^{page 29 note 1} Déchelette, J., Les Vases céramiques ornés de la Gaule romaine, Paris, 1904.Google Scholar

^{page 29 note 2} Dr. D. B. Harden, Professor J. M. C. Toynbee, Professor Donald Atkinson, and Miss M. Bimson (British Museum Laboratory) are sincerely thanked for their help; and also all who have allowed sherds to be included here or have provided photographs. The Museums concerned are recorded in the Lists. Mr. Wilfred Dodds made the drawings, and I am much indebted to him for undertaking this arduous task.

^{page 29 note 3} Terra sigillata is not employed here because the strict meaning of that term does not include the plain wares and refers only to vessels ‘decorated with little figures’ Cicero, Verr. II, 4, 14, § 32, sigillati scyphi—metal (silver) cups.

^{page 29 note 4} Déchelette, ii, 167–234, ‘Vases à reliefs d'applique’.

^{page 29 note 5} Two moulds were used to make the South Gaulish jugs; cf. Hermet, , La Graufesenque, pls. 4, 5, nos. 13, 15, 18.Google Scholar

^{page 30 note 1} Oswald, and Pryce, , An Introduction to the Study of Terra Sigillata (London, 1920), p. 230.Google Scholar

^{page 30 note 2} The beaded border around the appliqué type 46 is decorative (cf. the patera handles, Déchelette, ii, pl. VII) and the beads are not, as Déchelette supposed, the impressions made in the clay by an original metal vessel from which the type might have been taken. Professor Toynbee informs me that she has not seen any Roman silver or bronze vessels with applied and riveted designs. The effect would have been distinctly like a patch.

^{page 30 note 3} Bimson, Mavis, Antiq. Journ. XXXVI (1956), 200–4, ‘The Technique of Greek Black and Terra Sigillata Red’. Unintentional variations occurred in a modern experiment described by Miss Bimson (203), for three vessels were fired at the same time, and two were slightly reduced and the third became black.CrossRefGoogle Scholar

^{page 31 note 1} There is a photograph of a second-century kiln at Lezoux with numerous clay pipes; cf. Chenet, G. and Gaudron, G., La Céramique sigillée d'Argonne (1955), fig. 40.Google Scholar

^{page 31 note 2} Antiq. Journ. xxxiv, 67–68, where it is described as Castor ware.

^{page 31 note 3} Arch. Cant. lxix (1955). 71.Google Scholar

^{page 31 note 4} Curle, J., Newstead (1911), pp. 254–6Google Scholar; Kenyon, K. M., Leicester (1948), p. 120Google Scholar; Oelmann, , Niederbieber (1914)Google Scholar, Taf. 2, types 29–38; Ludowici, Rheinzabern: examples of barbotine red samian cups were reproduced in O. & P., pls. LXXIX, LXXX. Pottery very like British barbotine Castor ware was made in northern France in the later empire.

^{page 32 note 1} Déchelette, i, LIBERTVS: 282, 28, 29 from Lezoux; 283, 60 (no. 2); 283, 64 (no. 5). BVTRIO: 257, 17, 25; and 296, 17. PATERNVS: 290, 43 (no. 7).

^{page 32 note 2} Déchelette, ii, 173: AVITI, MAGNANI MANV M, MARCI, M[A]R[TI]ALIS, MATERNI, PRISC[, PVL[, SAT[, SEVERINI SVRIV, ]ATV [MA]NV.

^{page 32 note 4} Déchelette, 12, 16, 18, 43, 46, 51, 57, 69, 75, 80, 100, 104, 105, 108, 131, 154 (no. 12 from Wroxeter).

^{page 33 note 1} Déchelette recorded 134 types and J. A. Stanfield added 19 more.

^{page 33 note 2} Oswald, F., ‘The decorated work of the potter Butrio’, J.R.S. xx, 71–77.Google Scholar

^{page 33 note 3} Déchelette, i, 283, 54, found at Vichy; reproduced in O. & P., pl. XXI, 7, Form 65. A white slip contains very little or no iron.

^{page 33 note 4} O. & P., p. 137: ‘many of his figure-types can be traced back to prototypes in metal’.

^{page 34 note 1} Toynbee, J. M. C. and Hume, I. Noel, Arch. Cant. LXIX (1955), 73–74.Google Scholar

^{page 34 note 2} My thanks are due to the excavator, Mr. E. D. Bailey, and to Mr. A. Warhurst (Northampton Museum), for allowing me to examine the samian ware from this pit.

^{page 35 note 1} Nos. 1 and 3 are from drawings by the late J. A. Stanfield.

^{page 37 note 1} Hölder, O., Die römischen Thongefässe der Altertumssammlung in Rottweil (1889), Taf. xxii, 1.Google Scholar

^{page 42 note 1} Professor J. M. C. Toynbee and Professor I. A. Richmond have examined the sherd and kindly made these comments.