Late Viking-Age Dragon-Slayers — Two Unusual Urnes-Style Brooches from Sjælland, Denmark

Abstract

IN 2003 AND 2007, TWO UNUSUAL VIKING-AGE BROOCHES were discovered during metal-detector surveys on Sjælland (Zealand), Denmark. The brooches depict a horseman killing a serpent-like animal with his spear. The stylised beast is clearly associated with the Urnes style, dating to c ad 1050–1100, and the meaning of the motif is discussed in the context of early medieval dragon-slayer narratives and imagery. The acceptance and transformation of Classical and Byzantine iconographic traditions of equestrian saints has a long history in continental Germanic and Scandinavian art. In the Migration and Merovingian periods these images were incorporated into a pagan iconography, but the Urnes style was used in contexts which were definitively Christian, and it is possible that the two Danish brooches represent a very early depiction of Saint George in Scandinavia. Together with other finds, they add new perspectives to our understanding of the adoption of Christian iconography and narratives in Scandinavia.

Résumé

Chasseurs de dragons de la fin de l’ère Viking — Deux broches originales dans le style Urnes à Sjælland, au Danemark par Anne Pedesen et Sigmund Oehrl

En 2003 et 2007, deux broches originales de l’ère Viking ont été découvertes lors de campagnes de détection de métaux à Sjælland (Zélande), au Danemark. Les broches représentent un cavalier qui tue avec sa lance un animal ressemblant à un serpent. La bête stylisée est clairement associée au style Urnes, daté d’environ 1050–1100, et la signification du motif est évoquée dans le contexte d’histoires et d’images de chasseur de dragon remontant au début du Moyen-Âge. L’acceptation de traditions iconographiques classiques et byzantines de saints équestres et sa transformation sont très anciennes dans l’art continental germanique et scandinave. Durant les périodes de migration et mérovingienne, ces images ont été incorporées dans une iconographie païenne, mais le style Urnes était utilisé dans des contextes sans aucun doute chrétiens; il est possible que les deux broches danoises portent une représentation très précoce de saint George en Scandinavie. Avec d’autres artefacts, elles apportent un nouvel éclairage sur ce que l’on sait de l’adoption de l’iconographie et des discours chrétiens en Scandinavie.

Zussamenfassung

Spätwikingerzeitliche Drachentöter – Zwei ungewöhnliche Fibeln im Urnes-Stil aus Sjælland, Dänemark von Anne Pedesen und Sigmund Oehrl

Bei Metalldetektoruntersuchungen auf Sjælland (Seeland), Dänemark, wurden 2003 und 2007 zwei ungewöhnliche Fibeln aus der Wikingerzeit gefunden. Sie stellen einen Reiter dar, der mit seinem Speer ein schlangenähnliches Tier tötet. Das stilisierte Untier ist eindeutig dem Urnes-Stil zuzuordnen, der ca. 1050-1100 n. Chr. datiert. Die Bedeutung des Motivs wird im Kontext frühmittelalterlicher Drachentöter-Überlieferungen diskutiert. Die Rezeption der antiken und byzantinischen Ikonographie der Reiterheiligen hat in der kontinentalen germanischen und skandinavischen Kunst eine lange Tradition. In der Völkerwanderungs- und Merowinger-/Vendelzeit wurden diese Bilder in die heidnische Ikonografie aufgenommen, der Urnes-Stil hingegen begegnet in eindeutig christlichen Kontexten. Möglicherweise zeigen die beiden dänischen Fibeln eine sehr frühe skandinavische Darstellung des Heiligen Georg. Zusammen mit anderen Funden eröffnen sie neue Perspektiven für unser Verständnis der Übernahme christlicher Ikonografie und Vorstellungen in Skandinavien.

Riassunto

Ammazzadraghi di epoca tardo-vichinga: dall’isola danese di Selandia due insolite fibule nello stile di Urnes di Anne Pedesen e Sigmund Oehrl

Nel 2003 e nel 2007, durante le ricerche con cercametalli nell’isola danese di Selandia (Zelanda), furono rinvenute due insolite fibule di epoca vichinga. Le fibule ritraggono un cavaliere nell’atto di uccidere con la lancia un animale serpentiforme. L’animale stilizzato è chiaramente associato allo stile di Urnes risalente al 1050-1100 d.C. circa e il significato dell’immagine viene discusso nel contesto delle narrative e delle raffigurazioni altomedievali del cavaliere ammazzadraghi. L’adozione e la trasformazione delle iconografie classiche e bizantine di santi cavalieri hanno una lunga storia nell’arte germanica continentale e in quella scandinava. Nel periodo delle migrazioni e nel periodo merovingio, tali immagini vennero incorporate nell’iconografia pagana, ma lo stile di Urnes fu impiegato in contesti definitivamente cristiani ed è possibile che le due fibule danesi costituiscano una delle prime raffigurazioni di San Giorgio in Scandinavia. Insieme ad altri ritrovamenti, esse aprono nuove prospettive per la nostra comprensione dell’adozione in Scandinavia dell’iconografia e delle narrative classiche.

acknowledgements

For kind help, explanations, and advice concerning the Novgorod pendants and their dating, we would like to thank Vlad Sobolev, St Petersburg. For translations of Russian research literature, we sincerely thank Sofiya Oehrl, Göttingen. We would also like to warmly thank Dirk Steinforth for thorough proof-reading of our text.

Notes

3 Dobat 2013.

4 Museum Archives 2003:127; National Museum Archives Næstved Museum Archives 2003:127; National Museum Archives 49/2004, inventory no D887/2004.

5 Museum Vestsjælland Archives; National Museum Archives 52/2007, inventory no D228/2007.

6 Watt 2009; Helmbrecht 2011, 167–8, fig 43b.

7 See Baastrup 2009.

8 National History Museum of Latvia (Riga), inventory no A 9964:1585.

9 Zem$\overline{ı}$tis 1998a, 591.

10 Ibid, fig 1:2.4.

11 Ibid, 593.

12 Statens historiska Museum Stockholm, inventory no SHM 6620:2; see Hårdh 1976, no 71; Fuglesang 1980, cat no 34.

13 Gräslund 2006, 122–6.

14 See eg Pedersen in press.

15 For a summary and introduction to the literary sources, see eg Page 1990, 69–77, and Ney 2017; also eg von See 2006. Although often referred to as a dragon, the beast was a giant serpent as shown at Ramsundsberget, cf Graham-Campbell 2013, 146. For the art motif see Blindheim 1973a; 1973b; Margeson 1980; Düwel 1986; 2005; Oehrl 2013a; Bertelsen 2015; Ney 2017.

16 Södermanland, rune stone no Sö 101 — Brate and Wessén 1924–36, 71–3, pl 48. See also Düwel 2003.

17 Bergendahl-Hohler 1999, cat no 114. Now kept in the Museum of Cultural History, Oslo University, Norway.

18 For the churches from Lunde and Nes in Telemark, see Tuulse 1968, 93; Blindheim 1973a, 31, figs 20–1; Blindheim 1973b, 17, 25, 28, figs 24–5. In Denmark, the killing of Fáfnir has been identified in a few Romanesque stone carvings, eg a tympanum from Tømmerby church (Kaspersen and Thøgersen 1974), and a baptismal font from Astrup (Oehrl 2013a, 373–8, figs 9–10). There are many other uncertain or even unlikely images of Sigurðr, most of them included in Blindheim 1973a and 1973b (see Düwel 1986, 247–9 for critical comments). In addition, some medieval Icelandic images (in particular a wooden panel and a book illumination, see Guðmundsdóttir 2017, 30–7) and even a depiction on a Spanish church building are considered (Oehrl 2013a, 371–8, fig 8).

19 Fried 1989, 438–70; Callahan 2003.

20 Kretzenbacher 1958.

21 See Book of Revelation verses 12:7–9 and 20:1–3.

22 Schiller 1990/91; Schaller 2006; for the fettering, see also Oehrl 2011, esp 148–9, 257–8, figs 266–8.

23 Andersson 1953, 81–5. As suggested by Lang (1982, 59) and others, and argued in more detail by Düwel (1986, 264–71), an important reason for the success of the cult (and iconography) of Saint Michael in the North was its similarity to the Sigurðr legend.

24 Düwel 1986, 264; Oehrl 2011, 236–7.

25 Düwel 1986, 264–5; Gschwantler 1998, 744–5; Oehrl 2011, 236. Denmark: DR 212, 380, 398, 399, 402; Sweden: G 203, U 478. For the Danish examples, see also Jacobsen and Moltke 1941/42, 686 sv Michael and 991–3 sv Sjæleformler.

26 N 404 — Olsen 1941–1960, 213.

27 For early medieval Scandinavian depictions of Saint Michael as psychopomp and weigher of souls, see Oehrl 2011, 236–42.

28 Sahl: Christensen and Johannsen 2018, 2827. Øster Starup: Stubbe-Teglbjærg 2009, 128. Skåne: Lang 1982, 58, figs 2–3.

29 Karlsson 1980; Nordanskog 2006. The fittings have been dated very differently, from the early 12th to the beginning of the 13th century (Karlsson 1980, 3). However, recent dendrochronological analysis shows that at least the wooden door itself was manufactured around 1275 (Nordanskog 2006, 141).

30 Runverket Vg 202 — Svärdström 1970, 356–9; Düwel 1986, 265.

31 Dixon et al 2001; Moe (1955, 20) dates them to about 1020, Karlsson (1983, 84) to the early 12th century.

32 Karlsson 1983, 84, 165–6, figs 189–90; Dixon et al 2001.

33 Schaller 2006.

34 Depictions of dragon-slayers on foot without halo or wings also occur in medieval art. They could represent Saint Michael or any legendary knight, but cannot be identified with certainty. Scandinavian examples are the iron fittings from Kungslena in Västergötland from c 1200 (Karlsson 1990, fig 8), one of the stone carvings from Rydaholm church in Småland from the 12th century (Stubbe-Teglbjærg 2009, 143), as well as the baptismal fonts from the Græsted (Jutland) and Lihme (Sjælland) churches (Nancke-Krogh 1995, 112, 136).

35 Lewis 1973, 54–5, fig 31; Vierck 1981, fig 5.5; Deckers 2001, 50, fig 6; Eger 2008, 6, fig 7.

36 Eger 2008, 6–7, figs 1, 4a–c.

37 Concerning the equestrian saints in general and their most famous representative, Saint George, see eg Strzygowski 1903; Roosval 1924; Lucchesi Palli and Braunfels 1974; Schwarz 1972; Kretzenbacher 1983; Rohner 1995; Hahn et al 2001; Heyden forthcoming.

38 Lewis 1973, 55, fig 32; Parlasca 1982; Hahn et al 2001, cat no I.15; Barthol 1963, cat no 77; Effenberger 1976, 208–9, pl 47. See also Frenschkowski forthcoming.

39 Several examples are kept in the Museum für Byzantinische Kunst in Berlin (Staatliche Museen Berlin) and the Kunstgewerbemuseum in Berlin; see Holmqvist 1938, pl 35:1; Holmqvist 1939, pls 20:4, 22:6.

40 Deckers 2001, 50.

41 Holmqvist 1939, 110–28, pls 20:4, 22:1–6; Lewis 1973, 54, fig 28; Böhner 1982, 103–8; Brune 1999; Deckers 2001; Fingerlin 2010, 34–46; Stein 2012, 218–20; Iamanidzé 2016.

42 The Sousse wall tiles are found at the Sousse Museum (Fourlas 2008, no 12) and Archäologische Sammlung München (Hahn et al 2001, no II.31). The Skopje fragment is held at the Museum of Macedonia (Walter 2003a, pl 25; 2003b, 98, fig 12.6; Fourlas 2008, fig 39, no 11; Heyden forthcoming, fig 4.) There are many more ancient Byzantine depictions of Theodore killing the dragon. In these, however, the saint is usually fighting on foot rather than on horseback. See Fourlas 2008 on these objects and the general topic.

43 Fourlas 2008, no 13, with literature.

44 The piece was kept in Berlin but has been lost since the end of World War II. See Wulff 1909, no 1128; Holmqvist 1938, pl 22.4; Braunfels-Esche 1976, fig 76; Svanberg and Qwarnström 1993, 18, fig 11.

45 Sammlung C.S. München. Hahn et al 2001, no II.32A–C.

46 Hahn et al 2001, 150.

47 Menghin et al 1987, 277, 293, figs 14–15, cat no VI,47a; Eger 2008, 6, fig 8b.

48 Iamanidzé 2016, fig 14 and passim.

49 Strzygowski 1918, 290–1, figs 329–30; Lucchesi Palli and Braunfels 1974, 370; Walter 2003a, xi, 127, pls 27–28; Iamanidzé 2016, 184–5, figs 31–2.

50 Walter 2003a, fig 27; 2003b, 99, fig 12.8.

51 Strzygowski 1918, fig 329.

52 Staatliche Museen Berlin, Preußischer Kulturbesitz, Skulpturenabteilung. Braunfels-Esche 1976, 77; Svanberg and Qwarnström 1993, 20, fig 12.

53 Hahn et al 2001, 138–66; for a general summary of the cult and worship of Saint George in the Byzantine East, see Blazudakis 1992. See also Svanberg 2011, 67–119.

54 Münchener Staatsbibliothek, Mirakelsammlung clm 14473. Hahn 2001, 78; Haubrichs 2001, 57. There is a record in a 12th-century manuscript from the monastery St Emmeran in Regensburg, but this seems to be directly based on a Greek source.

55 Forrer 1912; 1927, 756–67, fig 565; 1934, 232, fig 70; Holmqvist 1939, 117, pl 23:2; Böhner 1982, 106–7, pl 26:2; Quast 2012, 22, fig 20.1.

56 Phalerae, ornamental disks or bosses, were worn on the heads or breasts of horses or by men. Fingerlin 1976; 2008; 2010; 2012.

57 Quast 2012, 11, 20–5, fig 11.

58 Périn 1989, 35, fig 13; Böhner 1982, 107, pl 27.4–5; Quast 2009, 334, fig 4.2; 2012, 24, fig 20.5.

59 Baum 1937, 74, pl 8:14; Hauck 1957, 10–11, fig 1; Drack 1979, fig 29; Böhner 1982, 107, pl 28:1; Quast 2009, 334, fig 3.4.

60 Laur-Belart 1943; Quast 2009, fig 3.4; 2012, fig 20.7.

61 Böhner 1982, 103–5, pl 14–19.

62 Werner 1979, 461–5, pl 34.

63 Holmqvist 1938, 69–70; Holmqvist 1939, 110–28; Holmqvist 1952; Holmqvist 1977, 197–8; cp Böhner 1991, 714–16.

64 IK 65 — Hauck et al 1985–1989, 121, pls 77–8.

65 Concerning the Christian influence on the iconography and meaning of the Migration-period gold bracteates see von Padberg 2011 and Hauck 1986.

66 Böhner 1982, 103–5; Quast 2002, 269, 275; 2009; Fingerlin 2010, 42–6.

67 Ellmers 1970, 217–19; Quast 2002, 257–8; Pesch 2015, 94. Karl Hauck interpreted the rider as a Nothelfer, a dioscuric helper in time of need; Hauck 1983, 439.

68 Arne and Stolpe 1912, 13, pl 6:1; Helmbrecht 2011, cat no 1090, figs 6g, 11f–g; Beck 1964, 9–12, 19–31.

69 This motif is frequently associated with the Old Norse god Óðinn and his two ravens. When seeking a new meaning and re-interpretation of the Mediterranean motif in its Scandinavian context, however, it seems equally possible to link the Vendel image with the so-called ʻbeasts of battleʼ topos, which is very popular in Old Norse poetry (Beck 1964; Oehrl forthcoming a). Numerous accounts of battles mention the wolf, the raven and the eagle, as they feast on the slain, eating the corpses on the battlefield and drinking the blood of dead warriors in the aftermath (Beck 1970, 55–67; Honegger 1998; 2017; Jesch 2002). Nonetheless, it must be emphasised that a horseman killing a serpent does not correspond to any account in Old Norse literature, neither in the myths of Óðinn nor in the imagery of Skaldic poetry.

70 More examples of equestrian saints with birds: Holmqvist 1938, 79, pl 35:1; Holmqvist 1939, 123, pl 20:4; Effenberger 1976, cat no 384; Böhner 1991, 708–9, pl 66; Brune 1999, 266, fig 2.10; with angels or victoriae instead of birds: Holmqvist 1939, 115, pl 22:6; Lewis 1973, 49, fig 28; Böhner 1991, 708, pl 65:2.

71 Picture stone no U 1055. See Jansson and Wessén 1953–1958, 323–6; Ahlberg 1978, 16–17, no 8.

72 Lindqvist 1941/42 vol I, fig 5; vol II, 110, fig 481; Lindqvist 1962; cf Lamm and Nylén 2003, 29. The horseman mentioned here was discovered only in 2013, published by Oehrl 2017, 98–103, and discussed in detail by Oehrl 2019a, 105–20. See also Oehrl forthcoming b and c.

73 Oehrl 2017, 98–103; 2019a, 109–20; 2019b, 161–7; forthcoming b; forthcoming c. The discs on the early Gotlandic picture stones, occasionally encircled by serpents or decorated with short lines or spikes looking like a kind of corona, are commonly regarded as celestial bodies, representing cosmological concepts (eg Ellmers 1981; 1986, 342–50; Hauck 1983; Andrén 2012; 2014). As a result, the Sanda monument is divided into two cosmological spheres, the world above and the world below. The fight between the equestrian and the beast takes place in the underworld. This observation goes well with the common interpretation of the rowing boat on the early picture stones of Gotland, that the vessel that carries the dead to a transmarine world of death (eg Hauck 1983, 546, 577). Whether the tiny tree at the border between the two cosmological spheres could therefore be regarded as a kind of world tree (and if so, whether there might be a connection to the world tree Yggdrasill) remains uncertain. The beast directly beneath the roots of the tree would, on this background, be reminiscent of the dragon Níðhǫggr who, according to Grímnismál stanza 35 (Neckel and Kuhn 1983, 64), is sitting under the world tree, gnawing at its roots. In Vǫluspá, stanza 39 (Neckel and Kuhn 1983, 9), Níðhǫggr is devouring the corpses of the dead at the shore of the netherworld (Nástrǫndr). On a side note, it should be mentioned that the motif of a cosmic, world-encircling serpent is recorded also in Christian, particularly in Egyptian-Coptic tradition (Oehrl 2013b), which seems to indicate further (Eastern) Mediterranean influence on the iconography of the Sanda monument.

74 Holmqvist 1953, 11.

75 Barthol 1963, cat no 325, 351; Lewis 1973, 49–51.

76 Lindqvist 1941/42 vol I, 91–3; Holmqvist 1953; Oehrl 2013b.

77 Lindqvist 1941/42 vol I, figs 141, 144; vol II, 47, fig 353. The observations and interpretations summarised in the following are described in more detail by Oehrl forthcoming c.

78 Lindqvist 1941/42 vol I, 120–1.

79 Summarising Oehrl 2019a, 18–21.

80 Dalton 1924, 389–90, pl 17:3; Bonner 1950, pl 14:294–295; Garam 2001, pls 31:3, 18:4.

81 Kantorowicz 1998, 122–3; Küchler 1989.

82 Dorigny 1891; Besson 1909, 141, fig 81; Bonner 1950, pl 17:324; Menzel 1955 with pl 4:1–2; Kantorowicz 1998, 121–4, pl 33 fig 33a, pl 34 fig 34.

83 Kantorowicz 1998, 122, pl 34 fig 35.

84 Lindqvist 1941/42 vol II, 47.

85 See Holmqvist 1938, 79; Holmqvist 1939, 123–4.

86 Baum 1937, 81, pl 17:46; Drack 1979, figs 15a–b; Böhner 1982, 107, pl 26:1; Quast 2009, fig 4.1; 2012, 22–4, fig 20.4; Fingerlin 2010, 43, fig 28.

87 Jost 2014; 2017, 271–83, fig 12; 2018.

88 Lewis 1973, esp 45–9; Brune 1999, esp 120–2, 190–2.

89 Oehrl 2006, 29, 30–2; Jansson and Wessén 1949–1951, 206–27, figs 128–30, pl 45.

90 Haubrichs 1984; Hahn 2001, 79.

91 Roosval 1924, 53, pl 6:1; Pesant 1992, 20; Svanberg and Qwarnström 1993, fig 14.

92 Spörlein 1992, 24; Kiesewetter 1992; Pfister 2001, 65.

93 Hahn et al 2001, 184, cat no VIII.1. The specimen presented in the catalogue is from 1116. The seal was used in Bamberg from the 11th until the 15th centuries. See also Zimmermann 1992.

94 Bibliothèque Nationale Paris; Cod lat 2251, fol 51v Dufour 1972, 77; Haubrichs 1984, 384; Hahn 2001, 79.

95 Svanberg and Qwarnström 1993, 20–1, fig 13; Vigi 2000; Svanberg 2011, 55, fig 32.

96 Svanberg and Qwarnström 1993, 23, fig 16, fn 20, 27. See also Keyser 1927 (1904).

97 For medieval images of St George in Scandinavia, most of them dating to the 14th and 15th centuries, see Kjær 1990; Svanberg 1990; 1997; 2011; Svanberg and Qwarnström 1993.

98 Svanberg 1990, 30; Svanberg and Qwarnström 1993, 29–30, 207, fig 23; see also Hallbäck 1972, 142, figs 104–9.

99 Svanberg 1990, 29, fig 1; Svanberg and Qwarnström 1993, 29, 206; see also Söderberg 1971.

100 It is not always easy to differentiate between the dragon-slayers Saint George and Saint Michael. Occasionally, Michael is seen without wings, and George is shown off his horse, which seems to mix up both iconographies.

101 Nordanskog 2006, 183–5, fig 53.

102 Svanberg and Qwarnström 1993, 20.

103 For a discussion of the typology and chronology, see Sobolev 2017a; 2017b. Further examples in Kroker 2011, 30, 110–11, nos 206–10.

104 Sobolev 2017a, 544.

105 All information about the dating of these pendants is taken from Vlad Sobolev, St Petersburg (see fn 105 and pers comm).

106 Regarding the basic characteristics and dating of the style, see for instance Wilson and Klindt-Jensen 1966, 160; Wilson 2008, 336–7; Graham-Campbell 2013, 133–57. See also Karlsson 1983, 71–80, 164–8, and Fuglesang 2001, 172–94.

107 Bonde 1993, 234.

108 Fuglesang 1998; Gräslund 2006.

109 Bonde 1993, 312.

110 Fuglesang1981, 103–6; Graham-Campbell, 2013, 186–90.

111 Lønborg 1994; Bertelsen 1994. For Lund/Uppåkra, see Bergman and Billberg 1976, 206–8; Hårdh 2010, 265–8, fig 10; for Ribe and Aalborg, see Søvsø and Jensen 2020.

112 In modified form, the Urnes style also remained in use in 12th-century Ireland, at least until the 1130s, but probably even later. In particular, the Royal Cross of Cong, an exquisite processional cross combining different early medieval styles (Insular, Romanesque, Urnes) should be mentioned here, dated by an inscription to 1123 (Murray 2014).

113 For Scania, see Helmbrecht 2011, 167, fig 43b. For Latvia, see Zem$\overline{ı}$tis 1998b, 104, fig 3:1.

114 For further examples dating to around 1100 or the early 12th century, see Roosval 1924, pl 2:6, 3:4; Frings and Willinghöfer 2010, 325–6, cat no 446.

115 Archäologische Staatssammlung München. Hahn et al 2001, 153–4, cat no II.44.

116 Landesmuseums Mainz, Germany (inventory no O,1730); Frings and Willinghöfer 2010, 187, cat no 91.

117 Restle 1967 vol 1, 121–3; vol 2, figs 32, 156, 246–7; Braunfels-Esche 1976, fig 20; Walter 2003b, 99–102, summarising several of those depictions in Cappadocia. See also Svanberg 2011, 99–104, figs 46, 76–9.

118 Deckers 2001, 51, fig 10.

119 Iamanidzé 2016, fig 31.

120 Fuglesang 2001, 191.

121 Grinder-Hansen et al 2013, 145–7. The Cross of Cong mentioned above and the bronze crozier-head from Þingvellir in Iceland, dating to about 1060 (Eldjárn 1971, 67, figs 3a–b), are also ornamented in the Urnes style.

122 Thompson 1975, 30–1.

123 Summarising the types of Christian inscriptions: Düwel 2008, 142–7, and Oehrl 2011, 99. See also Gschwantler 1998.

124 Segelberg 1972, 162.

125 For more detail see Oehrl 2015. Four-legged beast- or dragon-like animals are the main figurative motif on late Viking-Age rune stones. It has been argued that these Ringerike- and Urnes-style quadrupeds should be regarded as a stylised beast of prey, ie a lion or a wolf. Both pagan and Christian links have been considered and remain possible. Either way, the meaning of the beast motif revolves around elements of eschatology. The same ʻamalgamationʼ of pagan and Christian motifs seems to be the context of several depictions of Urnes-style serpents eating their own tail. Concerning this material and its interpretation, see Oehrl 2011 and 2015.

126 Book of Revelation 2:7, cf 22:2 and 22:14. Oehrl 2015, 481–2, fig 25; see also Fuglesang 2001, 182–3, who calls it ʻthe lambs adoring the crossʼ. The carver may have seen depictions like that on a pilgrimage to Rome or Jerusalem.

127 Bauerreiß 1938; Domagalski 1990, 36–8.

128 Bauerreiß 1938, 7–37; Bertelsen 2006, 41–4.

129 Ibid, 35–40.

130 Christiansson 1995; Oehrl 2015, 489, fig 47; For serpents on rune stones and the Fall of Man, see also Bertelsen 2006, 35–40. An early image of a winged serpent/dragon can also be seen on the rune stone from Åbyggeby (U 1142), signed by the carver Åsmund.

131 The image represents the text of Psalms 90(91):13: Super aspidem et basiliscum calcabis conculcabis leonem et draconem (‘Upon the asp and the basilisk you shall tread, and you shall trample the lion and the dragon’). Tweddle at al 1995, 191–2, Ills 232–4; Oehrl 2015, 468–9, fig 15.

132 Oehrl 2015, 482; Bergman 1948, 82–84; Weber 1972, 332.

133 Gräslund 2014.

134 Pedersen 2001; 2014; in press.

135 Wittkower 1939; Schneider and Stemplinger 1950; Wehrhahn-Stauch 1968; Oehrl forthcoming a.

136 Oehrl 2011, 16 fn. 23, 32, 223–4, 227.

137 Oehrl 2006, 123–4; 2011, 239–40, fig 128.

138 Ibid, 117–18, figs 322, 365.

139 Ibid, 111–17, figs 352, 353, 357.

140 Jansson and Wessén 1949–1951, 207; Oehrl 2011, 31–2; 2015, 471–2; Gräslund 2014, 29.

141 Holmqvist 1939, 113; Barthol 1963, fig 351.

142 In other cases, a big cross is located in the same position as the Söderby horseman, directly above the beast, presumably conveying the same message — the victory of Christendom over the Devil.

143 Braunfels-Esche 1976, 197–8; Svanberg and Qwarnström 1993, 17; Hahn 2001, 77; Fourlas 2008, 519–25 (catalogue).

144 Helmbrecht 2011.

145 Odenius and Kilström 1960, 269; Svanberg 1990, 29; Svanberg and Qwarnström 1993, 29.