^{page 27 note 1} xxvii, 210 ff.

^{page 27 note 2} li, 130 ff.

^{page 27 note 3} Les vases céramiques ornés de la Gaule romaine (1904), i, 181Google Scholar.

^{page 27 note 4} Oswald does, however, mention moulds in ‘buff micaceous clay’ (op. cit., p. 214). Among easily accessible French publications, only Mme Mathonnière's paper in Gallia xix (1961), 57Google Scholar, describes the ware at all clearly, apart, that is, from Vertet's publications cited below. The ware is described by J. Martin in a paper important for its illustrations (‘L’évolution des vases sigillées de Lezoux au I^er siècle de notre ère’, Bull. hist. et scient. de l’Auvergne, lxii (1942), 181–209Google Scholar, see p. 185), but without mention of mica. This paper is apparently not available in Britain, and I am consequently much obliged to M. Bernard Hoffman and to Dr. Grace Simpson for access to a copy of it.

^{page 28 note 1} For ‘Caerleon ware’ see now G. C. Boon, ‘Legionary ware at Caerleon?’, Arch. Camb., 1966, 45–46.

^{page 28 note 2} As suggested by Vertet, , Gallia, xx (1962), 380Google Scholar. Cf. also the use of the word Arretinum to describe both South- and Central-Gaulish products (ibid., p. 375, refs.), of which the Lezoux example DURA/FARE may be quoted (Gallia, xix (1962), 64Google Scholar, fig 24).

^{page 28 note 3} Technique of manufacture: Blanc, A., Rev. archéol. de I’Est (1964), pp. 267 ff.Google Scholar; Rev. archéol. du Centre, iv (1965), 21 ff. I am indebted to M. Blanc for opportunities of discussing this question.

^{page 29 note 1} Nos. 2, 3, 4, 8, and 12 are partly burnt. The proportion of burnt to unburnt pieces seems larger than normal with other types of sigillata.

^{page 29 note 2} Information from M. Blanc.

^{page 30 note 1} Les vases céramiques ornés, i, 178.

^{page 30 note 2} Gallia, xx (1962), 351 ff.Google Scholar; Actes du 88^ième Congrès national des Sociétés Savantes (1965), pp. 105 ff.-I here acknowledge the great help received from M. Vertet in connexion with the present paper. Not only have I had occasion to lay his published works in fee, but I have had the benefit of discussions with him, and he has further supplied me with unpublished material (pl. 1) of his own discovery.

^{page 30 note 3} Camulodunum (1947), pp. 174 ff. I must own to some doubts as to whether this case can be accepted in toto. If it may apply at Camulodunum, it may not apply elsewhere; and if it applies to la Graufesenque and Banassac—certainly the latter—the most ancient factory at Montans may have to be excluded. For a Loeschcke 8 of Montans-make at Silchester, see May, Silchester Pottery (1916), pls. 111, 12 and its stamp AC in wreath, lxxxia, i.

^{page 31 note 1} The opportunity of including a little-known specimen of Tardo-Italic ware from Silchester is here taken for comparative purposes (p. 42).

^{page 31 note 2} Rev. archéol. du Centre, ii (1963)Google Scholar.

^{page 31 note 3} I here acknowledge the generous help received from Mr. Hartley on general matters connected with the subject of this paper; on specific points also. He has further lent me representative shards from his excavations at Lezoux, and has read a previous draft of this paper.

^{page 32 note 1} Not identified as Lezoux. Not stated to be micaceous. ‘Tile-red fabric; orange or yellow-red slip; closely similar to red “Belgic” ware.’ Clearly Lezoux: note the ‘spidery’ gadroons, cf. Vertet, Actes, figs. 1, 10, etc., Curie, fig. 17, 2.

^{page 32 note 2} The nearest South-Gaulish-if it is South Gaulish-parallel to the Lezoux trumpets and scheme appears on the vrvoedi (? –]vrvoed) bowl, Knorr 1919, Textbild 28; Knorr 1952, Taf. 61; Oxé, Schumacher-Festschrift (1930), Taf. 36, 9.

^{page 34 note 1} If potters trained in the correct procedures had migrated to Lezoux, they would have used them, A consideration of the characteristics of the fabric (p. 28) shows that they were not used.

^{page 34 note 2} See p. 28, n. 2.

^{page 34 note 3} Examples could be multiplied of instances where names imprinted in moulds, etc., differ from those imprinted on the resultant bowls. For South Gaul, see Hermet, La Grauf., pp. 276–7, pl. 106; for Central Gaul, Stanfield-Simpson, Central Gaulish Potters (1958), pl. 27, 324, with Terrisse, Rev. archéol. du Centre, ii (1963), 285Google Scholar.

^{page 34 note 4} Probably because the upper surface was wiped before firing, a procedure tending to assist the alignment of the siliceous materials parallel with the surface. See Blanc, op. cit. p. 28, note 3 (i), 276, note 2. The poor detail of much Banassac ware is probably also due to bowls being wiped before firing. Certainly, many Banassac moulds are much sharper than much of the production would suggest: ovolo ‘A’, for example, with acorn-shaped tongue (Déch. i, 118, fig. 79), is really toothed in the upper lobe (from personal examination).

^{page 35 note 5} I am greatly indebted to Professor Howard Comfort for the following information. Vit (u) lus Naevi is recorded from Pozzuoli (21 exx.), Arlaines, Trier, Neuss, Vechten, and Haltern. Vit(u) lus alone occurs at Pozzuoli (25–30 exx.), Tarragona, Aquileia (5 exx., probably a Po valley potter), Orange, Nîmes, Le Garros, Olsbergerhof, Mainz, Neuss, and Xanten. Oxé had observed that the ‘Silchester’ Vitlus (of which he recorded in the forthcoming Oxé-Comfort List only the Dr. 29) was an early Gaulish potter and nothing to do with the Italic one. Examples from Hüfingen, Mainz (4 exx.) and Xanten (see Oswald & Pryce, pl. 26, 3 c) may be associated with the ‘Silchester’ potter but it would be necessary to examine the vessels concerned to form a definite opinion. The Xanten stamp, at least, is different. Oxé also drew attention to some grey Belgic plates (Form Cam. 16AB) from Trier published by Koethe, , Oxé Festschrift (1938), p. 92Google Scholar, Abb. 2, no. 91, p. 105, with the framed stamp VITLV, probably in imitation of the stamp of the Italic or Arvernian potters.

^{page 36 note 1} Titos falls into the same category of names as Vitlus (see Loeschcke, Taf. 29, 2 5 2, and pp. 18 5–6 for a Puteolan Titus: Professor Comfort adds that, in various forms of stamp, this potter is known at Pozzuoli (80 exx.), Rome, Cherchel, Carthage, Lyon, Mainz, Xanten, and Haltern). Most Lezoux potters of the early phase have Gaulish names, as noted by Déchelette, i, 179-a list which can now be revised and extended. Titos of Lezoux was first recorded by Martin, 188 [internal stamp].

^{page 40 note 1} Loeschcke suggests that these dishes were used for cooking large flat cakes; the deep lustrous slip corresponding to the ‘non-stick’ surface of modern kitchen-ware.

^{page 42 note 1} My thanks are due to Mr. D. T.-D. Clarke, F.S.A., for sending representative samples from Colchester for comparison.