Szabó Miklós, Petres F. Éva: Decorated weapons on the La Tene Iron Age in the Carpathian Basin. (Inventaria Praehistorica Hungariae 5; Budapest, 1992)

II. DRAGON-PAIR SWORDS

Fig. 18 La Tène, iron scabbard 4 with transitional J. V. S. Megaw has also called attention to finds from Slovakia (Dubnik: Cat.nos 83-88; Pis 8688).

162 The scabbard from Ameglia, grave 11 (Fig. 19) can be assigned to the beginning of the 3rd century B.C. on the basis of a cup from the "atelier des petites estampilles"

: this date is also supported by a kylix from the same burial which can be assigned to the Phantom Group.

163 This would suggest that De Navarro's Type I dragon-pair evolved from the zoomorphic lyre (Type II) in an Italo-Celtic context under strong impulses from the Mediterranean: the bronze openwork mounts of a helmet from Selinus definitely merits consideration in this respect.

164 However, this can by no means be considered a final solution to the problem since it is exactly the research into dragon-pair swords that best illustrates how new finds radically modify earlier views. The last word in this debate can only be expected after a systematic analysis of grave finds which might finally prove dragon-pair motif (after the original by M. Ecsedy) that dragon-pair swords reflecting a transition between Types II and I are definitely earlier than the canonic series of Type I scabbards and that the Italian finds precede all other swords in terms of absolute chronology.

165 Grave 102 from Dürrnberg which can be assigned to the Münsingen E-F/H horizons on the basis of its fibula, and which corresponds to the Swiss La Tène lb and early Ic in the traditional chronological framework must definitely be mentioned in this context. In other words, there exists the hypothetical possibility that the Type I dragon-pair sword belonging to this grave is earlier than the Italian specimen representing the transition between Types II and I.

166 The other difficulty encountered when trying to prove hypotheses of motif history is that non-canonic variants occur rather frequently among dragon-pairs and that owing to the "inter-Celtic currency" nature of the design it is not easy to locate workshops, and, conse-