attested manufacturing centre, and the easiest for us to examine.  The ware directly embodies the Celtic tradition.  It is based, indeed, on classical elements, foliated scrolls, hunting scenes, and occasionally mythological representations (Figs. 15, 16).  But it recasts these elements with the vigour of a true art and in accordance with its special tendencies.  Those fantastic animals with strange out-stretched legs and backturned heads and eager eyes; those tiny scrolls scattered by way of decoration above or below them; the rude beading which serves, not ineffectively, for ornament or for dividing line; the suggestion of returning spirals; the evident delight of the artist in plant and animal forms and his neglect of the human figure—­all these are Celtic.  When we turn to the rarer scenes in which man is specially prominent—­a hunt, or a gladiatorial show, or Hesione fettered naked to a rock and Hercules saving her from the monster[4]—­the vigour fails (Fig. 17).  The artist could not or would not cope with the human form.  His nude figures, Hesione and Hercules, and his clothed gladiators are not fantastic but grotesque.  They retain traces of Celtic treatment, as in Hesione’s hair.  But the general treatment is Roman.  The Late Celtic art is here sinking into the general conventionalism of the Roman provinces.

[Footnote 1:  For the New Forest ware see the Victoria Hist. of Hampshire, i. 326, and Archaeol.  Journal, xxx. 319.  The Brough brooches have been pointed out by Sir A.J.  Evans, whose work on Late Celtic Art is the foundation of all that has since been written on it, but have not been discussed in detail.]

[Footnote 2:  Victoria Hist. of Northamptonshire, i. 206-13; Artis, Durobrivae of Antoninus (fol. 1828).]

[Footnote 3:  For the Belgic ‘Castor ware’ see the Belgian Bulletin des commissions royales d’art et d’archeologie (passim); H. du Cleuziou, Poterie gauloise (Paris, 1872), Fig. 173, from Cologne; Sammlung Niessen (Koeln, 1911), plates lxxxvii, lxxxviii; Brongniart, Traite des arts ceram., pl. xxix (Ghent and Rheinzabern).  M. Salomon Reinach tells me that the ware is not infrequent in the departments of the valleys of the Seine, Marne, and Oise.  The Colchester gladiator’s urn mentioning the Thirtieth Legion (C.R.  Smith, Coll.  Ant., iv. 82, C. vii. 1335, 3) may well be of Rhenish manufacture.]

[Footnote 4:  This, or the corresponding scene of Perseus and Andromeda, is a favourite with artists in northern Gaul and Britain.  It occurs on tombstones at Chester (Grosvenor Museum Catalogue, No. 138) and Trier (Hettner, Die roem.  Steindenkmaeler zu Trier, p. 206), and Arlon (Wiltheim, Luciliburgensia, plate 57), and the Igel monument.  For other instances see Roscher’s Lexikon Mythol., under ’Hesione’.]

[Illustration:  FIG. 14.  FRAGMENTS OF NEW FOREST POTTERY WITH LEAF PATTERNS. (From Archaeologia).]