Ritualized Mesolithic Hoarding in Southern Scandinavia An Under-Recognised Phenomenon

Department of Archaeology and Heritage Studies, Aarhus University mathias.jensen@cas.au.dk Intentionally deposited groups of artefacts, here classified as hoards, form a relatively under­ studied aspect of the southern Scandinavian Mesolithic. Here analysis of 124 southern Scan­ dinavian Mesolithic hoards is used to further the concept of ritualization, applying a holistic approach to the observed variability and patterning in their biographies. Contrary to the common assertion that hoarding began in the Neolithic, the results indicate that hoarding practices can be traced back to at least the Early Maglemose and extend throughout the Mesolithic. A catalogue of studied hoards is included in the supplementary online material, as well as a separate catalogue of use­wear analysis findings from a subset of the hoards.


Introduction
The hoarding practices of the southern Scandinavian Mesolithic have re ceived little archaeology attention (Karsten & Knarrström 2003:127;Lars son 1978:163-164). In general, the focus has rather been on the identifica tion of waste disposal areas, which along with scatters and activity areas, are largely seen as the detritus from everyday life. This stands in contrast to research into the Neolithic, Bronze and Iron Ages, in which depositional practices, especially those deemed as ritual hoarding, have been the focus of intense discussions (Karsten 1994;Levy 1982;Müller 1886;Nielsen 1977;Rech 1979;Worsaae 1866). Research on Scandinavian Mesolithic ritual depositional practices has largely sidelined hoarding, being dominated in stead by mortuary finds, deposition of single finds, intrasite patterning of particular forms of material culture, and unusual artefact assemblages found at a few key sites (e.g. Larsson 1988Larsson , 2003Karsten 1994:166-170;Koch 1998:157;Hansen 2003;Karsten & Knarrström 2003;Nilsson Stutz 2003;Andersson et al. 2004:138;Carlson 2008:156-165;Toft 2009;Molin et al. 2014;Petersen, E.B. 2015;Pedersen & Petersen, E.B. 2017;Sørensen, S.A. 2020;Hallgren et al. in press). These diverging research trajectories, both within Mesolithic depositional research and between Mesolithic and Neo lithic hoarding research, have limited our understanding of the scale and na ture of Mesolithic hoarding practices. This has likely contributed to a wide spread acceptance amongst many archaeologists that such hoarding prac tices are a phenomenon first witnessed in the Neolithic (Solberg 1989:284;Sørensen, L. 2014:129). Thus, the gap between our perceptions of the Meso lithic and the Neolithic has further increased. However, I would argue that this is more a symptom of the relatively underresearched nature of the Mes olithic hoards and the different terminology used to describe and interpret Mesolithic deposits, rather than a reflection of the actual material remains.
The present paper seeks to bridge the interpretive frameworks and dis cussions of previous Mesolithic and later prehistoric research by focusing specifically on hoarding in Mesolithic southern Scandinavia. The archaeo logical material drawn upon comes from Denmark, SchleswigHolstein in northern Germany, and southern Sweden. The dataset includes archaeologi cal evidence dating from throughout the Mesolithic, in particular material from the Maglemose (circa 9500-6400 BC), Kongemose (circa 6400-5400 BC) and Ertebølle periods (circa 5400-4000 BC). The dataset and analysis presented here represents the most extensive attempt so far to describe and understand Mesolithic hoarding, with 124 southern Scandinavian Meso lithic hoards included in the multiscalar analysis 1 .
The main objectives are to ask research questions on the foundational characteristics of southern Scandinavian Mesolithic hoarding: what kinds of objects and materials are found in the hoards, and what temporal and spatial variability is present? How were the objects treated prior to and dur ing deposition, and in what environments and contexts were the hoards deposited? How should the hoards be interpreted?
The following section discusses definitions and analytical approaches to southern Scandinavian hoards used within this paper with reference to previous research. Next is a summary and critical assessment of the state oftheart understanding of Mesolithic hoarding, stressing especially the conceptual legacies of wider approaches to Mesolithic and Neolithic stud ies. This is followed by a discussion of practice theory and the concept of ritualization, which has already been successfully used to interpret and ap proach other forms of depositional practices in the Mesolithic, especially burials. These insights are intended to frame the applicability of ritualiza tion in relation to Mesolithic hoards, in contrast to commonly assumed ritualprofane dichotomies.
Next, I present an overview of the general characteristics -including both commonalities and variability -based on the quantitative and qualitative analysis of 124 hoards (Figure 1 and appendix 1). This compilation is over seven times larger than the largest prior study (Larsson 1978), allowing for the identification of patterns and variability in Mesolithic hoarding prac tices. The analysis includes an examination of lifehistories of the objects included in the hoards and the treatment of the hoards prior to and during their deposition. These different analytical lenses are intended to examine various temporal, social and spatial scales of the hoarding practices -from individual objects and hoard biographies to the longterm continuities and changes of the practice over the entire southern Scandinavian Mesolithic. Finally, based on these empirical results I explore several possible interpre tive scenarios for Mesolithic hoards, including why the concept of ritual ization may provide the best means of approaching, understanding and in terpreting these hoards.
Hoards have often been treated in isolation and without extensive com parative analysis. This restricted research makes it difficult to assess general patterning and variability, and it effectively hinders attempts at a compre hensive interpretation of these practices. In the present paper, I aim to ad dress some of these issues by using consistent terminology and systematic criteria to identify and classify Mesolithic hoards.
The term hoard is used in this paper, rather than cache, depot or de posit, as it is intended to as a bridge between the Mesolithic and Neolithic research traditions, where hoard is more commonly applied. However, the term is not meant to imply continuity between the Mesolithic and Neo Table 1. Frequency of particular interpretations/classifications used for Southern Scandinavian hoards (data from references in Appendix 1). Many hoards are interpreted in different ways, within the same text or between different texts, so hoards and their respective interpretation/ classification are counted each time they are mentioned in any of the referenced texts.

Interpretation/classification Number
Ritual (sacrifice, offering, votive) 78 Neutral (depot, deposition, deposit) 76 Find description (collection, found together, toolkit) 36 Cache (store, storage) 33 Hoard 27 Lost (accident, dropped, leister-set) 8 Waste 5 Grave (cenotaph and emptied grave) 2 Belongings of a crime victim 1 lithic or a priori notion that these deposits were strictly profane or ritual in nature. Hoard is regularly used in wider European research for material interpreted as profane storage as well as ritual or sacrificial offerings (e.g. Bradley 1990;Levy 1982).
In this paper, a hoard is defined as an intentional single deposit contain ing multiple objects found in a discrete cluster (Figure 2). Deposits that in clude debitage, unworked faunal or flora remains or human remains are, however, excluded. This is done to distinguish hoards in a strict sense from other depositional practices, such as production or consumption waste, ac cidental losses or mortuary remains, and due to the difficulty of determining the level of intentionality behind deposits of organic unworked raw mate rial in particular. It should be noted, however, that the a priori exclusion of these materials is not to imply that deposition of these materials could not be related to hoarding. Rather their exclusion was for analytical reasons and to prevent a categorization too broad to be analytically useful. Thus, the categorization of the included deposits as hoards is intended as an op erational analytical definition, not necessarily representing an emic one.
Data were collated from published reports and museum accession re cords; many of the hoards presented here have not previously been pub lished. A large number were found during old excavations or during peat digging or agricultural activities. This legacy material suffers from numer ous sourcecritical caveats, chiefly a lack of contextual information. This represents challenges regarding the classification and interpretation of these assemblages. On some occasions (N=16), I relied on prior interpretations that had classified a given deposit as a hoard or similar (for example depot, offering, or cache). These 16 hoards vary in composition, context, and lo cation, but fit the observed patterning of the more certain hoards.
The catalogue was analysed both quantitatively and qualitatively to iden tify the general characteristics as well as the variability of southern Scandi navian Mesolithic hoards: the object composition, observable continuities and changes of the composition at different temporal and spatial scales, the predepositional lifehistory of the objects, the peridepositional treatment of the objects and the assembled hoards, and finally the depositional con texts in which these hoards are found. As part of the biographical analysis of these objects, a subset was subjected to detailed macro and microscopic visual examination. This examination primarily consisted of usewear anal ysis using either a Dinolite AM 4815ZTL portable USB micro scope or a standard Nikon eclipse LV150 metallurgical microscope, depending on the accessibility of the material, supplemented with macroscopic observations of the objects themselves or published images.

A brief history of research into southern Scandinavian Mesolithic hoarding
The first southern Scandinavian Mesolithic hoards were found as early as the midnineteenth century with a slow trickle of similar deposits identified over the next 150 years. However, many of these early hoards, especially of blades, were originally considered Neolithic or even Bronze Age (Ebbesen 1982:21;Karsten 1994:95-97;Nielsen 2017:126-127;Salomonsson 1957). The earliest publication that specifically includes Mesolithic hoards was by Rydbeck (1918) on bog and dryland hoards held in the Lund University Historical Museum. However, only four possible hoards of Mesolithic axes are briefly noted, whereas the main focus was on the much larger number of Neolithic wetland hoards. After Rydbeck's (1918) study, Mesolithic hoards went largely overlooked, except for the occasional brief note in site reports or regional archaeological publications (Avnholt 1944;Henriksen 1976:80;Mathiassen 1943Mathiassen :69-91, 1959Salomonsson 1968:263-268). Spurred on by some of his discoveries at Ageröd, Larsson (1978) dedicated a short section in the Ageröd I:B and I:D site report to discussing 14 Mesolithic 'depots' found in Scandinavia. Just over one page in length, until the early 2000s this text represented the most extensive discussion of this phenom enon anywhere in Europe (Karsten & Knarrström 2003:91-101;Sjöström 2004:43-44). Several other researchers, including Hammarstrand Dehman and Sjöström (2009:19-20), Karsten (2001:125-126, Karsten & Knarr ström 2003, Kjällquist et al. (2016:256-259), and Petersen, E.B. (2015:77-79) all also mentioned, albeit briefly, Mesolithic hoards found during their excavations and contextualized them against findings from other sites. Table 2 summarizes the main texts discussing southern Scan dinavian Mesolithic hoards.
Based on some of these more recently discovered hoards, especially from the RönneholmAgeröd bog complex, the assumption that such hoards dated to the Bronze Age or the Neolithic (Salomonsson 1957;Ebbesen (Sjöström 2004:44). Re cently, a few researchers have even stated that hoards were not uncom mon during the Mesolithic (Sjöström & Hammarstrand Dehman 2015:17;Kjällquist et al. 2016:270;contra, Solberg 1989:267;Koch 1998:158). Meso lithic hoards have also been mentioned alongside other Mesolithic practices interpreted as ritual (Bradley 1998;Strassburg 2000;Koch 2004:333-335;Toft 2009:614-620;Sørensen, S.A. 2017a:38;Pedersen & Petersen, E.B. 2017:237-258). Occasionally, Mesolithic hoards are used to contextual ize Neolithic as well as later prehistoric depositional practices (Karsten 1994:166-170;Bradley 2017:72, 108), mostly in the service of an argument for the longue durée of such practices and, not least, a degree of continuity between Late Mesolithic and Neolithic hoarding practices. Beyond the isolated studies mentioned above, Mesolithic hoards are nei ther robustly incorporated into our general understanding of the southern Scandinavian Mesolithic (for example : Price 1985;Larsson 1990;Søren sen, S.A. 1996;Grøn 2003;Blankholm 2007, Sørensen, L. 2014Astrup 2018;Sørensen, M. et al. 2018), nor have they hitherto been directly com pared to Mesolithic mortuary practices. In fact, and despite the recent in crease in interest in Mesolithic hoards and hoarding, many researchers state that such depositional practices are either entirely unknown prior to the Neolithic (Solberg 1989:267;Sørensen, L. 2014:129) or are 'almost non existent' (Grøn & Sørensen, L. 2018:960) and 'did not become a definite, important feature of the rural life of the Danish people before the Neo lithic way of life was established' (Koch 1998:158). The reluctance by some researchers to acknowledge that hoarding was practiced in the Mesolithic has been argued to be at least partly be due the difficulty of demarcating Mesolithic hoards from other forms of more quotidian practices that took place in similar contexts (Larsson 1978:164). For example, the wetland lo cations, in which hoards are often found, were seemingly a part of every day life in the Mesolithic, compared to the Neolithic, where such areas are seen as being more liminal and ritual in nature. Furthermore, more gen erally Karsten and Knarrström (2003:127) have previously noted that 'no systematic or penetrating study of ritual finds from this time has ever been carried out; the discussion has had the character of small forays concern ing individual sites or objects'.

Exploring ritualization in the Mesolithic of southern Scandinavia
In contrast to Mesolithic hoards, Neolithic and Bronze Age Scandina vian hoards have been investigated extensively. Their interpretation has remained largely consistent, generally seen as either profane storage in tended for later retrieval or as a permanent ritual deposition (see Karsten 1994:9-31 andBerggren 2010:44-104 for extensive overviews of Scandi navian hoarding and depositional research). The concept of ritualization has seen relatively little use within Scandinavian hoarding research (how ever, see Larsson 2004;Sørensen, C. et al. 2020), but it has been success fully applied elsewhere, for example, to other Scandinavian Mesolithic and Neolithic practices (Nilsson Stutz 2003;Berggren & Nilsson Stutz 2010;Berggren 2010. Ritualization as explored here has its roots in practice theory applied to the study of rituals in ethnographic settings by Bell (1992) and subse quently within Scandinavian Stone Age research (Nilsson Stutz 2003;Berg gren 2010). In brief, this comes from the perspective that all actions lie on a spectrum of ritualization (Bell 1992:140-142;Berggren 2010:379). Thus, practice theory provides an alternative framework for understanding prac tices that moves beyond Durkheim's (1915) dichotomy between ritual and profane, now widely appreciated to be largely a postenlightenment Western construct that is neither ethnographically nor historically attested (Brück 1999). Instead, the line between ritual and profane life is either blurred or nonexistent in many societies. Thus, rather than dividing practices into strictly profane or strictly ritual, from this practice theory perspective the focus is on identifying and understanding the role of ritualized practices and the actions and processes that make up a practice (Bell 1992).
The emic meaning of practices is not only contextually and culturally dependent, but can vary between participants and between different per formances of the same practice (Bell 1992;Bloch 2005;Keane 2008:111). Thus, discussing the meaning of a practice is of less importance than fo cusing on the attributes of the practice, and how and when these change, as well as the societal role and function the practice may have (Berggren & Nilsson Stutz 2010:176). In contrast to focusing on the meaning, ritual ized practices are instead are seen as establishing 'relations between peo ple, things, places and anything else that plays a part in the act' (Berggren 2010:380). This marks a key aspect of practice theory in terms of how rit ualization relationally integrates and differentiates not only practices but also communities of participants, as well as the places and the things in volved (Bourdieu 1977:120;Bell 1992:125).
In Bell's analysis, mechanisms of differentiation and integration are key attributes of ritualized practices: they are the very means by which a practice becomes ritualized, and their strategies of ritualization may set them apart from as well as connecting them to other societal practices (Bell 1992:74-93, 125). Strategies of ritualization represent individual or groups of ac tions, and aspects of those actions within a longer performance constitute the entire practice. Many of the strategies of ritualization which Bell and other anthropologists have focussed on would leave no archaeologically identifiable traces; this would for example be the performance by particular people, the use of specific gestures, movements, sounds or words, and the use of specific clothing, food or other substances (Bell 1992:90, 204-207). Thus, much of the archaeological use of these concepts, especially in Scan dinavian Stone Age research, has instead relied on either focussing on the use of certain forms of material culture, specifically axes (Larsson 2004;Sørensen, C. et al. 2020) or particular places or features, such as the offer ing fen of Hindbygården (Berggren 2010) or special transitional times such as death and burial (Nilsson Stutz 2003).
In this paper, the biographical approach, practice theory and the con cept of ritualization and its prior archaeological uses (Nilsson Stutz 2003;Berggren 2010), are used to guide the analysis and interpretation on all ob servable lifestages of the objects and the assembled Mesolithic hoards. In this way the application of these approaches is pushed further back in time (both chronologically and in an object biographical sense) and is applied to a largely overlooked prehistoric practice. These perspectives are used to explore whether ritualization provides a likely explanatory scenario for the observable empirical results and the practice as whole, rather than focus sing on the interpretation of individual hoards.

Characterizing southern Scandinavian Mesolithic hoards COMPOSITION
A wide variety of object types are found in the Mesolithic hoards. The large numbers of hoards containing blades, axes, partially worked bone/antler, or cores/nodules are noteworthy ( Figure 3). This suggests that generic tools (blades or axes) or raw material (flint nodules and cores, or partially worked bone and antler) were most commonly deposited. Occasionally, more ac tivityspecific objects were also hoarded, such as bone points, beads and arrowheads. The domestic tools so common on Mesolithic sites such as scrapers, burins and borers are rarely or never included in the hoards, sug gesting a selection process underlying the composition of the depositions, yet this selection process appears not to be dictated by either functionality or time/material investment. Likewise, some patterning may also be pre sent in the types of objects generally found in singleobject type or mixed hoards (Table 3). For example, it is notable that bone points, blades and partially worked antler/bone are rarely deposited with other object types, whereas cores/nodules, hammerstones, flakes and to an extent axes are comparatively often deposited with other object types. Thus, the selection of objects deposited together also appears to have been influenced by nor mative ideals regarding the role, importance or relationship different ob jects had with each other.
In general, the hoards consist of quotidian objects, although a few con tain rare or even unique artefacts. One interesting example is an unpar alleled large conicalshaped chalkcovered flint nodule that was found to gether with two extremely long bone points in the refuse layer at Tågerup. These bone points have been interpreted as a possible tattooing needle and a clothes/hair pin (Karsten & Knarrström 2003:91). Likewise, it is also pos sible that the apparent bead hoards, all found in wetlands, should be con sidered as having a special significance. They contain beads in such great numbers that similar concentrations are only known from later burials. For example, 21 tooth beads from a variety of different animals were found in a bog (Sørensen, S.A. 2017b:226) at Skamstrup. Near another Magle mose settlement at Øgårde, 98 tooth beads were found in a tight cluster (Mathias sen 1943:91), and on the outskirts of the settlements of Ageröd IV and Ageröd V, a Kongemose hoard of 33 hazelnut beads was placed next to 13 blades and a large core (Larsson 1983).
Some of the hoards contain nonlocal objects or material. One of the most dramatic examples is the deposit of two Rössen shoelast adzes and a flat stone axe found in a pit during peat digging at Udstolpe (Lomborg 1962). These axes likely originate from Thuringia or Lower Saxony (Sørensen, L. 2014:129), and had thus travelled over 300-500km before they were depos ited during the Late Ertebølle on Lolland. At least three other hoards from Snyggatorp (Salomonsson 1957), Flaekkemagle (Fischer 2004a) and Rönne holm FP 878 (Sjöström 2011:14-16) contain nonlocal material -flint in the first two hoards and shell beads in the latter -indicating the movement and curation of material across regions. Two of these hoards (Snyggatorp and Rönneholm FP 878, see Table 5) have macroscopically visible evidence of use, suggesting that they were not transported and curated solely for depo sition, but rather their deposition was the end phase of a longer uselife. In contrast, it has also been noted by Larsson and Sjöström (2013:494-495) that many of the blade hoards found in the wetlands at Rönneholm may have transported for deposition, as they were not produced on any of the known sites in the area and rarely show evidence of wear.
A few hoards, all dating to the Maglemose or Kongemose, contain un usually large tools. The aptly named site of Flaekkemagle (big blades) in particular demonstrates this, as hoard of 13 blades that were 18-20cm long was found tightly bundled together (Fischer 2004a:30). A similar hoard, dating to the Early Mesolithic, was found wave dispersed in the refuse layer at Norje Sannusund and contained 37 blades up to 18cm long (Kjällquist et al. 2016:256-259). The aforementioned Maglemose hoard from Rönne holm 8 that was found deposited away from the main settlement area, con tained 108 large flint blades up to 15cm long, many of which are of excep tional quality and show no traces of use (Sjöström 2004:28). A few bone point hoards also contain unusually long points, such as that at Horne Terp.
Here, five bone points, all circa 30cm long, were found bound together dur ing peat digging in a bog (Andersen, S.H. 1978). Two similar hoards were found within possible fishing areas at Øgårde 9 and 14, one hoard contained five points 14-26cm long (the shorter bone point has evidence of resharp ening) and the other hoard from Øgårde 14 contained three bone points 20-28cm long, respectively (Andersen, K. 1983:165-166). In addition, the three flint picks (one 44cm and two 29cm) found together at the settlement at Sjöholmen, represent some of the longest flint objects in Sweden (Karsten & Knarrström 2003:94). Although these hoards contain unusually large ob jects, other than perhaps the hoard from Sjöholmen, they were likely fully usable objects and thus stand in contrast to the oversized, presumably cer emonial axes found in some Neolithic hoards (cf. Sørensen, L. 2014:176).

CHRONOLOGICAL AND REGIONAL VARIABILITY
In most cases, a chronological assignment is only possible by either typologi cal dating of the artefacts found in the hoards themselves or those found at an associated settlement. In a few instances, radiocarbon dates are available from the associated settlement, but uncertainties as to the contextual asso ciation of hoard and settlement remain. Furthermore, many settlements may have seen repeated occupation over extensive time spans. In all 16 hoards could not be typologically dated more precisely than to the Mesolithic, and in some cases the hoards contain only artefacts that could date to either the Mesolithic or the Neolithic. These latter hoards have been omitted from chronological analysis. In many cases the difficulttodate hoards contain either blades or trindøkser (roundbutted pecked stone axes) as stray finds. composition. The top row shows the chronology of all hoards, the lower rows show the chronology for different object types. The dating for the hoards is often either typological or based on the associated sites; thus many hoards may have broad date ranges. The time line is divided into 100year increments. Where a hoard has a date range over, for example, five centuries, each century is assigned a value of 0.2, as the one hoard is divided by five. In this way the colour scales indicate the number of hoards per century. The colour scales are nearexponential, to accentuate differences between periods containing only few hoards and periods where hoards are more significantly common. > Based on the remaining 108 hoards that are either typologically datable or come from dated sites or contexts, there is notable chronological vari ability in the frequency as well as the composition of the deposits. Figure 4 shows that hoarding was least frequent in the Early Mesolithic. This rar ity continues into the earlier Kongemose. At circa 6000 BC the number of hoards increases, peaking at the end of Kongemose with 43 hoards that likely date to this period. The number of hoards then drops again during the early phases of the Ertebølle only to grow once more during the middle and Late Ertebølle. The reason and significance of this variable frequency of deposition is unclear, but it is likely due to multiple taphonomic as well as societal factors. In addition, the increase of deposition in the later Magle mose to the Late Kongemose is skewed by the relatively large numbers of hoards found at RönneholmAgeröd that date within this phase.
Links may be made between some of the chronological changes in the hoard composition and wider changes within societal structure. The de crease in bone point hoards after the Maglemose could reflect a general change in fishing practices and settlement locations in the Kongemose and Ertebølle towards coastal regions, as well as an increasing reliance on terres trial products rather than lacustrine fishing (Schilling 1997;Astrup 2018). Lacustrine bone point depositions may also be less exposed to dispersion than those on coastal sites, so the data may be skewed by such taphonomic conditions. General changes in lithic technology may explain the possi ble shift away from microliths in the Maglemose to an increased focus on blades in Kongemose hoards, which matches the greatly increased reliance on larger blades in the Kongemose compared to the microlithic industries in the Maglemose (Sørensen, S.A 2017a:37-38). Finally, the growing focus on axes starting in the Late Ertebølle may suggest that local Mesolithic prac tices were influenced by contact with Central European Middle Neo lithic scouting groups that likewise hoarded axes, in line with previous studies (Karsten 1994:166-170;Koch 1998:158;Sørensen, L 2014:129). By the same token, these results reveal that hoarding of axes was known from the ear liest Maglemose and occurred, albeit sporadically, throughout the Meso lithic. This challenges earlier studies that have suggested axe hoarding to have been introduced by contacts with external Neolithic groups. Notably there is a shift in the focus towards axes, rather than the introduction of an entirely new practice in the later Ertebølle.
Regional variability in the composition as well as number of known hoards is also evident in the Mesolithic hoards (Table 4). Some of this may be a product of regionalised research history; in particular, the Swedish dominated Mesolithic hoard research, as well as the relatively large num bers of hoards found in wellstudied regions or sites such as the Rönneholm Ageröd complex, Vedbaek fjord and Åmosen. Nonetheless, this does not  fully account for the regionalised composition of hoards; note especially the proportion of hoards that contain blades, bone points and arrowheads. The regionalisation of hoarding is mirrored in tendencies of both micro and macroregionalization in other aspects of the Mesolithic; see for example microlithic technology (Larsson 1978;Andersen, S.H. 1983;Petersen, P.V. 1984;Blankholm 1990) and art (Toft 2017) varies across southern Scandi navia in the earlier Mesolithic. In the Late Mesolithic there is observable regionalisation in different types of material culture including Tshaped ant ler axes, bone rings, bone combs, Limhamn and flint flake axes (Petersen, P.V. 1984), ornamentation (Andersen, S.H. 1980(Andersen, S.H. , 1986Nash 1998;Toft 2017:259-271) and pottery (Sørensen, L. 2015).
In addition, a degree of 'localisation' -site or microregional pattern ing -is also observable. For example, all of the hoards from Siggård (three hoards), Lystrup (two hoards) and Husted Mose (two hoards) contain blades, and an unusually large proportion of the hoards containing blades were found at the bog complex at Rönneholm (53 per cent compared to 20 per cent across the remaining hoards in southern Scandinavia). In con trast, only two out of the more than 16 hoards found at various sites around Åmosen contain blades. Instead, at these sites, the deposition of osseous objects (bone points, tooth beads and partially worked bone) is strongly represented. Where a site or microregion contains multiple hoards, lo calised traditions can be discerned. This pattern of localisation suggests that certain persistent practices were often tied to particular sites, areas or groups (Barton et al. 1995). As the hoards in question were often found in sites with multiple occupation phases and have not been precisely dated, the temporality and longevity of these localised traditions must unfortu nately remain unclear.
The variability suggests that although hoarding practices did occur throughout the southern Scandinavian Mesolithic, they were not static but varied temporally and spatially. Such variability is to be expected given the long duration and the relatively large study area, and it likely reflects changes in for example population structure, mobility or subsistence, econ omy, material culture or even ideological differences. However, in contrast to this expected variability, the observed localisation suggests that certain practices were closely linked to particular locales or were normative within the local groups. Thus, an interplay between notable variability on the one hand but also multiscalar patterning is observable. This patterning is sug gestive of driving forces behind the composition as well as the depositional contexts, and it therefore implies that these practices are not isolated occur rences but seem to be dictated by underlying normative ideals and habits, that are present at different spatial and temporal scales.

PRE-DEPOSITIONAL LIFE HISTORY
To understand Mesolithic hoarding practices more fully, the analysis of predepositional lifehistories (including: the uselife and the treatment of the individual objects as well as the entire assembled hoard) constitute a key aspect of this study. In the present analysis, 15 of the 124 hoards were subject to usewear examination. Those, as well as prior usewear stud ies (Salomonsson 1957;Arbman 1954:6;Karsten 2001:126;Knarrström 2001;Sjöström 2004Sjöström :44, 2011Berggren 2007:116-117) show that Mes olithic hoards often contain used objects (Table 5 and Appendix 2). The degree of use and the type of material worked varies significantly, both between different hoards and between different objects found within the same hoard. Some of these objects have little usable life left, due to exten sive prior use or extensive damage (see below), suggesting that they were not being cached for later use. Some of the objects within a hoard showed no usewear traces, appearing entirely pristine and useable. These unused objects are however, often deposited with used objects. Whilst combining objects with different uselives may have been a feature of the hoarding practice, there is no observable 'idealised biography' (Fontijn 2002) for the objects, nor an idealised combination of biographies that were deposited together in these hoards.
Many Mesolithic hoards contain fragmented objects, such as blades, axes, bone points and bones. In some cases, this may be explained by post depositional disturbances including ploughing, peat extraction or excava Figure 5. Examples of usewear traces, including striations and edge damage on two of the axes (under) found in a hoard from Maglelyng XL, Zealand with two core axes, two flake axes and three cores (above). A flint nodule from this hoard is not pictured as it does not appear to have been handed in to the National Museum of Denmark. tion damage. In others, the damage may have occurred in prehistory, as part of manufacturing processes or during use. There are also indications that some objects in the hoards were damaged prior to deposition, to an extent that they likely would not have been usable. One of the Kongemose hoards from Tågerup offers an example: a core axe with a heavily burnt flint pick found lying parallel in the dryland area of the settlement (Karsten & Knarrström 2003:91). Of the nine axes found at Porskjaer Bakke, two were found as fragments (Klaus Hirsch pers. comm.). Several other hoards also contain axes with edges so heavily damaged that their future functionality would have been compromised. Some objects appear to have been inten tionally fragmented. For example, the four core axes that were part of the Late Maglemose hoard found in a pit at Arreskov Sø all had their edges re moved by repeated flaking parallel to the edge ( Figure 6). This left the axes likely too short to have been resharpened any further, and their edges too irregular and too obtuse to be used as axes. This modification, prior to the deposition of these objects, rendered them useless as axes, a treatment that is mirrored at the site of Ageröd I:B, Scania. This hoard, which was found in the refuse layer, contained 41 fragmented as well as intact microliths tightly bundled together (Larsson 1978:67; Figure 7).
Due to the very careful excavation and stratigraphic recording of this hoard, it was possible to determine that the microliths had been broken, likely using percussion (based on Jennings 2011), prior to being wrapped in some sort of organic container and deposited on the periphery of the settle ment (Larsson 1978:67). Seemingly intentional destruction of objects can also be seen at Bjällvarpet, where in a pit that lay underneath a hearth, a severely fragmented grinding stone was deposited with two rounded peb bles placed on either end (Hernek 2005:272) likewise, two intentionally fragmented stone pestles were found deposited on a hearth at Timmerås (Hernek 2005:268).
As some objects seem to have been intentionally fragmented and some even are unusable when deposited, it is unlikely that these hoards were used as storage for later use. Some object types, such as blades, were used very briefly but would have taken very little time or effort to produce, speaking against a need for storing large quantities of these objects.

DEPOSITIONAL TREATMENT
One of the most observable depositional practices is for objects in Mes olithic hoards to have been carefully arranged. Some objects are found stacked together, as seen in the flint cores and nodules at Nørre Sandegård (see Figure 2) and the bone and antler hoards found in the refuse layer at Ringkloster (Andersen, S.H. 1975:19, Figure 5a, b, 1998:28, Figure 12) and Ageröd V (Larsson 1983:79, Figure 49). Other objects are found lying parallel to each other, such as the two roundbutted stone axes found at Sjövreten (Welinder 1977:47, Figure 29). A similar arrangement is seen in particularly wetland hoards with worked metapodials such as Ageröd V, see In some hoards objects were apparently deposited standing vertically. Examples of this, according to the finders, are the core axes and pick from Sølund (Kaj F. Rasmussen pers. comm.), two core axes from Dagstorp (Rydbeck 1918:7), some of the blades in a hoard from Husted Mose (NM A 48298-A 48302, National Museum of Denmark) and two bone points from Siretorp (Montelius 1917:107, Figure 46). In situ photos of the hoard from Arreskov Sø show one of the axes standing vertically, whereas two other axes lying parallel to each other but with the axes facing opposing directions with their edges facing up (see Figure 2). Even more striking is the hoard of twelve greenstone roundbutted axes from Hasselfors placed, apparently, in the shape of a sun (Hermansson & Welinder 1997:70). The hoard of flint nodules from Maglemosegårds Vaenge ( Figure 8) and a hoard of flint blades from the shellmidden Doverodde may have been arranged in a circle, with the latter placed around a large stone (Klaus Hirsch pers. comm.). At Svaerdborg, half of the microliths in a hoard were in a radiat ing fan shape (Henriksen 1976:80).
One of the most common arrangements found in hoards is tight packing of objects. This is particularly common in blade hoards, where the blades  (Figures 8 and 9). This is seen in in situ photo graphs of some blade hoards, but also likely in a few other examples of blade hoards based on their find descriptions. For example, the hoard from Snyg gatorp was described by the finder as 'lying together like the segments of an orange and sorted according to size' (Salomonsson 1957). The blades in the hoard from Revlen XI were said to be found lying neatly on top of each other next to two parallel elk and red deer bones (Andersen, K. 1983:94), and an Ertebølle blade hoard from Skal is described in a similar way, with five blades lying neatly on top of each other (Simonsen 1952:214).
The tight arrangement of objects in many hoards, including the afore mentioned blade hoards, suggests that they were wrapped or bound in an organic material (Larsson 1978:70;Salomonsson 1968). A few hoards had such wrapping remains still preserved at the time of discovery, such as 13 large flint blades from Flaekkemagle (Stafford 1999:70), five bone points from Horne Terp (Andersen, S.H. 1978), and a bundle of split red deer and elk bones from Ageröd V (Larsson 1983:79) were originally wrapped with plant material such as bast and birch bark. Similarly, the 20 or so bone points from Garbølle Mose were apparently wrapped in hide (NM A 42158-42159, Danish National Museum), and a hoard of microliths were found packed in a broken bone in an area of Maglemose flint at Kongemosen by Anders Fischer (pers. comm.). Although comparatively few hoards have their wrappings or containers preserved, it is deemed likely that Mesolithic hoards were often originally wrapped or placed in containers, given the tight clustering of many of the hoards.

DEPOSITIONAL CONTEXT
Throughout the Mesolithic, hoards are primarily found in wetland con texts (58 per cent) or within settlement areas (65 per cent). Often these two contexts overlap, as most Mesolithic settlements are near bogs, lakes, rivers and along the coast. Unfortunately, the published records of these hoards are often not reliable for determining if the wetland areas were wet or dry at the time of deposition or how accessible they were. Furthermore, as many hoards have been found in or near the wetland edges of settle ment, differentiating between these two depositional context categories is almost impossible. Chronological variability in depositional context is observable ( Figure  10). The reasons behind these changes are unclear. However, in some cases it may be due to overrepresentation of certain objects deposited during particular stages of the Mesolithic and preferentially found in certain con texts (see below), such as hoards of bone points dating to the Maglemose and one Kongemose found in wetland fishing areas ( Figure 4 and Table 6). Temporal changes in depositional context can also be impacted by local ised trends. However, the increase in extramural hoards, particularly axes (six out of eight in the later Ertebølle) may be attributable to influences from contacts with Central European Neolithic groups (Kaufmann 2012;Pétrequin et al. 2012).
The extramural hoards are frequently made up of stray finds, discov ered during peat digging or agricultural work. It cannot be determined if these represent hoards placed away from settlements or if settlement ma terial around these hoards was not observed or collected. There are only two examples of excavated hoards that were found away from any known settle ments: the four flint picks from Tissø (Fischer 2004b) and two used core axes from the Late Ertebølle hoard of Hindbygården (Berggren 2007). In addition, some of the blade hoards from Rönneholm, have been consid ered extramural (Larsson & Sjöström 2013:494-495), but as they are found in the vicinity of settlements their context is more questionable. Neverthe less, given the large number found away from any known settlements, the results do support the notion that hoards were occasionally deposited ex tramurally in the Mesolithic. < Figure 10. Chronological variability of the depositional contexts of southern Scandina vian Mesolithic hoards. The top row shows the chronology of all hoards, the lower rows show the chronology for different depositional environments and contexts. The dating for the hoards are often either typological or based on the associated sites; thus many hoards may have broad date ranges. The timeline is divided into 100year increment with the col our scales indicating the number of hoards per century. The colour scales are near expo nential, to accentuate differences between periods containing only few hoards and periods where hoards are more significantly common.
The composition of the extramural hoards differs from those found within settlements (Table 6). Many object types, in particular blades, cores and partially worked antler/bone, are primarily found in settlement areas, whereas bone points and axes are found in settlements or extramurally in roughly equal proportions. These results may further suggest that, rather than just being discrete depositional events, the treatment of objects through their deposition was guided by relatively widespread accepted principles.
Although the exact depositional context is unknown for a large number of these hoards, they appear to be placed in a wide variety of different set tlement contexts. Generally, throughout the Mesolithic there is a preference for settlement hoards to be placed in apparent refuse areas, on the surface of a settlement, and also often in pits (Table 7). The numbers of hoards found in refuse areas could suggest that these areas served more variable purposes than often assumed; this is in line with other discussions of such areas (see also Karsten 2001:144;Carlsson 2008:164-173;Sørensen, L. 2014:129).
Mirroring the localisation seen in the composition of the hoards, there also appears to be localisation in the depositional context. For example, all the hoards found at Lystrup (Søren Andersen pers. comm.), Bökeberg III (Karsten 2001:125-126), Ringkloster (Andersen, S.H. 1975(Andersen, S.H. :19, 1998, Ageröd V (Larsson 1983:79-81) and two of the three hoards from Tågerup (Karsten & Knarrström 2003:91-97) were found in the refuse layers. Whereas at Ulkestrup Lyng (Andersen, K. et al. 1982:42, 98) and Øgårde (Andersen, K. 1983:30, 165-166) the hoards are mainly placed away from main settlements, in areas interpreted as fishing places. The three hoards from Skal were found on the surface of the settlement, and at Rönneholm Ageröd the hoards were often found away from the main area of the set tlement or indeed away from any known settlement.
Various forms of wooden markers may also have been used. For exam ple, at the Maglemose site of Ålyst, eight large flint nodules were found in a posthole of a hut (Casati & Sørensen, L. 2012). Similarly, at Timmerås, two roundbutted stone axes were found near to two postholes (Hernek 2005:273) and at Maglelyng XL two vessels were found lying up against a wooden post in the bog (Koch 1998:157). Furthermore, a few hoards, from Bøgebakken (Avnholt 1944:56), Rönneholm 9 (Sjöström 2004:33), Husted Mose (NM A 48298302, Danish National Museum), and a hoard of flint nodules from Maglemosegårds Vaenge (see Figure 8), have been found next to a preserved tree or roots. These forms of marking would have acted as a physical reminder of the events surrounding the deposition and the lo Table 7. Depositional context of Mesolithic hoards found associated with settlements. Number  81  22  17  15  13  5  3  3  3  1   Percent  65  27  21  19  16  6  4  4  4  1 cation of the hoard. These may been important if the hoard was intended be retrieved and/or if the act of hoarding or the hoard itself was symboli cally important.

Ritualized Mesolithic hoarding
The significant increase in the number of recognised and analysed hoards in this study has resulted in a more coherent understanding of southern Scan dinavian Mesolithic hoarding practices. Although there are various biases (Larsson 1978:164;Sjöström 2004:43) and difficulties with data qualitymost commonly derived from older and at times poorly published excava tions -there is enough evidence to argue confidently that hoarding prac tices took place as early as the Early Maglemose, continued throughout the Mesolithic, and were common enough to be considered a marked feature of the southern Scandinavian Mesolithic. The results discussed here indicate a degree of continuity between the Mesolithic and Neolithic, particularly in the predominance of the axe hoards (see also Karsten 1994:166-170), with a shift to more extramural hoards in the Late Mesolithic and into the Early Neolithic. Within particular attributes of Mesolithic hoarding practices, there is some degree of patterning and also high levels of idiosyncrasy (Table 8). This variability might initially seem to complicate interpretations of the practices, with some indicating a more profane interpretation of either stor age or waste disposal, and others indicating a ritual role. For example, the propensity for hoards to be found within or close to settlement areas, and to contain raw material or useable objects, fits our expectations of profane caching: easily retrievable, usable material and everyday objects (Binford 1979;Schiffer 1987:78-93;Galan 2007:77-79). Other hoards, like those found at Ringkloster (Andersen, S.H. 1975:19, Figure 5a, b, 1998, fit what is normally assumed to be evidence of waste disposal: found in refuse layers and containing utilised or partially worked material. Yet, the careful arrangement, the wetland depositional context of many of the hoards, as well as the damaged, even intentionally destroyed, nature of some objects are not easily reconciled with a profane interpretation. Instead these attrib utes fit a classic archaeological interpretation of ritual (e.g. Levy 1982:17-25). These 'ritual' aspects fit how a number of Mesolithic archaeologists have interpreted other forms of depositional practices as part of some sort of tripartite Mesolithic cosmology -especially focussing on the wetland and liminal nature of the depositional contexts and a particular importance of axes (e.g. Bergsvik 2009;Glørstad 2010:229-247;Blinkhorn & Little 2018). However, the aforementioned seemingly 'mundane' characteristics of many of the Mesolithic hoards do not fit neatly into common conceptions of Mesolithic cosmologies. In sum, we are left with a somewhat conflict ing picture of the hoarding practice, if we try to fit it into either category. This conclusion is hardly surprising, given that such a dichotomy be tween ritual and profane appears to be a largely a postenlightenment west ern construct, that it is neither ethnographically nor historically widely at tested (Brück 1999). Thus, the Mesolithic hoards are seen here to be more in line with the concept of ritualization (Bell 1992), with this ritualized practice taking place in quotidian contexts and using quotidian materials that are often usable, well used or raw material.
Such ritualized practices may be archaeologically identifiable by mate rial differentiation from other practices (Berggren 2010:379-380). Large scale deliberate structuration of wider practice is considered evidence of one such strategy, visible through 'material cultural patterning' (Garrow 2012). This may be observable in the localisation patterning of some traits of Mesolithic hoarding (for example in composition and depositional con text) as well as longue durée of similar traits within the hoarding practice. For instance, there appears to be patterning in the type of objects that are included within the hoards (especially blades and axes). There is also pat terning, in the types of objects deposited as either single object type (par ticularly bone points) or in mixed hoards (such as cores, hammerstones and flakes). In contrast, other objects, such as burins, flakes, scrapers and borers, are conspicuously absent or rarely found in hoards, even though they are common finds in settlements. This could suggest that strategies of ritualization focused on specific object types were in place, related to the differentiated use of object types in different spheres of Mesolithic life.
There also seems be patterning within the object biographies. The over whelming majority of the analysed hoards contain objects that were not only used, but had remarkably different uselife histories, in terms of cu ration, treatment and use. This suggests that the uselife and the combina tion of objects with different biographies appears to have been a key part of the practice -a feature that may be more difficult to fit with a storage/ based interpretation of these hoards. Instead, I argue that the differences in the objects' uselives may have served as a means of individualizing the ritualization of each hoard, reflecting different forms of socialtemporal spatial relationships formed during the objects' lives, which became objec tified and transformed through their participation in the hoarding practice (Bell 1992:216;Berggren 2010:280;Baires & Baltus 2016;Baltus 2018;BjørnevadAhlqvist 2020).
Strategies of ritualization might have taken place at different levels of Mesolithic hoarding practices. Mesolithic hoards may have been part of a ritualized tradition, and they may also become ritualized on an individual level. Individual actions may have ritualized the hoard: the bundling or careful arrangement of some hoards, the intentional fragmentation of ob jects, perhaps the marking, as well as the assemblage of objects with par ticular and divergent biographies. Thus, the patterning and idiosyncrasies in Mesolithic hoarding practices might be explained by strategies of ritu alization on a multiple scales, combining prior traditions with a high de gree of individualization and flexibility as a part of the performativity of the practice (Berggren 2010:379-380;Berggren &Nilsson Stutz 2010:176).
Ultimately, the meaning of a given ritualized practice will largely lie out side the grasp of archaeology; it is anthropologically acknowledged that often the meaning of a practice may not be known within a given society (Bloch 2005) or may vary between different members (Keane 2008:111).
The theoretical framework employed here makes no attempt at identify ing any specific meaning of a practice, rather the function of the practice is discussed. Ritualized practices are instead known to act as mechanisms for relationship construction, memorialization, and for increasing group sociality (Bell 1992;De Boeck 1995;McCauley 2001;Atran & Henrich 2010;Peterson 2013;Xygalatas et al. 2013;WatsonJones & Legare 2016;Hobson et al. 2018). These functions may not be intended or even realized by the participants, but they are important byproducts of the performance of such practices (Hobson et al. 2018). Thus, combining practice theory and ritualization with these cognitive science insights allows us to discuss the strategies of ritualization employed during the Mesolithic as well as the role of the ritualized practices. From this perspective, the production, use, curation, treatment, accumulation and subsequent deposition of these hoards are seen to have enchained the objects, people, and moments in time and places together with these hoards, acting as mnemonic devices, place making entities and mechanisms for increasing social cohesion.

Conclusions
The present study represents the first largescale analysis focusing on south ern Scandinavia Mesolithic hoards and the first to analyse this material in detail. It is now possible to understand the general characteristics and vari ability within the southern Scandinavian Mesolithic hoards and the prac tices behind them. Hoarding as a practice is not merely a Neolithic phe nomenon, rather it stretches back to at least the Early Maglemose and can be identified throughout the entire Mesolithic (contra Solberg 1989:284;Sørensen, L. 2014:129). Southern Scandinavian Mesolithic hoards vary significantly in terms of their composition, biographies, and depositional treatment. At the same time, they also show notable objectbased pattern ing as well as temporal, regional and even localised structuration. Some of the variability may be due to wider societal differences and demographic dynamics, whereas other differences may indicate localised traditions of certain groups at certain times.
Rather than representing the culmination of strictly profane or ritual ac tions, the hoards appear to emphasize the inseparable nature and intercon nectedness of these spheres in Mesolithic daily life. The use of ritualization to interpret the everydaytype materials and contexts stands in contrast to the prior Scandinavian research into ritualized practices that has focused on more extraordinary sites (e.g. Hinbygården, see Berggren 2010), ex traordinary lifestages (such as death and burial, see Nilsson Stutz 2003), or spectacular objects ('ceremonial' Neolithic axeheads, see Sørensen, C. et al. 2020). The more quotidian nature of the Mesolithic hoarding thus stands in notable contrast to the wellknown and often more spectacular Early Neolithic hoards. Yet based on the expanded data presented here, similarities and continuities of the hoarding practices across the Mesolithic-Neolithic transition are becoming more apparent.
The roles of ritualized practices such as Mesolithic hoarding are inter connected with the various traits of the practice -in particular the impor tance on object biographies (Kopytoff 1986) -as well as innate aspects of human cognition (Hobson et al. 2018), rather than being contextually or culturally dependent. From these perspectives ritualized hoarding is seen as integrating and differentiating not only the practice, but also the com munities of participants, the places and the things involved with the prac tice -through the construction, maintenance, modification and demarca tion of relationships (Bell 1992:130).
The hoards presented in this paper likely represent only a fraction all Mesolithic hoards that existed or have been found, given the biases that may have impacted the available material. It is nonetheless now clear that hoarding was indeed an important, but often overlooked, feature of the southern Scandinavian Mesolithic. The analysis of these hoards provides key insights into the treatment and perception of objects, materials and practices throughout the Mesolithic and how the intersection of objects, time and space helped shape Mesolithic worldviews and social relation ships.

Acknowledgements
I would like to thank the following numerous colleagues for information, access and permission to publish many of the hoards in this paper, and with out which this paper may not have been I would also like to thank Helle Juel Jensen for her help with use wear analysis and insightful discussions and Clara Fischer Stephansen for her help creating the distribution map of Mesolithic hoards. Special thanks to Jorgen Jensen for creating an algorithm that facilitated a new means of producing the timelines used in this paper. I would like to express my grat itude to Laura Ahlqvist, Hazel and Jorgen Jensen and Felix Riede for re visions and many valuable suggestions. Finally, I would like to thank the two anonymous reviewers and the editors for their useful comments and suggestions that improved the quality of this paper. Arreskov Sø Four flint core axes, whose edges have been removed. Three of the axes were found lying next to each other, with one placed vertically.
Western Denmark 8 Bjällvarpet Two rounded stones were placed with an intentionally broken grinding stone. The oval stones were placed on either side of the grinding stone.
Southern Sweden 9 Björkeröds fällad 72 blades, 32 are retouched and only 8 have no use damage. Southern Sweden 10 Bøgebakken An antler tine on a stone next to a flint core axe, one flint flake axe and a hammerstone. Eastern Denmark 20 Gøngehusvej 7 Two flint nodules and a possible anvil stone in the bottom of a pit.
Eastern Denmark 21 Gøngehusvej 7 12 blades found neatly bundled together near to an "alleged burial" and under the cultural layer. Some of the blades were pristine and others had been water-rolled prior to deposition. Present author -macro and low power microscopic observations using the Dino-lite microscope Both axes have macroscopically visible polish, although the extent of this polish varies significantly between the axes. One axe appears to have been re-sharpened and other axe is broken near the butt and the edge.
Present author -macro and low power microscopic observations using the Dino-lite microscope Two of the axes had possible hafting polish, one had had been resharpened after an edge fracture, two of the axes had slightly rounded edges, and one had been reused as a hammerstone leaving an extremely rounded edge. Based on the photo supplied by the finder, the last axehead has a typical fracture on the centre of the edge suggestive that it also was used.
Present author -macro and low power microscopic observations using the Dino-lite microscope and high power microscopy using the Nikon Eclipse microscope Most of the blades had evidence of use, but only two blades and the flake scraper had diagnostic traces. One blade appears to have been held in hide and used on an unknown material causing so-called polish-23 to form. The other blade had traces of use to cut a hard material and the scraper had been used to work dry-hide.

Present author and Helle Juel
Jensen -high power microscopy using the Nikon Eclipse microscope Four of the blades had evidence of use, but only one of these was diagnostic, which showed evidence of being held and used on dry-hide as well as on much harder material like wood or bone/antler.

Present author and Helle Juel
Jensen -high power microscopy using the Nikon Eclipse microscope At least two of the blades had been used on dry-hide, the other blade had undiagnostic use-wear traces. The core scraper was not analysed.

Present author and Helle Juel
Jensen -high power microscopy using the Nikon Eclipse microscope Several blades were water-rolled, the other half are 'pristine'. Petersen, E.B. 2015:79 Too patinated for use wear analysis and no macro-wear. Present author and Helle Juel Jensen -high power microscopy using the Nikon Eclipse microscope Too patinated for use wear analysis and no observable macro-wear. Present author and Helle Juel Jensen -high power microscopy using the Nikon Eclipse microscope None of the axes have any edge damage. Arbman 1954:6 The flint core axe has no traces of use, unclear if the pressure flaker was used.