You are on page 1of 21

See discussions, stats, and author profiles for this publication at: https://www.researchgate.


A ritual site with sacrificial wells from the
Viking Age at Trelleborg, Denmark

Article · October 2015
DOI: 10.1080/21662282.2015.1084730.


1 67

4 authors, including:

Anne Birgitte Gotfredsen Karin Margarita Frei
Danish Natural History Museum The National Museum of Denmark


Lars Jørgensen
The National Museum of Denmark


Some of the authors of this publication are also working on these related projects:

Bronze Age Wool Economy View project

The NOW-project. Living Resources and Human Societies around The North Water in the Thule Area,
NW Greenland View project

All content following this page was uploaded by Anne Birgitte Gotfredsen on 03 June 2016.

The user has requested enhancement of the downloaded file.

Danish Journal of Archaeology

ISSN: 2166-2282 (Print) 2166-2290 (Online) Journal homepage:

A ritual site with sacrificial wells from the Viking
Age at Trelleborg, Denmark

Anne Birgitte Gotfredsen, Charlotte Primeau, Karin Margarita Frei & Lars

To cite this article: Anne Birgitte Gotfredsen, Charlotte Primeau, Karin Margarita Frei & Lars
Jørgensen (2015): A ritual site with sacrificial wells from the Viking Age at Trelleborg, Denmark,
Danish Journal of Archaeology, DOI: 10.1080/21662282.2015.1084730

To link to this article:

View supplementary material

Published online: 15 Oct 2015.

Submit your article to this journal

Article views: 5

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

Download by: [Copenhagen University Library] Date: 20 October 2015, At: 05:43

The National Museum of Denmark. mounts. University of Copenhagen. Furthermore. Denmark Anne Birgitte Gotfredsena*.ku. The settlement. It was to the eleventh century by metal detector surveys. Frederiksholms Kanal 12. Keywords: human sacrifice. This study presents new cross-disciplinary investigations focusing on three sacrificial well-like structures (47. DK-1350 Copenhagen K. Charlotte Primeaub. ment only 300 m east of the promontory where the fortress 2011). Price et al. Environmental Archaeology and Material Science. the rivers Tude Å and Vårby Å in southeastern Zealand 4 ha. silver ingots. hereby providing important and at that time new lished survey carried out by the regional museum by zoological information on our common domesticates and magnetometer has also indicated activity areas and traces on husbandry and hunting practices at a Viking Age of other constructions. strap the same publication. paganism Background graves could indicate that the individuals came from the The circular Viking Age fortress of Trelleborg on the Slavonic areas in the southern Baltic area and thus provid- island of Zealand in Denmark is well known as a military ing further support for the results delivered by the stron- fortress built by Harold Bluetooth in AD 980/981. has produced more than 200 finds from the seventh with a view towards the Great Belt (Figure 1). p.doi. Nørlund also dealt fortress consisted of individuals of both foreign and local with this unique material in his publication. Email: abgotfredsen@snm. The site tium analyses. is of utmost importance for the understanding of the However. Several excavated in the years 1934–42 by Nørlund from the are high-quality objects. However. Denmark. 50 and 121) from the pre-Christian Viking Age at Trelleborg.1084730 A ritual site with sacrificial wells from the Viking Age at Trelleborg. which covered c. 2100 Copenhagen Ø. it is important to note that when the fortress beginning of the Danish state formation and has therefore was built in 980/81 there already existed a larger settle- been intensively investigated (Nørlund 1948. Strontium analyses indicate ritual features and finds on the promontory from the period that the buried garrison population in the cemetery by the prior to the construction of the fortress. Trelleborg is situated at a promontory bordered by was erected (Figure 1).org/10. University of Copenhagen. cDepartment of Conservation and Science. Degerbøl published the faunal mate. Some of the axe types in the fied the features as forming a sacrificial site from the *Corresponding author. and he identi- origin (Price et al. The strontium isotope results of the four children point to local provenance. animal sacrifice. The fortress and succeeds the military garrison on the promontory. DK-1220 Copenhagen K. Øster Voldgade 5-7. blót activities. weights.2015. Department of Forensic Medicine. and they indicate the presence of National Museum who a few years later published a an elite residence. The National Museum of Denmark. Frederiks V’s Vej 11. Altogether intentional offerings deposited while still enfleshed and interpreted to have served as propitiatory sacrifices to honour or appease the gods and to ensure © The Partnership of the Danish Journal of Archaeology 2015 . In sists of brooches and pendants of silver and bronze. Well 47 produced a young he-goat and well 121 a hindlimb of an above-average-size young horse. the find material con- comprehensive treatise on the results (Nørlund 1948). a large part of a young cow and a large dog. The large settlement both predates military garrison (Degerbøl 1948. Denmark. 2011). strontium isotopes. the results of each well seem to pair up in a systematic way pointing to that the children might come from two different key surrounding areas at Trelleborg. Ny Vestergade 11. Denmark (Received 21 April 2015. accepted 12 August 2015) Downloaded by [Copenhagen University Library] at 05:43 20 October 2015 The promontory facing Storebælt with the well-known circular Viking Age military fortress of Trelleborg erected by Harold Bluetooth in AD 980/981 seems to have been an important ceremonial space prior to the erection of the fortress and contemporary with a nearby high status settlement dated to the seventh to the eleventh century. DK-1471 Copenhagen K. Danish Journal of Archaeology. Among others. It is functioned as a military garrison for a short period and obvious to link this settlement with a number of unique probably only a few decades. dDepartment of Research and Exhibitions. dirhams. etc. This research provides new information that enlightens the formation processes underlying accumulation of cultural deposits in features such as ritual wells. 2015 http://dx. 241ff). Ancient Cultures of Denmark and the Mediterranean. Denmark. Two of the sacrificial wells (47 and 121) included the only skeletal remains of four children hitherto recovered from Danish Viking Age wells. Karin Margarita Freic and Lars Jørgensend a Natural History Museum of Denmark. An unpub- rial. in the period prior to Christianity. ritual well. bLaboratory of Biological Anthropology. pre-Christian Viking Age. for example.1080/21662282. the three wells contained animal remains of primarily domestic livestock partly representing consumption waste from either profane or ritual meals deriving from.

121 and 123) were clearly Purpose of the study associated with special horseshoe-shaped trench features As already suggested by Nørlund. Lidar scan showing the location of the Trelleborg fortress on the promontory bordered by the rivers Vårby Å and Tude Å. we will elaborate further on the cleaned and contained only few finds. 41f). there is During the excavation of the fortress. two entrance posts Viking Age by re-examination of especially wells 47. and to improve our knowledge of the offering rituals that However. the feature was probably associated with uncer. 50 and a row of posts in the trench marked the construction. two wells or well-like structures from the ninth (Nørlund 1948. well 20/36 gr. The present excavated at the contemporary residential complex at Lake investigations focus on the finds from the pre-Christian Tissø (Jørgensen. Map: Danish Ministry of the Environment and Cultural Heritage Agency. p. Nørlund 1948. In well 123. In our paper. 36ff) (Figures 2 and 3). key in the upper fill layer (Nørlund 1948. As we p. unpublished). p. The settlement thus existed both before and after the fortress. 40f). p. In and an antler comb from the tenth century were found particular. an axe from the to the tenth century.B. more or less con- highly interesting ritual features and finds. the finds attest the (1948. p. health conditions and possible inju- and discussed recently (Henriksen 2015. the building were placed in the fill of the well. it indicates that this well also belonged a more ordinary assortment of animal remains was to the period before the construction of the fortress. 50 probably can associate well 123 with the adjoining horse- (= well 50) including skeletal parts of an adult person and shoe-shaped ditch. Nørlund suggested that prior to the erection of the fortress at least five out of six wells may have served as sacrificial wells. Nørlund noticed also the possibility that well 125 was not filled with ritual several wells and pits partly due to their unusual contents. Downloaded by [Copenhagen University Library] at 05:43 20 October 2015 Figure 1. 47 and well 12/37 gr. took place at the promontory. ries. well 21/36 gr. fragments of soapstone vessels. The two other wells 125 and 123 seem to have been p. In the fill layers of well for example. 40). 2 A. 243ff. and partly their con. Gotfredsen et al. p. described in more detail by Degerbøl (1948. The human remains have been examined to provide an Other circular ditch features have also been presented overview of the ages. Well 117 attracted attention by holding children’s skeletons and is clearly older than the fortress as the supporting posts of some nearly complete animal corpses (Degerbøl 1948. the taphonomical approach aims to shed . and 121 in order to shed light on their formation process The trench construction was not associated with a well. period before the fortress was erected (Nørlund 1948. To the east of the fortress is marked the extent of the settlement area (Trelleborggårde) producing rich metal detector finds from the seventh to the eleventh century. temporary with the closing of the wells. 44). A similar trench feature has recently been importance of the area as a ceremonial space. p. 36ff). depositions during the function. Here. 243f). However. tain activities that took place inside the trench feature. a few iron fragments struction and location (1948. 201ff). p. Also. Moreover. 117. Viking period was found close to the bottom and an iron 121 (in the following designated wells 47 and 121). human skeletal parts. of which three (117.

2009. being the leftovers of ritual feasts and/or ordinary meals (cf. Consequently. it is of outmost importance to study the natural destruction Moreover. It was unclear . Danish Journal of Archaeology 3 Downloaded by [Copenhagen University Library] at 05:43 20 October 2015 Figure 2. in order to understand the pro- analyses of the four young individuals from two of the cesses of deposition and subsequent destruction and loss. a horseshoe-shaped trench and several pits and/or postholes is located immediately west of the north gate. Gotfredsen forthcoming) The three wells or special features such as wells or so-called shafts (e. Well 47 was located within house 4 in the southeastern Nilsson 2003.and post-mortem practices but also include the difficulty in distinguishing between animal bones contextual considerations. It was oval and Morris 2011) may offer better opportunities. and human-made modifications inflicted on the bones complete goat and a part skeleton of a cow have been (Serjeantson 1991. the archaeozoologi- the human and animal remains. Russell. it is essential to pay attention to the spatial measured from the surface of the subsoil. It is thus difficult to evaluate the archaeozoological evidence of ritual activity in settlement Materials and methods waste dumps. Magnell 2012). 41f).90 × 1. 2010. The six wells and the four horseshoe-shaped ditches are marked. Serjeantson and building complex (Nørlund 1948. were treated post-mortem. whereas special areas such as bogs. p. A possible ritual area with a special building. a semi. Magnell and Iregren conducted in order to identify the time relation between 2010.8 × 2. light on whether it can be stated how the humans died and distribution and contexts of faunal remains (cf.g. Since beliefs measured 1. funerary sites (Magnell and Iregren 2010. It was 2. Morris 2008. Moreover. new 14C analyses of the four children. p. 50). Plan: Nørlund (1948) with additions. Nørlund’s plan of the excavation. Additional strontium isotope 2012. Lauwerier 2004). wells were conducted in order to reveal their provenance.20 m at the bottom. cal part of the present study will focus on taphonomic The study of animals in ritual use is often hampered by analyses. Grimm 2008.35 m at the surface and narrowed down and associated practices can shape the zooarchaeological to 0. peri.13 m deep as assemblage.

Gotfredsen et al. and various iron objects.4 m. weaving weights. However. The deepest lying object was the tortoise brooch in a depth of 1. spindle whorls. well and the adjoining trench are cut by the building wall fig. 4 A. Medicine. macroscopically as well as using CT images for teeth mately 3. The age of the indivi- level was found fragments of a knife. complex (Nørlund 1948. semi-finished antler for combs. Photo: Roar Skovmand. Pl. XXVII 1a–c). Bone lengths were mea- Traces of wooden planks were documented. XVIII 4–6). etc. rial from wells 50 and 121 is currently curated at the plex (Nørlund 1948. Slavonic pottery (Nørlund 1948. Downloaded by [Copenhagen University Library] at 05:43 20 October 2015 Figure 3. bowl handle. whet. Trelleborg. Both bronze bowl. a dog. The animal bones1 were examined and identified by use of stone. 108b). The faunal bone analysis fragments of pottery. the adults too little material was available for sex Well 121 was located in the northeastern building assessment. and for the fortress period.B. two whetstones. The finds were. Bone growth was of a possible sledge runner. p. two cows.10 m and from a depth of 1. It was circular with a diameter of 3–3. p. all located in the upper fill layers. The human bones were examined at the Laboratory of Nørlund (1948. however. The well contained skeletons of two 4-year-old children and parts of three horses. Sex cannot be determined for (1948. and the material from well 47 is on exhibition in diameter of 4.59 m. The well was approxi. 39) suggests that the well was in function during children as young as the material present here. a knife fragment stone setting or planks were detected (Nørlund 1948. 39ff) dates the well to the period before Biological Anthropology (LBA). well as they derive from the secondary fill material. a red deer and a peregrine falcon. 3 m below the surface of the subsoil. a bone skate and iron fragments were found. four pigs.4 × 1. At the bottom sured using an osteometric board. The the comparative collection of the Natural History Museum . iron fragments.6 m deep measured from the top of the subsoil. The bones were examined down to measuring 1. none of the finds seem p. In the upper fill layers. University of Copenhagen. fragment method according to Ubelaker (1989). The sacrificial well 121 and the associated circular foundation ditch during the excavation in 1937. whether the shaft was in fact a well since no traces of lower layers provided three belt buckles. National Museum. p. 41). The finds seem The human bone analysis primarily to be related to the closing of the well. Among the artefacts in the fill was a tortoise brooch to be associated with the primary function as a sacrificial from the tenth century (Nørlund 1948. it had a LBA. The ditch shows weak traces of posts or planks. development and pathologies. a duals was determined from dental development by the quern stone. fragments of and thus from the period before the fortress. Department of Forensic the fortress.75 m and reached a depth of c.5 m it narrowed the museum at Trelleborg. three sheep. Nørlund Primeau et al. 42). Pl. 39). arrow head. three wooden bowls. p. several wooden fragments and examined according to Scheuer and Black (2000) and a wooden spoon (Nørlund 1948. (2012). At the surface. The human mate- Well 50 was located in the southeastern building com.

. aged and pathology was recorded. (Ubelaker 1989. scaled to our needs.2 mode on a VG Sector 54 IT mass spectrometer equipped which was also seen for the human skeletons from the with eight faraday detectors (Department of Geoscience Trelleborg burial ground (Price et al.98% single rhenium filaments. species distribution. Consequently. The and Natural Resources Management. Strontium samples were bones was undertaken. 42) that the 7 ml Teflon beakers (Savillex™ Savillex.2 ml stem volume charged with intensively pre-cleaned exposure (Shipman et al. taphonomy. We found no the solutions were dried down on a hotplate at 80°C. Moreover. The definitions of cut mark types some overarching archaeozoological principles such as followed Binford (1981. Both skeletons were fairly complete and well pre- Corporation. Table 1 provides an den Driesch (1976).5 ml 30% H2O2 (Seastar. After drying. (1992). this colouration is France) resin. joints and semi- den Driesch and Boessneck (1974). Minnesota. skeletal ele. though fragmented (Table 1 and Figure 5A). Special focus rial is presented by each of the three wells. It was also noted (Nørlund 1948. In contrast. p. The samples were dissolved in a 1:1 This current examination found that the younger child mixture of 0. 1984). small pieces of remains of the two children are described as found near enamel were removed by means of a small diamond the bottom of the well. care has been taken not to sample dentin material along that they were articulated when they were deposited in the with the enamel. Danish Journal of Archaeology 5 of Denmark. The remained visually clear. For the strontium isotope analyses of human remains. 106ff. bones of the younger child. p. Eden Prairie. 66). human skeletal remains of two children were ultrasonically in ultrapure (MilliQ™ EMD Millipore found. This is mainly evident mesh 50–100 SrSpec™ Sr-specific ion chromatographic on the tibia of the older child and some of the unfused hip extraction resin (Eichrome Europe Laboratories. table 4. Sidney BC. Results and discussion Downloaded by [Copenhagen University Library] at 05:43 20 October 2015 bone surface alterations such as gnawing. 2011. and it is reasonable to presume due blade saw and/or with chromium steel pliers.5 ml 6 N HCl (Seastar. Canada) and 0. p. evidence of heat exposure of the bones as mentioned in Enamel samples were taken up in a few drops of 3 N Nørlund (1948. 99. The values for calculating height at overview of the representation of the human and animal the withers followed the guidelines recommended by von remains with respect to isolated bones. archaeological finds related to the wells. Extreme to their completeness. Amounts of 1–5 mg were weighted into well. Age determination was based on mandibles with teeth in situ or isolated mandib- ular deciduous premolars (dp4) or molars (M3).2ϭ). portion of the bone and side. Darmstadt. The elution recipe essentially followed more likely to stem from the burial environment and the . Some of the bones do have a HNO3 and then loaded on extraction columns with a reddish brown colour which could be ascribed to heat 0. USA). no was eluted/stripped by pure deionized water and then the systematic registration of the in situ position of the animal eluate was dried on a hotplate.5 µl of a Ta2O5–H3PO4–HF activator observations in the field. p. Seastar Chemicals was 4 years old (range 3–5 years).1 mm according to the definitions by von rial and the large species variation. burning presentation of the animal bones is organized according to and butchery marks. the investigations conducted on human weathering were assessed. Samples were mea- lated in the laboratory. University of Copenhagen. Further. with a and species choice due to the large amount of bone mate- precision of 0. Germany) water until the water served. which is equivalent to previous ples typically decomposed within 5–10 min. after which examination (Sellevold et al. The bone mate. The tooth enamel from each individual was mechanically pre-cleaned Well 47 with a dental drill and subsequently repeatedly washed In well 47. 131). Copenhagen). p. p. dard gave 87Sr/86Sr = 0. bones which are solution and directly loaded onto previously outgassed considered to be articulated joints were actually rearticu. The results of the strontium isotope and radiocarbon dating are presented The strontium isotope analysis in Table 2 and Figure 4. The human mate- scored according to Behrensmeyer (1978). selection of body part Measurements were acquired by a digital calibre.710236 ± 0. The sam. bones had been exposed to fire. a tooth from each individual was sampled (either a decid- The human bone material uous molar or a permanent molar). the was on man-made bone alterations. as stated by Nørlund (1948. 7 years old (range 5–9 years) based on dental development Seastar Chemicals Inc. 1984.04). 481). Sr rial was recovered by shovels and not sieved. However. that by Horwitz et al. 42). 42). University of skeletal elements were identified to species. Sidney BC. for example. although Nørlund made a few dissolved in 2. although for horses complete skeletons in the three wells as well as other May (1985) was followed.000010 (n = 10. The weathering stage was and faunal bone material are presented. trampling and In what follows. Canada).. The faunal material had suffered a sured at 1250–1300°C in dynamic multi-collection substantial taphonomic loss since the time of recovery. 5 ng loads of the NBS 987 Sr stan- ment. Also. The bones were sexed. and the older child was Inc. Bruz. p.

iron proximal ends of a right ulna Black-headed Single bone fragments. this could not be ascertained in the present analysis. On top of the dog found in lower layers of fill portion of vertebral and children column1. right hind leg and lower limbs1 of three individuals Note: 1Designates that based on the bone elements published by Degerbøl (1948) it was assumed that the bones had been associated. a bronze children aged 4 and 7 years. Table 1. context and associated finds. . Placed near the bottom at a Dog Single bones of individual 1 spindle whorls and fragments depth of 2 m One heal joint and single of weaving weights found in bones of individual 2 the fill Pig Single bones of three individuals Cattle Single bones and foot joint1 of one individual Goat Semi-complete skeleton of Downloaded by [Copenhagen University Library] at 05:43 20 October 2015 one individual Horse Single bones of one individual Well 50 Animal remains Species Type of deposit Placement Associated finds Fragments from the calvaria and Swan Single bones Fragments of a knife. bone skate and 4 years. a quern stone. Well 47 Animal remains Human remains Taxon Type of deposit Placement Associated finds Semi-complete skeletons of two Jackdaw Single bone All near the A tortoise brooch. Census of the contents of the three Trelleborg wells showing type of deposit. a knife individuals fragment and iron objects Cattle Single bones and skull.B. Gotfredsen et al. an arrow head. however. Slavonic with some fragmentation. right hind leg of two individuals Sheep Single bones of three individuals Horse Single bones and portions of the neck1. semi- skulls and few postcranial individual finished antler for combs. fragment of a individual possible sledge runner. carved individual Dog Single bones of one bowl handle. Fragments from the one individual children iron fragments found in upper calvaria from an adult Red deer Single bone layers of fill individual Pig Single bones of four Three belt buckles. two whetstones. depth of 2 m pottery. and radius from an adult gull three wooden bowls. Red deer Single bones wooden fragments and a Pig Single bones of four wooden spoon placed at the individuals bottom level Cattle Single bones of three individuals Sheep Single bones of one individual Horse Single bones of one individual Well 121 Animal remains Species Type of deposit Placement Associated finds Semi-complete but fragmentary Peregrine falcon Single bones of one Fragments of pottery. parts of two children aged Dog Semi-complete skeleton of On top of the whetstone. 6 A. bottom at a bowl.

The three datings from well 47 indicate that the formations processes behind the depositions reach back to the eighth to the ninth century. Strontium isotope results from the four children. . Figure 5. Basic skeletal drawing after Helmer (1987).3 (Bronk Ramsey 2013) and IntCal13 (Reimer et al.71084 13 Note: Errors reported are within-run (2ϭm) precisions of the individual runs.2. 24 bone elements were reassembled from 38 bone fragments. 87 Well no. (A) Census of the skeletal elements present of the two children of well 47. To the left is the younger child (c. In total. Radiocarbon datings from wells 47 and 121 calibrated with OxCal v4. No butchery marks except for the possible blow to the frontal bones were observed.71023 20 121 (1) First molar 0. Skeletal elements that are hatched represent elements which are greatly fragmented and hence not precisely identified. Downloaded by [Copenhagen University Library] at 05:43 20 October 2015 Figure 4. Tooth description Sr/86Sr Error ± ppm 47 (young child) Second deciduous molar 0.71088 21 47 (older child) First deciduous molar 0. Danish Journal of Archaeology 7 Table 2. 7 years). (B) Census of the skeletal elements present of the young he-goat of well 47. 2013). 4 years) and to the right the older child (c.71019 27 121 (2) First molar 0.

p. diseases which are not expressed osteologically rial. 39). and outside the south gate under chronically ill health. Basic skeletal drawing after Helmer (1987). Park and Richter 1953. p. 1985. There is no evidence of trauma. The examination revealed a may have been transported from elsewhere. cut marks or dismem. 8 A. 41). have been broken by force. 2009). incomplete and fragmented. As such there are no indications of a 125 (Nørlund 1948. from the right side (n = 11) (Table 1). Both children had reached appropriate around the Trelleborg fortress. revealing found. . The chil. p. well number growth for their age. whereas light shaded elements were reported by Degerbøl (1948) and may have been articulated. all and as such have no evidence of violence. Skull fragments from both children were present and some few postcranial parts (Table 1 and Figure 6A). it was found that the fractures only that there is no visible signs thereof. human skeletal remains of two children were viously defined by Frei and Frei (2011. The heal joint exhibited dismemberment marks (astragalus and calcaneus). Dark shaded elements were reanalysed in the present study. Several bones on the ulna and radius resemble post-mortem breakage were broken. however. According to Nørlund. this does age could not be more precise than ‘adult’. (C) Census of the skeletal elements of the young adult cow from well 121. Both skeletons were well preserved though that the children are most probably of local origin. fractures resemble post-mortem breakage. described that the bones were found in the fill of the well. variation in colour between the bones is probably due to the calvaria and an upper fragment of both lower arm bones variations in the microenvironment of the well. p. signs of pathology. The material only consists of nine fragments from were found stratigraphically. The strontium isotope results (Table 2) of the two Well 121 children fall within the Danish bioavailable range pre. This is also mentioned as being the case elsewhere cannot be excluded. Sex could not be determined from these fragments and Downloaded by [Copenhagen University Library] at 05:43 20 October 2015 berment on any of the bones present. p. for example. Basic skeletal drawing after Helmer (1987). (A) Census of the skeletal elements present of the two children of well 121. Goodman and Rose 1990. Further. To the left is skeleton 1 (c. There were no not mean that violence or dismember did not take place. Figure 6. 4 years) and to the right skeleton 2 (c. Gotfredsen et al. 41) described Well 50 that both children were complete when thrown in the In well 50. No bones exhibited butchery marks. (B) Census of the skeletal elements of the adult dog of well 121. In well 121. for example. that is. There are no records of how the two children found. 2013). hence containing random bone mate- However. According to dren were examined for growth disturbances expressed as Nørlund (1948. 4 years). single Harris line on the younger child.B. including the bones of the skull. However. the bones should therefore not be Macadam et al. A total of 34 bone elements were retrieved. Nørlund (1948. Skeletal elements that are hatched represent elements which are greatly fragmented and hence not precisely identified. 244). the bones from the lower arm enamel hypoplasia and Harris lines as well as cribra orbi. The cause of death cannot be determined. Hillson ascribed too much importance as the construction material 2000. There was no backfill in the form of sediments cleaned up from other further evidence of the above-mentioned pathologies. hood disease (Harris 1931. 39) further talia which could be a sign of dietary deficiency or child. the bridge at the inner moat (Nørlund 1948. Stuart. wells or the bog area. Nørlund (1948. when the bones were dry. Walker et al. a few skeletal remains from an adult were well.

Symmetry and ontogenetic age of the bones were considered in the MNI calculations. There are no signs of pathology or hypoplasia and Harris lines as well as cribra orbitalia. 15.) 2 1 Peregrine falcon (Falco peregrinus) 2 1 Black-headed gull (Chroicocephalus ridibundus) 1 1 Jackdaw (Corvus monedula) 1 1 Dog (Canis familiaris) 11 2 3 1 34 1 Red deer (Cervus elaphus) 10 1 1 Pig (Sus domesticus) 12 3 16 4 22 4 Cattle (Bos taurus) 8 1 21 3 131 2 Sheep (Ovis aries) 6 1 4 3 Goat (Capra hircus) 38 1 Horse (Equus caballus) 2 1 9 1 291 3 Total identified 72 68 105 245 Mammals unsp. based on dental develop. pigs predominated (Kveiborg and Ritchie 2013). The com. number of identified specimens. The results reveal two periods of physiological and four avian species (Table 3). 2011. Nørlund (1948) does not men- children as there are no indications of the postcranial tion these two fragments in the publication. It is The adult bones found in well 121 consist only of uncertain whether the postcranial elements are from both two parietal fragments. 41) differs from this assessment or those of well 50 and as such not given any further whether a significant amount of bones have gone missing emphasis. which ment. This was the case for many of the Age cannot be asserted any further and neither can bones from the actual cemetery from the Trelleborg the sex of the individual. 1 The number of discarded vertebrae and ribs was not informed. that is. Although whole skeletons of indications of chronically ill health. It is possible that unknown whether the previous definition of completeness the fragments were considered secondary deposits like by Nørlund (1948. U. 131). Danish Journal of Archaeology 9 This current examination found both children to be previously defined by Frei and Frei (2011.) 11 1 12 Grand total 83 68 106 257 Notes: NISP. Diseases which are not expressed osteologically cannot be excluded. that is. however. The skeletal fragments are from an adult. cut marks or dismember. when the bones were dry. vious examination (Sellevold et al. The cause of death dog and goat inflated the number of identified speci- cannot be determined. the skull. bones representing more than one individual. played Table 3. all fractures resemble post-mortem whereas the two other wells with children’s skeletons breakage. 17). The children were resemble post-mortem breakage. 241). p. ing that the children were most probably of local origin as ment (Ubelaker 1989. One of the children (skeleton 1) had in The faunal material addition two teeth examined histologically in a previous The three wells provided remains of seven mammalian study. The bird and mammal remains found in the three wells at Trelleborg as presented by taxon. reveal- 4 years old (range 3–5 years). and horse. concluded for well 47. Well 47 Well 50 Well 121 Total Taxon NISP MNI NISP MNI NISP MNI NISP Swan (Cygnus sp. 66). p. including the bones of is in agreement with the species distribution of well 50. minimum number of individuals. There are no showed a different pattern. 1984. which is equivalent to pre. p. p. This bones (Degerbøl 1948. mens (NISP) counts. since the excavation. Møhl (unpublished sources) noted them at the pleteness of the children at the time of deposit cannot be Zoological Museum. The faunal material published by Degerbøl (1948) with new material sorted from the human bones in 1981 and 2013 at LBA included. in well 121. children fall within the Danish bioavailable range especially in well 47. However. Several bones were broken. 1998. The edges of the fractures Stronghold (Price et al. (Mammalia sp. the minimum number of indivi- The strontium isotope results (Table 2) of the two duals (MNI) distributions still showed that dog. The Trelleborg site as stress within a week of birth as well as six periods the a whole showed a preponderance of pig and cattle following year (Alexandersen et al. when the Downloaded by [Copenhagen University Library] at 05:43 20 October 2015 examined for growth disturbances expressed as enamel bones were dry. trauma. In recent excavations is not uncommon on material of archaeological origin. It is included in the original publication. It is uncertain why they were not confirmed from the remains preserved today. There is no evidence of trauma. p. The examination revealed no indications of the above. . p. at Trelleborg Enge north and west of the stronghold. 2013). 482). MNI.

p. which indicate that part of the pit at Popovo in Karelia (Bridault 1992) which points content of this well had been subaerially exposed for towards a strong symbolic value of these elements. Well structures of (unpublished sources) originally to have been present. The some time prior to deposition (Table 5). Horse was represented by most body a blow to the frontal bones (Figure 5B and Figure 9). All sheep symbolic significance but also represent primary butch. Noe. Further. the cow is assumed origin- to some extent horse and cattle. than bones of Mesolithic Europe deposited jaws and feet in a ritual of the two other wells. bones due to the waterlogged clayey calcium-rich sedi- nected to life-giving powers (Cooper 1993). four damaged ones from sheep but most fragmented with proportions of whole bones varying none from horse except for two lower molars from a between 25% and 36%. Rudebeck 2010). 243). were represented by ally to have been deposited as a part skeleton since large young animals (Figure 8 and Supplementary Table 1). p. and in ments of the wells. mandibles predominate in many Pre-depositional events ritual contexts (Nilsson 2003. especially pigs and ovicaprines and (Figure 6B). 245). Møhl postcranial goat bones had a relatively low proportion of . Elk hunters damage. 243). 2009. forthcoming). Gotfredsen et al. p. whilst the likely that the dog was deposited as a whole skeleton domestic livestock. Skulls including mandibles may have had a between 46–90% and 50–80%. Denmark. Data from Table 3. in particular. Dog Pig Cattle Sheep Goat Horse NISP n = 71 MNI n = 8 NISP n = 55 MNI n = 10 NISP n = 102 MNI n = 13 0% 20% 40% Percent 60% 80% 100% Well 47 Well 50 Well 121 Figure 7. and The dog and cow in well 121 were considered to have Järrestad. Selection of specific body parts Skulls and. for example. the Except for the butchered dog bones of well 47. also showed higher relative been whole skeletons by Nørlund (1948. cattle bones were hardly overlooked during excavation (Figure 6C). jaw bones may have served as a symbol of the sacrificed animals and in addition were easily transported from elsewhere to the place of deposition (Magnell and Butchery practices and fragmentation Iregren 2010. p. In addition. Downloaded by [Copenhagen University Library] at 05:43 20 October 2015 relatively larger roles (Figure 7). At the entire Trelleborg site. Scania. Magnell and Iregren Very little diagenetic degradation had occurred to the 2010. Horse bones did not exhibit ery waste needing to be disposed of. butchery marks (Table 6. although one left metatarsal from well 50 was extent sheep were primarily represented by meaty parts worked cranially presumably to make a bone skate. Sweden.B. The parts. 41): ‘The frequencies of horse remains (but not dogs) as opposed cow lay on top of the dog and the children. On the contrary. contemporary sites such as Tissø. the wells produced two mandibles from see Supplementary Table 2). It seems very remains to be solely from adult animals. although bones in well 50 exhibited a slightly well-known practice since the Mesolithic (cf. The age animals that it was safe to say that they were interred distribution of the animals in the three wells shows dog whole into the well’ (author’s translation). respectively. The three samples showed low average Scandinavia the deposition of mandibles has been a weathering. and so to other structures at the respective sites (Nilsson 2003. whilst vertebrae and ribs were reported by U. especially in wells 50 semi-complete he-goat of well 47 was most likely killed by and 121 (Table 4). 398. gnawing and trampling. one from cattle. pig and only intact skulls found were the goat cranium in well 47 cattle bones overall exhibited the highest frequencies of and the dog and cow crania in well 121 (Degerbøl 1948. The of upper front and upper hind legs. much was preserved of the children as well as the two table 1. 10 A. Gotfredsen et al. pig bones were the pig. Pigs and to some cut marks. The relative frequencies of the domesticated mammals of the three wells at Trelleborg as shown by NISP and MNI. p. bones occurred as fragments. crania are con. for more thorough descriptions. In traditional cultures. higher frequency of weathering and other pre-depositional Nygaard and Richter 1990. whilst cattle and horse bones varied mandible.

. p. The semi-complete dog skeleton of Norn Urd (Carlie 2002. for instance. wells played a central role as present paper to discuss in depth the different kinds of sacred places for people during the Iron Age (see wells or well-like structures (for ritualistic uses of wells. adult: permanent teeth in function and epiphyseal lines fused. juvenile: replacement of deciduous teeth with permanent dentition and many epiphyses still unfused (size of the bones smaller than the adult bones). whereas well 47 may not originally Place of deposition have been a well. For instance. well-preserved bone surface Snorre’s Edda describes how wells were situated adjacent and complete lack of butchery marks indicate that the meat to the roots of Yggdrasil. world (cf. Green 1992. (A–C) Age distribution of the livestock animals of the three wells at Trelleborg as combined from dental eruptions and wear (Supplementary Table 1) and fusion data of the postcranial skeleton. well 50 and possibly also well 121. well 121 had intact bones without butchery traces The so-called ritual shafts are structures that penetrate (Figure 6B). whole elements. Age categories follow Schmid (1972). 288. however. late fusing epiphyses still open or with visible epiphyseal lines (size of the bones almost the size of adult bones). recent damage. outside the scope of the Water and. 83). The smooth. p. which was caused by mainly Nilsson 2003. Näsström 2002. c. butchery marks were observed deep underground and form a line of communication on more than half of the fragmented dog bones of well 47 between the living of the earth and the dead of the under- (Table 6). 118). p. Danish Journal of Archaeology 11 (a) Well 47 adult subadult juvenile 3 2 Number 1 0 Dog Pig Cattle Sheep Goat Horse (b) Well 50 adult subadult juvenile 4 Downloaded by [Copenhagen University Library] at 05:43 20 October 2015 3 Number 2 1 0 Dog Pig Cattle Sheep Goat Horse (c) Well 121 adult subadult juvenile 4 3 Number 2 1 0 Dog Pig Cattle Sheep Goat Horse Figure 8. Such structures were often wells which had gone out of use. One well was of particular (or some of it) may still have been attached to the bones at importance and connected to the goddess or so-called the time of deposition. 88). 50%. p. p. 674. in particular. subadult: replacement of permanent dentition finished and epiphyses nearly ossified. 2009. It is. In contrast.

0 1 25.0 Cattle 1 25. see also Haasteren and Groot 2013). 1994. 95ff).4 2 9. 1 The number of discarded vertebrae and ribs was not informed. 12 A.6 2 25. Frequencies were calculated from NISP counts of the re-examined sample: well 47: jackdaw (n = 1). 2009. Table 5. Well 47 Well 50 Well 121 Skeletal element Dog Pig Cattle Goat Horse Dog Pig Cattle Sheep Horse Dog Pig Cattle Sheep Horse Calvarium 1 3 2 1 1 Mandibula 2 2 1 1 1 2 1 2 3 Caninus 2 Premolar/molar 1 2 Axis 1 1 V. dog (n = 34). dog (n = 3). dog (n = 11). 2 Total 11 12 8 38 2 3 16 21 6 9 34 22 13 4 29 Notes: The faunal material published by Degerbøl (1948) with new material sorted from the human bones in 1981 and 2013 at LBA included. Gotfredsen et al. Still.3 1 12.0 1 14. red deer (n = 1). pig (n = 8). lumbalis 2 2 Vertebra unsp. horse (n = 2). Many1 Costa 1 15 Many1 Scapula 1 2 2 Humerus 1 1 2 6 1 2 2 1 1 Downloaded by [Copenhagen University Library] at 05:43 20 October 2015 Radius 2 2 4 3 3 2 1 1 Ulna 1 1 2 2 5 Metacarpus 1 2 1 2 2 Sacrum 1 Pelvis 2 1 2 3 3 Femur 1 1 1 1 2 Patella 1 Tibia 3 1 2 2 1 2 6 1 3 Fibula 1 Metatarsus 2 1 3 5 2 2 Astragalus 1 6 2 1 Calcaneus 2 1 2 1 1 1 3 Tarsus 2 Metapodial II/IV 1 Phalanx 1 3 1 1 8 1 1 Phalanx 2 2 1 1 Phalanx 3 1 1 Phalanx unsp.5 2 50. horse (n = 3).B. Well 50: swan (n = 1). Fabech 1991. Nilsson and ceremonies in the early Iron Age changed through 2003.g. 288. sheep (n = 2). pig (n = 7). Gnawing Trampling Slight weathering Well 47 Well 50 Well 121 Well 47 Well 50 Well 121 Well 47 Well 50 Taxon N % N % N % N % N % N % N % N % Swan 1 100. thoracica 1 1 5 V. cervicalis 3 6 V.9 6 27.0 1 12. p. Table 4. The animal remains of the major domesticates of three wells at Trelleborg as distributed by skeletal element.1 1 1.0 Total 1 1.9 2 28. p.3 Notes: Slight weathering designates weathering stage 1 sensu Behrensmeyer (1978).8 3 5.2 5 22.6 4 50. horse (n = 0). black-headed gull (n = 1).0 Pig 1 14. cattle (n = 4).0 Red deer 1 100. sheep (n = 4). Well 121: peregrine falcon (n = 2). p.5 2 28.7 1 1. cattle (n = 10). pig (n = 7). 97. The widespread time to be settlement bound in the late Iron Age and practice of using bogs and wetlands for ritual sacrifices Viking Age (e. ritual depositions . Census of bones with surface alterations shown by number and frequency. goat (n = 38). cattle (n = 1).8 2 2.

however.0 1 14. continued to take place in wetland areas during the Viking The species choice Age (cf.7 6 10.1 1 1. possibly within the horseshoe-shaped features In rare cases. Monikander 2010.9 1 14.3 Cattle 1 4. 1-year-old male goat from well 47.0 1 25. none of the re-examined bones could be Downloaded by [Copenhagen University Library] at 05:43 20 October 2015 confirmed to have been subjected to fire.3 2 25. 86. Magnell and Iregren 2010).0 3 42.7 2 9. 2 Degerbøl (1948. 1 Frequency of burning was based on NISP counts from Table 3. Burning1 Cut/chop Impact scars 2 Well 47 Well 50 Well 121 Well 47 Well 50 Well 121 Well 47 Well 50 Well 121 Taxon N % N % N % N % N % N % N % N % N % Black-headed gull 1 100 Dog 6 54. p. hence none was recorded for well 47.3 Total 0 3 4. Currently. Wild species and fantasy . the wells.0 3 37. e.0 7 12. Figure 9.3 1 14. the cultic rituals and disposal of waste from bones and red deer calvaria and antler fragments were the these acts had been carried out nearby inhabited areas.1 5 22. Zachrisson 2014). p. Census of bones with surface alterations inflicted by humans as shown by number and their frequency.5 1 1.5 2 50. wild species were also used (see.g. Frequencies of cut/chop and impact scars were calculated from NISP counts of re- examined sample see explanation for Table 5.5 Red deer 1 10. 44). thus being in accordance with the general temporal trend. (Nørlund 1948. This is in line with most Iron Age and Viking Age ritual The presence of semi-complete skeletons indicates that the contexts which show a preponderance of the typical live- actual killing of the animals probably took place close to stock species (Nilsson 2009. Photo: Anne Birgitte Gotfredsen.8 1 9. 343) reported the majority of bones to have been subjected to mild burning. Skull of the c. only remnants of wild species in the three wells (Table 3).0 Pig 2 12. At The major domesticates predominate whilst a few avian Trelleborg.8 Notes: Burning is slight burning and charring. Magnell 2012.7 1 1. Danish Journal of Archaeology 13 Table 6. Large parts of the forehead were missing which indicate that it was felled with a blow to the forehead before it was thrown into the well. p. 296). the skull is at a permanent exhibit at the museum at Trelleborg. p.

valued for their meat. p. The jackdaw humerus female goddess Freya. 249) and represent consump- (Gotfredsen 2014). pork together with horse meat was the a deliberate deposition of a swan wing could not be con. p. The cow may healthy without pathological signs on the bones and with thus have served as a propitiatory sacrifice to the gods. possible war dogs. p. 14 A. cattle remains comprised by Hatting (1978. Animal body parts used in burial rituals referred to those elements that were important in the animal’s relations with Pig. Wilkens 2006. In the Trelleborg wells. Thus cattle were associated with were found articulated and devoid of butchery marks and life-giving powers. It cannot be stated rituals (Jennbert 2006. 244f) advocated for a (Morey 2010. The pigs in was practised by members of the elite using the Trelleborg the wells conformed to the age of pigs from the site as a area prior to and/or during the occupation of the fortress whole (Degerbøl 1948. Thus an interpretation of 32). Tanngnjóstr and Tanngrísnir pulled Thor’s chariot as Furthermore. 1985. The dog symbolized mythological animal. c. 136f). 63 cm at the withers and presumably a male hind leg. 62). pigs was exemplified by the myth of the male pig ited with its feathers still attached. p. 2009. only a Särimmer being slaughtered and its meat cooked each broken distal swan carpometacarpus exhibiting trampling evening to feed the warriors of Valhalla (Faulkes 1995. Consumption of dogs is a wide- regional variation. For instance. in with fertility mostly in connection with Freyr but also the Downloaded by [Copenhagen University Library] at 05:43 20 October 2015 the present case the ability to fly. De Grossi Mazzorin and Minniti 2006. p. p. a wing and a leg bone from a and all but one were from juvenile and subadult indivi- male peregrine falcon in well 121 indicate that falconry duals close to the optimal age of slaughter. p. 89). The three Trelleborg wells contained exhibiting possible traces of feather removal was retrieved disarticulated bones of pig. In Celtic religion. p. it is regarded as a general Indo-German described in the Gylfaginning of the Prose Edda by phenomenon that dogs acted both as conductors and Snurri Sturluson (Faulkes 1995. spread practice. p. tion waste. the two male goats not only death but also healing and fertility (Green 1992). goat in well 47 was. Cattle. The juvenile he- guards at the entrance of the underworld (Gräslund 2004. primarily meat-rich portions. Old Norse mythology mentioned cattle in the crea- sized dog skeletons found in bogs. creatures were significant for identity and cultural trans. preferred meat for consumption at the blót meals firmed. 61–63 cm at the withers. summarized status. that is. the pig was associated humans (Schanche 2000. p. 89). c. 295. Sheep did not play a prominent role in the tant role in mythology and also connected to water and Old Norse mythology whilst the goat was an important spring cult (Gräslund 2004. ible (well 50) and a part skeleton of a young adult cow The dog in well 121 was a rather tall and slender indivi. p. 1 year old and thus most likely killed in spring/ in the Old Norse religion this guarding dog was named early summer probably by a blow to the forehead Garm (Jennbert 2006. Unfortunately. tion myth where they gave milk to the first gods Ymer and Röekillorna. written sources symbolic significance of the jackdaw wing bone in well 47 described how at certain times dogs were butchered and and a swan wing (ulna and carpometacarpus) in well 50. Gotfredsen et al. 11). although a nearly intact jaw bone from a boar (well 47) may have been selected for a deliberate Dog. for example. 22). but some meat was was preferred depended on the purpose. domesticated species were of particular importance In stark contrast to the whole dog in well 121 are the in rituals at houses and farms. p. Scania and Valmosen near Rislev. Still. and even within the same society dogs may both receive ritualistic burials and serve as food Avian species. Bure (Faulkes 1995). were corpses (Nilsson 2009. p. a mand- be large. dismembered at the heal joint (Figure 6C). only moderately worn teeth. 171. Which domestic species whether the meat was actually eaten. In the Old Norse mythology. were also reported to firstly of assumed consumption waste. The stratigraphically close position to the .B. p. traces and abrasion was retained. Ovicaprines. as evidenced from epiphyseal fusion p. Large pointer. milk therefore considered as having been deposited as whole and hide. p. secondly. Degerbøl (1948. p. as well as being useful draft animals. Furthermore. at its prime. large dog devoid of butchery traces in well 121 seems to formations and played an important role in the Eddas be in accordance with the sacrificial use of dogs during the whilst domesticates were mentioned to a lesser degree. Sacrificed dogs included in highly estimated domesticates indicating both wealth and Danish boat-graves and weaponry bog finds. including the skull (well 121). in wetland and at death butchered dog joints from well 47. prob- (evidenced from its prominent crista sagittalis). Zealand. The deposition of an entire (Figure 9). c. The articulated upper meaty dual.138). the dog was an esteemed creature with an impor. from well 50. and data. 171).133). Fowler 2004. Whole dog skeletons and skulls occurred in bogs and deposition. In ancient Rome. Moreover. Cattle. eaten at ritual meals (cf. An intact well-preserved black-headed gull ulna (Faulkes 1995). It was ably was deposited while still enfleshed. ritual context and stripped off the bones.73. The meat-producing powers of was well preserved and may have been deliberately depos. Iron Age. 138). 90). wells throughout the Iron Age in present-day Denmark and Scania (Nilsson 2003.

pit houses. meatless extremities as well as meat-bearing parts. p. sheep were A sacrificed young he-goat in well 47 killed during Downloaded by [Copenhagen University Library] at 05:43 20 October 2015 represented by lower frequencies in wells than in. for spring may be seen as a propitiatory sacrifice to honour or example. con- between different world realms in the belief systems of stituted a resource loss to the society. these two wells. that is. p. In addition. as an indication of horse from the royal residential complex at Lake Tissø. contrast. Additionally. Sheep were depending on the The three wells dated to the eighth/ninth to tenth centuries region widespread domesticates and much more abun. for example. fragments of meaty Due to post-depositional events such as heavy breakage of upper limb bones and mandibles which may be regarded the children’s bones and post-excavation loss of bones. In 2003. (Gotfredsen et al. animals. At Tissø blót activities. 196. One fine example is bones of. they do not stallions or bulls at the Freyr’s cult or winter blót. p. . such a practice cannot be Gotfredsen et al. Magnell and Iregren 2010. Magnell 2012). However. In connection to the ritual site at colt killed during spring/early summer. represent meal leftovers. Danish Journal of Archaeology 15 two children supports the notion that the juvenile goat Conclusions was a sacrificial animal. Hence. Lindängelund. Jennbert 2006. forthcoming) and Järrestad. dual function as a sacrifice to the gods and a conductor for Nilsson 2003. During the Viking found in the two wells (47 and 121) point to local Age. It is. 17% at Trelleborg being the remains of profane and/or ritual meals such as Enge than was evidenced from the three wells. there seems to be a pattern. for example. 301. Germanic tribes was already described by Tacitus (cf. Even male animals suitable for breeding. Although food leftovers from horse may 30 km north of Trelleborg (Jørgensen et al. These represent to date the only ticated mammals in wells 50 and 121. it was well 121 (1) have strontium isotope ratios a little less the practice to select and sacrifice the best and healthiest radiogenic in the order of 87Sr/86Sr ~ 0. the youngest child from well 47 and child from tant role as sacrificial animal in the blót meals (see Nilsson well 121 (2) have strontium isotope ratios of ~0. p. p. The periods have been excavated (Carlie 2013).7102. which was repeated in (seen altogether). 138). as formation and passage to other worlds (Loumand 2006). 100). seem to point to exactly the same area. Magnell there is a ritual meaning that involves the sacrifice of and Iregren 2010. table 2). 89f. 445). the strontium isotope results of each well pair up. Sheep comprised only 4% and 11% of the domes. 2009. and wooden artefacts Odin. dant than goats at Scandinavian Iron Age sites Two wells (47 and 121) contained children’s skeletons. as fertility. p. (Gotfredsen 2004. The sacrificed animals showed evidence of holy animal of Odin and Freyr and closely connected to having been deposited while still enfleshed and hence. whether it were boars. diagram 15) but occurred in some represented consumption waste of livestock species higher frequencies. the horse was seen as a symbolic creature of trans. hindlimb of a young large-sized presumed stallion and large parts of a cow placed together with the children Horse. According to Ström (1985. skeletons or limbs. whilst Christian belief system as described in the Eddas and the large-sized high prestige male dog may have served a evidenced in the archaeological record (Carlie 2002. 177). Horses dragged the sun over the sky and played an impor. 2009. 300). The strontium isotope results of the four children Bliujienė and Butkus 2007. forthcoming). respectively. At well 121 derived from an unusually tall individual. Horse bones occurred in high relative frequencies in plausible that the children came from two different sur- wells and wetlands sacrificial sites but played an insignif. perhaps because sheep had a more appease Thor and to ensure fertility. In the Trelleborg wells children from certain nearby areas. 2009. The deposition of a ordinary usage (Nilsson 2003. c. well 121 provided a metatarsal from a have been documented. They Danish Viking Age wells with human remains of children. have been present in well 121. the wells contained a variety of animal remains of which Kveiborg and Ritchie 2013. The significance of horses as an intermediary regards the animals normally used for consumption.7108. flesh being eaten. a number of wet areas and wells from different at the site as a whole (Degerbøl 1948. c. The horse hindlimb in sites has also been documented in earlier periods. rounding areas at Trelleborg. Sheep were of minor importance in could not be stated how the children died. close to the town of Malmö in southern 149 cm at the withers as compared to a mean of 140 cm Sweden. 246) considered the charred horse ological excavations in Scandinavia. Ritual wells have turned up in connection with archae- Degerbøl (1948. all body parts of horse occurred. In a number of deposition of an above-average-sized young horse (stal. may also be interpreted as propitiatory sacrifices. therefore. 2014. One could thus speculate if icant role at other sites (see Nilsson 2003. Degerbøl 1948. 85. Nilsson 2009. unearthed a variety of archaeofauna and human remains. only c. Magnell and Iregren which seem to be deliberate depositions of semi-complete 2010). wells from the period 150 BC–AD 300 deposits contain- lion?) may be interpreted as a fertility sacrifice for Freyr or ing remains of humans. that is. p. The horse played a prominent role in the pre. well 47. provenance. p. the oldest child from well 47 and child from p. 244. p. The use of wells as ritual ascertained in the present analysis. though all values point to local origin. 241. it as butchery waste. It was the the children. culture deposits at the Trelleborg site (cf.

hall also in the tenth century contained a stave built house montory are not the only ones in the area. assessment of bone element and sometimes portion of the bone. Apparently. bones (n = 53) was sorted from the human remains at tion with the construction of the fortress. At the large residence at Fugledegård at Lake the stray metal finds indicate that we are dealing with a Tissø. tion of cultural deposits in features such as ritual wells. Trelleborg we note an area just inside the northern gate Consequently. The cross-scientific analyses presented in this article demonstrate that further steps and new knowledge can be Funding gained within an interdisciplinary collaboration between The authors are grateful for the financial grant received from A. A few years ago. in the ninth to tenth century. contemporary with the fortress. who was thought that it coincided with the introduction of rather specific as regards species identification. however. and articulated vertebrae (n = 10) of the semi-complete . Therefore. a total of with or without the consent of the local elite. closed down 152 bones could be examined or re-examined in the present the old ritual site.P. these evaluations were applied to the assemblage fortress. This is important for our understanding of 1. particularly in order to Møller og Hustru Chastine Mc-Kinney Møllers Fond til almene clarify the formation processes underlying the accumula. Fig. 92f) suggested that the house could be cow was found on top of the children in well 121. applied on the entire bone (cf. The house and adjoining features seem to assemblage (n = 257). 16 A. we had a religion which in part seems very vigorous and whose Notes rituals seem to be firmly embedded in the mental world of the people. Today new archaeological excavations ciated with a settlement of higher status. However. In 1981. In his publication. it is important to note that it is closely asso. Those bones were not seen by Degerbøl. only 48% (n = 98) of the original bone sample (n = 203) published by Degerbøl (1948) could the transition from the ‘pagan’ to the Christian religion. The structure and indicate that Nørlund’s first interpretation could very well organization of this settlement is still unknown. the present analysis encompass analyses where a probable stave built house is associated with one regarding frequencies. Degerbøl (1948. p. but the building and the associated features at The distance to the find location for the sword is uncertain Trelleborg could very well constitute a similar ritual com- but is just outside the map in Figure 1. where so far five Kristian M. fig. Formaal to the research project ‘Pre-Christian Cult Places’. and Christianity. p. we would perhaps have the bone identifications undertaken by U. Supplemental data Christian religion on the Scandinavian sites. the presumed pre-Christian ritual area south of the complex of some status. Møhl. age and skeletal part distributions of the horseshoe features and several pits and/or postholes and occurrence of charring. 137ff). the authors wish to thank two anonymous reviewers for their much appreciated comments on this paper. p. be correct. The two with dredging of the river Tude Å. natural sciences and archaeology. Based on the archaeological traces of the pre. he rejected this interpretation as only vertebrae. The ritual features on the pro. Figure 2). Slagelse. one from well because as it is situated at the gate facing the large wet 50. During the present study. a ninth-century sword buildings are not exactly similar in their construction was found upstream of Trelleborg (Nørlund 1948. also Nørlund 1948. be found at the Zoological Museum–University of Trelleborg constitutes an important site with regard to Copenhagen ZMUC or at exhibit at the museum at these power political aspects as we witness the closing of Trelleborg. Nørlund (1948. However. perhaps not least vertebrae. The finds so far verted to Christianity. 244) reported on only 13 cattle bones plus many ‘gate house’ with an uncertain function. 7). contemporary ritual sites have been documented in the kindly sampled the animal bones for radiocarbon dating. In connection measuring 7 × 7 m (Jørgensen 2014. landscape encircling the royal residence (Jørgensen 2014. This indicates that the house and the features are actually re-examined in the present study (n = 152). 2. a small sample of animal an important regional pre-Christian ritual site in connec. Trelleborg. Natural History Museum of Denmark. Except for a dog’s axis. p. they were used both for gift and communion offerings. p. from the duck bones (Gotfredsen 2014). p. least see that. Instead. he called it a (1948. area to the north. We know that this event took place study. details. Gotfredsen et al. Further. Undated human plex that served the Trelleborg garrison where the majority skeletons have been found in the bog area just east of the of the inhabitants with all probability not yet had con- fortress (cf. and respect the gate entrance and the plank-built street into the therefore. a skull and some limb bones of the cow were apparently brought back to the laboratory since Degerbøl based on too weak arguments. well 47. Gregersen. A new political LBA and sent to ZMUC. 159). 243–45) based his publication on around AD 980.B. One additional peregrine falcon was reidentified and military power clearly entered the regional scene and. however. a ‘temple’. we can at Supplemental data for this article can be accessed here. 41) stated that a skeleton of a dog and a Nørlund (1948. indicate that systematic surveys and investigations would reveal further ritual sites and perhaps shape a ritual land- scape similar to the one that have emerged in connection Acknowledgements Downloaded by [Copenhagen University Library] at 05:43 20 October 2015 to the elite complex at Lake Tissø. if we look at the plan of moreover information on fusion stage and charring. Weathering. trampling and butchery trace assessment was not undertaken by Degerbøl. 4).

2010.K. 1990. Danish Journal of Archaeology 17 dog from well 121. Animal sacrifices and deposits in Carlie. 2006. Dog sacrifice in quity. 89–99. With a contribution on the anthro- A. eds. International Journal of National de la Recherche Scientifique. A. In: A. In: N.B. P. V. OxCal 4. Møhl. eds. British ritual shaft at Springhead. More. Gård och kultplats. Available from: http://c14. M. 1994. J.S. Ltd. doi:10. Museum of Denmark gives the impression that only identifi. A. recovered ribs and vertebrae were not Fabech. 1998.. A. K. In: L.pdf or symbolic interaction. 1991. 30. Klaipeda: The 1096-8644 Instutute of Baltic Region History and Archaeology.S. Archaeological Sites. 9–12 September 2002. 1981.. Routledge. domestic ungulates. Animals and their products in trade and Viking Age sites in http://www. Beitrage zur Archaeologie und Fortidsminder. In: P.. Leibzig: Thames & Hudson. 1993. Randsborg. In: P. American Journal of Weapons. Arkæoeologiske [Accessed 3 December 2014] Gotfredsen. 55. A.R. The stored animal bone assemblage of Copenhagen: Akademisk Forlag. 2008. 147–160... Ageing and Sexing Animal Bones from pdf. 54–75. 2011.M. Skånska regioner. Nørlund. ed. Proceedings of the conference at the Swedish Institute the ancient world: a ritual passage?. 23–36.. Sacred and pro- animal bones separated from the children’s skeletons at the fane. The use of tooth wear as a guide to the age of Available from: http://www. 109. 33. (Trans. Applied Geochemistry.. London: able portions of the major long bones with articular ends and Evryman. Alexandersen. 2004. J. ritual and LBA in 1981 lend support to the notion that those skeletal religion.. Snorri Sturluson Edda. 326–340..1017/S0079497X0001519X the Romano-British sanctuary of Springhead. R. 2003. 1995. Bliujienė.2. Årbog for Køge Museum. Ancient men and modern myths. Routledge. 2013. Proceedings of the Prehistoric Society.. Eastern Denmark.1002/oa. Grant. C. lished sources). 1992. et al. eds. J. forthcom- Skrifter No.. forthcoming. Binford. Kent (UK). 2013. 26..M. Fowler. Bones.. A. BAR British Series. C. Armed men and their riding Goodman. Dogs in graves – a question of symbo- symbols. Animals in Celtic life and myth. A. ed. eknoglerne fortæller. The geographic distribution of strontium isotopes in Danish surface waters – a base for provenance studies in archaeology. PECUS. 169–183. In: Funen. U. Bejenaru and D. southern Sweden.2010. Boye and U.B. 8. Kalmergården.. Special Issue Paper. Proceedings of the 9th conference of Green. 365–377. Jørgensen. 56.2013. The status of elk during the Mesolithic. 10. Folklore. C. Gotfredsen. October 1991. A. D. London: Academic Press. Sheep in the Iron Age: a method of Anthropologie.M. and Frei. Dyreknogler fra Trelleborg. island of Bornholm (Denmark) – a base for provenance tion from bone weathering. Bridault. doi:10.H. Tusen år av kultur och pological material. 2013. Trelleborg.. Osteoarchaeology. Wiley & Sons. Snyder and E. study. Langenweissbach: Oldskriftselskab.O. Vol. Gotfredsen. Vol..folklore. Copenhagen: PNM – Publications from the of ritualization in archaeological remains from National Museum. Thrane.. London: the international council of archaeozoology. 151– A bird for all occasions: the use of birds at doi:10. 653–679. 40.. J. R. studies in archaeology and agriculture. In: L. Cooper.. Stockholm: Riksantikvarieämbetets Kroppedal.1016/j. 2014. Kent (UK). 6.apgeochem.1080/000163598423009 Sr isotopes from surface waters and soil extracts over the Behrensmeyer. ed.isvroma. C. hydrology and agricul- Downloaded by [Copenhagen University Library] at 05:43 20 October 2015 References ture. Lund Hansen.. Man and animal in anti- De Grossi Mazzorin. 09. ed..12.C.. inhumation graves of the Roman Iron Age in Zealand and gar på en yngre jarnåldersgård i Hjärup. The Viking manor Carlie. A. London: ments as well as jaws with teeth in situ were retained. Anthropozoologica. In: L. major domesticates presently kept at the Natural History Vol. 2007. and Butkus. L. In: . and associated histological structures. ing. sydvästra Skåne. IV (1). Dogs and people in social. Birds in subsistence and culture at Grant. Gesellschaft für Archäozoologie und Prähistorischen Ewbank. 2004. Applied. förlag.007 Bliujienė. 2004. In: A. 1992. and Rose. Papers of large-sized mammals were disregarded ( other easily identifiable portions of the main skeletal ele. whereas smaller fragments eds. Carlie. Archaeology and ritual: a case study on traces at Lake Tissø. eds.1002/(ISSN) Kazakevičius. 14–19. ed. Booty sacrifices in Southern Scandinavia: a retained. elements in some numbers had originally been present in the Fabech. weaponry and man: in memoriam Vytautas Physical Anthropology. 62–66. Det Kongelige Nordiske Prähistorischen Anthropologie. Garwood. A dog’s life: animal bone from a Romano- Degerbøl. An illustrated encyclopaedia of traditional Gräslund. and Jørgensen.. Nielsen. Lindängelund. A. 4 (2). Riksantikvarieämbetet. 241–265.006 gic sites in Sweden and Denmark.-S. Rome: Swedish Institute. L. 1982. The archaeology of Gudme and Lundeborg. Aspects of teeth from archaeolo. A. Reading society from the cultural land- wells. Nordiske Benecke.M. 1978. In: P. doi:10.. small-sized mammals were kept. K. 2002.. The archaeology of personhood. 38. Archaeologica Baltica..1016/j.. ed.. Paris: Éditions du Centre Serjeantson. and H. Taphonomic and ecologic informa. unpub. and Frei. Copenhagen: Nordisk Forlag. exchange. Om bruket av offerhandlin. A. J. et al. arch. 150–162. ed. M. et al.J. A. Grimm. 59–110. J.B. economic 167–176. 423–26. Frizell. Oxford: Oxbow Books.). and Minniti. K. Acta Odontologica Frei. 1948. Arkeologiska undersökningar Gotfredsen. doi:10. apgeochem. The Geographic distribution of Scandinavica. in Rome. doi:10. Some ribs and vertebrae occurring amongst the reassessment. samhälle i förändring.M. Paleobiology. Møhl noted in his identification lists that all bones scape. Durham. Chichester: John Bronk Ramsey. working..2367 ox. lism? In: B.. Assessment of systemic horses as a reflection of the warrior hierarchy in Western physiological perturbations from dental enamel hypoplasias Lithuania during the Roman Iron Age. 24..M. Proceedings of a conference on archaeology. Geochemsitry. 91–108.. Frei.. 49–68. presented at a conference at Svendborg. Henriksen. Jernalderpladsen Østervang – hvad dyr- 95–116. Oxford: Oxbow Books.C. South Scandinavia between sacred and political of bird species and special mammal species such as dog and power. 1964. Faulkes. August 2002.B. A. C. In: L. Grimm. A.

R.L. Våld och vatten. Højbjerg: Jysk Age fortress at Trelleborg.. a social bond. 1985. 223–250. 195–207. Thesis (PhD).J. M. 135–140. Arkæologisk Selskab. 8–13. A. Old Norse Nilsson. and Riksantikvarieämbetet Arkeologiska undersökningar. Robinson. 24.C. 69 (3).. In: religion in long-term perspectives. Who was in Harold Bluetooth’s army? Aggersborg. Oxford: Age calibration curves 0-50. Old Norse religion in 2008). Trelleborg. 69–74. O. Veitstu Hvé Blóta Skal? The 16. et al. 50.G. n° 1. 85. Saunders.. J. Kulturmiljö Halland. Nye fund fra Society.1127/ the zooarchaeology of ritual. In: D. In: Himlingøje-Gravpladsens Høje.P.S.. Extraction and Ion Exchange. materiality in Scandinavian death rituals. and C. 206–297. Behaviour behind bones: Anthropologischer Anzeiger.A.T. ed. and Ritchie. eds. 313–336. 186–199. Lund: Nordic Academic Press. 368–382. and Richter. T. D. nemgang af knoglemateriale fra Trelleborg. The Iron Age Uppåkra. Current Swedish Archaeology.. 4000 BC to AD 1550. E. The heroized dead. in environmental archaeology.. Stockholm: Norstedts. and Richter. In: A. c. Rapport Strontium isotope investigation of the cemetery at the Viking over undersøgelserne 2007–2010. 55. 2010. Stockholm: Stockholms Henriksen. People. Série B: Monikander.. New York: Wiley-Liss. Fascicule 4. Brinkhuizen eds. skeleton. and Dietz. ed. In: A. A novel Näsström. Durham. K. Järrestad. Det Kongelige Nordiske Oldskriftselskab.B. 2014. Odense Bys Museer. 1985. Church. Für Säugetierkunde.16947 Legge. P. 2006.F. Harris. Nationalmuseets Arbejdsmark 2014. 9. In: B. 2010. 2015.T. Jennbert. 51.M. and Iregren. Carlie. In: Nordiske Fortidsminder. 18. Kill-off Patterns in sheep and goats: the mand- Naturforchenden Vereines in Brünn. Experimentation and reconstruction Proceedings of the northern worlds conference.. Senda. Van Neer. Seventeen wild boar Scandinavia in the 6th–11th century. Gulløv. Skrifter. The Swedish Archaeological Hatting.-M. A. J. W. animals. Sóa.. 2012. Northern worlds – landscapes. Halmstad: Press.. Solvent Norden. Tro och offer i det förkristna strontium-selective extraction chromatographic resin. 1987. In: Nordisk Forlag. Häst och hund i fruktbarhetskult och blot. 66–72. Førkristne kultpladser – ritualer (1). Stockholm CNRS/APDCA. and Wing. M. and C. F. 249–286. 2000. J. Fiches descriptives pour les relevés d’ensembles Pferden – ein immer noch aktuelles problem. Våtmarkskult vid Downloaded by [Copenhagen University Library] at 05:43 20 October 2015 mammifère. 287–309.. and A. E. Copenhagen: Park. Järnalderens rituelle platser. Journal of the Kveiborg. C. C. S. Radiocarbon. Hillson. 1973. 476–489. Biological anthropology of the human Southern England and Yorkshire. J. Re-examining associated bone groups from R. Groningen: Barkhuis Gronigen University Library. ed. Cambridge: Cambridge University Press. Widerristhöhe und Langenknochenmasse bei Helmer. 1978. D. Dobat. K. Oxford: Oxbow Books. Cambridge: Bronze Age economy. In: M. Denmark. doi:10. Copenhagen. Dental pathology. Blóta. 1885. rituelle aktiviteter ved havet. Das Verhältnis der Ossa longa zur Hopkins Hospital. mandibles from Sludegårds Sømose: offald or sacrifice?. IntCal13 and Marine13 Radiocarbon cil of archaeozoology.1259/ ritual killing and burial of animals: European perspectives. In: H. 187–195. Ervynck. bird working group in Groningen. E. 13. Sacred cows or old beasts? A taphonomic childhood: the correlation of histological and radiographic approach to studying ritual killing with an example from appearances in clinical and experimental conditions. A method for estimating age of Danish animal: roman depositions and offerings. doi:10. B. 1992. Zeiler. L. long-term perspectives.. 561–588. Jones O’Day. Centre de recherches archéologiques du Skedemosse under jänåldern..J.B. 28–30 November 2012. U. Sweden. 317–333. E. W. 0007-1285-4-47-561 Oxford: Oxbow Books. Fyrkat og Price. IV Jørgensen. The horse and its role in Icelandic burial archaeology. P. et al.. Zoologisk beskrivelse af dyreknoglerne fra sb. Lines of arrested growth in the long bones in Magnell. Desse and N.. 0003-5548/2012/0168 Proceedings of the 9th conference of the international coun. 129–150. In: A. A. In: Stockholm: Riksantikvarieämbetet. Prummel. Skelethöhe bei den Säugetieren. ed.. Hatting. Anatolian Studies. ed. R.. 2010.. Bulletin of the John Koudelka.. May. . 23 August–27 August Jennbert. Excavations at Grimes Graves Norfolk Reitz.55. medieval sub-adults based on long bone length. (2008). Cambridge University Press. 127–153. 2009. 234–248. In: A. Antiquity.000 years cal BP. Dogs: domestication and the development of 201–216. Morris. mechanism of their development. et al.. August 2002. eds. In: 1972–1076. 1948 ed.1017/S0003598X00067880 Lauwerier. Lund: Nordic Academic A.S. interactions and dynamics. Fiches d’ostéologie animale pour l’archéologie.E. N. ed.4. 1992. Blot. Katzenberg and S. doi:10. Zooarchaeology. 18 A. 2. PNM Publications Symposia of the Association for Environmental from the National Museum. Horowitz. Huvudgård i centralbygd.P. doi:10.A. Kystens kultpladser – vikingernes Universitet. Jørgensen. Studies in Archaeology & Archaeology. 1990. 4. L. 2013.. Chiarizia.. Andrén. Magnell. Andrén.. 2015. Morey. and D. In: Animals. 130–134. 2006. Copenhagen: og tro i yngre jernalder og vikingetid. 23.. (2004). Bournemouth University. eds. Vol.. 93. Særtryk Af Old Norse Blót in the light of osteological remains from Frösö Antikvariske Studier.. mythology and society.J.. 22. Nørlund. Oxbow Books. Birds in Loumand. 1869–1887. Proceedings of the 6th meeting of the ICAZ practices. Jennbert.. History Vol. doi:10. L. eds. S. Gotfredsen et al. Desse-Berset. Analys av djurben. Kongens Borge. Verhandlungen des Payne. E. ed. eds. 07366299208918107 Söderberg. 1.C. 1931.A. H. et al. London: British Museum Press. Forhistoriske hunde i Danmark. 2013. T. In: S.1080/ Nilsson. The economic and non-economic Primeau. 175–187. Studies in Archaeology 52. environment and the Cambridge Manuals in Archaeology... Andrén. 2014. 2003. AD 200–1000. Zooarkæologisk gen. 2011.. 1953. Zeitschrift osseux animaux. Raudvere. status and identity. 2002.. 10. Reimer. D. The British Journal of Radiology. Raudvere. 81–99. Norse religion and ritual sites in Noe-Nygaard. Pluskowski. O. 1999.. 2012. 281–303.. ibles from Aşvan Kale. jernalderen. L. In: J. Copenhagen: National Museum..2458/azu_js_rc.. Jämtland. Dyr I Natur Og Museum. religion. Sweden. K. E.C. Transverse lines in bones: the National Museum. K. British Institute of Archaeology in Ankara.

D. 1989. Oxford: Oxbow Books. I trästodernas skugga. Foster. New York: Cambridge University Press. 2011. morphol. Durham. Amsterdam: Elsevier. 2006.. American Journal of Physical Anthropology. 11.. Danish Journal of Archaeology 19 Rudebeck. Lutbron och de förkristna dödsoffren i sjön bokaren. Danebury: an Iron Age hillfort in Hampshire. The biography of wells: Russell.J. the Low Countries. erfarenheter Ubelaker. A.M. 391–398. 73. Dogs and people in social. doi:10. economic or symbolic interaction. A guide to the measurement of Girji OS. Från ben till boskap. Kritische Schmid.v139:2 bones and teeth: an experimental study of color.. A.1468-0092. eds. J. crystal structure and shrinkage. More eds. Vretemark. Ström.. Graver i ur og berg.L. E.B. MA: Harvard University. D.. Shipman.. doi:10.. B. M. Museer. 66. J. 2014.. Proceedings of the 9th 4403(84)90013-X conference of the international council of archaeozoology. The sacrifice of dogs in ancient Italy. The excavations 1979–1988: the Finds. August 2002.. Kungliga Gustav Adolfs tative of childhood conditions.. Uppland: Saga Oc Sed. Serjeantson. Nilsson and E. 25–51. Cunliffe and C. Osteology..x hypothesis. 307–325. DC: Taraxacum Press. Malmö: Malmö analysis. et al. London: från Skara. In: B. View publication stats . P. The causes of porotic hyperostosis and Age and Roman Britain. Age man in Denmark. Hansen. San Diego: Academic Press. doi:10.. Serjeantson. djurhållning met utganngspunkt i medeltida benmaterial Downloaded by [Copenhagen University Library] at 05:43 20 October 2015 Volume 5. work- Archaeological Science. Wilkens. and Jørgensen. Available from: http://jalc.idno=0402a02 og religion fra forhistorisk til nyere tid. eds. Monumentala möten Physical Anthropology. P. M. Peabody Museum Scheuer. 2014.A. 109–125. A..pdf?c=jalc. P. 85–107. and Black. interpretation. Atlas of Animal Bones. 1972. Ravens and crows in Iron Walker. 1984. Copenhagen: Det Kongelige Nordiske Sonderdruck Aus Säugetierkundliche Mittteilungen. 1985. Council for British Archaeology. 83–251.. Porotic hyperostosis: represen. Human skeletal remains: exavation.. Anmerkungen zur Widerristhöhenberechnung aus Sellevold. 660407 Arkeologiska och forhistoriske världar. T. Res. Social zooarchaeology: humans and animals a functional and ritual life history...00360.und frühgeschichtlicher Tierknochen. doi:10. Burnt 139.. American Journal of Akademiens Årsbok. Developmental Juvenile Bulletin I. Skrifter Från Länsmuseet Skara (Skara). Tro och sed i förktisten tid. Fält. och stenåldersplatser i sydvästra Skåne. U. The bird bones. Cambridge. A. Händelser vid vatten. N. Akademiförlaget. 123–137. M. G. 2013.1111/j.. and Schoeninger. 2010. 30 cribra orbitalia: a reappraisal of the iron-deficiency-anemia (1). et al. Stuart-Macadam. 1991. 192. In: L. 2009. B.. 1976. Journal of Archaeology in in prehistory. E. D... Om näcken vid Books.1002/ajpa. Kosthåll och Pole.. 1997. 2000. 69–91. Stockholm: Scandinavian University Zachrisson. Oldskriftsselskab. and Groot. ing. 325–348.2010. Oxford Journal of Archaeology. Iron Längenmassen vor. Rudebeck. Journal of Snyder and E.1016/0305. Karasjok: Davvi von den Driesch. In: B. 2000. 497–81.1330 I neolithiceringens tid. J. 1984.. 1974.1002/ajpa. ogy. 2012. Nordisk hedendom. and Morris. 4 (2).. L. animal bones from archaeological sites. Samisk gravskikk nl/cgi/t/text/get-pdf318a. von den Driesch. and Boessneck.. E. Washington. 22 (4). van Haasteren. 1985. S. Schanche. Rep. 25.