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.

in the amphitheatre; at Gorsium graves lay close to the larger basilica (Data - mostly unpublished - summarised in Kiss 1981, 177-178; Bona 1987, 120-121; cf. Barkóczi-Salamon 1984, 180-184; and Fitz 1960, 158160 and 1980, 23 on Gorsium). Of actual 'Germanic' settlement traces in this period we can note only the (unpublished) sunken featured buildings at Mohács, Szák and Fenékpuszta; at the latter a sunken feature hut and possible post-built structures lay within the 6th century cemetery group immediately east of the horreum. This latter instance at least verifies German settlers within the walls of a Roman site (Bona 1964, 8283;Barkóczi 1968,288). A.D. 526-568: Longobards in Pannónia Presumed Suebic finds persist even after the first phase of Longobard settlement in Pannónia. This western Germanic tribe had slowly migrated sout-easrwards from the lower Elbe (modern Saxony) since the 3rd century: in the 450s they are attested in Bohemia; in 489 they moved into Rugiland opposite Noricum Ripense after the Rugian defeat by Odoacer; in с 500 they extended their control across the Danube between the forts of Comagenis and Vindobona (Vienna), thus getting their first taste of (former) Roman soil - though on the basis of the Vita Severini few of the Romanised natives will still have been around then. In 508-510 the Longobards overwhelmed the Heruls to take Herulia, chiefly comprising lands in Moravia, but most likely also including a chunk of Pannónia (Carnuntum and lake Fertő). In 526 they took control of the Suebic kingdom, certainly in northern Pannónia and perhaps nominally north of the Danube. The Aquincum hoard should relate to this takeover. How far south the occupation of 526 extended is unclear, but it may have reached the Balaton area. Subsequently, in 547/8 the Byzantine emperor Justinian 'bestowed upon the Longobards, the towns of Noricum and the fortresses of Pannónia' (Procopius, Gothic Wars, III, 33), a donation chiefly to counter the threat of the Gepids. We can take this to mean Pannónia Savia and Secunda with a formal recognition of their existing holds in the north. To facilitate this move the Longobards may have quitted their lands over the Danube and those in Noricum Ripense; indeed most cemeteries in these areas are claimed to terminate by с 550 (cf. MENGHIN-PÖLHORN 1987, 21-34, 49-57; Bona 1987, 124-126). The wording of Justinian's donation is in itself sufficient to show both towns and castella as ongoing concerns: the Norican towns are generally equated with Celeia and Poetovio, on the fringes of Savia, vital for defending access into Pannónia: the 'fortresses' of Pannónia need not refer to specific sites, although we cannot exclude military forethought on the part of Justinian as regards securing Byzantine troop movement westwards to Italy. Justinian counted the Longobards amongst his (paid) allies and they in their turn seemed keen to be loyal and share in Roman culture. They provided mercenary troops to the Byzantine army in Italy and some also fought in the East. In each case the Longobards became acquainted with the mode of warfare based on the possession of fortified centres. These of course already existed in profusion in Pannónia and given that the Danube formed a border with Gepidia, a reuse of certain fortresses seems logical. Certainly with their occupation of northern Italy in 568-569 they rapidly established a series of urban-based dukedoms and at the same time maintained the network of fortified sites in the Alpine zones, best documented in Friuli (Christie 1990, 100-103). Again, however, the difficulty lies in demonstrating such site reuse: in Italy we are reliant on stray finds or burials within or around documented Longobard seats (e.g. MENGHIN-PÖLHORN 1987, 149-156); only rarely, as at Brescia, is 6th-7th century Longobard domestic activity visible within a town, while for the 8th century churches provide our main guide (BROGIOLO 1991). Pannónia, in contrast to Gepidia, as yet lacks Germanic style sunken huts of Longobard date: some scholars argue that this signifies the reuse or adaptation of surviving Roman structures, although direct proof of this is absent. While stray finds from within Roman sites at least hints at Longobard activity, these may denote scavenging rather than fixed settlement. Excavations near the forum at Scarbantia, for example, revealed a series of late houses of 4th-6th century date, including one timber house set on stone footings: a 6th century Longobard S-brooch came from the presumed destruction level of this building (Gömöri 1986, 351-352, with fig. 20). Yet we know of a bishop Vigilius at Scarbantia at least until 579 (Tóth 1974a; Gömöri 1986, 355), and so destruction of this house cannot mean destruction of the whole town. Vigilius is, however, the sole bishop documented for Longobard Pannónia. His seat's 'survival' may be explicable if, as Bona argues, lake Fertő in NW Pannónia represented an enclave of 'Roman' settlement, combining Herul and Suebic groups (Bona 1964, 86-88; 1987, 125126; but cf. 1976, 31-32). Longobard period burials in this zone (of Hegykő type) indeed exhibit provincial traits (e.g. burials without gravegoods; needles, rings) as well as pre-Longobard Germanic elements (e.g. chip-carved brooches, large bow brooches). Bona has long pointed out the attachment of Longobard cemeteries to Roman castra and castella along the Danube and to other Roman sites within the province, notably around the main lakes (cf. fig. 4) (1964, 1976, 3133; 1987, 126). Three main cemetery types are identified: Hegykő type, noted above; Szentendre type, named after the cemetery located 2 km from Roman Ulcisia Castra; and Vörs-Kajdacs type, datable from the 540s and representing an inward and southward spread of Longobard control. Throughout, a strongly martial 309