Sometimes later, perhaps around AD 200, Noviodunum was promoted to the rank of municipium.8
Near the other auxiliary forts, civilian settlements of lower profile are attested. At Sacidava the site of the civilian settlement is not yet identified, but numerous inscriptions were found in the wall of the fort, set up by veterans who dwelled somewhere around.9 Six km north of Sacidava, in the neighborhood of the modern village of Rasova, a horreum of the military type was excavated.10 Since among the bricks used for this building some bear the stamps Leg. V Ma., Leg. XI Cl.P.F. and Leg. I Ital., it is certain that this was another settlement linked with military activity during the first half of the second century AD. Downstream of Rasova, Axiopolis would seem by its name to have been a civitas of old Greek origin, without a military link, and in fact Ptolemaios mentions it among other poleis (Ptol. Geogr. 3.10.5).
The civilian site of Capidava is still not located, though some tumuli of the cemetery extending around the fort have been excavated.11 Nevertheless, an inscription informs us of a territorium Capidavensis ruled by a quinquennalis (ISM V, 77) that is, organized to reproduce, on a reduced scale, the municipal administration. At Cius a vicus Verg[ob]rittiani led by a magister who was veteranus legionis V Macedonicae (ISM V, 115) is attested. Unfortunately one cannot determine whether this was the proper civilian settlement of the followers of cohors I Lusitanorum, or merely a village inside the area subject to the jurisdiction of this military unit. Aegissus, originally a native fort, was garrisoned by the Romans quite early, during the first century AD (ISM V, 286). One inscription that probably should be dated to the second century AD mentions a territorium A[eg(yssensis)] with an ordo decurionum, indicating a pseudo-municipal administration as for the territory of Capidava.12 Further southeast, at Murighiol, quite probably the antique Halmyris, beside the fort manned by the fleet a vicus classicorum sprang up, administratered by a magister and inhabited by cives Romani who in this instance, were obviously veterans of the fleet.13
At the end of this quick survey of the state of the Danubian frontier zone, some general remarks are required. As expected, the names of the sites of the forts and the civilian settlements related to them were overwhelmingly of pre-Roman origin, mostly Geto-Dacian. In the southern part of the frontier, there was a concentration of place-names ending in dava, characteristic of the Geto-Dacian hill-forts, indicating that the Roman army on its arrival in this region found a lot of local tribes dwelling in fortified sites according to their traditional habits. In the limes area, however, there is evidence only for a few, very small native sites surviving until the middle of the second century AD14 and for some Dacian individuals, who were in any case connected with the Roman army (IDRE II, 332; 336; 338). On the other hand, all over the frontier region there was a compact, cosmopolitan population depending upon the soldiers, sharing the Roman civilization, speaking solely Latin (at least in official circumstances) and living in communities of varying Roman legal status, among which two of the three Roman towns of Dobrudja stood out. As mentioned above, the only exception to this rule seems to have been represented by Axiopolis – apparently a civil site without any military connection, probably founded by Greek-speaking people though so far the site has produced Latin inscriptions exclusively.15 Even in the cemeteries at Capidava and Noviodunum, the only ones on the limes excavated to some extent, only graves of specific Roman provincial types were found including a few funeral assemblages containing military equipment, which was normal considering that some of the deceased were auxiliary soldiers.16
On the seashore the old Greek city-states of Callatis, Tomis, Histria were already established. Another much smaller Greek town was probably Argamum, mentioned in the horotesia of Histria (ISM I, 67-68) and located at Capul Dolojman, but the shortage of evidence prevents any further comments on it.
During the first-third centuries AD, the three major Greek towns of the Dobrudja were civitates peregrinae, i.e. self-governing communities without Roman status. Since the foedus between Rome and Callatis that has been partly preserved attests that the latter was a civitas foederata (ISM III, 1), for the other two cities one has also to consider every alternative status: civitates stipendiariae or civitates liberae et immunes. Nevertheless, the specific juridical status of each city-state was not of paramount importance as in practice the differences between distinct categories of civitates peregrinae had already begun to fade away from the first century AD.17
In Callatis, all the inscriptions set by individual inhabitants are in Greek except one Latin dedication to Trajan and cives Romani consistentes Callatis led by a quinquennalis perpetuus (ISM III, 83). According to the epigraphic evidence, at Tomis and Histria the population still spoke predominantly Greek, but in each of them one encountered a tribe of the Romans (ISM II, 256; I, 142). A considerable part of these Roman citizens, some of them Latin speakers, were veterans. And in both these towns are attested a lot of Latin speaking soldiers on duty or buried in the place of their residence or origin. As Tomis was the largest coastal Greek city of Moesia Inferior, it is no surprise that it produced a far larger number of army-related persons than Histria.18 It may be concluded that among the three Greek city-states of Dobrudja, Callatis best preserved its heritage and was at the same time free of any significant Roman military presence. The special situation of Callatis could derive from its privileged juridical status, from a traditional behavior characteristic of the Dorian colonists or even from its lack of importance to the Romans due to its reduced size and/or remote location. Given the meagre surviving evidence it is impossible to be more specific on this issue.
The inner part of Dobrudja was predominantly a rural area divided into the territories of the military settlements and Roman towns along the Danube frontier and of the Greek cities on the Black Sea shore (Fig. 1). It was only in the extreme south of this zone that Tropaeum Traiani developed, the third town of Dobrudja with Roman status. The Roman settlement near the Adamklissi tropaeum sprang up contemporarily with/or immediately after the inauguration of the monument in AD 109, as attested by a dedication to the emperor by Traianenses Tropaeenses in AD 116/117 (CIL III, 12470). Several mixed vexillations were composed of soldiers from both legions I Italica and V Macedonica (CIL III, 142143; CIL III, 14433) among which the largest, 1500 troops strong, probably worked on the construction of the monument and possibly also the settlement.19 Later, during the reigns of Antoninus Pius and Marcus Aurelius, officers of legio XI Claudia offered dedications at Adamklissi which, together with other undated inscriptions set up by soldiers, prove a continual military activity on the spot (CIL III, 7483; 142141; 142146). Sometimes before AD 170, Tropaeum Traiani was promoted to municipal status (IDRE II, 337). The town had an ordo decurionum municipii Tropaei (CIL III, 7484 = 12461; III, 142144) and the usual series of municipal magistrates and priests: duumviri (CIL III, 142142; 142146 et al.), duumviri quinquennales (CIL III, 14437 = 12462), aediles, quaestores, sacerdotes.20
In the remaining inner part of Dobrudja, one met only rural administrative units, most of them of Roman type: vici and villae rusticae. So, according to recent statistics of all the settlements of the Dobrudja, except the towns, 26 + 2(?) vici are attested, compared with 4 +6(?) komai and 2 pyrgoi + 2 turres (one turned into a vicus).21 In addition, there are at least 68 villae rusticae, 27 cemeteries and 74 isolated graves.22
The vici had the standard Roman organization, i.e. they were led by one or usually two annually elected magistri, and when of larger size they also had a quaestor. Characteristically, most of them had Roman names: Quinctionis (ISM I, 324-341), Secundini (ISM I, 342-349), Casiani (ISM I, 369-370), Celeris (ISM I, 351-352), Clementianenses (ISM II, 134; 136; 191), Narcissiani (ISM II, 133), Nov…(ISM V, 233), Petra (ISM V, 240), Ulmetum (ISM
