Sometime in the period between Charlemagne's offensive of fall 791 (successfully repelled) and the autumn of 795, the Avar empire was shaken by a catastrophic civil war. The two leading figures, the kagan and the jugurrus, fell out and waged a bitter struggle against each other. In nomadic empires, the 'right flank' was generally commanded by a leader of lesser rank; in the case of the western part of the Avar empire, this was the tudun. Therefore it is likely that figure who ruled over the region east of the Tisza, the Maros region, and Transylvania was the higher-ranking jugurrus (cf. the Hungarian harka-kündü-gyula triad in the mid 900s). The civil war ended with the defeat of the jugurrus and his followers; he himself was killed. However, time was too short for the kagan to fully reimpose his authority. In the winter of 795–96, the Franks renewed their attacks, and the outcome would be the collapse of the Avar empire. It can be inferred from these events that the Avars themselves decimated their ruling class in Transylvania.
Between 802 and 804, the Avars suffered further grievous blows, perhaps in several waves: the army of Krum, the Bulgar khan, advanced northward into the Tisza region. The invaders put the kagan to flight and captured a host of Avar soldiers; years later, the latter would serve in the Bulgars' campaign against Byzantium.
There is no sign that Krum's military campaign touched Transylvania, nor that any part of the Carpathian Basin was occupied by Bulgars in the first decade of the 9th century. To the contrary, reliable, contemporary sources indicate that 'Dacia on the Danube' (presumably the Temes region) and Sirmia were occupied {1-265.} by Balkan Slavs who were fleeing the Bulgars and seeking the protection of the Franks. These newcomers were the Abroditas (abodriti-praedenecenti 'qui ... contermini Bulgaris Daciam Danubio adiacentem incolunt') and the Timocians (timociani).[19] The emergence of this protectorate caused friction between the Frankish and Bulgar empires, which failed in the period 824–26 to settle their differences by negotiation. Around 827–31, the Bulgar Khan Omurtag launched a broad offensive against the Central Danube region; his armies occupied Sirmia and East Slavonia as well as parts of the Tisza region that are yet to be identified (perhaps as far as Csongrád = černi grad = 'black castle'). The Bulgar conquest was confirmed in the peace treaty that Khan Malamir concluded in 832 with the Eastern Frankish empire. There is no mention of the Timocians after 825, but the Abodritas are known to have been once again subjected by the Bulgars. The Bavarian Geographus counted among the Bulgars' neighbours the osterabtrezi Slavs — i.e. the eastern Abodritas — who lived in 'Dacia'. Ninth century written sources offer no further information with regard to the eastern half of the Carpathian Basin. However, the events of 863 and 883 show that the Bulgars were quite able to cross the Great Plain and the Garam River to attack Moravia. It is therefore likely that they had military outposts in the Tisza and Maros valleys.
Only a single piece of information, dating from 892, evokes the Bulgars' rule in Transylvania. King Arnulf's envoys requested that the Bulgar Khan 'Laodimir' (Vladimir) 'not allow the Moravians to buy (and transport) salt' ('Ne coemptio salis inde Maravanis daretur').[20] This indicates that in the 9th century, the Bulgars had taken possession of some of Transylvania's salt mines and traded in the salt mined by their subjects.
Fortunately, there are other, archaeological sources that throw light on Bulgar rule in Transylvania: the recently identified traces, dating from the 8th and 9th centuries, of the Bulgar khanate. The excavations were conducted on the site of the ancient capital cities, {1-266.} Pliska and Preslav, and at numerous cemeteries as well. Dobrudja, which was under Bulgar rule at the time, has also yielded much useful archaeological data. The finds in Transylvania are consistent with those in Bulgaria dating from the 9th–10th centuries.
The most significant site is the cemetery at Maroskarna. This unusual graveyard, located on the left bank of the Maros, has been repeatedly damaged by flooding. Despite repeated salvage attempts, the only traces of funeral rites are the dishes and skeletons washed up by the river, but this does not exclude the possibility that the cememtery had also contained cremation graves. Some 30 intact dishes, all made on the wheel, are now preserved in a museum. Their shape and decoration cannot be traced back to the earlier, Avar and Slavic pottery found in Transylvania. All of them bear the distinctive features common to Bulgar pottery of the period: yellow and reddish-brown carafes with one or two handles (amphorae); pear-shaped mugs, with thick, 'ringed' brims, and either a smooth finish or deep, web-like decoration; slightly funnel-shaped pots; an unusual, cauldron-like pot with pierced handle; and others that evoke the 'Medgyes type' pots. That last object was a common find in Bulgaria, and Bulgar-Slav influence may be responsible for its dissemination in 9th-century Transylvania. Vessels identical to those from Maroskarna have been unearthed in large numbers at the major settlements (Pliska, Madara, Preslav, Kadiköj) and cemeteries in Bulgaria, as well was in Dobrudja. They spread from Bulgaria toward the Carpathians along the Olt and Zsil valleys, and perhaps also by other routes that led from Dobrudja and Silistra to Transylvania. Traces of villages dating from the 8th and 9th centuries — the period of Bulgar rule — have been identified at Ploieşti-Bucov; the Slav-Bulgar finds reveal the links between Bulgaria and Transylvania. The cemetery, consisting mostly of graves with skeletons, that was discovered in the Mostiştea valley, at Sultana, confirms that the Danubian Bulgars established outpost settlements as they advanced towards Transylvania.
{1-267.} As far as Maroskarna is concerned, the routes leading to the Bulgar motherland are of secondary importance, for the cemetery is far more than the marginal manifestation of a northward-spreading, archaeological 'culture'. It is, in fact, the cemetery of a population that had been collectively resettled, along with its material culture, from south of the Danube. This resettled population formed an 'island', for, judging from its traces, it inhabited only a small number of sites along the middle reaches of the Maros. Graves (with skeletons) yielded fine Bulgar vessels, including distinctive amphorae at Szászsebes (along with iron knives and sheep-bones) as well as at Kudzsir and Sebesán (along with jugs and pots). The intact vessels found at Gyulafehérvár Castle (amphorae and pots with smoothed-in decoration), Gyulafehérvár-Partos, and Magyarszentbenedek originally came from graves as well. At Oláhgorbó, two coffins holding skeletons were unearthed in what is probably a smaller cemetery. The dishes that once held food for the deceased differ slightly from those at Maroskarna, but they are still typical of Danubian Bulgar pottery, and a woman's beads and bronze earrings are 9th-century Bulgar models; sheep bones indicate that meat was buried with the deceased, and thus testify to the Bulgars' pagan rite. The fragment of a hand-made, earthenware cauldron — unique in Transylvania — that was unearthed at the Bulgar site of Oláhdálya-Troján cannot be dated precisely, for the settlement was still in existence in the 10th–11th centuries. Bulgar traces have been found along both banks of the Maros on strips of land 30–40 kilometres wide. A fine, ring-rimmed Bulgar pot with smoothed-in web decoration, found at Kézdipolyán, testifies merely to some contact with the Bulgar settlement area on the Maros or with the Lower Danubian region.
Further up the Maros, the cemetery at Csombord evokes yet another 'island' of settlement. So far, thirty-two graves have been excavated; aligned on an west-east axis, they date from the 9th–10th centuries. The grave goods included women's jewellery: {1-268.} four graves yielded 14 diverse, richly-filigreed pendants, while three others yielded beads, a half-moon shaped pendant (lunula), and a disc-shaped pendant ornament.
Jewels of the same type as those from Csombord and Oláhgorbó have been found only in Bulgaria, almost exclusively in cemeteries dating from the 9th–10th centuries. Bulgarian archaeologists judge that both the Csombord and Maroskarna cemeteries are of Bulgar origin.
In all likelihood, the Maros valley was occupied by a detachment from Khan Omurtag's army. Perhaps it was a unit that veered away from the Tisza in the direction of Transylvania; one of its chiefs, Tarkan Onega(bon) of the Küviar clan, drowned in the Tisza. Or, perhaps, it was another detachment that came up along the Olt River. Following the conquest, which must have occurred around 830, the Bulgars established settlements along the Maros. The settlers came from the right bank of the Lower Danube, in Bulgaria. Their material legacy — jewellery, and dishes, complete with potters' marks on the bottom — is entirely the product of Danubian Bulgars. The settlers, who enjoyed military protection (grave with spurs at Tatárlaka), were charged with the task of putting back into production the salt mines at Marosújvár, Mezőakna, Sóvárad, and Torda, and to organize the transport of salt on the Maros River. Some of the shipments, towed along the Tisza to Csongrád, went to the Moravians, but most of them were directed downstream towards Belgrade. Since the bulk of the salt was probably shipped to the Bulgar khanate on the Lower Danube, the settlers had easy access to Bulgar products and luxury articles (jewellery). None of the finds indicate that they changed their way of life.
There is no information in historical sources regarding the background of the settlers. The Bulgar empire on the Danube had a mixed population that consisted principally of Bulgaro-Turks (whose language began to acquire Slavic characteristics in the 9th {1-269.} century), various Slavic tribes, and the Balkan Latins (Vlachs). One of the fundamental goals of the early Bulgar khanate's empire-building policy was to resettle the various ethnic groups. In the case of Transylvania, the answer lies in burial rituals, for the graves contained skeletons; and there is general agreement that in Danubian Bulgaria's 'pagan' cemeteries of the 8th–9th centuries, these graves held the remains of Bulgaro-Turks. To be sure, cemeteries of this type were less numerous than those that cremation and double rites identify with the Slavic and local Balkanic population; but they reflect the ornate, 'nomadic' attire and wealth of a dominant group. It is therefore likely that Bulgaro-Turk soldiers and their families — the ethnic group that gave the empire its name — were present in Transylvania, a militarily and economically insecure border zone.
There has been linguistic speculation that such toponyms as Brassó, Krassó, Barca, and Barót have a Bulgaro-Turkish derivation. Since the Barcaság bears no traces of Bulgar rule, this hypothesis lacks archaeological confirmation. However, it is possible that, both in the Tisza region and Transylvania, the Bulgar overlords relied on the remnants of another ethnic group: the Onogur-Bulgars (Wangars), who had moved into the region in the same period as the Danubian Bulgars. In this case, the Bulgar-type toponyms of the Barcaság might well be of Onogur-Bulgar origin. A more plausible contention is that the name of the Karas-Krassó River, in the Temes region, is derived directly, or via the Slavs, from Danubian Bulgar. Vessels of the 'Maroskarna type' — archaeological traces of Bulgars — have been found at Temesvár and Parác, as well as in the area of Maros-Aranka (Nagyszentmiklós, Makó, Deszk).
The Romano-Byzantine castle of Singidunum had white stone ramparts, and, in the 9th century, the Bulgar conquerors named it 'Belgrad', meaning 'White Castle'. In Bulgaria, the predominance of the Slavic language can be dated from the death of the last Turkish-named khan, Omurtag (831). Archaeological finds (quantities of Maroskarna-type pottery, including fragments with {1-270.} smoothed-in web pattern) indicate that the Bulgars' central settlement in the Maros region was located within the Roman, white stone walls of Apulum; and it was evidently the Bulgars who named the town Belgrad. For centuries, the Slavs who lived in the region continued to use that name, which probably explains its adoption into modern Romanian (Bălgrad).
There is some significance in the very fact that the sparse traces of the Bulgaro-Turks who had been resettled into Transylvania do not reveal much about their relations with the local Slav population. The influence of the Danubian Bulgars' more advanced techniques may be detected in the pottery ('Dridu pottery-craft') that dates from the time of Medgyes-type urn cemeteries, and it was still visible in the Szilágynagyfalu group. The aftereffects can be noted in the vessels that were found around Slavic stone fireplaces and Hungarian earth ovens at the 10th-century site in Maroskarna.