The four horses of an Iron Age apocalypse: war-horses from the third-century weapon sacrifice at Illerup Aadal (Denmark).

Dobat, Andres S. ; Price, T. Douglas ; Kveiborg, Jacob 等

[ILLUSTRATION OMITTED]

Introduction

Since the late nineteenth century, more than 30 so-called weapon sacrifices dating from the period between the pre-Roman Iron Age and the Migration period (c. 350 BC-AD 600) have been discovered and investigated in the area of modern-day Denmark. While varying considerably in size and composition, all finds include various types of weapons, elements of personal equipment and other objects. Historical and archaeological parallels suggest that the finds represent the votive offerings of a victorious army to some sort of war god.

Four different offerings from the period between the third and the fifth centuries AD have been identified and excavated at Illerup Aadal, a long stretch of river valley on the eastern side of the peninsula of Jutland, south-west of the modern city of Aarhus. The majority of the artefacts belong to the oldest, and by far the biggest, of the various offerings (Illerup Aadal A) dating from the early third century. Today, this particular offering can be regarded as one of the largest and most complex sites of its kind (Ilkjaer 2000). The assemblage from Illerup Aadal A alone comprises 350 shields, 366 lances and 410 spears, at least 100 swords, 11 sets of riding gear, 124 sets of strike-a-lights and other elements of the warriors' personal equipment, as well as a complete set of different tools for wood- and metalworking. The find therefore paints a complex picture of an army consisting of at least 400 men organised within a strict hierarchical structure with complex functional differentiation and associated with a well-developed logistical apparatus (Ilkjaer 1990, 1993, 2001 ; von Carnap-Bornheim & Ilkjaer 1996a; Dobat 2008).

In the course of the sacrificial ritual, the majority of the artefacts underwent a process of systematic destruction before being deposited on the bottom of the shallow lake of Illerup Aadal, either as wrapped bundles dropped from boats or simply thrown into the lake from its southern shore (Ilkjaer 1990: 13-27, 2002).

[FIGURE 1 OMITTED]

Although the different artefacts obviously represent individual combatants, there is no evidence at all for the sacrifice of humans or war-related deaths. The only living creatures that apparently played a role in the sacrifice were at least four horses and a single bovine (see also Rowley-Conwy 1980). The skeletal remains of these animals are extraordinarily well preserved due to the favourable conditions in the peat layers of the Illerup Aadal river valley (Figure 1). Flence, they provide a rare view of cavalry as a central aspect of Iron Age military infrastructure in Northern Europe. Furthermore, the horses from Illerup Aadal also provide a medium for the application of new scientific methods such as strontium isotope analysis, which has proved to be a powerful tool for the investigation of the origins of archaeological material. Tooth enamel from two of the horses has been analysed for strontium isotopes. The results afford new perspectives on the origin of the army from Illerup Aadal A.

The horses in the context of the Illerup Aadal deposit

The four horses were excavated close to the southern shore of the Iron Age lake; horses 1 and 2 were found within the core area of the major deposit of Illerup Aadal A, around point 59,50/47,00 and point 62,00/47,00 of the excavated area (see plan C in von Carnap-Bornheim & Ilkjaer 1996b). Lance- or spearheads and other pieces of weaponry and personal equipment were found lying together with the animal bones, documenting the connection of these two animals with the major sacrifice of Illerup Aadal A (dating to around the year AD 210) (Figure 2). The circumstances are more ambiguous in the case of horses 3 and 4, which were both found in the periphery of the deposition at Illerup Aadal A, around point 131,00/85,00 8 (see 131/85 Plan 1 in von Carnap-Bornheim & Ilkjaer 1996b) and point 223,00/132,00. However, the similarity regarding ritually inflicted injuries on all four horses (see below) strongly suggests that they were all part of the sacrificial events. Horses 1 and 2 were accompanied by a single bovine; this animal will, however, not be further discussed in the present study.

[FIGURE 2 OMITTED]

With the exception of a few missing bone elements, the skeletons of horses 1 and 3 were almost complete. In the case of horse 1, the lower leg bones (metapodia and phalanges) were found in the immediate vicinity of the torso, suggesting that some displacement occurred during the decay of the carcass. Horse 2 completely lacks both rear legs including the pelvic girdle, while horse 4 lacks only the left rear leg. Some of the animals were partly disarticulated prior to their deposition in the lake. The complete lack of gnawing marks on the skeletal remains of all four horses indicates that the horses were deposited and covered by water immediately or shortly after the sacrifice. The isolated left hind leg (from the femur to the phalanges) of a horse was found at point 83,00/93,00 (see 83/93 Plan 1 in von Carnap Bornheim & Ilkjaer 1996b). It is probable that this leg belonged to either horse 2 or 4. Due to the lack of bone elements suitable for detailed aging and sexing, however, these bones are not included in the following discussion.

The Illerup Aadal horses are not unique. Several horse skeletons were found during the early excavation of the weapon sacrifices at Nydam in southern Jutland (Engelhardt 1865: 36-42). Here they can probably be associated with major offerings during the third and fourth centuries AD (Rau 2010). Bones from one or more horses were also recorded during the early excavations of the sacrifices at Vimose, which is partly contemporary with the Illerup Aadal A sacrifice, and at Kragehul (Engelhardt 1867, 1869: 30), both on the island of Funen.

The ritual killings

Just as with the other material components of the find, the horses show clear traces of seemingly violent 'destruction'. The skull of horse 2 (Figure 3) displays traces of at least 12 cuts from sharp-edged (but not particularly heavy) instruments, 10 wounds from pointed weapons struck or shot from various directions and a heavy blow to the forehead with a massive blunt instrument (a hammer or the back of an axe). Three injuries from pointed weapons were found on the palate bone inside the nasal cavity, showing that the weapons were driven with great force through the skull. Only the heavy blow to the animal's forehead, however, can be identified as most definitely fatal. Of the 18 paired ribs only two ribs (costae) from the left and three ribs from the right have no visible injuries. The impacts are dominated by perforations made by pointed weapons and cut marks made by sharp-edged weapons. The majority of the wounds have been inflicted laterally but some perforations were found on the inside of the ribs suggesting that the weapons (most probably arrows) penetrated the torso. The direction of the injuries indicates that horse 2 was struck, shot and stabbed several times from various directions. The grouping of two distinct types of wounds probably indicates that the wounds were inflicted by at least two individuals standing next to each other with two different types of weapons (probably a sword and a lance or spear) (Figure 4). The orientation of the injuries and the fact that they are found on both sides of the torso indicate that the animal was standing while the wounds were inflicted.

Horse 1 received a blow to the forehead similar to horse 2. The injuries inflicted on the torso are not as conclusive as in the case of horse 2; however, as with the latter, some injuries may have been inflicted while the animal was standing. The position and direction of the deep cut marks on the lower front legs of horse 1, by contrast, indicate that they happened when the horse was lying on its left side. The injuries from cutting weapons inflicted on the right radius, ulna and metacarpal bones are found on the lateral side of the bones while injuries inflicted on the left third metacarpal are found on the medial part of the bone. Injuries from pointed weapons on the dorsal aspects of some vertebrae also suggest that horse 1 was lying on its side when the wounds were inflicted. Similar injuries can be observed on horses 3 and 4.

[FIGURE 3 OMITTED]

The modifications observed on the horse bones, their anatomical position and finally the types of implements used in this process do not resemble those typically made during skinning, butchery or similar processes (Noe-Nygaard 1989). We cannot rule out the possibility that at least some of the injuries had already been received in the course of the military clash which obviously preceded the sacrifice at Illerup Aadal. However, the quantity, position and severity of the injuries makes it rather unlikely that at least the majority could have been received during military engagements. It should be added that there are no signs of healed injuries that could indicate whether or not the horses were long-standing warhorses injured during previous military clashes. As this has been documented for the weaponry (most of the weapons show clear traces of having been systematically destroyed prior to their deposition in the sacrificial lake), the majority of the cuts and injuries most probably have to be seen as the result of sacrificial rituals, i.e. received around the time of death (peri-mortem).

In attempting to reconstruct the events that unfolded during the sacrifice, it seems likely that the horses were led into the shallow waters along the edge of the Iron Age sacrificial lake. Here, they received a multitude of blows and thrusts from different kinds of weapons. The evidence suggests that at least some of the wounds were inflicted while the animals were standing, in other words while they were still alive. Even though there is no definite evidence for it, one can speculate that the creatures were held in place by ropes. The heavy blow to the forehead with an axe, as seen on the skulls of horses 1 and 2, may represent the coup de grace. Afterwards the assault continued with the dead animals being shot at and stabbed with different kinds of weapons. The skull of horse 2 was cut into pieces, while the lower legs of horses 1 and 3 were broken and partly cut off. The missing rear parts of horses 2 and 4 could be the result of similar events, where parts of the body were disarticulated and deposited at different positions in the lake.

[FIGURE 4 OMITTED]

Charger or nag?

All four horses have pronounced canine teeth indicative of the male sex. Based on tooth development and wear of the mandibular incisor teeth, horse 1 can be estimated to have been about 10 years old, while horse 2 may have been somewhat younger, probably around 7 years (estimates on the basis of Habermehl 1975). Horses 3 and for 4 are estimated to have had an age of 8-9 years and 6-7 years respectively (Table 1). In modern dressage, reminiscent of military training (Museler 1972), the basic preparation of a horse is normally concluded after the animal is 6 or 7 years of age.

For this study, all complete long bones (i.e. femur, tibia, metatarsus, humerus, radius and metacarpus) and shoulder blades (scapulae) were measured according to the guide lines given by von den Driesch and Boessneck (1974: fig. 1) and von den Driesch (1976). The final estimation of the stature (height) of each horse was based on the mean value of all measurements from each individual as recommended by Kiesewalter (1888), von den Driesch and Boessneck (1974: 335) and Ambros and Muller (1975: 45-47). According to the size of the long bones, the horses had an estimated height at the withers between 130cm and 141cm (Table 1). The Illerup Aadal horses would thus have been comparable with average or large-sized prehistoric horses and modern day Icelandic horses with height at withers between 120cm and 140cm. Compared to other contemporary finds of horses from Iron Age contexts in Scandinavia, horse 2 appears to have been a surprisingly large animal. In a supra-regional perspective, it can be categorised as belonging to the upper range of Iron Age horses in Central and Northern Europe (Bokonyi 1968; Reichstein 2003). One could speculate whether this might indicate the import of horses from either Eastern Europe or the Roman Empire. It is of particular interest in this respect that various elements of riding equipment found in southern Scandinavian weapon sacrifices show close similarities to riding gear from the northern Danube area or the Roman Empire (Lau 2009: 272, 276).

The use of bits is indicated by pathological changes on the mandibles of the two oldest horses (1 and 3), which can be associated with the use of hard metal bits. Horse 1 exhibits distinct 'bone spurs' on the diastema (the gap between the canines and the cheek teeth), whereas horse 3 has a very steep bevel and a distinct dentine/enamel abrasion on the vertical anterior/mesial edge of the second lower premolars (P2) (Bendrey 2007; Anthony & Brown 2011). Two different types of mouth pieces are present in the Illerup Aadal assemblage: two examples of simple iron-jointed snaffle bits and nine complete sets of curb bits with stiff mouthpieces (Figure 5). The latter are highly complex devices with the actual mouthpiece and the lower parts of the reins made of bronze and iron. Their functional principle is equivalent to modern curb bits, which are particularly common in Western-style riding, with a stiff mouthpiece and lever arms. The standard equipment used for war-horses in the Roman cavalry displays clear parallels (Junkelmann 1992). Such curb bits are considered highly specialised gear which, when used in action, affect several parts of the horse's head and mouth. Accentuating the pressure applied by the rider, they hold the potential to inflict severe pain on the horse, and thus gain absolute obedience and control over its movements (Junkelmann 1992: 15; see also Lau 2009: 287). The use of such harsh tools would have presupposed a long process of training, making the animal accustomed to this particular type of mouthpiece.

The famous fourth-century-BC works on horses and horsemanship by Xenophon (Peri hippikes and Hipparchicus; Marchant & Bowersock 1968 [1925]) and the treatise on cavalry tactics in the Art Tactica by Arrian (second century AD; Hyland 1993) provide a detailed view of the enormous effort that was involved in the education of horses intended for use on the battlefield in a Greek or Roman military context (Gillmor 1992; Hyland 1993). The horses deposited at Illerup Aadal had reached the peak of their individual physical and mental development and the pathological changes on the mandibles and premolars are indicative not only of intensive use, but also training with iron bits. Even though horse training in Iron Age Scandinavia most certainly followed different standards than in Classical Greece and Rome or in modern dressage, we may assume that these animals underwent a lengthy and intensive training regime, enabling their masters to use them in a battlefield environment that was stressful for both rider and horse.

[FIGURE 5 OMITTED]

War-horse or scapegoat?

We cannot take for granted that these horses originally belonged to the armies represented by the offering. Several alternative scenarios leading to the horses' inclusion as an element of the sacrifice at Illerup Aadal can be proposed. They may have been picked up in 'enemy territory following the arrival of the armies via ship or boat. Another possibility is that the horses from Illerup Aadal did not belong to the opposing armies at all, but were chosen among the local stock, belonging to the victorious community itself, and used only for the occasion of the sacrifice.

There are, however, several lines of evidence which suggest that the inclusion of the horses in the sacrifice reflects their original association with the defeated armies of Illerup Aadal A. First of all, the horses show the same traces of 'destruction' as we see on certain material elements of the offering, especially the military equipment and the riding gear belonging to the highest level of the military hierarchy. Another argument can be based on the fact that the headgear found at Illerup Aadal seems to have been made to fit individual horses, given that it lacks any ability for adjustment to cater to the varying sizes and shapes of the heads of different horses (Lau 2009: 285). Indeed, the image of a military leader bringing non-adjustable headgear into enemy territory in the hopes of finding a suitable horse on which to put it renders this scenario rather implausible.

The most decisive argument, however, for the direct relation between the armies and the horses from Illerup Aadal A are the specific requirements for horses in combat situations, which presuppose extensive training and practice. In actual combat, the acquaintance of the horse with equipment and rider would have proved essential (see the discussion above). An unknown horse, taken into possession somewhere in enemy territory, would simply not have done the job. We can at least regard it as most likely that the horses deposited in the great sacrifice of Illerup Aadal A along with military and personal equipment originally belonged to the warrior group represented by this find. Hence, it follows that they originated from the same regional contexts as their masters. It is noteworthy that the horses found in the context of the sacrifices at Nydam support this assumption. According to the original report on their recovery in the late nineteenth century, the horses were found with parts of their headgear still in position (Engelhardt 1865: 36). Furthermore, the horses at Nydam had also received similar treatment to the animals at Illerup, showing a multitude of injuries from different types of weapons (Engelhardt 1865: 40, figs. A-E).

Where did the horses come from?

From the very beginning of modern research on the Illerup Aadal weapon sacrifices, the origin of the armies that are represented by the assemblage has been one of the major topics of debate. It has been suggested that the provenance of the material (and, hence, opposing armies) for the Illerup Aadal A deposit is the southern part of the Scandinavian peninsula (i.e. modern-day Norway or western Sweden). This hypothesis, based on the stylistic analysis of archaeological material (Ilkjaer 1993: 374-85) and scientific methods (Ilkjaer et al. 1994; Ilkjaer 2001: 376), has constituted the background for a historical scenario in which one or a coalition of several hostile armies crossed the Kattegat (the sea between Jutland and Sweden) and attacked the Jutland peninsula, where they were eventually defeated by an eastern Jutish military coalition.

Recent debate has focused on the question of whether the sites mirror defensive or offensive military actions, in other words whether the sacrifices were conducted by the defending communities (who had been attacked by the armies represented by the finds) or whether the booty was collected by attacking armies after a victorious expedition into foreign areas and thereafter transported back home for later sacrifice in a local context. Both scenarios seem possible on the basis of the background of the archaeological evidence at most of the finds. For the Illerup Aadal assemblage, the considerable number of implements of military infrastructure, such as tools for wood- and/or metalworking (Dobat 2008) may suggest that it was in fact the 'sacrificed' army that had been on campaign in foreign territories. In order to include both possible scenarios, we will avoid speaking of the army from Illerup Aadal A as the attacking army and instead use opposing army.

The uniform appearance of the assemblage at Illerup Aadal A suggests that the combatants had their origin within a coherent cultural community. However, there are also artefacts (e.g. several strike-a-light sets) which point toward the Continent (southern Denmark or modern-day northern Germany) as the most probable place of origin of either the warriors or the artefacts, or both (Ilkjaer 1993: 378). Hence, it is possible that the army had a more or less heterogeneous composition, with different units recruited from different areas, or individual combatants from various places joining the coalition for the limited period of a single military expedition (compare Jensen 2008: 146; see also Price et al. 2011). The horses from Illerup Aadal are of particular relevance in this regard, as they relate to the military infrastructure and the leading elites. Tracing the origin of the horses thus enables us at least to narrow the original spatial context of the organisational and strategic backbone of the military alliance that made up the army of Illerup Aadal A.

In order to identify the possible geographical origin of the horses, tooth enamel from horses 1 and 3 was analysed for strontium isotopes. This was done in full recognition of the various methodological limitations of strontium isotopes as indicators of regional origin and despite the similarity of baseline values across the moraine landscape of southern Scandinavia. It was expected, however, that the analysis would contribute positively to the question of whether or not the army of Illerup Aadal A originated from the ancient crystalline geological zones north and east of the Kattegat (Norway or western Sweden), as has previously been proposed, or whether they came from somewhere in the moraine landscapes of southern Scandinavia (modern-day Denmark, northern Germany or southern Sweden).

The principles of isotopic provenancing rely on isotopes that exhibit geographic variation and are incorporated in an animal's skeleton via the food chain (Sillen & Kavanagh 1982; Price 2000; Price et al. 2002). An essential question regarding strontium isotope analysis, and a prerequisite for distinguishing between local or foreign origin, is the identification of the specific strontium isotope signal for the area in which the animal remains were found (Price et al. 2002). The local bioavailable isotopic signal of the place of burial can be determined in several ways: in human bone from the individuals whose teeth are analysed; from the bones of other humans or archaeological fauna at the site; or from modern fauna in the vicinity. This baseline information on isotope values across an area needs to be obtained in order to make useful and reliable statements about the origins of the archaeological remains under study (Price et al. 2002; Frei & Price 2012).

Denmark is characterised by a relatively young (geologically-speaking) and rather homogenous 'basement geology. About 50 per cent of Denmark is composed of Late Cretaceous--Early Tertiary carbonate platforms, and the other 50 per cent by marine clastic sediments, all covered by glaciogenic sediments deposited during the two last ice ages. The Quaternary glaciogenic sediments are composed of, among other things, various weathered Precambrian granitoids (gneiss and granite) from Norway and Sweden. Almost everywhere in Denmark, glacial deposits are the source of strontium isotopes for plants, animals and people. There is very little bedrock exposure anywhere.

Frei and Price (2012) reported strontium isotope ratios from samples of archaeological fauna as well as modern mice and snails from Denmark. In addition they compared these ratios with strontium isotope median values from human enamel samples from archaeological sites within Denmark. The faunal samples range from [sup.87]Sr/[sup.86]Sr = 0.70717 to 0.71185 with an average of 0.70918, and for the humans (including non-locals) values range from [sup.87]Sr/[sup.86]Sr = 0.7086 to 0.7110, with an average of 0.7098 (Figure 6).

Frei and Frei (2011) measured [sup.87]Sr/[sup.86]Sr in almost 200 samples of Danish surface water and found similar results. In all these datasets we observe a small difference in the baseline values between the west (Jutland) and east (Funen, Zealand and the southern islands) of Denmark. Accordingly, two slightly different baseline ranges are proposed for the bioavailable strontium isotopic values within Denmark. The western area ranges from [sup.87]Sr/[sup.86]Sr = 0.7079-0.7099, whereas the range for the eastern area is defined as [sup.87]Sr/[sup.86]Sr = 0.7089-0.7108.

Tooth enamel from the mandibles of horses 1 and 3 from Illerup Aadal was taken for the strontium isotope analysis. Results of the analysis are presented in Table 1. These two values clearly fall within the range for Denmark (between 0.70717 and 0.71185) and suggest that the animals may have been of local origin within the region of Denmark, southern Sweden, or northern Germany.

[FIGURE 6 OMITTED]

Conclusions

The analysis of the four horses in the context of the major weapon sacrifice at Illerup Aadal invites further discussion on three general topics: 1) the character and significance of the sacrificial rituals themselves; 2) the nature of cavalry and its use in Iron Age warfare; and finally 3) the old question of the origins of the armies represented by the offerings.

The sacrifice of the animals has to be seen in the light of the prominent role of horses in sacrificial rituals throughout the entire Iron Age and in very different contexts. Only rarely, however, do we see such a detailed view of the chain of events that unfolded during the sacrificial ceremony as in the case of Illerup Aadal. The osteological analysis suggests that the horse sacrifice was a highly dramatic scene, in which the creatures suffered a gruesomely violent death, involving splattering blood and the agonised screams of the poor animals being cut down by several men with multiple weapons. Together with the shrill sound of sharp metal on silver-plated ornaments, heavy axes splintering shields and lances, and spears and swords being smashed to pieces, the scene must have provided a highly evocative and dramatic spectacle, leaving a lasting impression in the shared memory of the sacrificing community, and facilitating the commemoration of the ritual and the battle which preceded it. In this way the horses were made to take a leading part in what can be seen as a recreation of the apocalyptic scenery of the battle, in which they had already played a crucial role.

The combined evidence of the skeletal material from the four horses and the tack that was used to ride them provides us with a complex image of cavalry and its use in Iron Age warfare. On the one hand, the extreme ornamentation of mountings, rivets and decorative strap-ends on head gear, saddle straps and breastcollars explicitly signalled the high social status of the riders as well as their prominent role within the military hierarchy. On the other hand, the horses also emerge as highly developed and effective war machines. Given that they were likely to have been well trained and were experienced in wearing highly specialised gear (enabling the rider to fight from horseback with his weapon in one hand, while at the same time keeping full control of the animal, even in the stressful context of a battlefield), the horses from Illerup Aadal would have spread fear and terror among any opposing fighting lines.

With the [sup.87]Sr/[sup.86]Sr values below the baseline of strontium isotope values in both western Norway and western or central Sweden, it is unlikely that the horses came from either of those regions. They could, however, have come from the Viken area in southern Norway. Here we find a very varied baseline, with the very lowest values on the regional scale being similar to the values of the Illerup horses. On the other hand, there is a good match between the two Illerup horses and the [sup.87]Sr/[sup.86]Sr baseline for southern Scandinavia. Values around 0.710 are commonly found in various places in Jutland, Zealand and Scania, suggesting that the horses might well have been born and raised somewhere in the moraine landscapes of modern-day Denmark. The data sample is not yet adequate, and the results are not sufficiently clear to support a reconsideration of the earlier theories on the origin of the army of Illerup Aadal A. Nonetheless, these new results open up new perspectives on this question and on the background of the battle that preceded the sacrifice. Instead of an 'external' conflict involving military coalitions from across the Kattegat, the Illerup Aadal sacrifice might just as well be seen as an expression of a more regional, internal conflict between various southern Scandinavian groups. Further analyses have to be conducted, however, to evaluate this hypothesis.

Received: 18 March 2013; Accepted: 8 May 2013; Revised: 11 July 2013

Acknowledgements

The authors wish to thank the two reviewers for their beneficial comments and ideas. We are grateful to Lars Foged Thomsen for his drawings of horses and horse-gear. Sara Heil Jensen (Moesgard Museum) provided valuable comments on the horse equipment from Illerup Aadal and its possible use.

References

AMBROS, C. & H.-H. MULLER. 1975. Zur Methode der Widerristhohenberechnung aus den Langenmassen der Extremitatenknochen beim Pferd, in A.T. Clason (ed.) Archaeozoological studies: 45-50. Amsterdam & Oxford: North-Holland.

ANTHONY, D.W. & D.R. BROWN. 2011. The secondary products revolution, horse-riding and mounted warfare. Journal of World Prehistory 24: 131-60.

BENDREY, R. 2007. New methods for the identification of evidence for bitting on horse remains from archaeological sites. Journal of Archaeological Science 34: 1036-50.

BOKONYI, S. 1968. Mecklenburg Collection, Part I: data on Iron Age horses of Central and Eastern Europe (American School of Prehistoric Research Bulletin 25). Cambridge (MA): Peabody Museum.

DOBAT, A.S. 2008. Werkzeuge aus kaiserzeitlichen Heeresausrustungsopfern: mit besonderer Berucksichtigung der Fundplatze Illerup Adal und Vimose (Jernalderen i Nordeuropa; Jysk Arkaeologisk Selskabs Skrifter 61). Arhus: Jysk Arkaeologisk Selskab.

ENGELHARDT, C. 1865. Sonderjyske og Fynske Mosefund. Band 2: Nydam Mosefund, 1859-1863. Kobenhavn: I commission hos G.E.C. Gad.

--1867. Kragehul Mosefund: 1761-1865: Et Overgangsfund mellem den celdre Jernalder og Mellem-Jernalderen (Fynske Mosefund 1). Kobenhavn: I commission hos G.E.C. Gad.

--1869. Vimose Fundet (Fynske Mosefund 2). Kobenhavn: I commission hos G.E.C. Gad.

FREI, K.M. & R. FREI. 2011. The geographic distribution of strontium isotopes in Danish surface waters--a base for provenance studies in archaeology, hydrology and agriculture. Applied Geochemistry 26: 326-40.

FREI, K.M. & T.D. PRICE. 2012. Isotopes and human mobility in prehistoric Scandinavia. Journal of Anthropological and Archaeological Sciences 4: 103-14.

GILLMOR, G. 1992. Practical chivalry: the training of horses for tournaments and warfare. Studies in Medieval and Renaissance History 13: 7-29.

HABERMEHL, K.-H. 1975. Die Altersbestimmung bei Haus- und Labortieren. Berlin & Hamburg: Paul Parey.

HYLAND, A. 1993. Training the Roman cavalry: from Arrians Ars Tactica. Dover: Alan Sutton.

ILKJAER, J. 1990. Illerup Adal 1: Die Lanzen und Speere (Textband). Hojbjerg: Jysk Arkaeologisk Selskab.

--1993. Illerup Adal 3: Die Gurtel (Textband). Hojbjerg: Jysk Arkaeologisk Selskab.

--2000. Illerup Adal--et arkaeologisk tryllespejl. Hojbjerg: Moesgard Museum.

--2001. Illerup Adal 9: Die Schilde (Textband). Hojbjerg: Jysk Arkaeologisk Selskab.

--2002. Den bevidste odelaeggelse i krigsbytteoffringer, in K. Jennbert, A. Andren & C. Raudvere (ed.) Plats och praxis: studier av nordisk fdrkristen ritual (Vagar till Midgard 2): 203-13. Lund: Nordic Academic.

ILKJAER, J., A. JOUTTIJARVI & J. ANDRESEN. 1994. Illerup Adal. Proveniensbestemmelse afjern fra Illerup Adal: etpilotprojekt (Illerup Adal Sma Skrifter 1). Moesgard: Jysk Arkaeologisk Selskab.

JENSEN, X.P. 2008. Vimose revisited--perspectives and preliminary results, in A. Abegg-Wigg & A. Rau (ed.) Aktuelle Forschungen zu Kriegsbeuteopfern und Furstengrabern im Barbaricum (Schriften des Archaologischen Landesmuseums-Erganzungsreihe 4): 137-50. Neumiinster: Wachholz.

JUNKELMANN, M. 1992. Die Reiter Roms. Teil III: Zubehor, Reitweise, Bewaffnung. Mainz: Philipp von Zabern.

KIESEWALTER, L. 1888. Skelettmessungen am Pferde als Beitrage zur theoretische Grundlage der Beurteilungslehre des Pferdes. Unpublished PhD dissertation, Universitat Leipzig.

LAU, N. 2009. Die Pferdegeschirre aus dem Thorsberger Moor. Studien zu germanischen Zaumzeugen und Sattelgeschirren als Zeugnisse der militarischen Reiterei im mittel- und nordeuropaischen Barbaricum der jiingeren romischen Kaiserzeit. Band I: text. Unpublished PhD dissertation, Christian-Albrechts-Universitat zu Kiel.

MARCHANT, E.C. & G.W. BOWERSOCK (trans.). 1968 [1925]. Xenophon. Volume VII (Loeb Classical Library 183). Cambridge (MA): Harvard University Press.

MUSELER, W. 1972. Ride here/Reitlehre. Kobenhavn: Aamodt.

NOE-NYGAARD, N. 1989. Man-made trace fossils on bones. Human Evolution 4: 461-91.

PRICE, T.D. 2000. Les isotopes du strontium dans les restes squeletiques. Etude des migrations de populations archeologiques. Les Nouvelles de lArcheologie 80: 29-34.

PRICE, T.D., J.H. BURTON & R.A. BENTLEY. 2002. Characterization of biologically available strontium isotope ratios for the study of prehistoric migration. Archaeometry 44: 117-35.

PRICE, T.D., K.M. FREI, A.S. DOBAT, N. LYNNERUP & P. BENNIKE. 2011. Who was in Harold Bluetooths army? Strontium isotope investigation of the cemetery at the Viking Age fortress at Trelleborg, Denmark. Antiquity 85: 476-89.

RAU, A. 2010. Nydam Mose. Diepersonengebundenen Gegenstande. Grabungen 1989-1999 (Jernalderen i Nordeuropa; Jysk Arkaeologisk Selskabs Skrifter 72). Hojbjerg: Jysk Arkaeologisk Selskab.

REICHSTEIN, H. 2003. Pferd. [section] 2. Zoologisch-Archaologisches, in H. Beck, D. Geuenich, H. Steuer & R. Muller (ed.) Reallexikon der Germanischen Altertumskunde 23: 29-35. Berlin: Walter de Gruyter.

ROWLEY-CONWY, P. 1980. Two horses and a cow from Roman Iron Age Illerup. Unpublished report for the Illerup Aadal project, Moesgard Museum, Hojbjerg.

SILLEN, A. & M. KAVANAGH. 1982. Strontium and paleodietary research: a review. American Journal of Physical Anthropology 25: 67-90.

VON CARNAP-BORNHEIM, C. & J. ILKJAER. 1996a. Illerup Adal 5: Die Prachtausriistungen (Texthand). Hojbjerg: Jysk Arkaeologisk Selskab.

--1996b. Illerup Adal 8: Die Prachtausriistungen (Grabungsdokumentation undFundliste). Hojbjerg: Jysk Arkaeologisk Selskab.

VON DEN DRIESCH, A. 1976. A guide to the measurement of animal bones from archaeological sites (Peabody Museum Bulletin 1). Cambridge (MA): Peabody Museum.

VON DEN DRIESCH, A. & J. BOESSNECK. 1974. Kritische Anmerkungen zur Widerristhohenberechnung aus Langenmassen vor- und friihgeschichtlicher Tierknochen. Saugetierkundliche Mitteilungen 22: 325-48.

Andres S. Dobat (1), T. Douglas Price (2), Jacob Kveiborg (3), Jorgen Ilkjaer (3) & Peter Rowley-Conwy (4)

(1) Department of Culture and Society, Aarhus University, Moesgard Alle 20, Hojbjerg, DK8270, Denmark

(2) Department of Anthropology, University of Wisconsin, Madison, WI53706, USA

(3) Moesgard Museum, Moesgard Alle 20, Hojbjerg, DK 8270, Denmark

(4) Department of Archaeology, Durham University, South Road, Durham DH1 3LE, UK

Table 1. The estimated sex, age and height at withers of the
four horses from Illerup Aadal A and strontium isotope ratios
of tooth enamel from horses 1 and 3.

                            Height at withers (in cm)

                 Age in                           [sup.87]Sr/
          Sex    years    Min.    Max.    Mean    [sup.86]Sr

Horse 1   male     10     123.7   133.0   129.8    0.710657
Horse 2   male     7      138.5   144.2   141.3
Horse 3   male    8-9     130.2   139.6   136.3    0.710357
Horse 4   male    6-7     128.8   139.1   135.0