¹ Arch. Journ. cxiv (1957), 10–27.Google Scholar

² Pottery Quarterly v (1958), 72–5Google Scholar.

³ Archaeometry iv (1961), 4–30Google Scholar; and Archaeometry v (1962), 80–6Google Scholar.

⁴ Mayes, P., ‘Medieval Pottery Teaching Unit’, Leeds, 1968 and forthcoming.Google Scholar

⁵ Bryant, G. F., Journ. Scunthorpe Mus. Soc. iii (1970), pt. 1, 1–16Google Scholar; Workers Educational Association (Barton-on-Humber Branch) i (1971), 1–20Google Scholar; and ‘Experimental Romano-British kiln firings’ in Detsicas, A. (ed.), Current Research in Romano-British Coarse Pottery (C.B.A. Research Report x) (1973), 149–60.Google Scholar

⁶ Ceramic Review viii (1971), 10–11.Google Scholar

⁷ Care has been taken to distinguish between pottery kilns and domestic ovens, the remains of which can be very similar during this period.

⁸ Woods, P. J., Excavations at Hardingstone, Northants., 1967-8 (Northamptonshire County Council, 1969 and forthcoming).Google Scholar

⁹ Publication forthcoming.

¹⁰ Corder, P., A Romano-British Pottery Kiln on the Lincoln Racecourse (University of Nottingham, 1950), 8–9 and fig. 2.Google Scholar

¹¹ Bulltn. Northants. Fed. Arch. Socs. (Northamptonshire Archaeology) viii (1973), 7.Google Scholar

¹² R. E. M. and T. V. Wheeler, Verulamium, A Belgic and Two Roman Cities (1936), 178-80 and pl. LVI, A.

¹³ Ibid., 180. Cf. also C. F. C. Hawkes and M. R. Hull, Camulodunum (1947), 347 (‘Native brick like that of Prae Wood was very rare and nowhere demonstrably structural. No stratified pieces were of Period I, and of the half-dozen or so assignable to Periods III-IV all were probably from ovens or kilns (one from pit Fio was vitrified along one edge); the thickness was from ¾ to 2½ in.’).

¹⁴ SP 8763 6786, Kiln IV. Information from Mr. P. J. Foster who has kindly allowed me to illustrate this pedestal in FIG. 5, D.

¹⁵ Dryden, H., Ass. Archit. Soc. Reports xviii (1885), 61Google Scholar and pl. v, 1. The Hunsbury pedestals were uncovered during ironstone quarrying on the north incline of Hunsbury Hill in 1875. Two complete props were found, bodi with a single, lateral perforation through the stem (average diameter 1-1¼ in. (3 cm)), and there were fragments from at least six more. Their height was 11¼ in. (29 cm) and their flattened ends measured 8¾ in. (22 cm) by 6 in. (15 cm). They were arranged in two circles each about 4 ft. (1·2 m) in diameter. The area between the props was filled with charcoal ‘and a space outside the circle also’. Associated with the props were parts of one or more fire-bars and a coin of Claudius II (268-70).

¹⁶ Information from Mr. G. B. Dannell and Dr. J. P. Wild.

¹⁷ D. P. S. Peacock, ‘The black-burnished pottery industry in Dorset’, in A. Detsicas (ed.), op. cit. (note 5), 64-5.

¹⁸ Cf. G. F. Bryant, in A. Detsicas (ed.), op. cit. (note 5), 149-60 and pl. VIII.

¹⁹ Part of the rear wall of the kiln (i.e. that farthest away from the stoking area) was raised by two or three courses.

²⁰ Samples of clay from Hardingstone were heated in a small gas-fired muffle equipped with a pyrometer and withdrawn at 50°C intervals between 500°C and 1000°C. The greatest correspondence of fabrics occurred within the range 750-800°C. For a preliminary account of these experiments see Woods op. cit. (note 8), 37 (Appendix D).

²¹ Mayes op. cit. (note 4), introduction.

²² Bryant (1971 (op. cit., note 5)), 2, and in A. Detsicas, (ed.), op. cit. (note 5), 150.

²³ Arch. Journ. cxiv (1957), 14Google Scholar, fig. 3. The Canterbury reconstruction shows a temporary dome built up on a framework of wooden rods. Bryant comments (Bryant (1971), 2, footnote), ‘None of the practising potters consulted by the author would care to risk their wares, or expect to produce grey ware, in a kiln built as the Canterbury reconstruction’, and this has been the reaction of similarly experienced potters consulted by the writer. But this does not mean that wattles were not used in the construction of Late Belgic and Romano-British pottery kilns. Indeed, the excavations at Hardingstone (Sites I and II) and Wakerley have produced clear evidence to the contrary. Amongst the kiln debris from these three sites there are several fragments showing impressions of wattles up to 1¼ in. (3 cm) in diameter. It is the elaborateness of the reconstruction, when applied to Romano-British kilns in general, rather than the component materials, that is in question. Although several thousand fragments of baked clay, from the fillings of kilns excavated in the Nene Valley during the past ten years, have been carefully examined by the writer, none would seem to be appropriate to the kind of domed covering shown in Jenkins' reconstruction. Those which have any curvature at all are tunnel-shaped and appear to have formed part of the flue arch. The great majority of fragments are flat, suggesting a more vertically-sided superstructure.

²⁴ The term ‘fire-box’ (cf. Hartley, B. R., Notes on the Roman Pottery Industry in the Nene Valley (Peterborough Mus. Soc. Occasional Paper No. 2 (1960), 11Google Scholar) would be more appropriate. By definition a stokehole is ‘a large hollow dug below ground level in which the fire was started, and into which the ash could be raked’ (Arch. Journ. cxiv (1957), 13Google Scholar) and Bryant adds ‘in which the potter stood whilst tending the fire’ (Bryant op. cit. (note 18), 149). The stokehole pits of Type IB and IC kilns are far too small to accord with such a description.

²⁵ Kiln III (Table I, 2).

²⁶ Woods op. cit. (note 8), fig. 13, 88.

²⁷ Kiln F9 (Table I, 5).

²⁸ Kilns I and II (Table I, 3-4).

²⁹ Woods op. cit. (note 8), 39 and fig. 15, 106. Identification by Mr. M. R. Hull.

³⁰ None of the pieces was sufficiently large to warrant illustration, but the majority seem to have come from channel-rim jars and larger storage vessels of Claudian or Neronian date.

³¹ Bryant (1973 (op. cit. note 5), 157) has suggested that long flues ‘would facilitate the establishment and retention of reducing conditions in the oven during the firing period, and they would certainly enable the potter to get a large amount of wood into the flue immediately prior to sealing’. Without disagreement with this statement, it should be recorded that there is nothing in the evidence from the Bozeat, Hardingstone and Weekley kilns to associate this element with the production of reduced wares. On the available evidence from the upper Nene Valley the most that can be said of these elongated flue-stokeholes is that, once the fire had taken hold, they would give a better and more easily controlled draught with less risk of serious fluctuations in firing temperatures.

³² Kilns I and VI.

³³ Particularly in the case of Kiln I.

³⁴ Kilns X and XI (M. R. Hull, The Roman Potters' Kilns of Colchester (1965), 3-5).

³⁵ Surrey Arch. Coll. li (1949), 29–56Google Scholar and Arch. Journ. cxiv (1957), 23–4 and pl. IV.Google Scholar

³⁶ This type of structure was misleadingly described by Corder (Arch. Journ. cxiv, 23-4) as a ‘horizontaldraught’ kiln. More recent work by Gunn and Bryant (Bryant (1973 (op. cit. note 5)), 149) has shown that through-draught kilns are not a practical proposition.

³⁷ JRS lix (1969), dated late first-century.Google Scholar

³⁸ Kilns I and VI (Bulltn. Northants Fed. Arch. Socs. i (1966), 8–9 and Table I, 6)Google Scholar.

³⁹ Kiln II (Table I, 7).

⁴⁰ I am grateful to Mr. H. C. R. Frost for allowing me to consult the notes, plans and photographs relating to this excavation.

⁴¹ Cf. the three-chambered kiln built at St. Ives by Mr. T. Matsubayashi and the Korean and Japanese bank or climbing kilns which can have as many as 20 separate compartments (B. Leach, A Potter's Book, London (2nd edn.), 1945, 184-7). See also Type IIC below.

⁴² Bulltn. Northants Fed. Arch. Socs. i (1966), 8.Google Scholar

⁴³ E.g. at Mucking, , Essex (Panorama (Journ. of the Thurrock Local History Society) No. 16 (Winter 1972-1973). fig 5).Google Scholar

⁴⁴ Kiln IV (Table I, 9).

⁴⁵ Kiln VII (Table I, 10).

⁴⁶ Woods op. cit. (note 8), fig. 14, 93 and 99.

⁴⁷ TL 163974 (Bulltn. Northants Fed. Arch. Socs. vii (1972), 24).Google Scholar

⁴⁸ Kilns I, II and V (Table I, 11).

⁴⁹ TL 163974 to 154978, Kilns III (F 4), F 23 and F 26-7 (Bulltn. Northants Fed. Arch. Socs. (Northamptonshire Archaeology) viii (1973), 7)Google Scholar.

⁵⁰ Bulltn. Northants Fed. Arch. Socs. v (1971), 26.Google Scholar

⁵¹ The haul road site, Kiln F 23 (Bulltn. Northants Fed. Arch. Socs. (Northamptonshire Archaeology) viii (1973), 7)Google Scholar.

⁵² The bollard pedestals of Kilns I and II (F 4), on the haul road site, were made of four circular fire-bricks luted together with clay (Bulltn. Northants Fed. Arch. Socs. (Northamptonshire Archaeology) viii (1973). 7)Google Scholar.

⁵³ Antiq. Journ. xxxiv (1954), 218–9.Google Scholar

⁵⁴ Bulltn. Northants Fed. Arch. Socs. (Northamptonshire Archaeology) viii (1973). 7.Google Scholar

⁵⁵ Kiln I (Table I, 12).

⁵⁶ Four pedestals similar to those illustrated in FIG. 5, F were found in situ in a late first- to early secondcentury kiln recently excavated by Mr. P. J. Foster at Ise Village, Kettering (SP 8860 7799). Kiln II at Needham, Norfolk (Arch. Journ. cxiv (1957), 17Google Scholar (fig. 6)) had two free-standing supports.

⁵⁷ In Kiln IV at Hardingstone (Site II) the impression given was that the pedestal had first been set in place and then the clay lining of the furnace bowl extended over its foot to fix its position. The subsequent removal of this support, presumably for further use in another kiln, had destroyed most of the floorlining (Woods op. cit. (note 8), 5).

⁵⁸ If the example from Hardwick Park, Wellingborough (FIG. 5, F) was symmetrical its height must have been at least 1 ft. 4 in. (0·4m). The central column at Weston Favell (Antiq. Journ. xxxiv (1954), 218–9Google Scholar and pl. XXIII, b) was 1 ft. 2 in. (0·4 m) high. Pedestals of the kind illustrated in FIG. 5, C-F were still being used in the Trajanic period (cf. Brassington, M., Antiq. Journ. li (1971), 40–1Google Scholar and pl. ix (Kilns Ie and If)).

⁵⁹ Cf. the Hunsbury pedestals (Dryden op. cit. (note 15), 61 and pl. v, 1). These had a single, lateral perforation through the stem.

⁶⁰ As at Elstow, Bedfordshire (TL 058476 to 058477) (Beds. Arch. Journ. iv (1969), 85Google Scholar and vi (1971), 69-71). The prototype of the permanent bollard pedestal may, perhaps, be seen in the rudimentary support found at Weston Favell (Antiq. Journ. xxxiv (1954), 218–9Google Scholar and pl. XXIII, b). This comprised three large pebbles, capped by a roughly circular clay plate about 9 in. (23 cm) in diameter. Apparently, the stones had been rendered with clay to form an integral part of the kiln.

⁶¹ As in Kilns I and II (F 4) on the haul road site, West Longthorpe (Bulltn. Northants Fed. Arch. Socs. (Northamptonshire Archaeology) viii (1973), 7)Google Scholar.

⁶² Brisay, K. de, Colchester Arch. Group Annual Bulltn. xv (1972), 23–43Google Scholar. The associated pottery was mostly of Late Belgic character.

⁶³ The Ise Village site (see note 56). The several stone slabs found in the furnace of this kiln were probably used to support the pottery load.

⁶⁴ The Harringworth Road quarry site (SP 940983), Kiln I. Similar evidence came from the second of the three kilns currently being investigated by the writer at the invitation of Mr. D. A. Jackson who is directing excavations for the Department of the Environment (Bulltn. Northants Fed. Arch. Socs. (Northamptonshire Archaeology) viii (1973), 17).Google Scholar

⁶⁵ SP 684469. Current excavations directed by the writer.

⁶⁶ Hartley op. cit. (note 24), 13.

⁶⁷ The differential firing of the upper and (or) lower surfaces of these rings shows where they have been in contact with similar objects immediately above and (or) below.

⁶⁸ P. Corder, The Roman Pottery at Crambeck, Castle Howard (Roman Malton and District, Report No. i) (1928), 43 and fig. 18 and Arch. Journ. cxiv (1957), 27Google Scholar and pl. v. A.

⁶⁹ Information from the excavator, Mr. P. J. Foster.

⁷⁰ Kiln I. Information from Mr. P. J. Foster.

⁷¹ E.g. at Little Chester, Derby (Brassington op. cit. (note 58), 41 and pl. IX, b). See also Arch. Journ. cxiv (1957), 14, 26Google Scholar and pl. v, E.