Petersen type W sword

In the next article about swords, we would like to introduce the not yet well-known Petersen type W. As far as we know, this type of sword is rarely reconstructed among reenactors. In the following article we will describe, map and reveal this interesting type.


Type W denotes a sword whose hilt consists of one-piece pommel and cross-guard cast from copper alloy. From the frontal look, the cross-guard is straight, with slightly rounded shorter sides. The one-piece pommel has a simple, semicircular shape with rounded edges and peened tang on the upper side of the pommel. When viewed from above, the shape of both components is lenticular, i.e. tapering towards the edges. Both copper-alloy components are hollow shells (see X-ray). All known components are characteristically decorated. The pommel is decorated by a series of lines that divide the pommel into four fields, which are usually filled with zigzag pattern. This pattern is well visible on the cross-guard as well. In two cases, cross-guards are decorated with concentric rings (Nedošivina 1991: 166), in one case the components are decorated by massive Borre-style knot decoration (Brøgger 1921: Fig. 10). The decoration appears to be cast, although the lines (both dividing and those around the edge) might have been additionally highlighted. Some pieces have a shiny gold finish. The division of the pommel follows the type E and stands very close to the types U, V and X. Type W can be dated to the 10th century. Norwegian pieces date back to the first half of the 10th century (Petersen 1919: 157). Other swords with copper alloy components date back to this period, especially the Petersen type O, which were influenced by the same trend. In Eastern Europe, type W can be found in graves dating to the second half of the 10th century (Nedošivina 1991: 166).

typ_WDetail of copper alloy W type sword components. Found in 1816 at an unknown Norwegian site (B998). Author:
Svein Skare, Unimus.

The W type hilt is always a part of the double-edged sword. Swords of this type have relatively uniform dimensions. The complete swords are 878-930 mm long, with the blade always around 745-760 mm long. The blades are 50–60 mm wide and are embedded in prepared grooves on the undersides of the guards. The blades are usually without inscriptions, although the swords from Timerevo (grave 100) has a clearly readable Latin letter C on the blade (Nedošivina 1991: 166). The guards are 80–100 mm long, 12–18 mm high and 17–22 mm thick. The preserved pommels have a length of 58–67 mm, a height of 34–37 mm and a thickness of 19–21 mm (Androščuk 2014: 79–80 and self-observation). The handles are 85–105 mm long, which corresponds to the average width of the palms and testifies to the customized production. We were only able to find four pieces with preserved handles; in three cases, the tang forming the handle is covered with wooden scales, which in the case of the sword from Breivold (T3107) is additionally coated with canvas and wound iron wire. The fourth case, the sword from Klepp (S2453), has an antler handle. As far as sheaths are concerned, they can be assumed to have taken on standard forms. The Klepp sword has a preserved wooden scabbard with a leather cover, while the Timerevo sword (grave 100) is covered with fragments of a wooden scabbard (Nedošivina 1991: Рис I.I; see the picture here). At the end of the sword from Šestovica, a copper alloy chape has been preserved.

Detail of a preserved sword hilt from Breivold (T3107).
Author: Ole Bjørn Pedersen, Unimus.

Sword components of the sword from Hundstad, Norway (C16699). Guard length 90 mm, pommel length 66 mm. Both components are decorated with Borre-style patterns. Brøgger 1921: Fig. 10  

To illustrate the anatomy of this type of sword, we will show four well-preserved examples:

typ_W1Bikavėnai, Lithuania. Overall length 930 mm, blade width 50 mm, handle length 105 mm, guard length 85 mm, guard height 18 mm. Wooden pieces of the handle. Photo and description: Kazakevičius 1996: 64–67.

typ_W2Östveda, Sweden (SHM 25370). Overall length 878 mm, blade length 743 mm, blade width 50–31 mm, guard length 100 mm, guard height 14 mm, guard thickness 22 mm, pommel length 60 mm, pommel height 37 mm, pommel thickness 20 mm, handle length 86 mm, handle width 20–26 mm, total weight 892 g. Photo and description: Androščuk 2014: 79, 337–338, Fig. 35.

typ_W3Šestovica, grave 42, Ukraine. Total length 890 mm, total length of hilt 145 mm, blade width 60 mm, guard length 85 mm, guard height 17 mm, pommel length 60 mm, pommel height 35 mm. The tip of the sword is covered with a chape. Photo & description: Androščuk – Zocenko 2012: 212, Fig. 151; Jana Korkodim, Wojtek Szanek.

kleppKlepp (S2453), Norway. Overall length 899 mm, total length of hilt 139 mm, blade length 760 mm, blade width 58 mm, guard length 100 mm, guard height 12 mm, guard thickness 21 mm, pommel length 59 mm, head height 36 mm, head thickness 21 mm, handle length 91 mm. By: Unimus.


Generally speaking, the type W does not have too much representation among European swords – we currently register 18 pieces. However, distribution is interesting and deserves attention. Not surprisingly, we know the highest number from Norway. While Petersen knew 8 Norwegian W type swords (Petersen 1919: 156), Hernæs already knew nine of them and this number still seems to be current (Hernæs 1985). Only four of them have partially preserved blades, the rest being hilt components. In Sweden, we know one sword and two hilt components (Androščuk 2014: 79). We also know two components – one pommel and one guard – from the Schleswig region, Germany (Geibig 1991: Tab. 164: 4-5). There are two representatives from Timerevo, Russia (graves 100 and 287), where we find one complete sword and one fragment of the pommel (Nedošivina 1991: 166–167, Рис. I.I). We know one complete sword from the Lithuanian site Bikavėnai (Kazakevičius 1996: 64–67) and one sword from the Ukrainian Šestovica (Androščuk – Zocenko 2012: 212).

Several hybrid pieces stand very close to the W type swords and they do not fit to standard typology. First and foremost, a fragment of a two-piece pommel cast from copper alloy, found  during a settlement excavation in Pohansko, Czech Republic, in 2015, has to be mentioned (Košta et al. 2019: 215-6, Fig. 57-8). This piece is decorated with typical W type zigzag decoration. Another hybrid piece is the sword from Latvian Bēnes Kaijukrogs, which uses H/I type shape, but the components are cast in copper alloy and decorated with a zigzag pattern typical of type W (Jērums 2012: 2. att). The third hybrid piece is a detector finding of a one-piece pommel from Ukraine, which corresponds to a typical pommel of type W, but instead of a typical decor, it is decorated with a pit decoration typical of type E (personal discussion with Sergei Kainov). Three Latvian pieces that were mapped by Artūrs Tomsons (2019: 70) could be labeled as close to W type; their shape corresponds to type V pommels, but they are cast in bronze and decorated with patterns typical for R-S types. Significantly, hybrid pieces combine elements of types that stand very close each other. The total does not include finds from the United Kingdom (Jakobsson 1992: 213; Żabiński 2007: 65), since all these swords have iron components and therefore do not meet the basic W type criterion.

The hybrid piece from Pohansko, Czech Republic (Košta et al. 2019: Obr. 57-8).

The hybrid piece from Bēnes Kaijukrogs, Latvia (Jērums 2012: 2. att).

The hybrid piece from Ukraine (source: Sergei Kainov).

Of the total of 18 pieces, 8 are swords or fragments thereof, while the remaining 10 are separate copper alloy components. W type swords have been found in 7 countries, so they are relatively scattered compared to total numbers. The main distribution area is Northern and Eastern Europe, where swords are located in important sites.

typ_W4W type sword distribution, according to Jakobsson (1992: 228).


In this chapter, we would like to present five successful reconstructions of the Petersen type W made by various European swordsmiths.

Blade width 68–40 mm. Weight 1540 grams, balanced 170 mm from the guard. Producer: Tomáš Zela, 2017.

Reconstruction of the sword from Šestovica 42. Weight 1200 grams.
Producer: Dmitry Chramcov, 2015.

Reconstruction of the sword from Šestovica 42 compared to the original.
Producer: Wojtek Szanek, 2016.

bobrProducers: Petr Floriánek, Radek Lukůvka, 2018.

Producer: Arma Epona.


This work exists thanks to the initiative of Tomáš Břenek from the group Goryničové, who owns the reconstruction made by Tomáš Zela. Since this type has not yet been seen on the Czech battlefields, there was a need to point out the finds and their distribution. We would like to thank every persistent enthusiast who was not discouraged by waiting. Sergei Kainov, who pointed to two Russian findings, and Ferenc Tavasz, who helped me with his advice, also have their merit in the article.

Here we will finish this article. Thank you for your time and we look forward to any feedback. If you want to learn more and support my work, please, fund my project on Patreon or Paypal.


Androščuk, Fedir (2014). Viking Swords : Swords and Social aspects of Weaponry in Viking Age Societies, Stockholm.

Androščuk, Fedir – Zocenko, Vladimir = Андрощук Ф. O. – Зоценко В. (2012). Скандинавские древности Южной Руси: каталог, Paris.

Brøgger, A. W. (1921). Rolvsøyætten : et arkeologisk bidrag til vikingetidens historie. In: Bergens Museums Aarbok 1920–21, Hist.antikv. Række 1, 3–42.

Geibig, Alfred (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.

Hernæs, Per (1985). De østnorske sverdfunn fra yngre jernalder : en geografisk analyse. Magistergradsavhandling i nordisk arkeologi – Universitetet i Oslo, Oslo.

Jakobsson, Mikael (1992). Krigarideologi och vikingatida svärdstypologi, Stockholm : Stockholms Universitet.

Jērums, Normunds (2012). Divasmeņu zobeni Zemgaļu apdzīvotajās teritorijās (5.–14. gs.). In: Arheoloģija un etnogrāfija XXVI, Rīga, 74-104.

Kazakevičius, Vytautas (1996). IX–XIII a. baltų kalavijai, Vilnius.

Košta, Jiří et al. (2019). Velkomoravské meče z Pohanska u Břeclavia okolí – nová revize. In: Památky archeologické CX, 173-235.

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

Nedošivina N. G. = Недошивина Н. Г. (1991). Предметы вооружения, снаряжение всадника и верхового коня тимеревского могильника // Материалы по средневековой археологии Северо-Восточной Руси, Москва: 165–181.

Tomsons, Artūrs (2019). Zobeni Latvijas teritorijā no 7. līdz 16. gadsimtam, Rīga.

Żabiński, Grzegorz (2007). Viking Age Swords from Scotland. In: Studia i Materiały – Studies and Materials, Acta Militaria Mediaevalia III., Kraków – Sanok: 29–84.

The Wallet from Iholm, Denmark

In March 2020, I was notified of an interesting find of a wallet from 11th century Denmark, which has not been paid much attention. This brief article will provide basic information on the entire find and the reconstruction attempts.

Circumstances of the find and its content

On a Sunday in 1853, an unnamed brick factory worker from Tåsinge, Denmark, went to uninhabited island of Iholm, which lies in the Svendborg Strait between the islands of Funen and Tåsinge, accompanied by his friends. As he was strolling and destroying molehills, he saw metal reflections in one of them. Inside, he found 15 coins that he had buried again due to fear of disaster and illness, he washed himself and went home. The man shared the information about the discovery with the master brickman, who went to the island with its owner. The find-place, located in the middle of the small island, was thoroughly examined within a radius of about 4 meters and a depth of 1 meter. At a depth of 20-25 cm below the surface, they found a silver treasure that was kept in a leather case, and other silver objects were discovered within a 30 cm radius. The treasure was collected and handed over to the National Museum in Copenhagen (Grundtvig 1948: 170; Skovmand 1942: 90, Cat. No. 32).

The main part of the treasure consisted of 475 coins, more precisely 3 Danish, 1 Norwegian, 17 Swedish, 238 Anglo-Saxon, 10 Dutch, 163 German, 2 Carolingian, 8 Czech, 1 Byzantine, 1 Persian, 27 Kufic and 2 semi-finished coins (Erslev 1875: 119- 120; Hauberg 1900: 165, Cat No. 45; Malmer 1966: 269, Cat No. 43). Duczko informs that the treasure includes two shield pendants with whirl motifs (Duczko 1989: 18). In addition, three fragments of necklaces were found in the treasure (Hårdh 1996: 48, 191), four complete bracelets and ingots (Nationalmuseet 2020; Trap 1923: 706). In 1989, around 100 coins and silver fragments were discovered near the site, so the total number was about 590 pieces of silver (Nationalmuseet 2020). The dating of the treasure is the first quarter of the 11th century. The catalog number of the treasure, which is partially exhibited in the National Museum in Copenhagen (room 23), is C NM 13594-608, C. 1837. The find-place is sometimes also referred as Yholm, Bregninge, Svendborg, Svendborgsund, Tåsinge or Taasinge.

The available literature has always descibed the numismatic part of the treasure (among others Brøndsted 1938: 382; Galster 1980: 65; Rasmusson 1937: 125-6; Schive 1865: 13; Wahlstedt 1930: 23, 28), while the leather fragments remain almost unnoticed by literature. The next chapter will therefore be devoted to the description of leather fragments.

Wallet remains and reconstruction

But not all the silver that glitters! The wallet in which the treasure was located was no less valuable and was a representative item. In the present state, it consists of two leather fragments, one of which is part of the wallet pocket and the other is a sewn application that has been gold-plated (Nationalmuseet 2020; Mannering 2017):

  • fragment 1: a piece of leather with approximate size of 6 × 4 cm, which forms the tip of a folding wallet. The side exhibited in the museum as the upper side is the upper side of the inner pocket. The two sides forming the tip are lined with holes for stitches. Apperently, the fragment of the pocket has been sewn to the second, supporting layer. The top of the fragment and the tip are torn apart. The dominating part of the fragment is a semicircle of holes, which was used to find the originally circular application. This application was positioned at the center of the width of the object, with the offset from the sides being smaller than the offset from the tip. Whether the wallet had one or two pockets facing each other is not known, but both variants are possible. The wallet in its original state probably exceeded the width of 6 cm, while the original length is unknown, but due to hundreds of pieces of silver it could be quite large. The closest contextual and shape analogy is the wallet from Roswinkel, Netherlands, dated to the end of the 9th century (Pleyte 1883; Gräslund 1984: Abb. 16.2). This folding wallet consists of a supporting layer and a three-part pocket embellished with a sewn leather application; it was used to store the treasure – 144 silver coins and a gold coin in a small wooden box (personal discussion with Bert Tessens). Other examples of folding wallets are 24 wallets from Birka (Gräslund 1984: 143-6), Sigtuna wallet (Sigtuna Museum 2019), a small four-piece wallet from Bringsverd, Norway (C23116; Rolfsen 1981: 117) and a small wallet from the Evebø grave, Norway, 5th century (B4590). The presence of a tip at Iholm wallet indicates that the find did not belong to the group of two-piece wallets with integral leather slider, such as those from Elisenhof (Grenander-Nyberg 1985: 234, 247, Taf. 76) and Gniezno (self-observation). With a high degree of probability, we can also exclude it would belong to the group of bags with metal components.


  • fragment 2: leather application originally of circular shape. The diameter of this application could be about 3-4 cm. The application consisted of an interwoven motif with a rim, with holes for stitches at the edge of the rim. The interwoven motif was apparently made up of two pieces that were perpendicular to each other, one piece consisting of a rim and strips connecting its two sides, while the other had loose ends, which were interwoven between the strips of the furst piece and inserted under the rim and stitched together with the rim. At this time, exact reconstruction of the application is not possible and it is necessary to wait for detailed analysis and publishing in print. However, similar motifs can be found on pendants and textile applications in Viking Sweden. The leather application was gilded with foil, which is still visible today. The closest analogy to this decorative method can be found in Birka, where all parts of folding wallets are interwoven with gilded leather strips (Gräslund 1984: 143-6). A similar find to those from Birka is a leather pouch lid from Frankish grave 10 in Cologne-Müngersdorf, which is interwoven with copper alloy wires (Fremensdorf 1955: 93, 137, Taf. 92.1-2). Gilded leather can also be found on the wooden knife sheath from warrior grave excavated at Prague Castle (Borkovský 1939-46: 127). Notker the Stammerer mentions that Charlemagne wore gilded leather shoes (De Carolo Magno, translated by Thorpe, p. 132).

Source: Mannering 2017.

Source: Nationalmuseet 2020.

Source: Fashioning the Viking Age 2019.

Wallet from Roswinkel, which is the closest analogy. Source: Gräslund 1984: Abb. 16.2.

As far as we know, two attempts have been made to reconstruct the waller which should be mentioned. The first of these was created in the Danish workshop Nichols Naturligvis. The overall look is great and the only details we can criticize is the size of the application, which covers too big space compared to the original, and the shape of the lower edge, which should be more spiked and probably without a strap. Generally, this attempt copies models from Birka. The workshop is very active in experimenting with the possible looks of the original interwoven motif, which was symmetrical, in their opinion.

Attempted reconstruction by Nichols Naturligvis.

The other attempt was made by Swedish reenactor Veronica Wik, who mounted the asymmetric application on a purse. The benefit of this reconstruction is the fact it reflects the larger capacity of the wallet, which should have been able to hold several hundred pieces of silver, as well as a greater offset of the application from the edge, which is more consistent with the original find. We must also appreciate the involvement of coins and hence the pursuit of a realistic concept.

Reconstruction attempt by Veronica Wik.

Since both versions are not ideal, me and reenactor and graphic designer Tomáš Cajthaml prepared two graphic designs that outlines the original appearance of the artifact in the best possible way – the wallet is folding, has only one strap, the application has the correct ratio to the rest and the offset respects the original composition. We used the shape of Roswinkel wallet, which we consider the closest shape analogy. The look of the application was taken from the attempt of Veronika Wic, although we are aware that none of the designs is 100% accurate.











Suggested drawn reconstructions of Iholm wallet.
Made by Tomáš Cajthaml.

Acknowledgments and conclusion

Wallet from Iholm is a rare specimen that complements the mosaic of purses, bags and wallets known from the Viking Age. In terms of decorating, it ranks among the top finds. It suggests that gilded leather was a more widespread phenomenon than previously thought. It is also probably the first wallet known from Viking Age Denmark, which will be appreciated especially by reenactors interested in the region who now have the opportunity to take this artifact into consideration. All this should serve as an appeal to the staff of the National Museum in Copenhagen, pointing out that the wallet has not yet been published.

Finally, I would like to thank Nichols Naturligvis for drawing my attention to this find. My thanks also deserve Veronica Wik. In the last, most honorable place, I would like to pay tribute to Tomáš Cajthaml, who quickly and unselfishly created great graphics, thanks to which this artifact can be appreciated by people from all over the world.

Here we will finish this article. Thank you for your time and we look forward to any feedback. If you want to learn more and support my work, please, fund my project on Patreon or Paypal.


Notker the Stammerer: De Carolo Magno. In: Two lives of Charlemagne: Einhard and Notker the Stammerer, translated by Lewis Thorpe, Harmondsworth 1967.

Borkovský, Ivan (1939-46). Hrob bojovníka z doby knížecí na Pražském hradě. In: Památky archeologické 42, 122-131.

Brøndsted, Johannes (1938). Danmarks oldtid, bind 3, København.

Duczko, Władysław (1989). Runde Silberblechanhänger mit punzierten Muster. In: Arwidsson, Greta (ed.). Birka II:3. Systematische Analysen der Gräberfunde, Stockholm, 8–18.

Erslev, Kristian (1875). Roskildes ældste Mønter. Studier til Dansk Mønthistorie. In: Aarbøger for nordisk Oldkyndighed og Historie, København, 117–187.

Fashioning the Viking Age 2019. Fragment of a leather purse from Yholm. In: Fashioning the Viking Age Project. Visited 19.3.2020, available from:

Fremersdorf, Fritz (1955). Das fränkische Reihengräberfeld Köln-Müngersdorf, Berlin.

Galster, Georg (1980). Vikingetids møntfund fra Bornholm. In: Nordisk Numismatisk Årsskrift 1977–78, 5–246.

Gräslund, Anne-Sofie (1984). Beutel und Taschen. In: Arwidsson, Greta (ed.). Birka II:1. Systematische Analysen der Gräberfunde, Stockholm, 141-154.

Grenander-Nyberg, Gertrud (1985). Die Lederfunde aus der frühgeschichtlichen Wurt Elisenhof. In: Szabo, M. – Grenander-Nyberg, G.- Myrdal, J. (eds.). Die Holzfunde aus der frühgeschichtlichen Wurt Elisenhof. Elisenhof Band 5, Frankfurt – Bern – New York.

Grundtvig, Sven (1948). Danske folkesagn, 1839-83: samling. Danske stedsagn, København.

Hårdh, Birgitta (1996). Silver in the Viking Age: A Regional-Economic Study (Acta Archaeologica Lundensia, Series in 8° Nr. 25), Stockholm.

Hauberg, Peter Christian (1900). Myntforhold og udmyntninger i Danmark indtil 1146, København.

Malmer, Brita (1966). Nordiska mynt före år 1000 (Acta Archaeologica Lundensia, Series in 8° Nr. 4), Lund.

Mannering, Ulla (2017). Skattefundet fra øen Yholm. In: Nationalmuseet – Prehistory. Visited 19.3.2020, available from:

Nationalmuseet (2020). Udsøgt læderpung med guldtryk. In: Nationalmuseet i København. Visited 19.3.2020, available from:

Pleyte, Willem (1883). Nederlandsche Oudheden van de vroegste tijden tot op Karel den Groote, Leiden.

Rasmusson, Nils Ludvig (1937). Kring de västerländska mynten i Birka. In: Från stenålder till rokoko, studier tillägnade Otto Rydbeck, Lund, 113–135.

Rolfsen, Perry (1981). Den siste hedning på Agder. In: Viking, Vol. 44, 112–128.

Sigtuna Museum (2019). Veckans föremål. In: Sigtuna Museum & Art. Visited 19.3.2020, available from:

Schive, G. I. (1865). Norges Mynter i Middelalderen, samlede og beskrevne af G. I. Schive: Med Indledning af C. A. Holmboe, Christiania.

Skovmand, Roar (1942). De danske Skattefund fra Vikingetiden og den ældste Middelalder indtil omkring 1150. In: Aarbøger for nordisk oldkyndighed og historie, København, 1-275.

Trap, Jens Peter (1923). Kongeriget Danmark, 4. Udgave, København.

Trap, Jens Peter (1957). Kongeriget Danmark, 5. Udgave : Odense og Svendborg Amt, København.

Wahlstedt, Axel (1930). Den svenska plåtmyntningens historia. In: Numismatiska meddelanden 25, 22-36.