3 Economic Specialization and Ethnic Consolidation: Northern Hunting Societies in the Iron Age and Early Middle Ages

Sacrificial Tradition and Religion during the Early Iron Age in South Scandinavia – with Special Reference to Settlement SacrificesSacrificial customs and religion during the Early Iron Age (500 BC–400 AD) has occupied archaeologists from the infancy of archaeology. Most would probably agree that the religion was primarily fertility related, originating as it was in the existing peasant society. The literature does not reflect any disagreement about the religion of the Early Iron Age being polytheistic and consequently concerned a variety of gods. However, it is still unknown how the religion was integrated in the everyday life, and under which conditions it was practiced.The research interest and the overall synthesis framework have especially addressed sacrifices in bogs and wetlands (for instance weapon sacrifices, bog bodies, deposited earthenware, anthropomorphic wooden figures, domestic animals, cauldrons, ring sacrifices, etc.). Strongly simplified, the existing consensus may be expressed in one single sentence: The overall society-related sacrificial traditions develop from being almost exclusively connected with wetland areas during the Early Iron Age (until c.400 AD) to being primarily connected with dry land after this time, cf. Fig. 1.The question is whether – based on the intense data collection over the recent decades – archaeology can or should maintain this very simple picture of the development of the sacrificial traditions and the religions during the Iron Age? Is it possible that we – rooted in for instance narrow definitions of sacrificial finds, habitual thinking, and a “delusion” consisting of the numerous well-preserved, well-documented, spectacular, and impressive finds of bog sacrifices – fail to see numerous forms of deposits, which (as opposed to the impressive finds of sacrifices in bogs) are hidden in the archaeological material?The settlements of the Iron Age have been excavated in large numbers over the recent decades, and it is the ritual finds from these localities that provide the background for this article.The ritual deposits from the settlements can be divided into two superior groups distinguished by the physical context. One comprises sacrifices made to constructions, which are characterized by being directly connected to a specific structure; the other encompasses settlement sacrifices that are to a higher degree characterized by an overriding affiliation to the settlement. The establishment of a sacrifice definition suitable for scanning the archaeological material for relevant finds is of vital importance. As the definition should not beforehand restrict the search through the material, it is important not to narrow the basis by concentrating only on the physical characteristics of the individual artefacts. The general idea behind the present presentation is that the different ritual dimensions of a society are internally connected as they function within the same overall conventions and, as a consequence, make up parts of a general mental structure, which can leave physically recognizable traces across the different ritual dimensions, cf. Fig. 2. This principal viewpoint creates a theoretical starting point for my work and the established definition of sacrificial finds: All intentionally deposited objects, which analytically show significant similarities as regards their physical appearance and/or their deposition context with other recognized ritual objects/contexts, and which are closely connected to these in time and space, should, when analysed, be considered sacrificial finds.The British religious historian, Ninian Smart, describes religion as consisting of seven thematically describing situations, which – albeit not completely unconnected – may be described individually:1) A dogmatic and philosophical dimension, comprising doctrine systems.2) A mythical and narrative dimension, comprising tales of the deities, of the creation, etc.3) An ethical and judicial dimension, comprising the consequences of the religion in relation to the shaping of the life of the individual.4) A social and institutional dimension comprising organisations and institutions that tie together the individual religious society.5) An empirical and emotional dimension comprising the individual’s experience of god and the divine.6) A ritual and practical dimension comprising prayer, sacrifices, worship, etc.7) A materiel dimension comprising architecture, art, sacred places, buildings, and iconography.As archaeologists, we have a very limited possibility of investigating the very thoughts behind the practiced religion. It is therefore natural to concentrate to a higher extent on the overall setting for it – the ritual dimension and the materiel dimension respectively. The ritual dimension and in particular its sacrificial aspect is traditionally divided into groups characterised by their significance level within the religion as such.1) The first and most “important” group consists of cult rituals. These are characterized by being calendar rites based on the myths of the religion or the history of the people, and by playing a part in the events of the year.2) The next group comprises transition rites (rite de passage), which follow the life cycle of the individual.3) The last group comprises rites of crises, which serve the purpose of averting danger, illness, etc.It is important to realize that the two first ritual groups are predictable cyclic rituals addressing the gods, the myths, and/or the people/the individual respectively. Only the third and least central group of rituals is determined by non-predictable and “not-always” occurring incidences. On this background, it becomes central to analyse, which category one is facing when one wants to assess its importance for the religion as such, in order to evaluate the primary character of the religion.In an attempt to understand the overall importance of a specific ritual practice, one cannot ignore a very complicated problem, which is to evaluate whether the sacrifices were practiced by single individuals or by a larger group of people as part of more common and society-supporting rituals. The issue of the relation between different sacrifice types and the groups causing these has been addressed repeatedly. Often, narrow physical interpretation frames as to who sacrificed what are advanced (i.e. Fig. 3). However, the question is how suitable are these very narrow and rigid interpretation models? As mentioned above, a sacrifice is defined by the intention (context) that caused it rather than by the specific physical form of the object!The above mentioned methodical and theoretical issues provide the background for the author’s investigation of the archaeological sources, in which he focused especially on the relationship between ritual actions as they are expressed in bog deposits and in burial grounds and measured them against the contemporary finds from the settle­ments.The analysis of the archaeological material is based on those find groups (sacrifices of cauldrons, magnificent chariots, humans, animals, metals, and weapons), which have traditionally been presented as a proof that society supporting and more community influenced ritual sacrifices were carried out beside the bogs.The examination of the material supports that sacrifices of cauldrons, magnificent chariots, humans, animals, and earthenware are found in both settlements and wetlands (Figs. 4-12), and that the deposits seem to follow superior ritual conventions, i.e. Fig. 2. The sacrifices were not made in fixed sacred places but in a momentary sacred context, which returns to its daily secular sphere once the rituals have been carried out. Often, the ceremony consists of a ritual cutting up of the sacrificed object, and the pars pro toto principle occurs completely integrated in connection with both burial customs, wetland sacrifice customs, and settlement sacrifice customs. Sacrifices often occur as an expression of a rite de passage connected to the structures, fields, or infrastructure of the village. However, the repeated finds of earthenware vessels, humans, and animals in both wetland areas and in the villages indicates that fertility sacrifices were made regularly as part of the cyclic agricultural world. This places the find groups in a central position when it comes to understanding the religious landscape of the Early Iron Age. In a lot of respects, the settlement finds appear as direct parallel material to the contemporary wetland-related sacrificial custom and so one must assume that major religious events also took place in the settlements, for instance when a human or a cauldron was handed over to the next world. Both the selection of sacrificial objects, the form of depositing, and the preceding ceremonial treatment seem to follow superior ritual structures applying to both funerary rites and wetland sacrifices in Iron Age society.Often, the individual settlement-related sacrificial find seems to be explained by everyday doings, as largely all sacrifice-related objects of the Early Iron Age have a natural affiliation with the settlement and the daily housekeeping. However, it is clear that if the overwhelming amount of data is made subject to a comprehensive and detailed contextual analysis, settlement related find groups and attached action patterns appear, which have direct parallels in the ritual interpretation platform of the bog context. These parallels cannot be explained by pure practical or coincidence-related explanation models!As opposed to ploughed-up Stone Age axe deposits or impressive bronze depots from the Bronze Age and gold depots from the Late Iron Age, a ploughed-up collection of either earthenware, bones, human parts, etc. are not easily explained as sacrificial deposits. However, much indicates that the sacrificial settlement deposits of the Iron Age were not placed very deeply, and so they occur in the arable soil of later times. We

Muldfjælsplovens tidlige historie – Fra yngre romersk jernalder til middelalder

Danerne og det danske kongeriges opkomst – Om forskningsprogrammet »Fra Stamme til Stat i Danmark«

TamdrupRoyal residence and memorial church in a new light
 Tamdrup has been shrouded in a degree of mystery in recent times. The solitary church located on a moraine hill west of Horsens is visible from afar and has attracted attention for centuries. On the face of it, it resembles an ordinary parish church, but on closer examination it is found to be unusually large, and on entering one discovers that hidden beneath one roof is a three-aisled construction, which originally was a Romanesque basilica. Why was such a large church built in this particular place? What were the prevailing circumstances in the Early Middle Ages when the foundation stone was laid?
 The mystery of Tamdrup has been addressed and discussed before. In the 1980s and 1990s, archaeological excavations were carried out which revealed traces of a magnate’s farm or a royal residence from the Late Viking Age or Early Middle Ages located on the field to the west of the church (fig. 4), and in 1991, the book Tamdrup – Kirke og gård was published.
 Now, by way of metal-detector finds, new information has been added. These new finds provide several answers, but also give rise to several new questions and problems. In recent years, a considerable number of metal finds recovered by metal detector at Tamdrup have been submitted to Horsens Museum. Since 2012, 207 artefacts have been recorded, primarily coins, brooches, weights and fittings from such as harness, dating from the Late Viking Age and Early Middle Ages. Further to these, a coin hoard dating from the time of Svein Estridson was excavated in 2013.
 The museum has processed the submitted finds, which have been recorded and passed on for treasure trove evaluation. As resources were not available for a more detailed assessment of the artefacts, in 2014 the museum formulated a research project that received funding from the Danish Agency for Culture, enabling the finds to be examined in greater depth.
 The aim of the research project was to study the metal-detector finds and the excavation findings, partly through an analysis of the total finds assemblage, partly by digitalisation of the earlier excavation plans so these could be compared with each other and with the new excavation data. This was intended to lead on to a new analysis, new interpretations and a new, overall evaluation of Tamdrup’s function, role and significance in the Late Viking Age and Early Middle Ages.Old excavations – new interpretationsIn 1983, on the eastern part of the field, a trial excavation trench was laid out running north-south (d). This resulted in two trenches (a, b) and a further three trial trenches being opened up in 1984 (fig. 6). In the northern trench, a longhouse, a fence and a pit-house were discovered (fig. 8). The interpretation of the longhouse (fig. 4) still stands, in so far as we are dealing with a longhouse with curved walls. The western end of the house appears unequivocal, but there could be some doubt about its eastern end. An alternative interpretation is a 17.5 m long building (fig. 8), from which the easternmost set of roof-bearing posts are excluded. Instead, another posthole is included as the northernmost post in the gable to the east. This gives a house with regularly curved walls, though with the eastern gable (4.3 m) narrower than the western (5.3 m).
 North of the trench (a) containing the longhouse, a trial trench (c) was also laid out, revealing a number of features. Similarly, there were also several features in the northern part of the middle trial trench (e). A pit in trial trench c was found to contain both a fragment of a bit branch and a bronze key. There was neither time nor resources to permit the excavation of these areas in 1984, but it seems very likely that there are traces of one or more houses here (fig. 9). Here we have a potential site for a possible main dwelling house or hall.
 In August 1990, on the basis of an evaluation, an excavation trench (h) was opened up to the west of the 1984 excavation (fig. 7). Here, traces were found of two buildings, which lay parallel to each other, oriented east-west. These were interpreted as small auxiliary buildings associated with the same magnate’s farm as the longhouse found in the 1984 excavation. The northern building was 4 m wide and the southern building was 5.5 m. Both buildings were considered to be c. 7 m long and with an open eastern gable. The southern building had one set of internal roof-bearing posts.
 The excavation of the two buildings in 1990 represented the art of the possible, as no great resources were available. Aerial photos from the time show that the trial trench from the evaluation was back-filled when the excavation was completed. Today, we have a comprehensive understanding of the trial trenches and excavation trenches thanks to the digitalised plans. Here, it becomes apparent that some postholes recorded during the evaluation belong to the southernmost of the two buildings, but these were unfortunately not relocated during the actual excavation. As these postholes, accordingly, did not form part of the interpretation, it was assumed that the building was 7 m in length (fig. 10). When these postholes from the evaluation are included, a ground plan emerges that can be interpreted as the remains of a Trelleborg house (fig. 11). The original 7 m long building constitutes the western end of this characteristic house, while the remainder of the south wall was found in the trial trench. Part of the north wall is apparently missing, but the rest of the building appears so convincing that the missing postholes must be attributed to poor conditions for preservation and observation. The northeastern part of the house has not been uncovered, which means that it is not possible to say with certainty whether the house was 19 or 25 m in length, minus its buttress posts.
 On the basis of the excavations undertaken in 1984 and 1990, it was assumed that the site represented a magnate’s farm from the Late Viking Age. It was presumed that the excavated buildings stood furthest to the north on the toft and that the farm’s main dwelling – in the best-case scenario the royal residence – should be sought in the area to the south between the excavated buildings. Six north-south-oriented trial trenches were therefore laid out in this area (figs. 6, 7 and 13 – trial trenches o, p, q, r, s and t). The results were, according to the excavation report, disappointing: No trace was found of Harold Bluetooth’s hall. It was concluded that there were no structures and features that could be linked together to give a larger entity such as the presumed magnate’s farm.
 After digitalisation of the excavation plans from 1991, we now have an overview of the trial trenches to a degree that was not possible previously (fig. 13). It is clear that there is a remarkable concentration of structures in the central and northern parts of the two middle trial trenches (q, r) and in part also in the second (p) and fourth (s) trial trenches from the west, as well as in the northern parts of the two easternmost trial trenches (s, t). An actual archaeological excavation would definitely be recommended here if a corresponding intensity of structures were to be encountered in an evaluation today (anno 2016).
 Now that all the plans have been digitalised, it is obvious to look at the trial trenches from 1990 and 1991 together. Although some account has to be taken of uncertainties in the digitalisation, this nevertheless confirms the picture of a high density of structures, especially in the middle of the 1991 trial trenches. The collective interpretation from the 1990 and 1991 investigations is that there are strong indications of settlement in the area of the middle 1991 trial trenches. It is also definitely a possibility that these represent the remains of a longhouse, which could constitute the main dwelling house. It can therefore be concluded that it is apparently possible to confirm the interpretation of the site as a potential royal residence, even though this is still subject to some uncertainty in the absence of new excavations. The archaeologists were disappointed following the evaluation undertaken in 1991, but the overview which modern technology is able to provide means that the interpretation is now rather more encouraging. There are strong indications of the presence of a royal residence.
 FindsThe perception of the area by Tamdrup church gained a completely new dimension when the first metal finds recovered by metal detector arrived at Horsens Museum in the autumn of 2011. With time, as the finds were submitted, considerations of the significance and function of the locality in the Late Viking Age and Early Middle Ages were subjected to revision. The interpretation as a magnate’s farm was, of course, common knowledge, but at Horsens Museum there was an awareness that this interpretation was in some doubt following the results of the 1991 investigations. The many new finds removed any trace of this doubt while, at the same time, giving cause to attribute yet further functions to the site. Was it also a trading place or a central place in conjunction with the farm? And was it active earlier than previously assumed?
 The 207 metal finds comprise 52 coins (whole, hack and fragments), 34 fittings (harness, belt fittings etc.), 28 brooches (enamelled disc brooches, Urnes fibulas and bird brooches), 21 weights, 15 pieces of silver (bars, hack and casting dead heads), 12 figures (pendants, small horses), nine distaff whorls, eight bronze keys, four lead amulets, three bronze bars, two fragments of folding scales and a number of other artefacts, the most spectacular of which included a gold ring and a bronze seal ring. In dating terms, most of the finds can be assigned to the Late Viking Age and Early Middle Ages.
 The largest artefact group consists of the coins, of which

Detektorfund og bebyggelse – Det østlige Limfjordsområde i yngre jernalder og vikingetid

A Holocene soil complex has been studied in six sections. In well-developed sections, a Holocene pedocomplex consists of five separate soils: a dark-grey forest soil (Subatlantic period), a dark-grey forest soil (Subboreal), separated by the Velbar cultural layer; weak Mollisol with podzolic features (Atlantic period), Sod-Podzolic soil (Boreal), Cambisol (Preboreal). The gran-size analysis perfomed has confirmed the morphological characteristics of the separate soils of the Holocene soil complex.
 The features of relic relief and soil cover in the studied area caused space structure of settlements. On the Late Pleistocene dune and it slope were located settlements. In the northern part of the river terrace on the most fertile soils the agricultural lands were situated. During almost entire Holocene in the southern part of terrace on the dunes pines woods were grows. Investigation of the relic dune near to the archaeological site showed that two phases sand accumulation were caused by forest clearance (Eneolithic and the early Middle Age).
 The stratigraphic position of cultural layers is related to soil horizons of the pedocomplexes and polygenetic soils. Based on changes in soil morphology in natural and human-transformed soils, the main evolutionary phases in anthropogenic transformation of elementary soil processes and soil covers have been revealed, particularly Non-agrarian (Paleolithic — Neolithic), Early Agrarian (Eneolithic — Early iron age), Transformational (Rome age — Early Middle age), Turbulent (Slaviс period), Degradation (Poland-Lithuanian period), Geochemical (New and Modern period — XIX—XX centuries).

The excavations in the Strzelińskie Hills, regularly conducted for more than 10 years led to discoveries of many interesting sites. In 2015 was examined the area of the Borowe Skałki located on the slopes of the Gromnik – the highest peak of the Strzelińskie Hills. In the vicinity of the rock outcrops were encountered traces of settlements from the Upper Palaeolithic, the Mesolithic, the Iron Age as well as the early and late Middle Ages. Also an iron battleaxe, the so-called bradatica was found associated with the Great Moravian infl uence. This artefact belongs to relatively rare fi nds. From the area of south-west Poland it is the seventh battleaxe of this type, and yet one of the best preserved. This fi nd sheds a new light on the other artefacts from the early Middle Ages in the Strzelińskie Hills, which allow to identify new links between this part of the Sudetes Foreland an the Great Moravian sphere of infl uence.

Bona in Bredestad: From a Viking Age magnate farm to a medieval estateThe remains of a magnate farm from the late Vendel Period and the Viking Age have recently been excavated in the parish of Bredestad in northern Smaland, Sweden. Thanks to the unusually rich source material it has, moreover, been possible to reconstruct a medieval estate that existed in the parish during the Early Middle Ages. The manor on this estate was Bona, which is a stone’s throw from the Viking Age farm.This article presents expressions of power in settlement and landscape from the Vendel Period when a magnate farm was established, to its function as a manor on a major estate in the Early Middle Ages. The article thus highlights the site continuity evidenced at Bona. It also elucidates how land use interacts with changes in settlement and how it was affected during the Migration Period crisis and in the late medieval agrarian crisis.The study combines three different types of source material: archaeological material, evidence from early maps and written sources, and pollen analysis. The selected types of source material are complementary,

Although hill-forts from the end of the Bronze Age and the beginning of the Iron Age associated with Lusatian culture appear in vast areas of modern Poland, they are absent in Pomerania beside the Lower Oder region. This scarcity is surprising, especially taking into account the relatively numerous appearances of hill-forts in Greater Poland, the region directly neighbouring Pomerania to the south. On the other hand, investigations conducted in the 1960s and 1970s to verify Pomeranian hill-forts described as originating from the Early Medieval and Medieval periods resulted in the detection of at least a dozen sites with material from the Late Bronze Age and the Early Iron Age. The aim of this paper is to present the problem of the supposed presence of Lusatian culture hill-forts in the central part of Polish Pomerania. It is highly probable that this kind of settlement played an important role in interregional contacts between eastern and western parts of Pomerania, together with Greater Poland and probably also Nordic Bronze Age zones. In a wider perspective, their role in the course and working of the Amber Road at the end of the Bronze Age should also be taken into account and investigated. It seems that new tools available for archaeologists, like Lidar data, modern geophysics and aerial photography, may provide new openings and new perspectives on research into this case study. Key words: Central Pomerania, Lusatian culture, Late Bronze Age, Early Iron Age, hill-forts, fortified settlements, trade/exchange routes, archaeological cartography. DOI: http://dx.doi.org/10.15181/ab.v24i0.1565

Botanical investigation of archaeological sites situated in the northwest of the region bounded by the rivers Maas, Scheldt and Demer (‘MSD region’), west of the city of Breda, has provided a great deal of evidence about the landscape and its use in the period between 2000 b.c. and a.d. 1500. From pollen analysis, it appears that this cover-sand area gradually lost its woodlands through human activity after the beginning of the Bronze Age (ca. 2000 b.c.). Patches of woodland did survive there, however, until the early Middle Ages. In contrast to the cover-sand area in the vicinity of ’s-Hertogenbosch and Oss-Ussen in the northeast of the MSD region, the first large heathlands in the Breda area did not evolve until the early Middle Ages. In late prehistory, land use in this area was not much different from that in the micro-region of ’s-Hertogenbosch and Oss-Ussen. In the Bronze Age, Hordeum vulgare ssp. vulgare (hulled six-row barley) and Triticum dicoccon (emmer wheat) were grown. During the Iron Age, Panicum miliaceum (common millet) and T. spelta (spelt wheat) were introduced, but these crops disappeared during the Roman period. The Roman period is remarkable because of the lack of any Mediterranean culinary herbs or exotic fruits. Only pollen of Juglans regia (walnut), found around the transition from the Roman period to the early Middle Ages, indicates the introduction of an exotic tree into the region. From the early Middle Ages onwards, Secale cereale (rye) was the most important cereal, which was grown as a winter crop. In the course of the Middle Ages, arable weeds of the Sclerantho annui-Arnoseridetum plant community appeared, which is associated with the continuous growing of rye.

Despite being a central element in traditional agricultural practices, shielings have rarely been subject to large-scale archaeological investigations in Norway. Combining results from development-led surveys and small-scale research projects, this study demonstrates how such an extensive archaeological material – despite inadequate contextual information and chronometric hygiene – may contribute to a refined chronology for transhumance and shielings in the county of Møre and Romsdal, and beyond. Radiocarbon dates from these investigations may relate to specific archaeological features or events or to more complex archaeological contexts. Albeit mixed and incorporated into palimpsest deposits, charcoal sampled for radiocarbon dating could still be treated as a remnant of significant past events. Varying means of statistical modelling may thus enhance our understanding of this chronometric material, revealing both continuity and change at various scales. Hence, I propose a development of transhumant practices across four phases, spanning from the Pre-Roman Iron Age to modern times. According to a definition that emphasises settlement, however, there is still no evidence of shielings until the Late Iron Age and Early Middle Ages.

Våben i vand – Om deponeringer i vikingetiden

Avlsten

The organization of the living space in the Romenska culture part of Vodyane hill-fort in Zmiiv district of Khakiv oblast is analyzed in the paper. The hill-fort occupies the edge of plateau and a high cape of the right bank of Uda River; it existed during two periods: the Early Iron Age — mid. 5th—4th cent. BC (the forest-steppe culture of Sсythian Age) and Early Middle Ages — the Romenska archaeological culture. In the latter period only the northern part of the site was inhabited (0,5 ha). The Severian neighboring community lived here for 50—60 years (within the 10th — beginning of 11th centuries). A total of 11 dwellings, 2 household buildings, 3 open heating systems (cookhouses), vault, 6 household pits (one of which was used for food storing) were discovered on the hill-fort.
 In the Severian time the hill-fort was built by rows/lines of houses from the southern east to the northern west with the consideration of the topographic level of each separate row. Three building lines are watched and the southern one was probably the earliest. First researchers recorded the entrance to the hill-fort on the west edge of the cape with a rampart. The absence of objects and, practically, of cultural layer in the south-western corner of the yard and along the considerable part of the western edge of the cape provide an opportunity to offer the organization of the living space of the settlement. From the entrance to the hill-fort, the road ran along the western edge of the cape, and the passes between rows of houses were located to the south from this road.
 At least 6 homesteads stand out on the hill-fort. 3 more homesteads can be defined based on the presence of cavities on the site surface. It points to the fact that the organization of the living space on Vodyane hill-fort in the end of the 1st millennium is defined by linear character (by rows) with elements of homestead building. It corresponds to the late period of the Severians’ development, when clan community was replaced by neighboring one.

Hillforts are conventionally regarded as a phenomenon of the Late Bronze Age and Iron Age of temperate Europe, with some sites being constructed or reoccupied in the post-Roman Iron Age or Early Medieval period. In broad chronological terms, 1000 BC to AD 1000 covers the two millennia of the ‘long Iron Age’ in which hillforts are a major field monument. The concept of enclosure nevertheless has a much longer ancestry, from at least the earliest Neolithic. Some enclosed sites of the Neolithic and earlier Bronze Age in central Europe may be located on elevated ground or on promontories and may involve palisades or earthworks around their perimeter, just like Iron Age hillforts, so that the question arises whether these should not qualify as hillforts. To argue that their topographic location, or the scale or layout of enclosure, is not indicative of a primarily defensive purpose will not do, because some Iron Age hillforts seem to be compromised on these criteria. Nevertheless, by not entirely rational convention, hillforts as a regular class of field monuments are generally recognized from the Late Bronze Age, when their appearance in central and western Europe coincides with an intensification in the quantity and number of types of weaponry and defensive armour associated especially with the Urnfield culture. There are a number of hillfort sites in Britain where there is underlying evidence of Neolithic occupation, including occupation that was originally defined by enclosing works of earth or stone. There is no question of claiming continuity of occupation from Neolithic to Iron Age, but since the earlier earthworks would almost certainly still have been visible—at Maiden Castle, for instance, where the earliest Iron Age hillfort follows almost exactly the extent of the Neolithic enclosure—there is every reason to suppose that the existence of earthworks that would have been recognized as ancient, even if they were not formally venerated as places of ancestors, may have encouraged choice of these sites. An alternative interpretation would be simply to assume that the same advantages of location that commended themselves to Neolithic communities coincidentally satisfied equally the requirements of their Iron Age successors. But in that event the earlier monuments, like the Hambledon Hill long barrow or the Foel Trigarn cairns (Plate 14b), would hardly have been accorded the respect by later occupants that their condition indicates they were.