Alba Regia. Annales Musei Stephani Regis. – Alba Regia. Az István Király Múzeum Évkönyve. 25. – Szent István Király Múzeum közleményei: C sorozat (1995)

Tanulmányok – Abhandlungen - Christie, N.: The Survival of Roman Settlement alog the Middle Danube: Pannonia from The tenth Century A. D. p. 303–319. t. XX–XXIII.

exhibit a high Germanic (i.e. non-Avar) content, even if burial was in Avar guise (with weapon and horse burial, with stirrups, bows and nomadic belts for males; with beaded necklaces and globular earrings for females). Kömye yielded 152 graves covering the period 568-650 and included 20 horse graves: Germanic features comprised shields, lances, axes, short and long swords, and belt fittings; anfhropological study revealed no Mongolid component. Here then we can argue for a composite settlement including Gepids, Sarmatians and possible Longobards and natives (Salamon-Erdélyi 1971; Menghin 1985, 91-93). At Kölked-Feketekapu. two cemetery zones were excavated: one contained at least 681 graves of 6th-8th century date, and another, made up of a cluster of small graveyards plus scattered burials, contained 366 graves and was set within and around a village containing 140 house plans (Kiss 1979, 1988). The village was planned; houses, of semi-sunken form, were с 3x3 m in size, with stone hearths inside; external baking ovens were present, as well as corn pits, wells, and drainage channels. In the outer cemetery early finds showed weapon, dress and pottery of Germanic character which became 'Avarised' over time; in the village area Germanic and Avar components are roughly equal, but with Avar material dominant from с 680. Kiss has argued for a sizeable survival of Gepids in the Carpathian basin existing under Avar rule, and in fact they are still attested as an ethnic group in the later 9th century. Kiss idetifies a strong Gepid component in the Germanic finds at Kölked and a persistence and even development of Gepid style artifacts (e.g. stamped pottery, iron buckles, brooch decoration, cleavers) elsewhere in Hungary (1987). Although Gepid finds continue in Siebenbürgen (the eastern part of former Gepidid) there is a clear absence of finds between here and the Danube, their former heartland. Rather, we find 'Gepidised' cemeteries occuring in Transdanubia, primarily in eastern Valeria. The likelihood is that there was an enforced transplantation of Gepid-German groups across the Danube after 568 to provide a buffer zone to the Avar centre (Kiss 1979, 190; 1987, 208-209). Their military role may be borne out in their finds and perhaps by their location: Környe and Kölked both lie on Roman roads in proximity to late Roman fortifications (Kölked is 1,5 km from Altinum). Other early Avar period burials have similar associations: Pilismarót (Castra ad Herculem), Szob, Szentendre (Ulcisia Castra) and Dunaújváros (Intercisa). In most cases too there is some correlation with preceding Longobard cemeteries, implying a similar settlement role. And yet the evidence at Kölked shows an Avar period village distinct from Roman Altinum. So does this mean that the fort was redundant or used merely as a refuge? In the case of towns the evidence is also variable: Sopianae is ringed by three Avar cemeteries set 2-8 km distant, each overseeing roads and presumably also an extant native population (Fülep 1973, 315-320); at Gorsium graves are known near the palace complex (FlTZ 1980, 23); and at Fenékpuszta they gather alongside the horreiim. At the latter finds show a mixture of native, Germanic. Byzantine and Avar influences, again suggesting only nominal 'pure' Avar settlement in Pannónia ^Müller 1987, 270-272; contra Barkóczi 1968, 307-309; Bona 1987, 129 notes the 7th century royal centre at Zamárdi to the south). The continued usage of the church at Fenékpuszta and probably those at Sopianae can also be noted, with Bona suggesting settlement of 'Romans' from captured Byzantine towns like Singidunum (1987, 129). But against this can be set the lack of reference to bishops, even at Scarbantia, in Avar Pannónia after 579. Late Avar cemeteries extend across much of former Pannónia; the western fringes, however, plus Savia and much of former Noricum lay in Slavic hands. The border zones exhibit cemeteries with both inhumations (Avar rite) and cremations (Slav rite) (e.g. Zalakomár, Pókaszepetk west of lake Balaton; Záhorská-Bystrica, Bratislava-Vajnory north of the Danube near Carnuntum). There are no signs that such borders featured fortified watchposts or marked a reuse of Roman limes stations. The rise of the Carolingian Franks from 750 brought pressure to bear on the Avars: the buffer of Bavaria fell in 788 and in 791 Charlemagne began campaigns against the Avars, attacking along the Danube and from NE Italy. Victory came in 803 with the surrender of the Avar Zodan, princeps Pannóniáé. Subsequently, despite insurrections, Carolingian rule extended over much of western Pannónia: Pannónia Inferior, south of the Drau, was held by the Duke of Friuli and overseen by the patriarch of Aquileia; Pannónia Superior from the Danube to the Drau was under the Duke of the East and the archbishop of Salzburg. The eastern border seems to have extended only so far as Carnuntum and Savaria in the north and Sopianae and Siscia in the south (Bona 1985, 151-153). East of these lay Avar vassal princedoms. The Bulgár invasion of SE Pannónia in 826-829 upset the balance, pushing remaining Avars north into Valeria and forcing the Carolingians to divide their lands into smaller more military units. The document known as the Conversio Bagoariorum et Carantanorum of 870, which provides rare details on the character of Carolingian Pannónia, omits these Avars; it does, however, refer to the towns of Savaria, Bettobia (Poetovio?) Mosapurc (Mosaburg = Zalavár, west of Fenékpuszta) and Quinque Basilicae (Sopianae-Pécs), each presumably with a set of Carolingian functionaries and a church community (Bona 1985, 149-150). Sopianae's name change is particularly interesting as it probably refers to the survival of five early Christian cult buildings (Fülep 1973, 321-324). Archaeologically virtually nothing is known of this carolingian interlude bar the signs of renewed activity at Fenékpuszta and the distribution of Carolingian type lances (Kovács 1978-79). It is doubtful that the 311