesia alkalies, titania and so forth. The excess silica is largely present as quartz but it may also occur as colloidal silica and silicates. (Scott 1929, 51)The track of glaze along the lower surface of the Bristol tipping muffle is consistent with the very fluid melt which results from direct application of lead oxide to the stem.