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.

century but with the arrival post-380 of'foreign' elements such as spindlewhorls. armrings, bone combs and poly- or octahedral earrings in female graves, and knobbed brooches and bronze buckles in male graves (SalamonBarkóczi 1978-79. 75-78. 81; Salamon-Lengyel 1980, 9398). The occurrence of Csákvár type finds at military sites like Brigetio, Castra ad Herculem (Pilismarót), Ulcisia Castra (Szentendre) and Környe bolsters this picture of federate settlement. In each case we can note the association with a late Roman (fortified) site and population. Of interest here is the last phase of activity at Pone Navata (Visegrád-Sibrik), apparently post-dating the desertion of the late watchtower. Excavations in the SE corner of the fortress revealed traces of three semi-sunken huts (Grübenhäuser) set over and within ruinous Roman structures, cut to a depth of с 40 cm and with post supports; hearths lay in two huts and associated finds included 'eingeglättete Keramik', 'barbarian' handmade pots, spindlewhorls, glass, a whetstone and two Valentinianic coins. Although there are no signs of subsequent occupation, the 10th-l 1th centuries saw house construction on the hill (SOPRONI 1985, 44-47). Elsewhere early 5th century building work is predominantly in dry-stone or features poor stonework bonded with earth or clay: this is evident at Vindobona and Carnuntum, where such walls overlay earlier foundations or ran across roads and alleys; civilising influences nonetheless persist in the form of underfloor heating in some buildings. In the same period burials begin to occur within the walls (SOPRONI 1985, 80-83; Kandier 1980, 85-90 - compare data for Noricum in Ubl 1982, 74-86). Similar pictures emerge at forts like Arrabona, Brigetio, Tokod, Aquincum and Intercisa, while Campona (SW of Aquincum) featured a wooden post building associated with burnished ware (SOPRONI 1985, 55-74, 83. Throughout, we should assume abandonment of exposed vici alongside the forts - Gabler et al. 1990, 20-21). The construction of chapels at such sites may also be an early 5th century event (Mócsy 1981; Barkóczi-Salamon 1984, 175-176). Coins extend into the reign of Honorius (395-423) and in a few cases (e.g. Carnuntum, Quadrata) into those of Theodosius II (408450) or Valentinian III (425-455), but overall suggest a cut-off coinciding with the Hunnic occupation of Pannónia (SOPRONI 1985, 84-85; Barkóczi-Salamon 1984, 179). It is difficult, on the basis of restricted archaeological investigation, to follow closely the fate of the inner fortifications and towns. The construction of fortifications such as Környe and Fenékpuszta and the reworking of the defences at Ságvár and Heténypuszta may, as noted above, date only from 380 and signify a new defensive scheme based largely around the mobile army. Internal features are poorly understood, but striking is the presence of large granaries. Barkóczi and Salamon have pointed out the frequent coincidence of churches (usually two) in close proximity to such horrea in both forts and towns, suggestive of coherent planning (1984, 169-177). At Keszthely-Fenékpuszta a massive granary of 45x28 m lay close to the west flank; immediately north lay Basilica 2, while Basilica 1 was sited near the north gate. 6th-7th century graves cluster around both granary and Basilica 2. At Gorsium two 4th century basilicas are known east of the palace complex; a horreum was subsequently built over the east rooms of the palace , with associated 5th-6th century graves here and around the larger basilica (cf. pl.l). Nearer the Danube, Tokod features two likely chapels which utilised tower interiors as their apses; both post-date the granary. At Pilismarót the horreum lies near two apsed buildings (though both are N-S oriented). Similar granary-basilica groupings are presumed at the towns of Savaria and Sirmium. Whilst the data are still limited, they may well indicate the increasing importance of the church in these desperate years, signifying how the church - in the absence of any official administration oversaw the collection and distribution of grain. Barkóczi and Salamon (1984, 176-178) refer us to the Vita Severini regarding the role of the church in maintaining urban life in Noricum. Of ordinary houses in the urban context our information is restricted to Scarbantia (Sopron), where recent excavations near the Roman forum demonstrated continuous habitational activity into the 5th and even the 6th century: five houses were identified extending over the buried south end of the forum, of generally rough stone build. Finds overall indicated that the forum was kept fairly clean until с 330 after which rubbish layers accumulated; coins document market activity until 380 at least, although with the forum no longer the symbolic town centre (Gömöri 1986, 347, 352, 365, 381). Again the picture compares closely with what is known of urban activity in Noricum Ripense. From Huns to Longobards. A.D. 427-526 Information for the hundred years between the end of Roman Pannónia and the Longobard occupation is scattered and imprecise. There is nothing substantial from anywhere to show a persistence of settlement and we can only point to stray finds of Hunnic, Germanic or even Christian character of 5th to 6th century date occurring near or within forts like Brigetio (Kiss 1981), Tokod, Pilismarót-Malompatak (SOPRONI 1985, 58-61, 105-106), Aquincum and Intercisa (Bona 1982, 188-189). The question here, as in later contexts, is whether such proximity denotes a physical presence of barbarian groups within the old Roman settlements, outside their walls, or in distinct sites away from them. Furthermore we must ask whether such associations imply the survival of a Romanised population, with the barbarian settlers implanted so as to ensure control of these natives. Possible indicators of this situation may be the discovery 307