The helmet from Tjele

The fragment of the helmet from Tjele. Author: Arnold Mikkelsen, Nationalmuseet. Taken from the catalogue of National Museum of Denmark.

In 1850, an extraordinary find was discovered by a young farmer in the forest called Lindum Storskov, near Tjele, Denmark. The find consisted of a set of blacksmith equipment – two anvils, five hammers, three tongs, sheet metal shears, two files, a wedge, two nail headers, casting bowls (with traces of tin and lead), a small touchstone, a set of scales, nine weights, five sickles, a key, three iron nails, an axe, two jingles, a spearhead/arrowhead, bronze wires, a lid of a box for scales, bone and bronze fragments of a casket, a mount of a drinking horn, iron fragments and pieces of a helmet (Leth-Larsen 1984; Lund 2006: 325). Thanks to local authorities, the set was sent to Copenhagen, where it was analyzed. The find was published three times – in 1858 (Boye 1858), then in 1939 (Ohlhaver 1939) and finally in 1984 (Munksgaard 1984; Leth-Larsen 1984).


Some other objects from the find from Tjele. Taken from Boye 1858: Pl. II–IV.

The helmet fragment is a very interesting object, that was originally interpreted as a saddle mount. It was Elisabeth Munksgaard, who expressed the theory about the helmet. Still, it is rather an overlooked artefact that was never studied in detail nor scientifically reconstructed. That’s the reason why this article was written.

Munksgaard sums up several important details:

This winged-shaped object is not a saddle mounting, but the eyebrows and nose-gueard of a helmet, made of iron and bronze. […] We are, unfortunately, not able to judge what the Tjele helmet looked like. There is not a trace of chain mail rest of the helmet, nor any iron plates fit for making up the rest of the helmet. But there are eight fragments of thin iron strips, about 1 cm broad and of varying length which might have been used for joining the plates together.” (Munksgaard 1984: 87)

More than detailed description, her article includes the comparison with the helmet from Gjermundbu. Since she considers the helmet from Gjermundbu to be the closest analogy, it is obvious she interprets the fragment as a part of a spectacle low-domed helmet. This type of helmets was used until 1000 AD (Munksgaard 1984: 88). The dating of the find from Tjele was corrected by Lund (2006: 325, 339), who claims the set belongs to the period between 950–970 AD. Tweedle (1992: 1126) assumed that the mask was multi-pieced; two ocular pieces were riveted to the nasal. The hole in the broader piece of the nasal could support this theory. Moreover, the mask from Kyiv shows the same feature.

The size of the mask is not convincingly given, but both Munksgaard and Tweedle suggest it is 12 × 7 cm (Munksgaard 1984: 87, fig. 4; Tweedle 1992: 1128, fig. 561). Just in the middle of eyebrows, at the base of the nasal, a hole for a rivet is placed. At least one decorated bronze strip was mounted on the eyebrows. It seems that entire eyebrows were symmetrically covered by bronze strips like this one. As a result, the mask was a distictive feature of the helmet, as can be observed in cases of other helmets too (Gjermundbu, Lokrume, Kyiv or St. Wenceslas helmet).

Regarding the construction, we can not say much. Munksgaard gives the information about eight fragments of narrow bands, which makes it possible to imagine that the helmet could have the similar construction as the helmet from Gjermundbu. The dome of the helmet of Gjermundbu is formed by four triangular-shaped plates. Under the gap between each two plates, there is a narrow flat band, which is riveted to a somewhat curved band located above the gap between each two plates. In the nape-forehead direction, the flat band is formed by a single piece, that is extended in the middle (on the top of the helmet) and forms the base for the spike. There are two flat bands in the lateral direction. Triangular-shaped plates are riveted to each corner of the extended part of the nape-forehead band. A broad band, with visible profiled line, is riveted to the rim of the dome. Two rings were connected to the very rim of the broad band, probably remnants of the aventail. In the front, the decorated mask is riveted onto the broad band.

The scheme of the helmet. Made by Tomáš Vlasatý and Tomáš Cajthaml.

The scheme of the helmet of Gjermundbu. Made by Tomáš Vlasatý and Tomáš Cajthaml.

Even though the mask from Tjele is just a fragment, we can not underestimate the meaning of this find. It broadens our vision about Viking Age protective gear, its decoration and the makers. Recently, two of my friends have tried to replicate the helmet fragment from Tjele. The reconstruction of the complete helmet is impossible, but I personally think that these both versions are decent and plausible tries that should be accepted by reenactment community.

First, let’s have a look on the work of Dmitry Hramtsov. The dome of this version is based on Vendel Period helmets. Since multi-pieced masks are typical for pre-Viking helmets, such a dome seems to be understandable. Metal bands are, however, much wider than those found in Tjele. The eyebrows are decorated with 14 bronze strips.


The second try is the helmet made by Konstantin Shiryaev and Maxim Teryoshin. In this case, the dome is based on the helmet from Gjermundbu. Konstantin used 16 bronze strips.


Boye, V. (1858). To fund af smedeværktøi fra den sidste hedenske tid i Danmark (Thiele-Fundet og Snoldelev-Fundet). In: Annaler for Nordisk Oldkyndighed og Historie, København: 191–200.

Leth-Larsen, B. (1984). Selected objects from the stock of the Tjele smith. In: Offa 41, Neumünster: 91–96.

Lund, J. (2006). Vikingetidens værktøjskister i landskab og mytologi (Viking Period tool chests in the landscape and in mythology). In: Fornvännen 101, Stockholm: 323–341.

Munksgaard, E. (1984). A Viking Age smith, his tools and his stock-in-trade. In: Offa 41, Neumünster: 85–89.

Ohlhaver, H. (1939). Der germanische Schmied und sein Werkzeug. Hamburger Schriften zur Vorgeschichte und Germanischen Frühgeschichte, Band 2, Leipzig.

Tweddle, D. (1992). The Anglian Helmet from 16-22 Coppergate, The Archaeology of York. The Small Finds AY 17/8, York.

The sword from Sarskoe Gorodishche


A drawing of the sword. Taken from Leontiev 1996 : 120, Fig 47:7.

Sarskoe Gorodishche (Hillfort on the bank of the Sara River) is one of the few settlements on the territory of ancient Russia, where a large amount of Scandinavian material culture occured. Both quality and quantity bears witness of not only trade contacts, but also of direct Scandinavian presence on the site. The most representative collection of Scandinavian objects is weaponry, mainly arrow tips, sword and seax chapes and a sword. The sword will be the topic of this article.

According to some sources, the sword was discovered on the slope of Sarskoe Gorodishche by D. N. Eding and D. A. Ushakov in 1930. However, the sword was firstly published A. N. Kirpichnikov in 1966, as a find from a mound (Kirpichnikov 1966: 80, No. 49). The sword was studied several times (Kirpichnikov 1992: 79, Leontiev 1996: 121; Kainov 2000: 252-256); nevetherless, in 2003, the sword was studied again and some new decoration was discovered. At the present time, the weapon is deposited in Architecture and Art Museum in Rostov (Ростовское архитектурно-художественное музей; catalogue number Р 10335, А- 92).

The lenght of the sword is 94.6 cm, the blade is 78.4 cm long. The blade has the width of 55 mm by the crossguard and 30 mm by the tip (30 mm far from the tip, respectively). The thickness of the blade by the crossguard is 5 mm. The fuller is 23 mm wide and 1 mm deep in the upper part of the blade. The crossguard (lower guard) is 90 mm wide and 20 mm high, while the upper guard (base of the pommel) is 80 mm wide and 20 mm high. The pommel has the height of 46 mm.


Description of sword parts, according to Peirce – Oakenshott 2002.


The sword from Sarskoe Gorodishche. Taken from Kainov 2011: 152, Fig. 10.

Both hilt and blade are very well preserved. The shape of the hilt belongs to the Petersen type E, which was very popular type with at least 123 examples in whole Europe (39 from Sweden, 31 from Norway, 20 from Finland, 15 from the Ancient Rus, 6 from Estonia, 6 from the former Prussia, 4 from Ireland, 1 from Poland; Kainov 2012: 19-21 and my personal observations). More correctly, the shape of the hilt should be classified as the subtype E3. This subtype is “represented by hilts decorated with oval pits arranged in trefoil or quatrefoil compositions” (Androshchuk 2014: 53; Kainov 2001: 57). To compare, Androshchuk lists at least 5 Swedish swords of the subtype E3 (ibid.). Until 2003, all studies had been pointing out that the sword from Sarskoe Gorodishche had been a typical example of this subtype, but after the examination, the sword showed to be rather unique. The reason is its decoration, which is not typical for any subtype of the type E. The decoration is why we should thing the sword forms “a separate variant of the E-type swords” (Kainov 2011: 149).


Four main types of pit decoration on swords of the type E (E1, E2, E3, E4). After Kainov 2001: 57, Fig. 4, taken from Androshchuk 2014: 52, Fig. 14.

In 2003, a diagonal grid of inlayed yellow metal wire was discovered on both sides of the pommel. The wire is about 1 mm thick. Such a decoration is very rare and the closest analogies – two swords from Gotland (SHM 16905, GF C 4778) – belong to the Mannheim sword type (special type 2), with not less than 20 examples dating from the second half of the 8th century to the beginning of the 9th century (Kainov 2011: 148).

What is more, the examination discovered the fact that pits situated on the central part of the pommel, upper and lower guards are not oval nor round, but square. To my knowledge, no other sword shows this type of pit decoration. These pits are arranged in a checkerboard pattern, sometimes quite uneven. Corners of pits are connected with grooves, which were probably empty and were punched after applying inlayed stripes from yellow metal. Inlayed stripes always occur in paires or threesomes between pits; they are uneven, with spaces ranging from 0.2 to 1.5 mm.

The upper guard and the pommel were separated with a helix from twisted wires of yellow metal. By the same method, the central part of the pommel was separated from side parts. The helix is stamped in order to form pearl-like balls (so called beaded wire). This method is rare on Viking Age swords, with only several known examples from Norway (C8598 – type E, B6685a – type H), Sweden (SHM 34000:942 – special type, SHM 34000:850 – type H/I), Denmark (C3118 – special type 1), Ireland (WK-5 – type K, WK-33 – type D) or France (JPO 2249 – type H). Ends of wire helix is hidden under the pommel.

Details of the hilt of the sword. Taken from Kainov 2011: Fig. 2-9.


Geibig’s typology of blades. Taken from Geibig 1991: 84, Abb. 22.

The blade belongs to the Geibig’s type 3, which is dated to period between 750 and 975 AD and is characterised by gently tapering blade with tapering fuller, blade lenght between 74 and 85 cm and blade width between 5.2 and 5.7 cm (Geibig 1991: 86, 154; Jones 2002: 22-23). On one side of the blade, there is an unique Latin inscription +LVNVECIT+, on the other side can be found the sign IᛞI (horizontally situated hourglass with two vertical bars before and after). These inscriptions are made by welding of simple iron rod on the surface. The method of welded inscriptions can be attested on dozens of European swords; the raw material varied from iron and steel rods to pattern welded material (see Moilanen 2006).

The most common welded Latin names on blades are Vlfberht, Ingelrii and Hiltipreht, while the less known are Atalbald, Banto, Benno, (C)erolt, Gecelin, HartolfrInno, (L)eofri(c), LeutlritNisoPulfbrii or Ulen. These names probably denote makers or workshops, since some names have the addition (me)fecit, “made (me)”. Among others, magical formulas occur sometimes (their shortcuts respectively), like SOOSO (“S[ALVATOR] O[MNIPOTENS] O[MNIPOTENS] S[ALVATOR] O[MNIPOTENS]) or INIOINI (I[N] N[OMINE] I[ESU] O[MNIPOTENS] …). As the result, the inscription +LVNVECIT+ (“Lun made”) denote the unknown maker Lun and the sign IᛞI is probably the shortcut for the formula In nomine Iesu (“In the name of Jesus”).


The inscription on the blade. Taken from Kainov 2011: 151, Fig. 4.

Regarding the dating of the sword, it is very complicated to date an untypical object like this one. Besides some exceptions, Scandinavian swords of the type E are dated to the 9th century, while Russian examples are dated to the 10th century (Kainov 2011: 149). So, the shape of the hilt can be dated to the 9th or 10th century. The diagonal inlayed grid on the sides of the pommel has analogies in the 8th and 9th century. The beaded wire was used in the same period, in the 8th and 9th century. The shape of the blade can be dated to the period between 750 and 975 AD. Mentioned Latin names were used from the 9th to 11th century. It seems logic to think that the sword from Sarskoe Gorodishche belongs to the transitional type between the Mannheim type (special type 2) and the type E (Kainov 2011: 149). The sword, or at least the blade, was probably made in the 9th century on the Continent and used until the 10th century by a man with strong connections with Scandinavia.


The complete sword. Taken from Kainov 2011: 150, Fig. 1.



The sword from Sarskoe Gorodishche has been recently (winter 2015 – spring 2016) replicated by famous Belorussian swordmaker, skillful crafter and my friend Dmitry Khramtsov (aka Truin Stenja). Even though I think the sword is the best copy of the found, I hold the opinion that the Dmitry’s version needs a short comment.


The method of “container” with inner parts braided with silver wire. Taken from Kainov 2011: 24, 28, Fig. 12, 15; Arendt 1936: 314, Fig. 2.

Regarding the sizes, the sword is true copy. The weight of the sword is 1370 grams, an average weight for a type E sword. The inscription was correctly done from iron rods. The handle was made from bog oak, which seems to be a good choice, as no traces of the organic handle survived. The upper guard and the pommel are hollowed, which is characteristic for the type E. Inlayed motives on the hilt (stripes and the diagonal grid) are made from copper alloy wire in the right manner. What is striking on this copy is the usage of silver wire grid in pits and grooves. This decoration is not known from any sword find and it seems like misunderstanding of a rare method used on several swords of types E and T from Sweden (Gräfsta [SHM 19464:6]; Birka grave 524 [SHM 34000:524]), Russia (Gnezdovo mound L-13; Ust-Ribezgno mound XIX and a sword deposited in Kazan museum) and Ukraine (Gulbishche) (see Androshchuk 2014: 53; Arendt 1936; Kainov 2012: 19-25). The method is described by Arendt (1936: 314):

“Both guard and the pommel form a kind of containers or coverings, which contain smaller but equally shaped parts. These latter [inner parts] were braided with silver wires and placed in the way that their crossings were just under the pits in containers.”

It seems that Dmitry based his version on some pictures of destroyed hilts, where the wire jutted out through damaged pits to the surface. However, I still think that Dmitry’s copy is the best version of the sword ever made and that Dmitry took the chance to fill rather illogical (and pattern destroying) grooves with more decoration. We should understand the version as a combination of outstanding replica and a free interpretation of the author.

If you wish to write to the author, please, use this email adress:


This article would never existed without the spectacular work of Dmitry Khramtsov, who inspired me and kindly send me photos in original resolution. All my thanks and respect also go to Sergey Kainov, who helped me with his best advices and answered all my bothering questions.


Androshchuk 2014 = Androshchuk, F. (2014). Viking swords : swords and social aspects of weaponry in Viking Age societies. Stockholm.

Arendt 1936 = Arendt, W. W. (1936). Ett svärdsfäste från vikingatiden. In: Fornvännen 31, pp. 313-315. Online.

Geibig 1991 = Geibig, A. (1991). Beiträge zur morphologischen Entwicklung des Schwertes im Mittelalter : eine Analyse des Fundmaterials vom ausgehenden 8. bis zum 12. Jahrhundert aus Sammlungen der Bundesrepublik Deutschland. Neumünster.

Jones 2002 = Jones, L. A. (2002). Overview of Hilt and Blade Classifications. In. Oakeshott E. – Peirce I. G. Swords of the Viking Age, pp. 15-24.

Kainov 2000 = Kainov, S. Yu. (2000). Меч с Сарского городища. / Сообщения Ростовского музея. Вып.Х. pp. 252-256. Online.

Kainov 2001 = Kainov, S. Yu. (2001). Еще раз о датировке гнёздовского кургана с мечом из раскопок М.Ф.Кусцинского (К вопросу о нижней дате Гнёздовского могильника) // Гнёздово. 125 лет исследования памятника. Труды Государственного Исторического музея. Вып. 124, pp. 54-63. Online.

Kainov 2011 = Kainov, S. Yu. (2011). Новые данные о мече с Сарского городища // Военная археология. Вып.2. Сборник материалов Проблемного Совета “Военная археология” при Государственном Историческом музее, pp. 147-152. Online.

Kainov 2012 = Kainov, S. Yu. (2012). Swords from Gnёzdovo. In: Acta Militaria Mediaevalia VIII, pp. 7-68. Online.

Kirpichnikov 1966 = Kirpichnikov, A. N. (1966). Древнерусское оружие: Вып. 1. Мечи и сабли IX– XIII вв.// АН СССР, Москва.

Kirpichnikov 1992 = Kirpichnikov, A. N. (1992). Новообнаруженные клейма раннесредневековых мечей // Fasciculi Archaeologiae Historicae. Fasc. V, pp. 61-81.

Leontiev 1996 = Leontiev А. Е. (1996). Археология мери. К предыстории Северо-Восточной Руси // Археология эпохи великого переселения народов и раннего средневековья. Выпуск 4, Москва.

Moilanen 2009 = Moilanen, M. (2009). On the manufacture of iron inlays on sword blades: an experimental study. In: Fennoscandia archaeologica XXVI: pp. 23-38. Online.

Petersen 1919 = Petersen, J. (1919). De Norske Vikingesverd: En Typologisk-Kronologisk Studie Over Vikingetidens Vaaben. Kristiania.