^{page 397 note 1} Brit. Arch. Rep. 15 (1975), 125–9.

^{page 397 note 2} One in colour-coated ware from Catterick, North Yorkshire, was found near a temple (Britannia, iv (1973), 280Google Scholar); and there is another from Wilderspool, Cheshire (Thompson, F. H., Roman Cheshire (1965), p. 85Google Scholar, fig.21, no. 1).

^{page 397 note 3} e.g. Römisch-Germanischen Kommission, Germania Romana, v (1930), 7, and Taf. iii. 2 (from Cologne).

^{page 397 note 4} Mackreth, D. F., Roman Brooches (1973), pp. 12Google Scholar and 15.

^{page 397 note 5} e.g. Owmby, Lincolnshire (Lines. Hist.& Arch., i (1966), 44, fig. 4a), and one in Devizes Museum (Catalogue of Antiquities …, part ii, 1934, 213, and pl. lxvii, no. 1, reg. no. 333a).

^{page 397 note 6} For continental examples see Jahrbuch des Romisch-Germanischen Zentralmuseums Mainz, i (1954), 234–5, and Abb. 10.

^{page 401 note 1} Generally, North, F. J., The Evolution of the Bristol Channel (National Museum of Wales, 3rd edn., 1964), pp. 59–63Google Scholar.

^{page 401 note 2} Savory, H. N., Antiq. Journ. li (1971), 296–7CrossRefGoogle Scholar.

^{page 401 note 3} Idem, Bull. Bd. Celt. Stud, xxvi, 2 (1975), 247–9Google Scholar.

^{page 401 note 4} Cf. e.g., Hawkes, C. F. C. and Hull, M. R., Camulodunum (1947), pl. 98Google Scholar, 176; from Corbridge, Arch. Ael. ser. 3, v (1909), 406, fig.21.

^{page 401 note 5} S. Locke, Mon. Antiq. iii. 1 (1970–1), 1–16, refs.; but recent work shows it to be mostly of pre-Roman accumulation.

^{page 401 note 6} North, op. cit., pp. 78–82.

^{page 401 note 7} C. Fox, Arch. Camb. lxxxii (1927), 44–66.

^{page 402 note 1} Boon, Silchester, the Roman Town of Calleva (1974), p. 75, fig. 16, I.

^{page 402 note 2} Ibid, and idem, Isca (National Museum of Wales, 1972), p. 127, n. 104, fig. 18.

^{page 402 note 3} Chronology of barracks, ibid., pp. 116–17.

^{page 402 note 4} Fowler, E., Arch. Journ. cxx (1964), 120–2Google Scholar, 151–2, fig. 2.

^{page 402 note 5} Burley, E., Proc. Soc. Antiq. Scot. lxxxix (1958), 137–9Google Scholar, 169.

^{page 402 note 6} Curie, J., A Roman Frontier Post … (1911), p. 337Google Scholar, pl. 92, 11; drawn, Kilbride-Jones, H. E., Proc. Roy. Irish Acad. xliiic (1935–1937), 398Google Scholar, fig. 31, 1.

^{page 403 note 1} Type E, a good specimen (like an unpublished Segontium one) from the Rath of the Synods, Tara, soon to be published by Mr. S. Caulfield; Type F, Kilbride-Jones, loc. cit., 404, fig. 5; 406, fig. 6.

^{page 403 note 2} The supposed associations of a Type F brooch with a second-century trulleus and buckle at Long-faugh, Forth, may well be doubtful, as most commentators claim; but the Porth Dafarch, Holyhead, specimen (Kilbride-Jones, ibid., 446, fig. 26, 82) has more interesting associations than some have thought, though not second-century: see W. O Stanley, Arch. Journ. xxxiii (1876), 131 and the following illustrations of pottery, including not only samian of Antonine date but a black-burnished ware flanged dish of the full third century. The coarser of the two furrowed brooch-hoop fragments illustrated much resembles that of the brooch from the interior of a wall of Kingsweston villa, erected c. 270–300, Boon, Trans. Bristol & Glos. Arch. Soc. lxxxvi (1967), 195–6. The other appears to be part of a Type E brooch, but the terminal is not clear in the cut.

^{page 403 note 3} Kilbride-Jones, loc. cit., 404, fig. 5. 16. The same design occurs on the back of another F brooch from the Ferwerd terp in Friesland (see E. Fowler, loc. cit., 102, fig.3, 2), supposed to come from a level of the sixth to seventh century. The closest parallel to the pin, however, is the silver example from Halton Chesters Roman fort, much better made; D. Smith, Arch. Ael. ser. 4, xxxviii (1960), 231, pl. 11,1.

^{page 403 note 4} Boon, , Bull. Bd. Celt. Stud. xxii. 3 (1967), 302–10Google Scholar, xxvi. 2 (1975), 237–40 for Welsh hoards, plus a new one (March 1975) from Llanedeyrn, Cardiff (1093 radiates, Valerian-Tetricus Junior),

^{page 403 note 5} V. E. Nash-Williams, The Roman Frontier in Wales (ed. M. G. Jarrett, 1969), pp. 70–3.

^{page 404 note 1} Collingwood, R. G. in , R. E. M. and Wheeler, T. V., Lydney (1932), pp. 102–4Google Scholar.

^{page 404 note 2} A. M. ApSimon, Proc. Univ. Bristol Spelaeo-logical Soc. x. 3 (1965), 230; Branigan, K., Trans, Bristol & Glos. Arch. Soc. xci (1973), 120–1Google Scholar.

^{page 404 note 3} An alternative explanation associates the pins with the arrival of the Deisi in south-west Wales. This event used to be dated c. 270, which would have suited our dating of the Margam pins; but, as Alcock has shown (Alcock, L., Arthur’s Britain (1971), pp. 123–4Google Scholar), such a date is far too early, and the end of the fourth century much more likely. It would also have to be supposed that the Deisian vessel concerned was driven far off course, to the east rather than the west side of Gower, in fact.

^{page 404 note 4} R. Valenti, Il Museo Nazionale di Zara (Itinerari dei musei e monumenti d’Italia), pp. 21, 27.

^{page 404 note 5} I am indebted to Dr. A. Bruckner for a sketch of this unpublished vessel.

^{page 404 note 6} Lancel, S., Verrerie antique de Tipasa (1967Google Scholar), no. 198, p. ix, 3, from tomb 93, Porte de Cesaree.

^{page 405 note 1} Jahrbuch des Bernischen Museums in Bern, 45/6 (1965/6), pp. 401, 425, Abb. 20. Reference supplied by Dr. Bruckner.

^{page 405 note 2} Yadin, Y., The Finds from the Bar Kokhba Period in the Cave of Letters (1963), 106–10Google Scholar, No. 12, fig. 40, pl. 29.

^{page 405 note 3} Ibid., fig. 50a.

^{page 405 note 4} J. and Alarcao, A., Vidros romanos de Conimbriga (1965), nos. 71Google Scholar, 2, 3. 74 has the tip of the rim decorated but no facet cutting on the underside.

^{page 405 note 5} Harden, D. B. and Price, J. in Cunliffe, B., Excavations at Fishbourne. II. The Finds (Society of Antiquaries, 1971Google Scholar), p. 334, fig.138.

^{page 405 note 6} I. M. Stead, Excavations at Winterton (forth-coming).

^{page 406 note 1} Trans. Bristol y Glos. Arch. Soc. (forth-coming).

^{page 406 note 2} Oxoniensia i (1936), 63, fig. 12, 1.

^{page 406 note 3} J. P. Bushe-Fox, 2nd report on the excavations … at Wroxeter (Soc. of Antiquaries, 1914), p. 20, fig. 12.

^{page 406 note 4} E. J. Hollingworth and N. M. O’Reilly, The Anglo-Saxon cemeteryat Girton (1925), p. 32, pl. xii.

^{page 406 note 5} For interim reports, cf. Arch. Cant, lxxviii (1963), 125–41, and following volumes; for the general area of the findspot, cf. ibid, lxxxviii (1974), 120, fig. 1.

^{page 406 note 6} Richter, G. M. A., The Sculpture and Sculptors of the Greeks (1950), pp. 289–90Google Scholar, fig.749 on p. 607.

^{page 406 note 7} Toynbee, J. M. C., Art in Roman Britain (1962), p. 149Google Scholar, no. 53, pl. 56.

^{page 407 note 1} Jones, M. U., ‘Saxon Pottery from a Hut at Mucking, Essex’, Berichten van de Rijksdienst voor het Oudheidkundig Bodemonderzoek, 19 (1969), 145Google Scholar.

^{page 407 note 2} W. Holmqvist et al., Excavations at Helgo, IV: The Workshop, 1972.

^{page 407 note 3} Lamm, K., ‘The Manufacture of Jewellery during the Migration Period at Helgö in Sweden’, Bull. Hist. Metallurgy Group, 7, no. 2 (1973), 1–7Google Scholar.

^{page 407 note 4} Arrhenius, B., ‘East Scandinavian Style 1—a review’, Med. Arch, xvii (1973), 26–42CrossRefGoogle Scholar.

^{page 408 note 1} Lamm, op. cit., p. 3.

^{page 408 note 2} Wainwright, G. and Spratling, M., ‘The Iron Age Settlement of Gussage All Saints’, Antiquity, xlvii (1973), 109–30CrossRefGoogle Scholar.

^{page 408 note 3} Raftery, B., ‘Rathgall, Co. Wicklow: 1970 excavations’, Antiquity, xlv (1971), 296–8Google Scholar.

^{page 408 note 4} Such as two Late Bronze Age mould fragments for casting pins among unpublished Ayles-ford, Kent, material in the British Museum (identified by Dr. M. G. Spratling).

^{page 408 note 5} As suggested by Savory, H. N., ‘An Early Iron Age Metalworker’s Mould from Worm’s Head’, Arch. Camb. cxxiii (1974), 170–4Google Scholar.

^{page 408 note 6} Jones, M. U. and Jones, W. T., ‘An Early Saxon Landscape at Mucking, Essex’, Brit. Arch. Rep. 6 (1974), 23Google Scholar, and Antiq. Journ. liv (T974), 192.

^{page 409 note 1} D. W. Harding, The Iron Age in the Upper Thames Basin (1972), fig. 9, pl. 80a, from Han- borough, Oxon.; P. J. Woods, ‘Types of late Belgic and early Romano-British Pottery Kilns in the Nene Valley’, Britannia, v (1974), figs. 3, 5.

^{page 409 note 2} Worsfold, F. H., ‘An Early Iron Age Site at Borden’, Arch. Cant, lxi (1948), 148–55Google Scholar.

^{page 409 note 3} Op. cit., 152.

^{page 409 note 4} Hodges, H., Artifacts (1964), p. 27Google Scholar.

^{page 409 note 5} Fagg, W. and Picton, J., The Potter’s Art in Africa (British Museum, 1970), pp. 9Google Scholar, 12, 13.

^{page 409 note 6} Wright, R. P. and Hassall, M. W. C., Britannia, ii (1971), 295Google Scholar.

^{page 410 note 1} D. R. Wilson, ibid., 277.

^{page 410 note 2} In the 1971 report the weight is shown as 9.3375 kg. and is therefore slightly in error.

^{page 410 note 3} Current Archaeology, 43.

^{page 411 note 1} Musty, John, Antiq. Jourtt. xlix (1969), 98–117CrossRefGoogle Scholar.

^{page 411 note 2} C. Keepax, F.Archaeol. Sci ii (1975), 145–50.

^{page 411 note 3} Leeds, E. Thurlow, ‘A Saxon Village at Sutton Courtenay, Berkshire’ (Second Report), Archaeologta, Ixxvi (1926–1927Google Scholar), pl. vi, figs. 2, 10, p. 72.

^{page 412 note 1} A more recent comparison is offered by Mr. Ian Millar of Epping, who recalls similar pots used t o keep coffee pots warm in Alexandria, Egypt, in the 1930s.

^{page 412 note 2} Kindly lent by Miss Elisabeth Crowfoot.

^{page 412 note 3} Figs. 4 and 5. British Museum MS. Reg. 10 E. IV, thirteenth-fourteenth century.

^{page 412 note 4} A pair of woolcombs from the middle to late Saxon site at Wicken Bonhunt, Essex, in the Ancient Monuments Laboratory, Department of the Environment. Miss Evison has identified a fragmentary example from a Viking grave at Harold; Beds. Arch. J. v (1970), 39–42Google Scholar.

^{page 412 note 5} By Manning, W. H., ‘Caistor-by-Norwich and Notitia Dignitatum’, Antiquity, xl (1966), 60–2Google Scholar.

^{page 412 note 6} Although called ‘combs’ they more nearly resemble brushes, with several rows of long iron ‘bristles’. No doubt this explains their alternative name of hackles. There seems to be no difference between combs for wool and for flax. P. D. C. Brown has made this suggestion for iron points found in Saxon levels at Shakenoak; Brodribb, A. C. C. et al., Excavations at Shakenoak Farm, near Wilcote, Oxfordshire (1973), p. 134Google Scholar.

^{page 413 note 1} Myres, J. N. L., Anglo-Saxon Pottery and the Settlement of England (1969Google Scholar).

^{page 413 note 2} Ed. Steensberger, M., Vallhagar-a migration period settlement on Gotland, Sweden, ii (1955Google Scholar), figs. 460, 461, 479, and p. 1118.

^{page 413 note 3} A history of the ancient chapel of Birch in Manchester Parish’ in Chetham Society, xlvii (1859), 50; ‘Here too has found a resting place the general’s sword. Its blade is of bluish steel, straight and of considerable length. It is inlaid with gold and inscribed on either side with maxims, religious and moral: “Vincere aut mori”; “Si Deus pro nobis quis contra nos ?” Then follows the date 1652, and beneath the date a trooper on horseback. This again is followed by the words “Achilles Graecus” and below these words is a delineation of Achilles himself. The other side of the blade is similarly inscribed: “Fide sed cui vide”; “Regere seipsum summa sapientia”; trooper on horseback; the words “Anibal Cartagus” followed by a portraiture of Hannibal’. The sword was purchased for the Armouries from Howard Ricketts Ltd., of Bond Street, when it was stated to have come from the collection of Wing-Commander Carill-Worsley.

^{page 413 note 4} Details of Worsley’s career may be found in Booker, op. cit. and also D.N.B.

^{page 413 note 5} The scabbard was fitted with a hook which is missing, and the grip of the sword has been damaged and restored in resin,

^{page 414 note 1} See Laking, G. F., The Armoury of Windsor Castle (London, 1904Google Scholar), No. 57.

^{page 414 note 2} Blair, C., James A. de Rothschild Collection at Waddesdon Manor: Arms, Armour and Base-Metalwork (National Trust, 1974Google Scholar), No. 36.

^{page 414 note 3} Illustrated in Seitz, H., Blankwaffen, ii (Brunswick, 1968Google Scholar), fig. 78.

^{page 414 note 4} Illustrated in Laking, G. F., Record of European Arms and Armour, v (London, 1922Google Scholar), fig. 1500. I have been unable to trace the present whereabouts of this sword. The Martin Harvey collection was sold by Hampton and Son on 18–19 July 1938 and Lot 466 may have been this sword, but the catalogue description is unclear and inaccurate.