^{page 262 note 1} For an Anatolian origin—a view he did not, however, long sustain: Childe, , PPS, vol. 14, 1948, 183–5Google Scholar. For an Aegean origin: Catling, , PPS, vol. 22, 1956, 102–25Google Scholar; later withdrawn, Antiquity, vol. 35, 1961, 118–22Google Scholar. ‘Mycenaean influences’ on the Boiu type: Nestor, Boiu, 1937, passim; Werner, , Congresso Firenze, 1950, 293–7Google Scholar. For a central European origin: Cowen, V CISPP, Hamburg, 1958, 207–14; Foltiny, , Arch. Aust., vol. 29, 1961, 84, 88–9Google Scholar, and several subsequent articles. Partly traversed by Müller-Karpe, , Germania, vol. 40, 1962, 259–69Google Scholar; followed by Snodgrass, , Early Greek Armour and Weapons, 1964, 205–8Google Scholar. The view long held in Germany and expressed by many writers, though in general rather than specific terms, has been for a ‘northern’ origin— northern, that is, in relation to Greece and the Aegean area.

^{page 262 note 2} A summary account and lists of both types are given in Appendix B.

^{page 263 note 1} 40 BRGK, 1959, 1–78Google Scholar; in particular, 10–14.

^{page 263 note 2} Acta Arch., XXX, 1959, 126–31, Abb. 36Google Scholar.

^{page 263 note 3} Willvonseder, , MBÖ, 1937.Google Scholar The work has been criticized for lack of rigour in method, but the results seem to have stood the test of time; except in his subdivision of the disc-headed pins, and his allocation of certain varieties to Br Ba (Torbrügge, 40 BRGK, 1959, n and 36, with refs.).

^{page 263 note 4} Lomborg, loc. cit., 131–3, Abb. 36.

^{page 264 note 1} The strongest evidence comes from Moravia: Spurný, , Pant. Arch., XLV, 1954, 357–77Google Scholar; idem., Chron. préh. de la Tchécoslovaquie, 1956, 103–4. Cf. also Holste, , Die BZ in Süd- und Westdeutschland, 1953, 30Google Scholar; Čujanová-Jílková, , Pant. Arch., LV, 1964, 1–81Google Scholar. Brief notice in English: E. and Neustupný, J., Czechoslovakia, London, 1961, 109Google Scholar.

^{page 264 note 2} e.g. Foltiny, , Régészeti Füzetek, vol. 4, 1957, 1–62Google Scholar. Useful summary with map by Kőszegi, AAH, vol. 12, 1960, 137–45Google Scholar, Abb. 2.

^{page 264 note 3} Mozsolics, , AAH, vol. 8, 1957, 142–6Google Scholar. On this point Bóna too seems in general agreement, AAH, vol. 9, 1958, 211–43Google Scholar.

^{page 264 note 4} Kőszegi, loc. cit., 137; Willvonseder, MBÖ, S.V. Keszthely.

^{page 264 note 5} Germania, vol. 37, 1959, 71–5Google Scholar.

^{page 264 note 6} The maximum variation accepted has been taken as ±2½ per cent, but the majority fall within much narrower limits.

^{page 265 note 1} AÉ, XIX, 1899, 242, 244, pl. v, 6.

^{page 265 note 2} Peake, The Bronze Age and the Celtic World, pl. vii.

^{page 265 note 3} Montelius, , Civ. Prim. It., 1895, 186, pl. 34, 20Google Scholar.

^{page 265 note 4} He nowhere refers to any of Reinecke's publications; nor does he reproduce the Keszthely sword, possibly because he wished to illustrate only pieces with the double spiral, or spectacles, ornament.

^{page 265 note 5} Childe, Danube, pl. ii, and passim; notably p. 399, notes 1–3.

^{page 265 note 6} Germania, vol. 15, 1931, 217–21Google Scholar.

^{page 265 note 7} The group of material from Nagy-Gáj illustrated in AÉ, XIX, 1899, 413, fig. 5, is not an associated group, but a collection of finds from a given area. It has, therefore, no chronological value. Nor, in fact, do either of the weapons noted there (any more than that from Sombor) have more than an indirect connection with the Sauerbrunn or Boiu types.

^{page 266 note 1} Sargetia, I, Deva, 1937, 155–214; here otherwise referred to sımply as Boiu.

^{page 266 note 2} Willvonseder, , MBÖ, 1937, 89–92Google Scholar, especially p. 90. The two authors were in touch during the course of their work, as Nestor expressly acknowledged (Boiu, notes 9, 27, 29–31). He also used two of Willvonseder's drawings to illustrate his own article (ibid., figs. 6 and 7). Willvonseder, on the other hand, who had in his own area no swords ornamented with the spectacles motif to consider, continued to treat the Sauerbrunn-Boiu material as an undifferentiated mass.

^{page 266 note 3} These datings will be more closely examined below.

^{page 266 note 4} Holste, , NMHessen, 29–30Google Scholar; Müller-Karpe, , HessF, 24–8Google Scholar, Abb. 14, 1. The observations by Werner, in Atti. Congresso Firenze, 1950, 293–5, are too generalized for the present purpose; see also below, p. 287, note 1Google Scholar.

^{page 266 note 5} Sprockhoff, , Offa, vol. 9, 1951, 25–6Google Scholar, Abb. 2, and Tf. 3, 5.

^{page 266 note 6} Holste, , VGS Bayerns, 1953, 5, Tf. 15, 6Google Scholar; Appendix 2, 44–5; Map, Tf. 17, 2.

^{page 266 note 7} Cowen, , Germania, vol. 32, 1954, 225Google Scholar.

^{page 266 note 8} Arch. Aust., vol. 29, 1961, 76–95Google Scholar, Abb. 1–4; AJA, vol. 68, 1964, 247–57, pls. 73-6CrossRefGoogle Scholar.

^{page 267 note 1} MAGW, XCI, 1961, 136–40Google Scholar, Tf. 7, 1; Sbornlk Brno, III, 1963–4, 269-75, Tf. lii, 1–3; Arch. Aust., vol. 38, 1965, 31–30Google Scholar, Abb. 1–3. The last article contains Foltiny's latest lists of the Sauerbrunn and Boiu swords, correcting many entries in his earlier versions; and is the first in which the real distinction between the two types is consistently applied. Though we have often exchanged information on individual pieces, it may be worth saying that our respective lists have been compiled independently. It is the more satisfactory to note how close they now are in content; how far they differ in articulation will be apparent from a glance at Schedule I.

^{page 267 note 2} As observed by Holste, , NMHessen, 29 and note 4Google Scholar.

^{page 267 note 3} Nestor, , Boiu, passim., esp. pp. 175–6, 189–90Google Scholar; Sauerbrunn tangs: 184, 200.

^{page 268 note 1} Boiu, 189.

^{page 268 note 2} ibid., 195.

^{page 268 note 3} e.g. Foltiny, , Arch. Aust., vol. 38, 1965, 29Google Scholar.

^{page 268 note 4} Following on correspondence with the present writer (as he has kindly acknowledged) Foltiny some while ago accepted the tang as the criterion of distinction between the Sauerbrunn and Boiu types. This was a step in the right direction; but the material demands an altogether more radical re-examination.

^{page 268 note 5} Using the decoration as a criterion will give no more certain results where, as so often, the engraving is badly worn.

^{page 270 note 1} Willvonseder, , MBÖ, 90Google Scholar.

^{page 270 note 2} Lomborg, , Acta Arch., XXX, 1959, 126Google Scholar.

^{page 270 note 3} Willvonseder, ibid., 359–60.

^{page 270 note 4} ibid., 90, 392–3, Tf. 28, 5–7.

^{page 270 note 5} Hereafter referred to (following Nestor) as Keszthely I. The sword illustrated in Hampel II, CXXXIV, 3 (pl. XIX, 5), to which we shall come later, is for the same reason called Keszthely II. See below, p. 285.

^{page 270 note 6} Ziegert, , West HGK, 1963, 18, 29Google Scholar, Tf. 10, Ztgr. 4.

^{page 270 note 7} Peroni, , Mem. Mus. Verona, XI, 1963, 49–77, 102–4Google Scholar. The association of greatest interest was that in Peroni's Grave 4, where the undecorated sword of Boiu type was recovered with a typical example of the Terremare type. The possible establishment of a clear-cut position for the latter within the Terremare Culture could have valuable consequences. The dagger of Sauerbrunn type came very probably from Grave 7, but without known associations.

^{page 270 note 8} Holste, , VGS Bayerns, 13–15Google Scholar, placed it in Br B1. Hachmann, , Ostsee, 136–40Google Scholar, in a long passage decides for Br B2. On general grounds I prefer the latter alternative. As the blade is a Sauerbrunn blade it does not much matter whether the Spatzenhausen hilt is dated Bi or B2; both dates are equally acceptable for the Sauerbrunn type. The Cascina Ranza hoard, which contains another hilt of the Spatzenhausen type, hardly concerns us (Hachmann, ibid., Tf. 59). The multiple S-shaped lines on the blades of the two daggers with solid-cast hilts are indeed probably derived from the blade-formation of the Sauerbrunn-Boiu group, but such reminiscences turn up in other contexts down to much later dates, e.g. on a dagger-blade in the Tenja hoard (Yugoslavia), probably of Ha A1 (but containing earlier material), and have no chronological value.

^{page 271 note 1} Uenze, , Vollgriffdolche, 25Google Scholar, lists twenty-seven examples from nine sites, and illustrates them all. Of these all but one are of the Italian type, and all but three come from Italian sites. Cf. Hachmann, Ostsee, Tf. 57, 11; 58, 2 and 6; Karte 17. The dog's-teeth on the Pichlern blade are omitted in Willvonseder's figure, MBÖ, Tf. 51, 3 as noted by Müller-Karpe, , HessF, 27, n. 19Google Scholar; and Holste drew them, SkB 9, 9.

^{page 271 note 2} One from Croatia, and one from the central Rhineland.

^{page 272 note 1} Boiu, 192–4.

^{page 272 note 2} Loc. cit., 192; quoting Montelius, , Chronologie, 1900, 128Google Scholar, and Forssander, , Ostskanditiavische Norden, 1936, chap. III, especially p. 184Google Scholar.

^{page 272 note 3} None of which is more than 350 millimetres long (including the hilt); Hachmann, Ostsee, Tf. 57, 1–3.

^{page 272 note 4} Both under 500 millimetres; ibid., Tf. 29, 10; 60, 3.

^{page 272 note 5} ibid., Tf. 24, 1; cf. 25, 1 and 3; 60, 2; 63, 2.

^{page 272 note 6} Other series can be picked out of Hachmann's wealth of figures. I have quoted only the best documented, simply to illustrate a principle.

^{page 273 note 1} (1) Battaglia, , BPI, 67–8, 1958–1959, 316, fig. 112Google Scholar. (2) Colini, , BPI, 29, 1903, 68Google Scholar, tav. VI, 8; Montelius, VCI, Tf. viii, 9. (3) Behn, Italische Altertümer vorhellenistischer Zeit, Kat. RGZM, No. 8, 1920, 27, Abb. 4. The two latter quoted by Nestor (Boiu, 195)—but with emphasis rather on their ornament than their form.

^{page 273 note 2} Boiu, 194–5.

^{page 274 note 1} Boiu, 194.

^{page 274 note 2} Holste, VGS Bayerns, 46, List 4, no. 3, Tf. 1, 3; Nestor, ibid., pl. i, 4.

^{page 274 note 3} Schmid, , Carniola, vol. 2, 1909, 126, Tf. 4, 5 and Abb. 16Google Scholar.

^{page 274 note 4} Both are conveniently illustrated on the same plate as the Kelibia hoard by Hachmann, Ostsee, Tf. 62.

^{page 274 note 5} Boiu, 194.

^{page 274 note 6} Mutatis mutandis the same could be said of other searches for origins throughout the whole field of prehistory!

^{page 274 note 7} The whole sequence fully documented above, p. 272.

^{page 274 note 8} e.g. Hachmann, Ostsee, passim. The same formation was adopted, but in relief, on the dagger-blades of the Kelibia type, including the ornamented form, as at Barca (L. Hajek in Kommission Nitra, 1958, 72, Abb. 12), and Nagy-Gáj (Hachmann, loc. cit., Tf. 62, 10).

^{page 275 note 1} We have already noticed the significance of a special variety of the motif for early connections in the case of the Sauerbrunn blade from Pichlern—supra, p. 271.

^{page 275 note 2} Keszthely I, Povegliano, Leobersdorf, Sauerbrunn; fig. 3, 1–4.

^{page 275 note 3} Pichlern, fig. 3, 5; Gospić.

^{page 275 note 4} From ‘Hungary’, face A; pl. XVIII, 4; fig. 3, 6.

^{page 275 note 5} Hochstadt, and ‘Hungary’ (supra.), face B; fig. 3, 7–9.

^{page 275 note 6} For Zell, see Hachmann, Ostsee, Tf. 57, 8; for Studenec, ibid., Tf. 62, 8. Further afield the same feature is to be seen on a sword of the Apa type from Rosenfelde (Pomerania), ibid., Tf. 24, 3; and on blades within the Sőgel ambience: Gokels, and Fur sgn., ibid., Tf. 9, 15, and 10, 9.

^{page 275 note 7} Battaglia, BPI, 67–8, 1958–9, 304, fig. 109b. Multiple-band frames are found also on many swords of the Apa and related types–Hachmann, ibid., Tf. 18, 13; 19, 3 and 16; 24, 1 and 2; 25, 1. All but two of all pieces quoted in this and the preceding note show also the shaded triangles, pendent from a horizontal bar below the butt ornament, which are found on all the major groups of triangular daggers with bronze hilts.

^{page 277 note 1} Cf. Müller-Karpe, , HessF, 26Google Scholar.

^{page 277 note 2} Boiu, 193–4.

^{page 277 note 3} Hachmann, , Ostsee, Tf. 58, 7 and 8Google Scholar.

^{page 277 note 4} The unlocalized example from ‘Hungary’ (Uenze, Vollgriffdolche, Tf. 51, 126) is probably Middle Bronze Age in date. The closely related round-butted dagger from Studenec in Slovenia probably (pace Hachmann, Ostsee, Tf. 2–3, and 62, 8) of the earliest Middle Bronze Age, also comes from an obviously significant area. Cf. Nestor, Boiu, 196.

^{page 278 note 1} Foltiny, in Arch. Aust, vol. 29, 1961, 82Google Scholar; AJA, vol. 68, 1964, 249Google Scholar; Sborník Brno, III, 1963–1964, 270Google Scholar; Arch. Aust., vol. 38, 1965, 26–30Google Scholar.

^{page 279 note 1} As noted long since by Sprockhoff, , Griffzungenschwerter, 7Google Scholar.

^{page 279 note 2} To describe these swords, therefore, as ‘cutting swords’ (Foltiny, , AJA, vol. 68, 1964, 250–1Google Scholar) is to mis-understand their true character.

^{page 279 note 3} This feature cannot be detected on one of the swords from Casier (Mus. Treviso, no. 78; Foltiny, , Arch. Aust., vol. 29, 1961, Abb. 1, 7Google Scholar); but as the blade is at present covered by a layer of sand corroded onto the surface that is hardly surprising.

^{page 280 note 1} It is difficult to be certain that these elements are in fact in all cases spirals, and are not sometimes composed of concentric circles (Samşud). It seems, however, clear enough that the spiral is the normal form. In some cases the spirals are wound in opposite directions (Császártőltés, Kojátky).

^{page 280 note 2} The design here seems oddly to anticipate certain elements in Celtic art.

^{paeg 281 note 1} The formal classification here of the two short-swords with extremely advanced designs from Samsud and Casier, in Class la and Ib respectively, may be misleading. So placed they certainly spoil an otherwise consistent pattern. As regards the Samsud piece in particular the original form of the tang may well, as Nestor has suggested, be irrecoverable. If so it should be rated as a Stump, without significance for a refined typology. The undoubtedly short tang of the Casier dagger, on the other hand, may be no more than a ‘throw-back’. The unique (and insensitive) design on the peripheral sword from Fahrenkrug must, if the last published drawing is to be relied on, be a provincialism.

^{page 281 note 2} Most recently and best in Arch. Aust., vol. 38, 1965, 26–8Google Scholar. But his List (a), of swords of his ‘Boiu-Keszthely type without flanges,’ still covers a number of varieties.

^{paeg 282 note 1} Boiu, 166–7, 170, 188–9.

^{page 283 note 1} Without provenance; formerly said to be ‘from Taranto’.

^{page 283 note 2} Martfű, pl. XIX, 3; fig. 5, 5. Bratislava, pl. XIX, 4. ‘Ashmolean’, pl. XIX, 2; fig. 5, 3.

^{page 283 note 3} The tangs of Bratislava and Donawitz are broken; but I am satisfied that on Bratislava an intact portion of the original upper edge remains. And though I have not handled Donawitz I have seen it, and reached the same conclusion.

^{paeg 284 note 1} It does, however, occur on several pieces with broken tangs of undeterminable type (‘Stumps’) any of which may originally have been of the classic form.

^{page 285 note 1} Hampel II, CXXXIV, 3 and 4.

^{page 285 note 2} Willvonseder, , MBÖ, 113, 116Google Scholar; Torbrügge, 40 BRGK, 1960, 38–40, Abb. 11 and 13Google Scholar.

^{page 285 note 3} Willvonseder, ibid., 90, 392–3, Tf. 28, 5–7.

^{page 285 note 4} That is the dating proposed by Willvonseder himself, though he does not give his grounds for so doing. ibid., 90.

^{page 285 note 5} Offa, vol. 9, 1951, 25–6Google Scholar.

^{page 285 note 6} I must own to having, on a previous occasion, overvalued both the weight of the evidence, and the degree of refinement in dating which it could be expected to yield. V CISPP, Hamburg, 1958, 208–9Google Scholar.

^{page 285 note 7} Peroni, , Mem. Mus. Verona, XI, 1963, Tav. 1, 5Google Scholar.

^{page 285 note 8} ibid., Tav. 1, 4.

^{page 286 note 1} Cowen, VCISPP, Hamburg, 1958, 209–11, figs. 2 and 3Google Scholar.

^{page 286 note 2} Mozsolics, , AAH, vol. 12, 1960, 113–23, Tf. lxix–lxxGoogle Scholar.

^{page 286 note 3} Compare Holste, VGS Bayerns, Tf. 6–7 with Tf. 9–12.

^{page 287 note 1} This dating, it must be said, does not agree with that adopted by Hachmann, who assigns the type to his import-stream of Horizon III, dated to Br B1 (Ostsee, 154). For this he relies, so far as I can see, on Werner's treatment in Atti CongressoFirenze, 1950, 293 ff.Google Scholar, which placed the type ‘close to’ the Mycenaean rapiers (Ostsee, 169). But the relevant section of Werner's study is, as we have noted (above p. 266), expressed in very general terms; several varieties of rapier are handled without differentiation as a single phenomenon, while the Leobersdorf and Keszthely graves are treated both as contemporary, and as containing the same type of sword. Werner's Abb. 1, comparing the spiral decoration on the butt of a Mycenaean rapier of the shaft-grave period with that on the Boiu sword, is quite insufficient to prove them contemporary without much stronger evidence; it does not even prove a relationship. Rejection of a direct link between the multiple running-spiral ornament on the butt of a single Mycenaean rapier, and the twin-spirals on the butts of Boiu swords, does not automatically involve us in accepting the element of ‘chance’ for an explanation (Werner, loc. cit., 294). A third alternative is possible, and is that which is here preferred. Nevertheless succeeding writers seem to have accepted this most slender of evidence as establishing the Boiu type in Br B1, with far-reaching consequences for the chronology of the Middle Bronze Age. So, besides Hachmann, Milojčič, Germania, vol. 37, 1959, 74, 78Google Scholar. See also Pittioni in Pauly-Wissowa, , RE d. class, Wissenschaft, Supplementband, IX, 1962, art. Italien, cols. 222–3Google Scholar.

^{page 287 note 2} Except for one stray in Germany, and the Povegliano cemetery, uninformative in the present context.

^{page 287 note 3} Some of those found actually in the rivers may well have been votive deposits; but by no means necessarily all, as Foltiny implies (Sborník Brno, III, 1963–1964, 272)Google Scholar.

^{page 288 note 1} The earlier view that the Boiu type was, somehow or other, at least in part a product of Transylvania is now, following the discovery of so much new material, out of the question. On the other hand, in the present state of knowledge Foltiny's attempt to identify the position of individual workshops in northern Italy (‘neighbourhood of Udine’, and ‘neighbourhood of Treviso’, Arch. Aust. vol. 29, 1961, 89Google Scholar) is frankly premature.

^{page 289 note 1} Foltiny, , Arch. Aust., vol. 29, 1961, 86–8Google Scholar; idem, MAGW, XCI, 1961, 137–40; idem, AJA, vol. 68, 1964, 251.

^{page 289 note 2} References as in previous note.

^{page 290 note 1} This conclusion is in fact the first of the two alternatives advanced by Nestor as possible explanations of the origin of the Boiu type. The second was ‘that it was the result of a process of evolution parallel to that leading to the creation of the Sauerbrunn type, and which, starting from the same traditions, had as its immediate local predecessors blades of the Nagy-Gáj type’ (Boiu, 197). Though he did not express formal acceptance of either, he clearly favoured the second alternative: partly because the loose running spirals on the Nagy-Gaj blades were the closest analogy he could find in central Europe to the twin-spirals on the Boiu swords; and partly because in his Povegliano II–Joschewa group he saw the first tanged swords in extra-Aegean Europe, the tangs deriving ultimately from the Mycenaean area, and being transmitted on the swords of that group to central Europe via the Adriatic route to be adopted on the Boiu type (ibid., 196–8, 202–5). But we have already seen reason to discount the significance of the Nagy-Gáj blades (above, p. 274); and we shall soon have reason to see that the Povegliano-Joschewa group is not itself a viable entity (below, p. 296, n. 3). We have also seen that the first dated appearance of the Sauerbrunn type in Br B1 precedes that of the Boiu type (and in an early ‘transitional’ form at that) in Br B2. All of which runs counter to the acceptance of Nestor's second alternative.

^{page 291 note 1} Boiu, 200. In consequence he was forced to seek a model elsewhere, and had to fall back on the expedient of creating his ‘Povegliano II–Joschewa Group’. See below p. 295.

^{page 291 note 2} Evans, Bronze Implements, fig. 318, from Lissane, 770 millimetres long.

^{page 292 note 1} Very nearly the same form, interrupted only at the extreme tip, appears also on the Lavrica dagger, but there with a blank centre (fig. 2). A frame of this shape is indeed a natural infilling of the space available. That the form found at Hochstadt was not repeated can only be due to the persistence of the tradition, certainly derived from the solid-hilted triangular daggers, of a horizontal line dividing the decoration of the butt from that of the blade.

^{page 294 note 1} Boiu, 197.

^{page 294 note 2} Kommission Nitra, 1958, 72, Abb. 12; a closely related piece from Hernadkak (Hungary), ibid., Abb. 13, 2. To these and to the two known to Nestor, from Nagy-Gáj and ‘Hungary’, conveniently illustrated by Hachmann (Ostsee, Tf. 62, 10 and 11), add now a closely similar fragment (butt with five rivet-holes, and upper blade) in the museum at Arad. Though the provenance is unknown it must come from the area of eastern Hungary or western Transylvania. I am most grateful to Prof. M. Petrescu-Dîmboviţa for information and a drawing of this piece.

^{page 294 note 3} Boiu, 198–9.

^{page 294 note 4} The provision of a pair of eyes on the prow of a boat, with precisely the same object, is too widespread in time and space (including the bronze age in Britain) to call for further comment.

^{page 295 note 1} Bulletin Univ. of London Inst. of Archaeology, no. 4, 1964, 177–85Google Scholar.

^{page 295 note 2} Germania, vol. 15, 1931, 217–19Google Scholar.

^{page 295 note 3} Boiu, 199–305.

^{page 295 note 4} See p. 296, note 3. Note also that Nestor's account involves placing the emergence of this so-called group before the emergence of the Boiu type. For such a sequence there is not the slightest evidence. Further, the elimination of several components of the group, and the addition of others more closely related to the Sombor and Smolenice swords, significantly changes the pattern of distribution from a supposedly ‘Adriatic’ one, to one that is frankly central European. These factors seem to dispose of Nestor's preferred, if tentative, derivation of the tang from Mycenaean sources transmitted into central Europe by the Adriatic route.

^{page 296 note 1} Sprockhoff's Types Ia and Ib.

^{page 296 note 2} Boiu, 202.

^{page 296 note 3} A number of the pieces included by Reinecke, and particularly by Nestor, do not appear to belong here under any circumstances. The short dagger (only 307 millimetres long) from Joseva, in western Serbia, shares indeed one of the characteristics of the Smolenice sword (the rounded central rib of the blade); but the form of the tang, the rectilinear splay of the butt, and the rivet-formula, all seem to prohibit so early a date, which is certainly not imposed by the associated urn. (M. Garasanin, 39 BRGK, 1958, 101–2Google Scholar, Abb. 22, with full refs). The sword designated Nagy-Gáj II (now Veliki Gaj), from the Yugoslav Banat, is a unique piece without associations; but the blade-section should not be earlier than Br D, and the engraved ornament opposite the ricasso is otherwise unparalleled before Ha A. The sword from Bela Crkva (Fehertemplom, Weisskirchen), in the same area, is perhaps a hybrid piece, but by definition must be included within the Boiu group; it is in any case an unassociated find (see above, p. 284). And such of the swords from Manaccore as can still be seen are so heavily restored as to impair any evidential value they may once have had.

^{page 297 note 1} For a normal trapeze-shaped butt associated with a pommel of the Au-type see Holste, VGS Bayerns, Tf. 15, 2, from Sopron.

^{page 297 note 2} The edges are a good deal decayed, but the original shape is assured by the form of the raised central portion of the blade.

^{page 297 note 3} Another piece, from Čičov, Slovakia, to which Willvonseder (MBÖ, 82) has drawn attention as a parallel, is not of this kind at all, but a cut-down fragment of a sword of the Hallstatt type.

^{page 297 note 4} The figure in AÉ, vol. 28, 1908, 263, fig. 3, exaggerates both the width and the degree of swelling in the upper part of the blade.

^{page 297 note 5} AÉ, loc. cit., figs. 1–16.

^{page 297 note 6} Such as we shall see again in the Smolenice grave.

^{page 297 note 7} Contrary to the impression given by a former arrangement in the Subotica museum, it appears that this piece is in fact an isolated find.

^{page 298 note 1} Information kindly supplied in correspondence. Dr Gabrovec will shortly be publishing this grave-find.

^{page 298 note 2} Class Ia: ‘Hungary’ (Hampel I, XX, 10). Class Ib: Martfű, Bratislava, and ‘Provenance Unknown’ (Ashmolean Museum).

^{page 298 note 3} MBÖ, 88–9, 419, Tf. 34, 2. Willvonseder rightly saw that the piece must once have had a tang. It is no less evident from his plate, where the beginnings of flanges are clear to see on each side of the remains of the butt, that the tang also had flanges.

^{page 298 note 4} Through relying on a faulty drawing I myself once erroneously assigned it to Type IIa! (36 BRGK, 1956, 124Google Scholar).

^{page 298 note 5} Both types occur also in the Sombor assemblage; and the seal-head pin again in Keszthely II. The Sombor and Smolenice swords were both dated Br B2 some years ago by Willvonseder (MBÖ, 89); these dates still stand, and have recently been kindly confirmed for me from the Hungarian side by Dr A. Mozsolics.

^{page 300 note 1} Holste, , VGS Bayerns, 18–21, 35–9Google Scholar. That the apparently earlier general emergence of large cutting blades within the tradition of the solid-hilted swords may have influenced the course of development of the flangehilted family is undeniable. But in view of the evidence here marshalled it appears most unlikely that the flangehilted family could have been an offshoot of the solid-hilted; it seems to have its own separate roots. The influence therefore, if present, must have been indirect.

^{page 300 note 2} Sold to Berlin from the Julius Naue collection in 1904, it was believed to have been found at Mantlach, Oberpfalz; but Torbrügge has shown the attribution to be valueless. It does, however, probably come from southern Germany. Torbrügge, , BZ Oberpfalz, 227–8Google Scholar, Abb. 17, 1.

^{page 300 note 3} It figured, with Smolenice, in the lists of both Reinecke and Nestor; and Nestor called it Povegliano II. Both, however, included it in this context mainly on the basis of a chronological assumption that has not stood the test of time. See above, p. 270. We retain it here on purely morphological grounds.

^{page 310 note 1} Karo, , Die Schachtgräber von Mykenai, 1930–1933, 200–6Google Scholar; Sandars, , AJA, vol. 65, 1961, 17–22, 27CrossRefGoogle Scholar.

^{page 311 note 1} Related pieces are known also from northern Greece (Perámatos, Grevena). Best account in Sandars, , AJA, vol. 67, 1963, 119–26, 145–6; pls. 22, 28CrossRefGoogle Scholar.