^{page 19 note 1} My colleague Professor A. J. Beattie very kindly provides me with notes on the literary sources: the words are known only from late lexicographers. The form κάρνον is given by Hesychius, with the meaning of ‘trumpet’, and assigned to the Galatians. κάρνυξ is given by Eustathius (1189 = 1139. 56–57) and in the scholion on Iliad, 18.219 in virtually identical passages presumably deriving from a common original. Both describe it as a trumpet with an animal ‘bell’ or mouth, with a leaden tube, and assign it to the Celts and the Galatians. Polybius in the second century B.C. (ii, 29. 6) describes Celtic war-trumpets, using the phrase βυκανητῶν καὶ σαλπιγκτῶν, either indicating two types of trumpet or more likely glossing an unfamiliar word by adding ‘and trumpeters’; Diodorus Siculus a century later (v, 30) and in a similar context refers to trumpets ‘peculiar and such as barbarians use’, but neither employs the word carnyx. The basic archaeological studies are those of Behn, F. (Mainz. Zeitschr. vii (1912), 36Google Scholar; Reallexikon, viii (1927), 357; Musikleben in Altertum … (1954), p. 145) and Déchelette, Joseph (Manuel d'Arch., Âge du Fer II (2nd ed. 1927), p. 683).Google Scholar

^{page 19 note 2} Arch. Journ. xci (1934), 103–5, 183, with full references to earlier publications. Cf. also Proc. Prehist. Soc. xvi (1950), 1–28 (swords and daggers); ibid. xxi (1955), 198–227 (anthropoid dagger, no. 48, Class G variant); Linc. Archaeol. & Architect. Soc. Reports & Papers, vii (1957), 9 and fig. 1 (sword and scabbard).

^{page 20 note 1} Phil. Trans. lxxxvi (1796), 395, and pl. XI.

^{page 20 note 2} Loc. cit., pl. XXI.

^{page 21 note 1} I am much indebted to our Fellow Mr. F. T. Baker of the City and County Museum, Lincoln, for his help and for supplying me with a photograph of the drawing, here reproduced (pl. VI). The volumes of drawings are referred to by Mr. Phillips (loc. cit., p. 156): they were at that time in private hands

^{page 21 note 2} The Gundestrup scene has often been illustrated since its original publication by Müller, Sophus in Nord. Fortids. I (1892), pl. VIGoogle Scholar: e.g. Brønsted, , Danmarks Oldtid, III (1940), 94, fig. 76Google Scholar; Klindt-Jensen, , Foreign Influences in Denmark's Early Iron Age (1950 = Acta Arch. xx (1949), 1–229)Google Scholar, p. 125, fig. 78 E. I am indebted to Dr. Klindt-Jensen and the National Museum, Copenhagen, for the photograph here reproduced (pl. VIIb)

^{page 22 note 1} Powell, T. G. E., The Celts (1958), pl. 48.Google Scholar

^{page 22 note 2} Behn, opp. citt.

^{page 22 note 3} Antiquity, xxvi (1952), 87 and pl. 1. Other Roman and Gaulish numismatic representations are quoted by Evans, , Coins Anc. Brit. (1864), p. 192.Google Scholar Most of the monumental representations are in Espérandieu: cf. Déchelette, op. cit., for references

^{page 22 note 4} For the date of the Gundestrup cauldron, see Jacobsthal, P., Early Celtic Art (1944), p. 2—Google Scholar ‘an East Celtic work of the first century B.C.’; Klindt-Jensen, , op. cit., pp. 119 ff.Google Scholar

^{page 22 note 5} Proc. Prehist. Soc. xxiv (1958), 44; pl. 1, nos. 1–5.

^{page 22 note 6} Behn, opp. citt.

^{page 22 note 7} I am indebted to Dr. Adolf Rieth of Tübingen for kindly obtaining the photograph here reproduced (pl. xa), and for drawing the sections given in fig. 2.

^{page 23 note 1} Das Federseemoor als Siedlungsland des Vorzeitmenschen (1929), p. 162, pls. XXXIV–XXXV.

^{page 23 note 2} Keim, J. and Klumbach, H., Der römische Schatzfund von Straubing (1951)Google Scholar; I have discussed the comparable British hoards in Proc. Soc. Ant. Scot. lxxxvii (1952–3), 1–50

^{page 23 note 3} Arch. xcvi (1955), pp. 222, 231, pl. LXXXIII.

^{page 23 note 4} Ulster Journ. Arch. XX (1957), 85.

^{page 24 note 1} Richborough IV, p. 113, pl. XL, no. 152.

^{page 24 note 2} Fox, C., Arch. Camb. Region (1923), pl. XVIII, 4Google Scholar; Llyn Cerrig Bach (1946), p. 21, pl. vb; Brit. Mus. Early Iron Age Guide (1925), fig. 169. Cf. Lexden, (though not so close), Arch. lxxvi (1927), pls. LIX, LX.Google Scholar Stanfordbury is dated to c. A.D. 50 by Hawkes, and Dunning, , Arch. Journ. lxxxvii (1930), 261.Google Scholar

^{page 24 note 3} Piggott, S. and Daniel, G., Picture-Book Anc. Brit. Art (1951), nos. 51–52Google Scholar, with references

^{page 24 note 4} Cf. Hencken, H., Journ. Roy. Soc. Ant. Ireland, lxxxi (1951), 6.Google Scholar

^{page 24 note 5} Proc. Prehist. Soc. xvi (1950), 1–28; pl. 11, right.

^{page 25 note 1} Proc. Soc. Ant. Scot. VII (1866–7), 334; J-Anderson, , Scotland in Pagan Times: Iron Age (1883), p. 118Google Scholar, repeats Smith's illustrations and gives a précis of the account. The head was not mentioned by Leeds, E. T., Celtic Ornament (1933)Google Scholar, nor by Childe, V. G. in Prehist. Scotland (1935).Google Scholar The first published photograph was in Piggott and Daniel, , op. cit., no. 61.Google Scholar

^{page 25 note 2} The photographs of the Deskford head here reproduced (pls. VIII and IX) are by Malcolm Murray, Department of Prehistoric Archaeology, University of Edinburgh.

^{page 26 note 1} Cf. the Gundestrup carnyx representations (surely boars and not wolves, pace Allen, , Proc. Prehist. Soc. xxiv (1958), 45)Google Scholar, and, among other examples, the Hounslow figurines (Piggott, and Daniel, , op. cit., nos. 66, 67)Google Scholar, and also other representations noted below

^{page 28 note 1} Among the numerous representations of boars in Celtic art may be noted the ‘standards’ among trophy-groups such as that at Orange (pl. viia); the Neuvy-en-Sullias bronze (Varagnac, et al., L'Art gaulois (n.d.), Sculpture, pls. 47, 48)Google Scholar; the reliefcarving on the Euffigneix figure (ibid., pl. 7; Powell, , The Celts, pl. 67)Google Scholar; the little bronze from Bata, Hungary (Jacobsthal, , Early Celtic Art, no. 371)Google Scholar; the Gundestrup silver reliefs already noted; that in bronze on the Rynkeby cauldron (Klindt-Jensen, ), op. cit., p. 111Google Scholar, fig. 68b,; and that on a silver bowl from Lyon, (Amer. Journ. Arch, lv (1951), 47).Google Scholar The boar also figures on Gaulish coins of various tribes (Lengyel, L., L'Art gaulois dans les médailles (1954), pls. x, XXIIIGoogle Scholar, XL, XLI, XLIV; Fabre, G. in Varagnac, et al., op. cit., Les Monnaies, pl. 5, nos. 15, 21).Google Scholar In the British Isles there are the Hounslow figurines mentioned above; the outline of the (lost) figure on the shield, Witham (Horae Ferales (1863), pi. xiv, 1Google Scholar; Brit. Mus. Early Iron Age Guild (1925), p. 101, fig. 113); the unlocated Irish bronze figurine (Henry, F., Irish Art (1940), pl. 4Google Scholarc); and the Romanized bronze from the burial, Lexden (Arch. lxxvi (1927), pl. LVIIIGoogle Scholar, fig. 4). Coin representations include those of Tasciovanus, Cunobeline, Dubnovellaunus, and the Iceni (Mack, , Coinage Anc. Brit. (1953), nos. 164Google Scholar, 183, 220, 223, 243, 245, 291, 299, 405–11, 434). The presence of pig bones in quantity in La Tè;ne grave offerings in Britain and in central Europe may also be noted: Arch. lx (1906), 265; Ambros in Benedík, Vlček, and Ambros, , Kelt. Pohrebiská na Juhozápadnom Slovensku (1957), pp. 295–305Google Scholar (Czech and German text). Dr. Anne Ross, with whom I have discussed this subject, draws attention to the deity Moccus attested by the Langres inscription for Gaul, and the Celtiberian boar-figures in central Spain (Powell, , op. cit., p. 146)Google Scholar, as well as to the place played by pork in the traditional Old Irish literature for feasts both heroic and supernatural. She also notes the association of boars both in early Welsh and early Irish sources with warfare and destruction on the one hand, and with fertility and agricultural prosperity on the other

^{page 29 note 1} Powell, Bourray, op. cit., pl. 68Google Scholar; Tarbes, and Garancières, , Varagnac, et al., op. cit., Sculpture, pls. 2–5, 61Google Scholar, and col. pl. opp. p. 65; Charterhouse, Leeds, , Celtic Ornament (1933), fig. 29bGoogle Scholar; Marlborough, , Antiquity, v (1931), pl. 11Google Scholar opp. p. 42; cf. also Rynkeby and Sophienborgin, Denmark: Klindt-Jensen, , op. cit., figs. 68a, 69a.Google Scholar

^{page 29 note 2} The type was first defined by Smith, J. A. in Proc. Soc. Ant. Scot, XV (1880–1981), 316.Google ScholarAnderson, , op. cit., pp. 141–55Google Scholar, figs. 115–35, describes and illustrates the following armlets of this type: Aboyne, Aberdeenshire (two); Auchenbadie, Alvah, Banffshire; Bunrannoch, Perthshire (probably); Castle Newe, Strathdon, Aberdeenshire (two); Drumside, Belhelvie, Aberdeenshire; Pitkelloney, Muthill, Perthshire (two); Seafield, Kinghorn, Fife; Stanhope, Stobo, Peeblesshire; and a single North Irish find from Newry, Co. Down. The unlocated armlet illustrated by Anderson, fig. 126, can be identified with that from Stichill, Roxburghshire, illustrated in Pococke, , Tour in Scotland 1760 (Scottish History Soc. 1887), p. 331Google Scholar, and there is an unpublished armlet of this type from Achavrail, Rogart, Sutherland, in the Duke of Sutherland's Museum at Dunrobin.

^{page 30 note 1} Proc. Soc. Ant. Scot. vi (1864–6), 13; ibid. XV (1880–1), 330. For the coin, see ibid. lii (1917–18), 234, and for the dating of the Scottish souterrains, see Ant. Journ. xxxiii (1935), 65; Antiquity, xxvii (1953), 219.

^{page 30 note 2} Proc. Soc. Ant. Scot. xv (1880–1), 316.

^{page 30 note 3} Ibid. lxii (1927–8), 246.

^{page 30 note 4} Cf. Perkins, Ward in Proc. Prehist. Soc. v (1939), 183.Google Scholar The boss ornament on the Stanhope pieces is that of Gillam's Type B (‘petal-headed’) dress-fasteners, in Flavian contexts at Newstead, and generally of the late first to the early second century A.D. (Roman and Native in North Britain (1958), pp. 80–81).Google Scholar

^{page 31 note 1} Corstopitum Museum, unpublished, from Site XLI (Arch. Ael. ³ ix (1913), 262, pl. v), deep in a trial pit (p. 239), and so Flavian–Trajanic. Professor Ian Richmond, in giving me this information, adds: ‘the only other possibility is that it got there in A.D. 139 while Antonine levels were being laid out, but I am confident it cannot be later than A.D. 139’ (in litt. 11th July 1948).

^{page 31 note 2} Proc. Prehist. Soc. XX (1954), 27–86.

^{page 31 note 3} Proc. Royal Arch. Inst., York vol. (1846), p. 10; Brit. Mus. Early Iron Age Guide (1925), pp. 138–42, fig. 158; Later Prehist. Antiqs. Brit. Isles (1953), pl. xx, 4; for dating, Wheeler, , Stan-wick Fortifications (1954)Google Scholar, passim. The British Museum photograph of the masks is reproduced here (pl. XIa)

^{page 31 note 4} Journ. Rom. Studies, xliv (1954), 49, pl. II, 2. Professor Richmond has kindly allowed me to reproduce his hitherto unpublished photograph of the head here (pl. XIb).

^{page 31 note 5} Later Prehist. Antiqs. Brit. Isles (1953), pl. XXIII, 2; Piggott, and Daniel, , op. cit., nos. 48, 49.Google Scholar

^{page 32 note 1} Ulster Journ. Arch. xx (1957), 95–102.