^{page 197 note 1} The thanks of the authors are due to the Council of the Society of Antiquaries of Scotland and to Mr. R. B. K. Stevenson, Keeper of the National Museum of Antiquities in Edinburgh, for their permission not only to examine and photograph the chamfrein, but actually to dismantle it; to Dr. H. J. Plenderleith, F.S.A., for permitting one of the horns of the chamfrein to be examined in the British Museum Research Laboratory; to Mr. Herbert Maryon, F.S.A., for dismantling the horn and for much expert advice; to Dr. A. A. Moss, for analysing paper from the horn; to Dr. J. G. Speed, of the Royal (Dick) School of Vetinary Studies in the University of Edinburgh, for advice on the original function of the chamfrein; to the British Non-Ferrous Metals Research Association, and in particular to their officer Mr. N. B. Rutherford, for technical reports on specimens from modern repairs to the object; and to Major-General Sir Walter Maxwell-Scott for permission to examine the replica of the chamfrein at Abbotsford.

^{page 197 note 2} The exact site is uncertain, but is presumably in the area of marshy ground marked on the O.S. maps as Torrs Loch (25/782622).

^{page 197 note 3} MSS. 912, fol. 80.

^{page 197 note 4} The stand is preserved in the museum at Abbotsford, and now supports the replica of the chamfrein referred to below. It is of turned mahogany with a carved cruciform base. The inscription reads ‘Found in the Earth/at Torrs/Parish of Kelton/Galloway’.

^{page 197 note 5} National Library of Scotland, MSS. Walpole Bequest, vol. 19, fols. 417–18.

^{page 198 note 1} [W. Mackenzie], History of Galloway (John Nicholson, Kirkcudbright, 1841), ii, Appendix, 70–71.

^{page 198 note 2} New Statistical Account, iv (Kirkcudbrightshire) (1845), p. 153.

^{page 198 note 3} P.S.A.S. vii (1870), pp. 334–41.

^{page 198 note 4} Anderson, J., Scotland in Pagan Times: The Iron Age (1883), pp. 112–17Google Scholar.

^{page 198 note 5} Leeds, E. T., Celtic Ornament (1933), p. 8Google Scholar, fig. 4.

^{page 198 note 6} Museum accession no. FA. 72.

^{page 198 note 7} On each horn the collar and flange were originally in one piece, but are now broken apart. For convenience the two parts are named and described separately.

^{page 199 note 1} For convenience the description of the object is related to its customary appearance. Thus in pl. lxx the front is to the left and the rear to the right, the dexter side being farthest from and the sinister side nearest to the observer.

^{page 199 note 2} The damage is shown in Train's drawing (pl. lxxi), where the accompanying legend reads: ‘C. The dark shade represents a piece broken off, since found’. Inquiry and search in the museum at Abbotsford have failed to reveal any trace or record of this fragment, though it is reasonable to suppose that, if found, it would have accompanied the chamfrein there.

^{page 199 note 3} It is shown below (p. 209) that this device cannot have been the present horns.

^{page 199 note 4} Imperial Standard Wire Gauge, the usual British method of expressing the diameters of wires and the thickness of sheet metal.

^{page 202 note 1} For convenience the sinister horn is labelled A, and the dexter B.

^{page 203 note 1} Fox, Llyn Cerrig Bach (1946), pp. 44–45 and pl. xxxi.

^{page 203 note 2} i.e. a line of minute zigzags, formed by rocking the graver from side to side.

^{page 204 note 1} Such holes are characterized by a smooth cylindrical bore, and by the presence of a slightly raised rim where the metal has been forced up by the pressure of the drill.

^{page 210 note 1} The cutting of the ends of the horns has been carried out, or at least finished, with a coarse file. The surface of the metal is still bright and uncorroded.

^{page 210 note 2} A hole in. in diameter was drilled axially down the centre of the ‘rivet’, and a left-hand-thread screw-extractor of appropriate size inserted.

^{page 210 note 3} Of two specimens kindly examined by the British Non-Ferrous Metals Research Association, one had 38 and the other 48 threads to the inch, with a common diameter of 0·09 in. (2·28 mm.). These measurements do not correspond to any of the modern series of screw sizes and threads, all of which were introduced later than 1840.

^{page 211 note 1} There is no reason to suspect Train himself. His account and drawing of the chamfrein appear to be a genuine record of its appearance at the time it came into his hands.

^{page 212 note 1} Loc. cit., supra, p. 198, n. 3.

^{page 212 note 2} The authors are indebted to Mr. Brian Hope-Taylor, F.S.A., for preparing the drawings in pl. lxxix.

^{page 212 note 3} A reliable guide to the type and size of head is given by the skulls of Celtic ponies from Newstead (Curle, Newstead (1911), 362–71, pls. xcv, xcvi). These belong to animals standing 10–12 hands high. Sir Cyril Fox in his well-known reconstruction of a Celtic chariot (Antiq. Journ. xxvii (1947), p. 118, fig. 1) assumes a height of 11½ hands for the ponies.

^{page 213 note 1} Curle, op. cit., 153–5, pl. xxi.

^{page 213 note 2} Curle, ibid., pl. xcv. 1.

^{page 214 note 1} This form of horse-trapping persists in the delightful straw hats, pierced for the ears, still occasionally to be seen in summer adorning the heads of cart-horses.

^{page 214 note 2} Germania xxx (1952), 189, pl. viii. 1.

^{page 215 note 1} Celtic Ornament (1933), pp. 6–11.

^{page 215 note 2} Llyn Cerrig Bach (1946), pp. 48–49.

^{page 216 note 1} Leeds, op. cit., Fig. 2a; de Navarro in Heritage of Early Britain (1952), 75 and fig. v. Fox has noticed the ‘unstable equilibrium’ of ‘force held in leash’ achieved in the triquetral patterns on the Llyn Cerrig shield-boss and on the Lambay Island scabbard-mount: we have here the equivalent in a symmetrical design (Arch. Camb. 1945, pp. 203–4; Advancement of Science, xxx (1951), p. 6Google Scholar of reprint).

^{page 216 note 2} Henry, F., Irish Art (1940Google Scholar), fig. 7a; J. Raftery, Prehist. Ireland (1951), fig. 267.

^{page 216 note 3} Raftery, op. cit., fig. 266.

^{page 216 note 4} de Navarro, loc. cit., pp. 74–75.

^{page 216 note 5} Cf. Fox in Advancement of Science, xxx (1951), pp. 6–7 of reprint.

^{page 216 note 6} Antiq. Journ. xxviii (1948), pp. 123–37Google Scholar.

^{page 217 note 1} Piggott, , Proc. Prehist. Soc. xvi (1950), p. 4Google Scholar; pl. ii.

^{page 217 note 2} Leeds, op. cit., fig. 31; Piggott, loc. cit.; Fox, Advancement of Science, xxx (1951); Wheeler, The Stanwick Fortifications (1954), p. 2: but surely hoard rather than chariot burials ?

^{page 217 note 3} Arch. Camb. 1948, pp. 24–44.

^{page 218 note 1} De Navarro, loc. cit. 74, who regards the Brentford piece as an import of Jacobsthal's style II from the Rhineland; Leeds, op. cit., figs. 1 and 7.

^{page 218 note 2} Raftery, op. cit., fig. 216.

^{page 218 note 3} Llyn Cerrig Bach (1946), 48; Newnham Croft in V.C.H. Cambs. i. 194, fig. 26; here reproduced as fig. 4.

^{page 218 note 4} An unpublished example. We are indebted to Mr. Martyn Jope for a drawing.

^{page 218 note 5} This ambivalence was perceived by Fox (op. cit. 49 n.).

^{page 218 note 6} Leeds, op. cit., fig. 5 left.

^{page 218 note 7} See below, p. 223.

^{page 219 note 1} Fox, op. cit. 44–45 and pl. xxxi.

^{page 219 note 2} J. A. Smith's drawings of these designs (as again given by Anderson (Scot. Pag. Times: Iron Age (1883), fig. 93) and elsewhere reproduced (e.g. Leeds, Celtic Ornament (1933), fig. 4)) are inaccurate and misleading.

^{page 221 note 1} Basically the asymmetric triquetra as on the Bugthorpe and Amerden scabbards; Fox in Arch. Camb. 1945, 206, 208–9.

^{page 221 note 2} Best illustrations in Jacobsthal, , ‘The Witham Sword’, Burlington Mag. lxxv (July 1939Google Scholar).

^{page 221 note 3} British Museum, Early Iron Age Guide (1925), fig. 115; Proc. Prehist. Soc. xvi (1950), 15Google Scholar, fig. 8A.

^{page 221 note 4} Yorks. Arch. Journ. xxxi (1933), p. 94.

^{page 221 note 5} Early Celtic Art, Grammar of Ornament no. 67 (a), For another Glastonbury survival of an early motif cf. Proc. Prehist. Soc. xviii (1952), p. 164, fig. 3, G. iii, which appears in almost identical form among the engraved patterns on the circular Wandsworth boss.

^{page 221 note 6} Cf. Fox, Arch. Camb. lxxxii (1927), p. 67, figs. 3–5, 9.

^{page 221 note 7} Watson in Antiq. Journ. xxvii (1947), p. 178.

^{page 221 note 8} The spirals on the Stichill collar, related to but not strictly derived from the foregoing, are discussed below (p. 233). The relation of the Broighter ‘snails’ to the earlier ‘hair-spring’ spirals was noted by Arthur Evans (Arch, lv (1897), p. 388).

^{page 221 note 9} De Navarro, Heritage Early Brit. (1952), p. 75. For relevant spirals in the Hungarian sword-style cf. Jacobsthal, Early Celtic Art, nos. 106–8 (La Tène), 113 (Cernon-sur-Coole), 115 (Kis-Köszeg); no. 116 (Bölcske) though with-out spirals is very close in feeling to the ornament on the Torrs horns. The torc from the Marne (no. 216), and a brooch from Germany (no. 323) again show a very comparable use of incised spirals: they recall especially the Sawdon brooch. Another Marne torc (from Courtisols, inadequately illustrated by de Baye and Déchelette (iv, fig. 515, no. 6)) may be related.

^{page 222 note 1} e.g. by Leeds, Celtic Ornament, p. 14.

^{page 222 note 2} Early Celtic Art, 75.

^{page 222 note 3} Early Celtic Art, no. 11. Cf. no. 144 for the use of a similar face-motif, but enmeshed in pattern, as at Torrs, and Jacobsthal in Amer. Journ. Arch. xlvii (1943), p. 312 on melancholy Celtic faces. Cf. also Saxl, and Wittkower, , Brit. Art and Mediterranean (1948), pl. 3Google Scholar.

^{page 222 note 4} Early Celtic Art, nos. 160, 201, 208, 295, 308, 309; the last three on brooches which may well have formed the most likely means of transporting the motif to Britain.

^{page 222 note 5} The anthropoid daggers from North Grimston, E. R. Yorks (Mortimer, Forty Years (1905), p. 356) and from an unlocated Yorkshire find-spot (Piggott and Daniel, Picture Book Anc. Brit. Art (1951), no. 37) are cases in point; even more like the Torrs face is the head on another anthropoid dagger from Chatenay, Marne (Gilardoni, La Naissance de l'Art (Lausanne 1948), pl. 70).

^{page 222 note 6} As on the Aylesford bucket (Kendrick, Anglo-Saxon Art (1938), pl. ii, no. 3) or from Welwyn (Piggott and Daniel, op. cit., no. 70).

^{page 222 note 7} For the significance of the isolated human head in Celtic art and religion cf. Lambrechts, L'Exaltation de la Tête dans la pensée et dans l'Art des Celtes (1954).

^{page 222 note 8} Jacobsthal, Early Celtic Art, no. 16.

^{page 223 note 1} We owe this suggestion to Sir Cyril Fox; the species was tentatively identified at Glastonbury (Glast. Lake Vill. ii. 635).

^{page 223 note 2} National Museum, Dublin; Wilde, Cat. Mus. R.I. Acad. ii (1861), p. 534, fig. 413. The cup is 5¾ in. diameter at the mouth.

^{page 223 note 3} The photographs are by Mr. A. E. P. Collins and are reproduced with his kind permission and the consent of the Government of Northern Ireland's Archaeological Survey.

^{page 223 note 4} Bushe-Fox, Excav. at Hengistbury Head (1915), p. 61 and pl. xxix. 6; pottery from Site 3, pl. x, nos. 10, 12, 13, 15, 16; pl. xvi, nos. 4, 17, 18.

^{page 223 note 5} The Red Hill, Long Eaton, bird-brooch is only a partial parallel, though of course ultimately related: it has sockets for coral inlays, but not in the eyes (Hawkes and Jacobsthal in Ant. Journ. xxv (1945), p. 117). The St. Catharine's Hill bird-mount (perhaps from a brooch) represents the older, Hallstatt, tradition of plastic bird ornament. (Hawkes et al., St. Catharine's Hill (1930), p. 127.)

^{page 223 note 6} Early Celtic Art, pls. 153–7; ducks on nos. 289, 291, probably 298, 306, 307; coral eye-studs on nos. 289, 291–3, 300. Cf. Déchelette, Manuel, iv, fig. 533, no. 7.

^{page 224 note 1} Early Celtic Art, 116–17; nos. 143, 147, 148. Cf. Coutil in Bull. Soc. Préhist. Franc. 1913, pp. 1–8.

^{page 224 note 2} As at Viksø, Denmark (also with horns: cf. p. 226 below); Norling-Christensen in Acta Arch. xvii (1946), pp. 99Google Scholar.

^{page 224 note 3} Toynbee, J. M. C. in Journ. Rom. Stud. xxxviii (1948), p. 24Google Scholar. Cf. Polybius iii. 62, on ‘Gallic panoplies, such as are worn by their kings when they fight in single combat’.

^{page 224 note 4} Details and illustrations most easily accessible in Minns, , ‘Art of the Northern Nomads’ (Proc. Brit. Acad. xxviii (1942), pp. 16–18Google Scholar and pl. vi).

^{page 224 note 5} As suggested by Dr. Jacobsthal in conversation with the writers.

^{page 225 note 1} Altertümer von Pergamon, ii, pl. 43; cf. Reinecke in Tschumi-Festschrift, 1948, 93. We are indebted to Professor Jocelyn Toynbee for drawing our attention to this important piece of iconography. Cf. her comments (on the visor-mask in the same relief) in Journ. Rom. Stud. loc. cit.

^{page 225 note 2} Jacobsthal has discussed this aspect of the Pergamon relief in Amer. Journ. Arch. xlvii (1943), p. 310Google Scholar.

^{page 225 note 3} Alföldi and Radnoti in Serta Hoffilleriana (Zagreb 1940), pp. 309–19 and pl. xxxi.

^{page 225 note 4} Arch. Ael. 3rd ser. v (1909), pp. 322, 344; fig. 11. We are indebted to Mr. John Gillam for the photograph here reproduced.

^{page 225 note 5} Illustrated e.g. by Lantier, Les Origines de l'Art français (1947), fig. 112.

^{page 225 note 6} The pair of large iron horns in the Sesto Calende cart grave (more than 18 in. overall) are perhaps, as Déchelette suggested, cart-trappings comparable with Jacobsthal's interpretation of the ‘chariot-horns’ of Waldalgesheim and La Bouvandeau (Montelius, Civ. Prim. It. i (1895), p. 317, pl. 62; Déchelette, Manuel, iii, fig. 275; Randall-MacIver, Iron Age in Italy (1927), 69, Pl. 14. 4; Jacobsthal, Early Celtic Art, 121; cf. Fox, Llyn Cerrig Bach, pp. 15–19).

^{page 226 note 1} Manuel, iv, pp. 661 ff.

^{page 226 note 2} Bull. Soc. Préhist. Franç. 1913, p. 1.

^{page 226 note 3} Déchelette, loc. cit., figs. 484, 485.

^{page 226 note 4} Brit. Mus., Later Prehist. Antiq. Brit. Isles (1953), pl. xviii. 1; Piggott and Daniel, Picture Book …, no. 53.

^{page 226 note 5} e.g. Bushe-Fox, Excav. Richborough, iv (1949), p. 131, pl. xxxvii. 131; Hawkes and Hull, Camulodunum (1947), p. 340, pl. ciii. 6. Cf. also the so-called Group VI chapes of the native scabbard series (Piggott in Proc. Prehist. Soc. xvi (1950), p. 22Google Scholar) which in view of their asymmetry may not be scabbard-chapes at all; the Ham Hill dagger chape (with a native anthropoid dagger (Ant. Journ. iii (1923), p. 149)), is, however, certainly knobbed and comparable with Roman examples.

^{page 226 note 6} The convention begins in the Late Bronze Age: the Viksø and allied horns are also knobbed, as are many on the Sardinian warrior-statuettes, and knobbed horns appear on a well-known series of ox-handled bronze vessels from Central Europe (von Merhart, Festschrift Röm. Germ. Zentralmus. Mainz (1952), ii, pl. 14). Cf. Hawkes, in Aspects of Arch. (1951), p. 192Google Scholar.

^{page 226 note 7} Proc. Royal Irish Acad. xxviii (C) (1910), p. 104. A study of these horns by Professor M. J. O'Kelly and one of the present writers (S.P.) is in preparation.

^{page 226 note 8} Raftery, Prehist. Ireland (1951), fig. 233. The horns belong to Raftery's third and latest stylistic phase of Irish Early Iron Age art (op. cit., p. 190).

^{page 226 note 9} Kossack, Stud. z. Symbolgut der Urnenfelder- u. Hallstattzeit Mitteleuropas (R. G. K. Forsch. Bd. xx, 1954), p. 56; pl. 14 and 26 (map).

^{page 226 note 10} Jacobsthal, Early Celtic Art, pp. 111–14; note partichape cularly the reference to silver-mounted drinking-horns of the Urus by Caesar (B.G. vi. 28), and add the representation of a pair of Gaulish drinking-horns on the triumphal arch at Carpentras, Vaucluse (Espérandieu, Recueil, i, p. 180, pl. 243; Lantier, Les Origines de l'Art français (1947), pl. 88 (better illustration)). Klein Aspergle horns: Jacobsthal, nos. 16, 17; ram's heads on brooches: nos. 308, 312, 314.

^{page 227 note 1} Stenberger, Fornvännen, xli (1946), p. 147; Voss and Ørsnes-Christensen, Acta Arch. (1948), p. 209; cf. Hawkes, Aspects of Arch., pp. 190 ff.

^{page 227 note 2} The earlier examples would be those from Valsgärde, Taplow, and Sutton Hoo; the later series has been commented on by O'Riordain, Proc. Roy. Irish. Acad. lii (1949), Sect. C, p. 66.

^{page 227 note 3} Our thanks are due to the authorities of the Museum, and especially to Mr. D. W. Wotherspoon who made the reconstruction with such skill.

^{page 227 note 4} We are indebted to Mr. R. L. S. Bruce Mitford for this information; the surviving Taplow horn is smaller. Jacobsthal notes a pair of horns of ‘Bison priscus’, containing residue of beer and mead respectively, from Skudstrup, Hadersleben (Early Celtic Art, p. 113).

^{page 228 note 1} Loc. cit., p. 75.

^{page 228 note 2} See list with references in Appendix.

^{page 228 note 3} Loc. cit., p. 74.

^{page 228 note 4} Ant. Journ. xxiv (1944), p. 123Google Scholar; Cf. Hawkes in loc. cit. xxvi (1946), p. 187.

^{page 230 note 1} Ant. Journ. vi. 277; Leeds, Celtic Ornament, p. 5; Jacobsthal, Early Celtic Art, p. 95 (P 471); Fox in Proc. Prehist. Soc. xviii (1952), p. 50.

^{page 230 note 2} Piggott in Proc. Prehist. Soc. xvi (1950), p. 5; Fox, loc. cit.

^{page 230 note 3} Summary Guide, Dept. of Antiqs. Ashmolean Mus. (1951), pl. xlvi; Piggott, loc. cit., pl. 11.

^{page 231 note 1} This revises the later dates in Piggott, loc. cit. following discussions with Professor C. F. C. Hawkes. The Minster Ditch (Oxford) scabbard, and the greater part of the Thames dagger-sheath series, should go with the first La Tène I brooches in Britain, somewhat antedating the art-styles here discussed.

^{page 231 note 2} Wheeler, Maiden Castle (1943), pl. xxviii, especially no. 6.

^{page 231 note 3} Early Celtic Art, p. 95. He assigns the St.-Pol-de-Leon ‘stamnos’ to the same date; these Breton pots must at all events be earlier than c. 56 B.C.

^{page 231 note 4} We are indebted to Mr. Martyn Jope for drawing our attention to this feature while the paper was in proof. He would compare the Keshcarrigan cup with that from the Colchester burial also containing an engraved mirror (Antiq. Journ. xxviii (1948), p. 136Google Scholar), and one might add the bowls from the Birdlip mirror-grave (Arch, lxi (1909), p. 332Google Scholar). Both would be early first century A.D.

^{page 232 note 1} Piggott, in Proc. Prehist. Soc. xvi (1950), p. 14Google Scholar.

^{page 232 note 2} Raftery, Prehistoric Ireland, fig. 216, and footnote, 218, above.

^{page 232 note 3} The motif of asymmetric dome with spirals (above, p. 218; fig. 3, no. 7) is particularly clear on the Castle Strange stone.

^{page 232 note 4} Cf. Mahr in Proc. R. Irish Acad. xlii (C), (1934), pp. 23–24.

^{page 232 note 5} Mortimer, Forty Years … (1905), Frontispiece.

^{page 232 note 6} e.g. by Françoise Henry (Irish Art, 6), following Déchelette, Manuel, iv. 1029.

^{page 232 note 7} Loc. cit., p. 45. He compares the ornament with Déchelette, Manuel, iv, fig. 661, which represents painted pottery of La Tène I from the Marne and Ardennes, of which the Saulces-Champenoises pattern (Jacobsthal's P426 in his Grammar of Ornament) is indeed comparable with Turoe and of style II. But confusing Déchelette's fig. 661 with his fig. 663 (the well-known pots from Plouhinec and St.-Pol-de-Leon) Raftery then goes on to speak in the text of ‘the painted designs on the Breton pottery’ and so to assign the Turoe craftsman to a Breton origin as well.

^{page 232 note 8} Armstrong, Cat. Gold Orn. R.I. Acad. (1933), p. 65, pl. xiii. 98; Raftery, Prehist. Ireland, fig. 234.

^{page 232 note 9} Anderson, Scot. in Pagan Times: Iron Age (1883), p. 136.

^{page 233 note 1} Brit. Mus. Antiq. Roman Britain (1951), p. 13, fig. 6; 28, pl. iii. Probably second century A.D.

^{page 233 note 2} Celtic Ornament, p. 110. The Thirst House armlet was associated with coins of A.D. 154–270; Samian of c. 100–250, and second century brooches (Vict. County Hist. Derbyshire i. 234). A version of this ‘N’ in ‘Mirror-style’ art is on the Balmaclellan crescentric plate as its central motif.

^{page 233 note 3} Claudian-Neronian, and their motifs of course also derived from metalwork; these pottery versions are likely to have been the intermediaries between the native craftsmen and the silver ware of, e.g., Hildesheim type. (Cf. Fox in Advancement of Science, xxx, Sept. 1951.)

^{page 233 note 4} Cf. Proc. Soc. Ant. Scot. lxxxiv (1949–50), pp. 129–35.

^{page 233 note 5} As argued in Proc. Prehist. Soc. xvi (1950), p. 16.

^{page 234 note 1} For discussion of the date of this bowl see p. 231. It is here retained in the earlier group on stylistic grounds.