The Noviodunum Archaeological Project (NAP) The programme of excavation and field survey at Noviodunum, which began in 2005, continued in July 2006 and ran for eight weeks. The project team consisted of: Kris Lockyear (Institute of Archaeology, University College London [IoA], director) Adrian Popescu (Fitzwilliam Museum, University of Cambridge, director); Timothy Sly (Archaeology, University of Southampton [...SU], director); Simon Davis, Agnieszka Bystron, Iris Rodenbeusch (Museum of London Archaeological Service, excavation supervisors); Elizabeth Popescu (Cambridgeshire Archaeology, excavation supervisor); Doru Bogdan (excavation supervisor); Michael Bamforth (IoA, excavation supervisor); Sylvia Kennedy (excavation supervisor); Mihaela Ciauşescu (Cluj, pottery specialist); Robin Symonds (l'Institut National de Recherches en Archéologie Préventive, pottery specialist); Gareth Beale (SU, computing specialist); Jane Sidell (IoA and English Heritage, environmental coordinator); Yvonne Edwards (IoA, faunal remains); Martyn Allen (SU, faunal remains); Nichole Doub (IoA, conservation); Kim Lockyear (logistical support); Elliott Rampley (IoA, photography); Richard Farrant (finds support), as well as students from the Institute of Archaeology, University of Cambridge, University of Southampton, Bucharest University and other institutions. One of the major aims was to excavate the wall found in the previous season on the south-facing slope of the fortress near to the SW corner of the Ottoman fort, within the Roman defensive walls in the anticipated position of a tower (Area 1). Since the geophysical survey failed to provide clear evidence for a tower in this position, a new trench was opened in the putative location in order to verify its presence and character. The result was, however, negative as the depth of the deposits proved too thick at the base of the steep slope. Three other trenches were opened to check the results of the geophysical survey, which suggested the existence of two parallel curtain walls. These confirmed the existence of a wall similar to the one excavated in Area 1, probably contemporary with those excavated by V.H. Baumann further to the east. The wall uncovered in two of the three trenches seems to form part of the same structure as the 2m thick wall (context 1065) excavated in the 2005 season (Lockyear 2006). Although the ashlar facing of the two sections of the newly excavated wall had been robbed, the mortar and rubble core survives. The section of the wall nearest to the Ottoman fort had a vertical clay pipe running through it (Fig. 1), the function of which is as yet uncertain. The final phases of activity are associated with the robbing out of the ashlar facing from the curtain wall to provide building material for the later structures on the site, as well as for modern houses. A 16th-century silver plated forgery of a German taler produced after 1577, recovered from the robbing trench of the wall with the clay pipe, offers a terminus post quem for one of the robbing phases. Most of the 2006 contexts excavated in Area 1, which is within the Roman defensive walls, derive from medieval settlement (ephemeral buildings and numerous pits) but truncated late Roman contexts, which were evident in the base of pits. Detailed dating of the medieval contexts awaits completion of finds analysis, but most appear to be 11th to 13th century. One of the significant finds of the season was a fragmented dolium (context 1206), from which a large collection of charred oak root/knotwood was recovered. Wheat and barley were noted in its fill and a number of sample residues from surrounding deposits contained charred lentils. Small finds consisted of coins (Roman to Byzantine) and a largely late Byzantine assemblage of glass bracelets, iron tools, bone and antler artefacts. Of particular note is a copper alloy finger from a Roman statue found re-deposited. A new area was opened in 2006 (Area 2), lying on the north-facing slopes of elevated land to the south of the site of the Roman fortress and civil settlement, c.1km from the excavations in Area 1. This excavation sought to demonstrate the character of extra-mural settlement and was positioned between a late Roman building of relatively high status (with constructional materials including marble and water pipe) and a number of contemporary tile kilns, both of these areas having been explored in previous work by staff from Tulcea Museum. The excavation trench lies within the confines of an extensive late Byzantine cemetery, spanning the 11th to 13th centuries A.D. A group of fourteen articulated skeletons was excavated in 2006, comprising eleven adults and three juveniles. The burials were found in simple, relatively shallow, earth cut graves orientated east to west with the uppermost graves having been disturbed and compressed by later ploughing. At least two phases of burials were apparent, the graves lying in clear parallel rows and occasionally intercutting. Few finds were directly associated with the burials, although a small iron knife was found beneath one individual indicating that it had been worn at the back of the belt. Preliminary examination of the human remains indicates that the adult group includes five females or possible females, with the remainder being males or possible males. Although small, the group shows a remarkably high incidence of dental pathology, including pre-mortem tooth loss. Such high levels of dental problems indicate high tooth wear (relating to diet) and poor mouth hygiene. Accidental or deliberate injuries were apparent on other bones, including one individual showing injuries consistent with repeated blows to the head. Overall, the pathologies indicate a rural population, undertaking hard manual labour - probably farmers. The burials overlay an area of building debris (including tiles) which may relate to buildings contemporary with the known structure to the west - this provides a clear target for examination in the forthcoming season. The ceramic assemblage has only been subjected to an assessment and was quantified by weight and count. A total of 6770 fragments of pottery weighing 110,249g were recorded. Most of the pottery is late Byzantine (Byz) and only 10.22% is Roman (R) and Late Roman (LR). The latter is residual as in so far no contexts of Roman period were excavated. The total amounts of the major types of pottery are detailed in the following table. Full identification and quantification by estimated vessel equivalents (EVEs, Orton 1993) was reserved for major key contexts (1214 and 1163) within the excavation. The 2006 excavations produced 13,572 animal bone fragments (excluding fish) distributed across 127 contexts (96 contexts Area 1; 31 contexts Area 2). Some 19.6% of the bones have been identified to species/family. The combined weight of bone from all contexts was 91.240kg from Area 1 and 56.278kg from Area 2. In Area 1 cattle (Bos sp) remains make up 7.5% of the total assemblage, with sheep/goat contributing 5.74% and pig 3.8%. Sheep size (shp sz) specimens (n=3854) which probably include sheep/goat and pig, far out-number cow size (cw sz) specimens (n=1350) which might include some deer and horse although these taxa are relatively uncommon. It may be concluded that amongst the livestock, sheep were more numerous than cattle. However, it is likely that cattle made the major dietary contribution, as the meat yield per animal is significantly higher than for sheep. Equids represent 0.16% of the assemblage and were most likely remains of small ass/donkey (Equus assinis); deer, were mostly red deer (Cervus elephas), and accounted for 0.43% of the assemblage. Other domesticates encountered were dog - 0.22% (Canis familiaris), cat - 0.03% (Felis cattus) and probably domestic fowl (<5%). The Area 2 assemblage differs significantly from that of Area 1 in the greater importance of cattle, a decrease in pig numbers and increased incidence of juvenile/young dogs. Cattle bones numbered 507 and comprised c. 14% of the total assemblage. In line with this cw sz specimens were considerably more abundant (n=1098). Other domesticated species were at similar relative levels to those found in Area 1, such as sheep, goat (Ovis aries/Capra hircus plus shp/gt - 6.6%, equids (0.19%) and dogs (0.25%), although pigs (Sus sp) were less common (2.22%). Thus it appears that cattle were the predominant domesticate in Area 2 as determined by number of specimens and by weight. The data suggest that wild species were hunted at roughly similar levels in both areas, although there are notably more red deer specimens in Area 1 than Area 2. It seems very likely that during the medieval period the people living at Noviodunum had a measure of social control over the surrounding local population, as witnessed by the organization of pastoral farming, animal breeding and butchery. Several examples of deliberate bone modification for cultural purposes were noted. Three ovicaprine astragali showed evidence of single holes drilled through the centre of the bone anterior-posterior. The use of ruminant astragali as game pieces, fortune-telling pieces or ritual/funerary items has a long history from across the world (Bar-Oz 2001; Koerper & Whitney-Desautels 1999; Gilmour 1997). The fish remains are proving to be one of the key aspects of the environmental programme and have already provided an unparalled assemblage. Pit fills in Area 1 in some cases appear to be more fish than soil. In such cases, residues have been sub-sampled. A key question is whether distinct fish processing areas of the site can be identified, as opposed to fish consumption and post-consumption waste disposal areas. To date, 130 contexts have been examined, with a total specimen count of over 70,000 fragments. The fish examined from five contexts, three of which were hand collected (1020, 1054 and 1076), and two (1117 <48>, 1083 <39>) were from sieved samples (1mm mesh). All are thought to be medieval in date. The following species were identified; sturgeon (Acipenseridae), pike (Esox lucius), Cyprinidae, carp (Cyprinus carpio), bream (Abramis brama), barbel (Barbus barbus), cf ide (Leuciscus idus), cf wels (Siluris glanus), perch (Perca fluviatilis), cf zander (Stizostedion lucioperca). The fish represented in these samples could all have come from local fishing in the Danube, with the possible exception of grey gurnard. At the end of the 2006 season the trenches were back-filled to preserve the integrity of the deposits. In 2007 it is hoped that the excavations will resume in both areas, as well as at a selected rural site in Noviodunum’s hinterland. The latter was identified during the field survey and will be examined in order to clarify its nature, and to provide pottery assemblages for comparison with the surface and excavated collections.