The Late Roman period (late second to late fourth century ad)

In twenty-seven cemeteries of the Roman period, 154 human-horse graves were found, but only just over half provided reasonably exhaustive information (Figure 1; Table 1). Most of these graves were found in western Lithuania, in the so-called western Lithuanian stone-circle graves cultural area and in its small peripheral zone in south-western Latvia. In the latter cultural area, graves of armed men of high social status with parts of horses alongside appeared in the late second to early third century ad (Michelbertas, Reference Michelbertas1986: 40–41; Bliujienė & Butkus, Reference Bliujienė and Butkus2007). In western Lithuania, human-horse graves emerged under the cultural influence of the Sambian-Natangian culture (Jaskanis, Reference Jaskanis and Antoniewicz1966, Reference Jaskanis2013: 253–55; Ibsen & Skvorzov, Reference Ibsen and Skvorzov2004). In the cemeteries of the Roman and Migration periods, parts of horses or whole horses were found in male graves only; here we refer to them as human-horse graves. However, in western Lithuania, unlike in the Sambian-Natangian culture, most often specially selected horse body parts were found in the graves rather than a whole horse. For the Roman period, informative data are available for only sixty-one graves, from which we note that the head of the horse was the most frequent element (forty-two cases), representing the largest group of symbolically sacrificed horses. The head and limb(s) were buried much less frequently (ten cases), and constitute the second-largest horse sacrifice group. In only five cases were the head, limb(s), ribs, and vertebrae present (third group); a whole horse was buried in four cases (fourth group). In western Lithuania, such instances of burying a whole horse, on the left of the individual, are known from the cemeteries of Reketė (grave 35) and Baitai (grave 31); it is likely two other destroyed graves of this type were found in the Aukštkiemiai cemetery (formerly Oberhof; grave 238 and grave found in 2015) (Navickaitė-Kuncienė, Reference Navickaitė-Kuncienė and Tautavičius1968: 165–66; Banytė-Rowell, Reference Banytė-Rowell2007: 29–30; Bliujienė & Butkus, Reference Bliujienė and Butkus2007: 111; Baltramiejūnaitė & Tamulynas, Reference Baltramiejūnaitė and Tamulynas2015: 136) (Figures 2 and 3).

Figure 2. Horse body parts found in Lithuanian burial sites of the late second to late seventh centuries ad.

Figure 3. Remains of a horse head found in a human-horse grave (57/1998) in the Lazdininkai cemetery.

The poor preservation of human bones means that the sex of the individuals buried was established on the basis of weapon sets and elements of weaponry (one to two, or even three, spears, an axe, a fighting knife, a shield, a shoulder-strap, etc.); some men were buried with weapons only (one or two spears and an axe), and ultimately some other men were buried with just a universal working tool and a weapon, i.e. an axe (Bliujienė & Butkus, Reference Bliujienė and Butkus2007: 98–104). It is specifically in such graves of armed men that horse parts were found. Human-horse graves spread quite fast, as the mid-third century ad marked the apogee of the custom: at that time, graves of this type account for 9.1 per cent of a total number of researched graves in the stone-circle graves cultural area (Bliujienė, Reference Bliujienė2013: 283–88). Starting in the second part of the third century, the number of such graves decreased consistently in western Lithuania until, finally, in the first half of the fifth century, the custom of burying an armed man with parts of a horse finally disappeared (Bliujienė & Butkus, Reference Bliujienė and Butkus2007: 105). In the Migration period and later, a horse was symbolized merely by a snaffle-bit in western Lithuanian cemeteries (Bliujienė & Butkus, Reference Bliujienė and Butkus2009: 152).

The graves of the Roman period found in western Lithuania can be divided into three main groups on the basis of the repeated correlation between man and horse (Bliujienė & Butkus, Reference Bliujienė and Butkus2007: 100–04). The first group includes graves where an armed man and parts of a horse were buried in the same stone circle and in the same grave, with the heads oriented in the same direction, mostly to the north-west. Graves where just one or two, or occasionally more, horse teeth were found are included in this group. It is most likely that in such graves bone decay and other destructive processes resulted in the teeth being the only remains of a horse's head. The second group consists of graves where an armed man and parts of a horse were buried in the same stone circle, but in different graves, only some 30 to 50 cm distant from each other. In other words, the parts of the horse were at arm's length from the armed man. The third group of human-horse graves consists of complexes where the ritual offering of a horse was surrounded with the graves of armed men. The groups of men buried with parts of horses indicate a community's efforts to distinguish between armed men buried with parts of a horse and men buried merely with weapons and elements of weaponry. Moreover, the graves of armed men with parts of a horse stood out for the wealth and abundance of their grave goods. The graves include miniature cups, a whetstone or a fire steel, a sickle, bronze or even silver jewellery, a set of female jewellery, Roman sestertii, an ornate wrought-iron bridle with noseband, snaffle-bits, curb bits, and spur(s) (Bliujienė, Reference Bliujienė2013: fig. 29.1–3). An ornate bridle was not just a decorative element: in battle, it facilitated control of the horse and protected its head from blows. The bridles, as well as other artefacts discovered in western Lithuanian cemeteries, have parallels in the Sambian-Natangian, Bogaczewo, and Sudovian cultural sites of north-eastern Poland and among the offerings found in the wetlands of Jutland (La Baume, Reference La Baume1944: 2–3; Wilbers-Rost, Reference Wilbers-Rost1994: 202, 204, 209; Bitner-Wróblewska, Reference Bitner-Wróblewska2007: 81; Nowakowski, Reference Nowakowski2009: fig. 5; Bliujienė, Reference Bliujienė2013: fig. 29.1–3, 189–88; Lau, Reference Lau2014: 24–66) (Figure 4).

Figure 4. Details of bronze bridles found in human-horse grave 4/1991 (ad 220–260) in the Lazdininkai cemetery. After: Bliujienė, Reference Bliujienė2013: fig. 29.3.

Figure 5. Human-horse relationships in Lithuanian burial sites of the late second to late seventh century ad.

In the western Lithuanian cemeteries of the Roman period, the relationship between man and horse was accentuated by elements of burial practice. The parts of the horse were more frequently placed on the left than on the right of the individual. The trend continued in the Migration period, when human-horse graves spread to other regions of Lithuania. It is conceivable that the burial of parts of horses on the left of the human was based on real-life practice: a right-handed man would mount his horse from the left (Figure 5). Horse mounting on the left is documented in iconographic sources: the horse mounting blocks in Far Eastern cultures were also depicted on the left side (Shuvalov, Reference Shuvalov, Sinitsyn and Kholod2014: 568).

While in other Lithuanian regions only single human-horse graves were found, in the burial sites of the lower reaches of the river Nemunas (Dauglaukis) and the Trans-Nemunas region (Liepynai), as well as in the south-eastern Lithuanian stone mounds (Moša–Naujasodžiai barrow), humans were buried with whole horses (Abaravičius, Reference Abaravičius1993; Jovaiša & Asadauskas, Reference Jovaiša, Asadauskas and Jovaiša1993) (Figure 6). Single burials where horse teeth have been found include the Pavėkiai barrow cemetery and the Naujasis Obelynas cemetery in the Samogitian highlands (a specific ethnographic part of western Lithuania) (Michelbertas, Reference Michelbertas1986: 66–67; Astrauskas, Reference Astrauskas1998: 178).

Figure 6. A horse found in barrow 2 in the Moša-Naujasodžiai barrow cemetery. Photograph reproduced by kind permission of G. Abaravičius.

The Migration period (fourth/fifth to seventh century ad)

In the period of the fourth/fifth to the seventh century, which in the eastern Baltic is equated with the Migration period (see Bliujienė, Reference Bliujienė2011: tab. A), Lithuania remained the northernmost part of Eastern Europe where men were quite frequently buried with parts of a horse, a whole horse, and even several horses. Such burial sites are scattered over the lower reaches of the Nemunas, the Samogitian highlands, and central and eastern Lithuania (Tautavičius, Reference Tautavičius1981; Kazakevičius, Reference Kazakevičius1991, Reference Kazakevičius1993: 42–45;Vaitkunskienė, Reference Vaitkunskienė1995; Bertašius, Reference Bertašius2002: 43–44, Reference Bertašius2009a: 54; Bliujienė & Butkus, Reference Bliujienė and Butkus2009: tab. 1; Bliujienė & Steponaitis, Reference Bliujienė and Steponaitis2009: 185–97). In the graves of that period, the quantity of sacrificed horse parts increased, and the burial of whole horses became more frequent. In the Migration period, the number of burial sites rose to twenty-seven, and the number of human-horse graves more than doubled (to 154 in total). However, only seventy-three graves provide information on the kind of burial practised. Human-horse burials in the same grave predominate, with the horse or parts of a horse most frequently placed on the left of the human (Figure 5). However, horse graves unrelated to specific individuals made an appearance, and the distribution of such graves changed. Human-horse graves disappeared from western Lithuania, while their number increased in the lower reaches of the river Nemunas, and they appeared in central and eastern Lithuania (Figure 1). A general observation about the period concerns the increase in horse body parts (head, legs, vertebrae, and fragments of ribs) selected for burial (forty-six cases) and of complete horses/horse burials (sixteen cases). Moreover, the instances where just the head of a horse was buried (nine cases) or the head and several leg bones (two cases) decreased (Figures 2 and 7). The head of the horse remained a constant part of the symbolic horse sacrifices.

Figure 7. Horse parts found in horse grave 6/3 in the Užpelkiai cemetery.

Besides inherited custom, some different relationships between humans and horses are reflected in their graves: the head and legs of a horse were placed at the individual's feet either on the left or the right side, or were placed on a coffin over the deceased's head (Vaitkunskienė, Reference Vaitkunskienė1995: 73–76). When a whole horse was buried, it was most frequently laid on its left side, sometimes slightly higher than the buried individual, and sometimes as if on a separate step formed while digging the pit (Šimėnas, Reference Šimėnas1999, Reference Šimėnas2000; Petkus, Reference Petkus2012) (Figure 8).

Figure 8. Smalininkai cemetery, human-horse grave 4. Photograph reproduced by kind permission of M. Petkus.

In the Migration period in Lithuania, horses were further buried next to well-armed men whose graves were rich in grave goods and contained indicators of their high social status (such as silver jewellery, imported weapons, and elements of weaponry). Several fifth to seventh century high-status graves are known to have contained parts of two horses or both parts of a horse and a whole horse. Such graves were found in the cemeteries of Užpelkiai (grave 66 and horse grave 5/2), Vidgiriai (graves 35 and 36), Kalniškiai (grave 39), Taurapilis (barrows 1, 4, 5, and 6), and Paduobė-Šaltaliūnė (barrow 17) (Tautavičius Reference Tautavičius1981; Kazakevičius, Reference Kazakevičius1991, Reference Kazakevičius1993; Šimėnas, Reference Šimėnas1999, Reference Šimėnas2000; Bliujienė & Steponaitis, Reference Bliujienė and Steponaitis2009).

At the end of the Migration period, the spread of the cremation rite changed the relationship between men and horses in graves to a degree, although burials in the same grave survived; either parts of a horse or a complete horse were buried. Moreover, there are graves where the remains of the cremated people were placed in a pit over the complete horse (Bertašius, Reference Bertašius2009a: 145). Graves of armed men cremated together with their horses appear as early as in the second half of the sixth century (in the Šereitlaukis cemetery, cremation of an adolescent, 10–15 years old) (Jankauskas, Reference Jankauskas and Zalepūgienė2013: 84; Piličiauskienė, Reference Piličiauskienė2013a: 79). At the end of the period under discussion, around the sixth/seventh century, the first horse graves unrelated to humans emerge: such burials were found in the cemeteries of central Lithuania, e.g. at Plinkaigalis (graves I to III) and Marvelė (grave 26) (Kazakevičius, Reference Kazakevičius1993: 42–45; Bertašius, Reference Bertašius2005: 90). The emergence of graves of horses buried separately from people, and their rapid spread, is related to the dissemination of the practice of cremation. Around the early seventh century, graves of horses unrelated to specific individuals also appeared in north-western Europe and Anglo-Saxon Britain (Oexle, Reference Oexle1992: 123, 139; Fern, Reference Fern and Pluskowski2005: 43–44).

Zooarchaeological data

The horse bones investigated were measured in accordance with von den Driesch's (Reference von den Driesch1976) methodology, and the age was estimated following the criteria established by Levine (Reference Levine, Wilson, Grigson and Payne1982) and Kolda (Reference Kolda1936). The withers height of the horses was determined by the length of the long bones, with the exception of the femur (Johnstone, Reference Johnstone2004: 161, 248), by means of the coefficient corrected and proposed by May (Reference May1985); the withers height of the horse discovered in the Paduobė-Šaltaliūnė barrow was established from the dimensions of its skull. Where both length measurements were present, the average of the two estimates was used. The metacarpal slenderness index was described according to the criteria listed by Brauner (Reference Brauner1916).

Horses of the late second to late fourth century

Although several human graves of that period containing parts of horses and whole horses are reported in the literature, the bones that have survived are very fragmentary, decayed, and not very informative. Lazdininkai is one of the central cemeteries in western Lithuania where thirteen human-horse graves were found. Only six fragments of skull and jaws were examined (graves 2/1991, 3/1991, 4/1991, 6/1992; 9/2000, 57/1998) (Figure 3). The age of the horses ranged between three and ten years. In some cases, only an approximate age could be established: one grave (grave 6/1992) contained a horse aged 3–7 years old. In the Anduliai-Ėgliškiai cemetery, the remains of a horse skull, the lower jaw, and the first two cervical vertebrae were found in grave 43 (2002). According to our data, the age of the horse was 7–8 years old, while previous research estimated it to be no older than 3.5 years (Baublienė et al., Reference Baublienė, Daugnora, Daugnorienė, Piličiauskienė, Trainienė and Veličkaitė2005: 292). The skull was possibly cut off below the second cervical vertebra; however, the bones were substantially decayed, and no traces of potential cuts could be recorded. Given the characteristics of the burial process in the Roman period and the poor preservation of the horse bones, we have almost no data on the horses’ withers height, sex, and other physical traits or pathologies. Only the horse discovered in a disturbed grave in the Aukštkiemiai cemetery in 2014 can be described more exhaustively (Baltramiejūnaitė & Tamulynas, Reference Baltramiejūnaitė and Tamulynas2015: 136). The remains of the skull, vertebrae, the front and hind limbs, and pelvic bones suggest that a complete horse must have been buried there. Its withers height was calculated as 122.2 cm, and it was older than five years (Figure 9; Table 2).

Figure 9. Horses' withers height in cm. For data, see Table 2.

Table 2. Morphological data of horse long bones from burial sites.

MC: metacarpus, RA: radius, TB: tibia, MT: metatarsus. Data for Plinkaigalis a, Plinkaigalis b, Marvelė, Degsnė-Labotiškės, Obeliai, Pakalniškiai from Veličkaitė, Reference Veličkaitė2006.

Judging from the sets of grave goods in the graves of the late second to late fourth century, it was the horses that were ridden by warriors that were buried, and this from the very beginning of the custom of burying them alongside humans. The examination of horse bones in neighbouring countries has indeed revealed specific pathologies caused by riding, but also by cart pulling, or other farming activities (Piątkowska-Małecka, Reference Piątkowska-Małecka2000; Serwatka, Reference Serwatka and Bitner-Wróblewska2007; Karczewska et al., Reference Karczewska, Karczewski and Gręzak2009; Kobryń & Świeżyński, Reference Kobryń, Świeżyński, Kontny, Okulicz-Kozaryn and Pietrzak2011: 159; Bendrey, Reference Bendrey2012: 144–47). For comparing the withers height of the horses found in Lithuanian burial sites, the data from the Sambian-Natangian, Bogaczewo, and Sudovian cultural sites, where whole horses were buried, are relevant. The withers heights of the horses buried in the Paprotki Kolonia cemetery of the Bogaczewo culture ranged from 115 to 127 cm and from 139 to 143 cm (Karczewska et al., Reference Karczewska, Karczewski and Gręzak2009), and those from the Sudovian barrow cemeteries measured 135 to 136 cm (Krysiak & Serwatka, Reference Krysiak and Serwatka1970; Serwatka, Reference Serwatka and Bitner-Wróblewska2007: 143–45). The withers of the horses of the Sambian-Natangian culture in the Roman period were 120 to 132 cm high (Zinoviev, Reference Zinoviev2009). The studies of horses in the Balt cultures as a whole showed that almost exclusively male horses were buried, but it is impossible to establish whether they were stallions from the material available (Piątkowska-Małecka, Reference Piątkowska-Małecka2000; Kobryń & Świeżyński, Reference Kobryń, Świeżyński, Kontny, Okulicz-Kozaryn and Pietrzak2011: 159). The horses of Roman and German warriors were mostly stallions (but mares were also present; see Johnstone, Reference Johnstone2004: 51). Nevertheless mares are known to have been used, and indeed make up a significant proportion of the horses from funerary contexts in Vendel-period Sweden (Fern, Reference Fern and Pluskowski2005: 66).

Therefore, one may assume that the same tradition existed among the Balts. They were mainly adult horses, fit for riding and farming activities, more often healthy than showing pathologies caused by exertion, trauma, or disease (Gręzak & Piątkowska-Małecka, Reference Gręzak, Piątkowska-Małecka and Bitner-Wróblewska2009; Karczewska et al., Reference Karczewska, Karczewski and Gręzak2009).

Horses of the fourth/fifth to seventh century

In the Migration-period burial sites, human-horse graves account for thirteen to twenty-two per cent of all the graves examined and for twenty-five to fifty-eight per cent of all the investigated male graves (Figure 10). In addition, previously unavailable information now includes the withers height and the sex of the horses (Figure 9; Table 2).

Figure 10. Relationship between excavated graves, male graves, and human-horse graves in the late second to late seventh centuries ad (data from Bezzenberger, Reference Bezzenberger1909; Vaitkunskienė Reference Vaitkunskienė1981, Reference Vaitkunskienė1995; Kazakevičius Reference Kazakevičius1991, Reference Kazakevičius1993; Šimėnas, Reference Šimėnas1999, Reference Šimėnas2000).

In the Užpelkiai cemetery, dated to the transitional stage between the late Roman and the early Migration period, twelve graves were found, i.e. almost forty-three per cent of all the investigated male graves. The remains of the horse found in grave 3/1 consisted of its front legs, scapula, skull, and the remains of the mandible (Figure 7). Its age was estimated as six to seven years old. In the Barzūnai, Rubokai, and Vidgiriai cemeteries in the lower reaches of the river Nemunas, parts of horses were recovered in as many as forty-six to fifty-eight per cent of the male graves investigated. In thirteen graves of the Vidgiriai cemetery, men were buried with a horse's head, while graves 35, 36, and 38 had two heads per grave. The age of these horses was six to nine years old. Meanwhile, in the cemetery of Viešvilė II, only one grave with the remains of a 3–4.5-year-old horse was discovered. These modest but informative statistics show that the number of human-horse graves in the cemeteries of the lower reaches of the Nemunas and the Samogitian upland were associated with these communities’ cultural contexts and were an integral part of the burial custom for males, while the horse as a manifestation of the buried man's social status and its relationship with warfare apparently remained in the background.

Meanwhile, in central Lithuania, the custom of burying men together with horses was only just appearing; only single human-horse graves have been found in the region (Kalniškiai, graves 36, 39, 139, 200, and 206; Plinkaigalis, grave 61; and Marvelė, grave 177). Out of seventy-six male graves excavated in the Kalniškiai cemetery, only five contained horses. Grave 39, however, had at least two horses buried at different depths (Kazakevičius, Reference Kazakevičius1991).

In the eastern part of Lithuania, in the Taurapilis barrow cemetery, four horse skeletons or their fragmented parts were found. The horses were buried in the graves of armed men. The withers height of the horse buried in barrow 1 was estimated to be around 128.8 cm, and it was over 3–3.5 years old. Barrow 4 contained a male horse, 5.5 years old and 130.8 cm at the withers. The metacarpal index of the horse was 15.9, which, following Brauner's criteria, classifies the animal as slightly massive-legged. In barrow 6, a horse was found with some still unfused epiphyses of the long bones: its age is estimated at 2–3.5 years old and the animal would have grown larger. The height of the horse, estimated by the length of the long bones, was 125.9 cm. The metacarpal index of 16.2 makes this animal slightly massive-legged. In barrow 5, the horse buried next to a military chieftain of the highest social level was two years old and 144–145 cm tall, according to previous studies (Barauskas, Reference Barauskas1971). As we have calculated the withers height of the horses found in other barrows as 7–9 cm smaller than those reported by Barauskas, one can assume that the withers height of the horse found in barrow 5 could have been 135–138 cm; in any case, it was taller than the other horses recovered in the barrow cemetery.

In the barrow cemetery of Paduobė-Šaltaliūnė in eastern Lithuania, the withers height of a 7–8-year-old horse in barrow 17, buried in a man's grave, came to 136–144 cm. The grave had been robbed in antiquity, but the horse burial was left undisturbed because the animal had no grave goods. In the cemetery at Plinkaigalis, three horse graves and four horse skeletons were recovered: they are the first horse graves not directly related to human cremations. We measured the metacarpal and metatarsal bones of three horses. The withers height of the first horse was 128.4 cm and the metacarpal index was 15.0. According to Brauner's criteria, it was a medium slender-legged animal. The withers height of a second horse was estimated to be 128.7 cm and its metacarpal index was 16.0, which classifies the animal as slightly massive-legged. A third horse, a 6–7-year old male from grave III B, was 136 cm at the withers and its metacarpal index of 14.9 classifies him as medium slender-legged. The average stature of the men buried in the Plinkaigalis cemetery was 174.83 cm, and among the richly furnished graves their stature could be as much as 176.89 cm (Jankauskas, Reference Jankauskas1998: fig. 1). This is tall for the period in question (even if today this would not be considered unusual); men had horses that would seem small in our eyes, but they were a perfect fit for them.