P1200415

Tłm fypołogy ha* betn deńaed mainly by rcferencc to pubfished materiał, and the fint drtaih of bit manufakturę could not be tonudered. In generał, the shanks of the A-Cbits atc ihoner and thicker ehan thoie of the 0-Fbits. And. taken in proport/on to the whote bit. the end-loops of the A-C* bito are generally larger than cm the D-F types. And bits of types D'Fulwa?* have a rein-ring m at least one of Chdr end-loops, whereas the other type* grnerafły do not.

Compare the bronie bone-bits associated wiih GundJJngen swords (type A : Lengenfeld f ig. 9.1 and Pfcdmłhce Fig. 9,2: type B. Platćnfce, Litomćricc Fig. 9,3, and Kissing Fig. 9,4) with the bronie horse-bitt found in wagon-gravcs (type I>. GroBcibstadt 1 Fig. 9.6. Kóngcn. Lhotlu. Tuchomćrice Fig. 9,7; type E: Pullach-Stid Fig. 9,8). The bito found with GOnćlmgen swords are of the sort with shorter and thicker shanks and larger end-loops; the bito found in wagon-grava are of the longer, thinner type with smal-ler end-loops and rein-ring*. Furthermore. GumJlingrn swords are only found with A and B type* whereas only Dand E types are found in wagon-gravcs.

Not wrprisingly. we find that when iron Mindelheim swords are associated with bronze horsc-bits. it is always with bits of the longer. thinner type with smaller end-loops and rein-ringi^M. Thus the same types of bronze bit* are found with Mindelheim swords and in wagon-gravcs (types 0 and E)^,5.

The map of Ha C bronze horsc-bits shows that the various bit types (A-F) are quite cvcnly spread within the bronze bit distribulion (Fig. 10). Only types E and Fhave a reutrteted disthbution. confined to Bayerisch-Schwaben or Bohemia. It seems. then, that the pattern detected arnong the assodations of the horse-bit types with Gundllngen swords. Mindelheim swords and wagon-graves cannot be exp!ained away as regional vanation. or the result of local workshop practices. On the contrary, it is morę likely that the types of bits associated with GOndlingen swords (A and B) were in use before the types associated with Mindelheim swords and wagon-graves (D and £).

In fact. bit types A to C are dosely related to lale Urnfield bit types whereas D, Eand F are not In particular. 0-F bits are always provided with a rein-ring in at least one of the end-loops, a practice unknown on Urnfield bits (Balkwill 1973). Furthermore, late Urnfield bits. Uke the type A. B and Obito, are generally of the kind with shorter, thicker shanks and larger end-loops. In contrast, the 0 and E bits are best compared to the iron bits of Ha C which, when well-preserved, often bear tight false twłsting, or have shanks with an angular cross-scction. have rein-rings, and conform to the type with longer, thinner shanks and smaller end-loops. These are the bits to be found in many Ha C wagon-graves. associated with the well-known series of rich bronze horse trappings which typify Kossack'* Ha Cl phasc.

Other element* of horse-gear have not often bcen found with GOndlingen swords. Apart from grave 5 at Plątamy. which has a yoke and rlch horse-gear. therc are phalerae from

Chavćria (umulus 16, Saulces-Champenoises and Kissing, and bronze side-pieces from Flatćnice grave 20, Pfedmćfice, Chaveria tumulus 16 and Lcngenfeld (the sword from Saulces-Champenoises is of iron. but seems related to the Gundlingen type). The grave from Litomćfice mcrcly has a smali iron loggie, an ‘omega-clip' and two iron rings (Zśpotocky 1964, 164. fig. 7,2-5).

Chavćria 16(Vuaillat 1977,99. fig. 63), Kissing (Czysz and Krahe 1980.36. fig. 11,7-8; 1983, 32, fig. 7,4) and Saulces-Champenoises (Flouest 1985) have similar bronze phale-rae, with a horizontal flange, Iow domed boss, and fiat central button or remnants of a central button. This form of phalera has bcen studied by G. von Merhart (1956), who concluded that they were developed from a late Urnfield type and continued in use during Ha C (for phalerae in Ha Cl contexts, see for example Bittelbrunn: Rest 1939, pl. 10, C; Mindelheim tumulus 7: Kossack 1959, pis 20, 5-10; 21). For this reason, the • typical phalerae of this group are not useful for fine chronology. But one of the Saulces-Champenoises phalerae has a form of knob which is typical for Ha Cl (Flouest 1985, fig. 4,8; compare Kossack 1954, 177, fig. 28, B13; 1959, pl. 60,17: Torbrfigge 1979, pl. 89. 10; Parę 1989, 89, fig. 9,4). And one of the phalerae from the Lcngenfeld grave (prob-ably associated with a Gundlingen sword) has a hemispherical rather than fiat knob, and a rib around the horizontal fiange (Torbrugge 1979, pl. 71,9). Both these features are unknown on Hallstatt examples, but very common on late Urnfield phalerae (von Merhart 1956, figs 3; 5-8). The same could bc said for the second pair of phalerae from Chavćria tumulus 16, bearing bossed decoration (Vuaillat 1977,99, fig. 63,2). This type of phalera is Ukewise unparalleled in the Hallstatt period, but belongs to the late Urn-field series of boss-decorated phalerae, best known from the lake-side settlements of Western Switzerland (von Merhart 1956; see for example p. 105, fig. 1; 106, fig. 2; and 107, fig. 3,4).

A Ha Cl datę for the Saulces-Champenoises grave is not only suggested by the phalerae, mentioned above, but also by remnants of a rich set of horse-gear: an openwork ‘Joch-schnalle', three hemispherical rein-knobs, numerous smali looped knobs (Flouest 1985). Only one other GOndlingen sword burial has rich Ha Cl horse-gear, Piańany grave 5 (Dvofśk 1933: rein-knobs with a central spike and loops around the rim, ring-footed hemispherical rein-knobs, toggles, smali hemispherical looped rein-knobs. all of bronze). Rich horse trappings are otberwise conspicuously lacking in all the other gra-ves with GOndlingen swords, but are commonly associated with iron or bronze Mindelheim swords, and are rightly regarded as charactericstk of Ha Cl.

Piańany and Saulces-Champenoises therefore indicate that Gundlingen swords were still used in Ha Cl. The fact that the Piańany sword was fitted with a chape of B type corresponds well with the late datę for the grave; as we expłained above, the B types came at the end of the chape development. One might think that B chapes are therefore

11