Generic selectors
Exact matches only
Search in title
Search in content
Post Type Selectors
Generic selectors
Exact matches only
Search in title
Search in content
Post Type Selectors

Early Medieval Sword from Maaseik, Belgium



8th-12th century swords can be considered extraordinary and poorly published finds in Belgium, and only 10 demonstrable specimens have been known to date, all of which have been collected from waterways. The text presented below represents the eleventh piece, apparently coming from the Meuse River from the site of Maaseik. The sword belongs to the so-called Altjührden type and can roughly be dated to the second half of the 8th or the beginning of the 9th century. The hilt components are decorated with an unusual tubular decoration that is known from only seven period swords from northwestern Germany, France, the Netherlands and Switzerland. A comparison of swords decorated with this form of decoration shows that they may be a precursor to swords with cross-organized pit decoration (represented mainly by Petersen’s types E and T and swords of the Curonian type), which were used throughout Europe for three centuries.


The study of early medieval swords in the territory of present-day Belgium suffers from the absence of an updated revision (see Geibig 1991 and Maczek 2016 for the most up-to-date lists). Until a thorough examination of archival material takes place, this deficiency presents an insurmountable task that will lead to the constant recycling of identical information. Knowing that they are certainly missing unpublished and unrecognized finds, the author team presents 10 convincing sword finds from the territory of Belgium that can be dated to the period of 8th-12th century. From the list, it is evident that the discoveries are strongly linked to waterways:

  • A sword found in the Scheldt River near the town of Melle in 1934 and stored in the Hasse collection in the Antwerp museum (Bjørn – Shetelig 1940: 124; Hasse 1935: 203; Müller-Wille 1978: Cat. no. 37). In addition to the textual description, we also have a very poor quality photograph (Hasse 1935: Fig. II) and a drawing (Hasse 1939: 103). It is obvious that this is a piece close to Petersen’s special type 8, i.e. a sword with a straight guard and a one-piece and slightly pointed pommel, which resembles a blending of pommels of types K and X (Petersen 1919: 111). Arbman assigned the sword to type X (Arbman 1937: 228), Maczek and Müller-Wille to type K (Maczek 2016: Cat. no. 3; Müller-Wille 1978: Cat. no. 37). It belongs to Geibig’s combination type 6 (Geibig 1991: Cat. no. 815). The dating can point to the 9th or early 10th century.

  • A sword found in the Scheldt River near the town of Temse and later deposited in the Raton collection in Paris (Geibig 1991: Cat. no. 1402; Hasse 1935: 203-4). According to the only available drawing, the hilt was equipped with a Brazil-nut pommel (Oakeshott type A) and a straight guard (Oakeshott type 1). In Geibig’s classification, the sword corresponds to combination type 16-I (Geibig 1991: 70-72). Analogical swords can be dated to the period from the second half of the 10th to the 12th century (Hošek et al. 2019: 255-6); Tackenberg dates the sword to the 11th-12th century (Tackenberg 1960: 19). On the blade, there was an equilateral cross and a variant of the inscription ULFBERHT (Müller-Wille 1970: Cat. no. 5; Stalsberg 2008: Cat. B2).

  • A sword found in the Scheldt River in the town of Dendermonde in 1898 and today stored in the Royal Army Museum in Brussels under inv. no. 12420 (Bjørn – Shetelig 1940: 124; De Prelle de la Nieppe 1907; Geibig 1991: Cat. no. 456; Hasse 1935: 203, Fig. VI; Müller-Wille 1978: Cat. no. 37). The sword corresponds to Petersen’s special type 2, which dates the sword to the late 8th and 1st half of the 9th century (Petersen 1919: 86). An almost identical and well-analysed parallel was found in grave no. 10 in Šlapanice near Brno, in a context dated to the 9th – the beginning of the 10th century (Hošek et al. 2019: 250-2). Other close parallels come from the grounds of the Palace of Westminster in London (Dunning – Evison 1961) and from grave 30 from site of Kvarnbacken, Åland (Kivikoski 1963: Taf. 9).

  • In 1884, a sword was found in the Scheldt River in the town of Heusden (today’s Destelbergen), which was subsequently stored in Ghent (Bjørn – Shetelig 1940: 124; Müller-Wille 1978: Cat. no. 36). The find is a textbook example of Petersen’s type L (Petersen 1919: 112-6), specifically Aksdal’s variant I (Aksdal 2017: Cat. no. 51). In Geibig’s classification, the sword corresponds to combination type 7 (Geibig 1991: Cat. no. 876). It can be assumed to be dated to the 9th, at the latest to the middle of the 10th century. Another L type sword was found in Wessem, Netherlands, near the Belgian border (Willems – Ypey 1985).

  • In 1920, a sword with straight, medium-length guards and a pyramidal pommel cap, divided into three lobes by means of vertical channels, was found in the Scheldt River near Rupelmonde. Typologically, the sword is close to Petersen’s special type 2 and can roughly be dated to the end of the 8th and the first half of the 9th century. It has a double-edged blade and a total length of 92 cm. The sword is known from a single source and a single photograph. In the 1960s it was in the private collection Lemmens (Ghellinck 1968: 28).

  • In the 1960, there was a well-preserved 96 cm long Petersen’s type X sword with a massive double-edged blade in the private collection T. H. M. Jageneau – de Lange (Ghellinck 1968: 28). The discovery was supposed to have taken place in the “dragon ship” near Appels, which is information that can no longer be verified. It cannot be ruled out that it was found in the Scheldt River. The dating of the sword points to the 9th-10th century.

  • In 1951/2, a sword was found during the dredging of the Meuse River near the island of Monsin in Liège, which is kept in the Museum of Ancient Arquebusiers in Visé. Inv. no. 108.bmp 901K. To the authors’ knowledge, the sword remains academically unpublished. It is in very good condition and the hilt components consist of straight guard with angular edges and a pyramidal head. It is very likely a sword of the Altjührden type, which makes it possible to date it to the end of the 8th or the beginning of the 9th century (Menghin 1980: 257). We thank Eddy Bruyère for the information provided.

  • In the Viking Museum in Hamme, there are a total of three early medieval swords bearing the inv. nos. VIK 169, VIK 170 and VIK 171 (Bogaert 1962; De Baere 2019). All of them were discovered in 1906 during the strengthening of a flood-damaged dam at the confluence of the Durme and Scheldt Rivers (the localities of Wijde Durme and the now non-existent Luizenbos island). All swords are double-edged. The first sword (VIK 169) corresponds to Petersen’s type K and is characterized by a relatively long crossgurd, low guards and a pommel cap with a massive central lobe. The find has shape parallels in the corpus of type K swords, for example the sword from Ludwigshafen am Rhein, Germany (see Bilogrivić 2009). The dating goes to the 9th or the beginning of the 10th century. The second sword (VIK 170) corresponds to Geibig’s combination type 12-I and is equipped with a curved crossguard and a Brazil-nut-shaped pommel. The blade is signed with an equilateral cross. For that reason, a dating to the second half of the 10th – 11th century can be suggested. The third sword (VIK 171) is a nice representative of Petersen’s type M, and its hilt is made up of straight guards without a pommel cap. Based on analogies, the dating of this piece can be determined between the middle of the 9th and the middle of the 10th century (see Vlasatý 2019).

Map 1: distribution of convincing finds of 8th-12th century swords from Belgium.

In addition to these pieces, there are other swords in the literature that do not come from the territory of Belgium, are difficult to verify or can be refuted. Some of the more significant specimens were selected:

  • In the Viking Museum in Hamme, there are also two other swords that are traditionally assigned to the Viking Age and which bear inv. nos. VIK 189 and VIK 201 (De Baere 2019). The sword VIK 189 has a disc-shaped pommel that can be dated to the 13th century at the earliest, while the sword VIK 201 has a hilt typical of the 6th-7th century.

  • There is one unpublished sword in the Grand Curtius Museum in Liège, which comes from the private collection of Malengret-Lebrun. Inv. no. GC.ARM.12a.1975.42706. It is a Petersen’s type H/I sword missing the pommel cap. The item is originally said to be from the British Isles, not Belgium. We are grateful to the Grand Curtius Museum in Liège for this information.

  • A sword from an unlocated site from the Scheldt River (Geibig 1991: Cat. no. 1043; Oakeshott 1999: Fig. 26). However, it does not necessarily come from Belgium, as stated by Košta (Košta 2021: Cat. no. 3502). If it is a real sword, it can stand close to the so-called Wallingford group (Evison 1967), i.e. the Anglo-Scandinavian variant of Z type swords (Androshchuk 2014: 84-6).

  • Petersen’s type X sword from the collection of Lambert of Oudenaarde, auctioned to Germany in 1926 (Geibig 1991: Cat. no. 916; Müller – Kölling 1990: 161, 361). Despite attempts to localize it to Belgium, the origin is unclear.

  • A sword with the inscription ULFBERHT, a Brazil-nut-shaped pommel and a curved crossguard, allegedly found in the Scheldt River, later deposited in the Christensen collection in Copenhagen (Bruhn Hoffmeyer 1954: 7, Pl. IVe). Corresponds to Geibig’s combination type 12-I. Despite attempts to locate the country of origin as Belgium (see Geibig 1991: Cat. no. 1115; Maczek 2016: Cat. no. 6; Stalsberg 2008: Cat. B1), the place of origin cannot be clearly established. According to Geibig, Matzek and Müller-Wille, the sword was discovered in the Meuse River (Geibig 1991: Cat. no. 1115; Maczek 2016: Cat. no. 6; Müller-Wille 1970: Cat. no. 5a).

  • Swords found in the Scheldt River in Wichelen (Rahir 1928: 188; Matzek 2016: Cat. no. 5). One of them allegedly came from the “Frankish era”. We lack any further details and pictorial documentation.

  • Two more swords were supposed to be found in the Scheldt River in Dendermonde, whose visualization is absent (Hasse 1935: 204; Maczek 2016: Cat. no. 1). One of the swords was supposed to be equipped with a triangular pommel with a copper alloy inlay, which sounds like a plausible description of a sword close to Petersen’s type H/I or special types 1 and 2. This is not contradicted by other parameters from the textual description. Similar swords are known from the neighbouring Netherlands (Maczek 2021).

  • A piece found in the Meuse River in Statte-Huy and kept in the Pauliac collection in Army Museum, Paris is included among the swords of the Viking Age in the literature (Hasse 1935: 204). However, it can be seen from the attached schematic drawing that it is a Behmer type VI sword, which can be approximately dated to the 6th-7th century (Behmer 1939).

  • The dagger and the pommel from Mechelen, which Hasse first believed to be from the Viking Age (Hasse 1935: 205), were later re-evaluated by the author himself as finds from other periods (Hasse 1939: 102).

  • Aksdal (Aksdal 2017: Cat. no. 52) names an L type sword from Dendermonde, which, however, does not exist and was created by a research error.

Fig. 1: an example of a sword from Belgium, the find from Dendermonde.
Source: Royal Army Museum in Brussels.

From the list presented, it follows that a convincing early medieval sword is a relatively rare phenomenon in the area of Belgium. For comparison, we can mention approximately 8th-12th century 60 swords that are known from the neighbouring Netherlands (Vlasatý 2020a). It is therefore our kind duty to present another, the eleventh sword find, which was allegedly found in the Meuse River near the town of Maaseik and which remains in a private collection. The aim of this text is to present the find to the academic and lay international public and to put it into context with the help of the latest literature.

Map 2: position of Maaseik on the map of Europe.

Description of the object

On January 22, 2005, the collector Guy Gramme of Amay, Belgium, offered Mr. Eugène Thirion a sword that was said to have been dredged from the bed of the Meuse River in Maaseik, Belgium. We do not know any more detailed information about the place or time of discovery. Mr. Thirion documented the sword by photographing it, measuring it and making a drawing of it, after which he handed the sword over to the collector Henri Lehance from the village of Ombret-Amay. In order to display the sword, Mr. Lehance attempted an interim restoration, which did not stop the corrosion of the artifact. He acquired an approximate reproduction and entrusted the sword to another collector, whose name is unknown to the author team. After the death of Mr. Eugène Thirion, co-author Gianni Gava contacted the heirs and obtained a folder of eight photographs and three drawings, which we would like to publish in this article with the consent of the family members. All provided photos can be downloaded by clicking on the following link:

The total length of the sword in its preserved condition is 90.2 cm, of which 72.4 cm is the blade and 17.8 cm is the hilt. The total weight of the sword in its preserved condition is 1.2 kg. The blade, which has a varying preserved width of 4.8-5.4 cm at the guard, gradually tapers and at a distance of 8 cm from the end, the preserved width of the blade is 4 cm. The center of the blade is filled with a fuller that extends from the guard to a distance of approximately 10 cm from the tip of the blade. The fuller narrows and is filled with patternwelded panels forming a SZ pattern (see Hošek et al. 2022: Fig. 44a). The maximum thickness of the blade reaches 0.5 cm. Typologically, the blade can be assigned to Geibig type 3, i.e. a slightly tapering blade with a tapered fuller, which can be dated to the period from the late 8th to the second half of the 10th century (Geibig 1991: 86; Jones 2002: 22-3). The proportions of the blade correspond to group b according to Košta and Hošek, which can be defined as medium-long and narrow blades typical of the early and late Carolingian period (Košta – Hošek 2014: 270).

The hilt is made up of straight guards forming a crossguard and an upper guard on which the pommel cap rests. The guards are faceted and a raised rib runs through their centers. The ends of the crossguard are rounded, the ends of the upper guard are angular. The crossguard is 10.5 cm long, 2.3 cm high and the overlaps blade by 2.7 cm on each side. From the grip, only the 10.5 cm long tang has been preserved to this day, which has a width of 3.2 cm at the crossguard and 2.1 cm at the upper guard. The grip undoubtedly consisted of a multi-part organic handle with subsequent wrapping (see Vlasatý 2020b; 2021). The length of the upper guard is 9.2 cm. The composite pommel is 5 cm high and the two components are separated by a contrasting copper alloy material that looks like beaded band or twisted wire from the outside, but in reality it could just as well be a plate that runs between the two components. The pommel cap has the shape of a low and slightly rounded pyramid, and it was apparently passed through by a tang that was peened on its top. The hilt components enable a closer chronological classification. Within Petersen’s typology, the sword can be assigned to type B (Petersen 1919: 61-3), specifically to variant B2 according to Androshchuk (2014: 42-4). In the context of western continental Europe, the names “Altjührden type” and “Immenstedt type” are used in the German literature for variants of the Petersen’s type B hilts, with the sword from Maaseik more likely to correspond to the Altjührden type (Stein 1967: 78-9). In Geibig’s classification, the sword belongs to combination type 1 (Geibig 1991: 25-31). Grave burials of B type swords are traditionally dated to the second half of the 8th century (Jørgensen 1999: 134; Menghin 1980: 257), but critical works of recent decades shows that in Scandinavia and Central Eastern Europe swords of this type were also deposited during the first half of the 9th century (Androshchuk 2014: 44-6; Hošek et al. 2019: 184-5, 205-6). In any case, a sword of this type can be considered a specimen of the early Carolingian horizon.

Fig. 2: drawing of the sword from Maaseik.

Fig. 3: details of the hilt components of the sword.

The hilt components are adorned with a very unusual form of decoration that makes the sword remarkable. Wide tubes of copper alloy were inserted into the core of the iron components, which were very likely filled with enamel or stones. On the guards, this tubular decor is organized in pairs placed in a square grid of two rows and three columns; the probable number of tubes on each side of the guards is 6. The pommel cap decoration is not easily legible due to corrosion, and if it was done identically on both sides, then there were 6 tubes on each side organized in three columns and five rows, with the central four following a non-rectangular grid.

Fig. 4: graphic reconstruction of the hilt.
Author: Diego Flores Cartes.

Decoration analogies

The sword from Maaseik fits into a regionally narrowly defined group that was used in the territory of today’s Belgium, France, the Netherlands, Switzerland and the northwestern part of Germany. The tubular decoration is shared by only seven other swords, which have not yet been understood as a single group and, despite their exceptional artistic quality, have never seen a collective publication. The swords are of different types (Altjührden, Mannheim, Mannheim-Speyer) which very likely coexisted in the 2nd half of the 8th and the beginning of the 9th century. In these swords, 0.15-0.2 cm long tubes made of a copper alloy (German: Buntmetallrundeln) of various diameters (roughly 0.2-0.95 cm outer diameter) are inserted into the iron core of the components. The tubes are applied in different grids (rectangular, cross) and are filled with enamel, stones or are full and lack filling with coloured material. Multi-coloured fills (red, green) could be applied to one sword. If coloured fillings were used, the fillings could be backed with gold foil. It can be assumed that such swords may have been accompanied by baldric sets decorated in a similar style (see Vierck 1980: 482).

Fig. 5: reconstruction of swords with tubular decoration.
From top, left: Maaseik, Bocholt-Lankern, Dorsten-Lembeck, Eggenstein-Leopoldshafen, Rheden / s’Hertogenbosch, Schortens, Souffelweyersheim, Thielle.
Author: Diego Flores Cartes. Higher resolution here.

Map 3: distribution of swords with tubular decoration.

The first specimen comes from grave 143 from the West German site of Bocholt-Lankern (Lehmann 2015: 387-8, Taf. 34–37). This sword belongs to the Altjührden type, so it makes a good parallel to the Maaseik sword. X-rays showed that the tang passed through both parts of the pommel and was peened at the top of the cap (Lehmann 2015: Abb. 126). The tubes have a diameter of 0.2 cm and are embedded about 0.15 cm inside the components (Lehmann 2015: 387-8). On the guards, the tubes are oriented in a regular grid of four rows. The number of columns on the crossguard is equal to 20, on the upper guard 18. The rows were alternately filled with green and red stone inserts. On the pommel cap, the tubes are located in five rows, 3-12 in number. Most tubes respect the edges of the individual pits, but some are made in such a way that one strip is twisted into an S or Ƨ shape and surrounds two pits at the same time. The find is now kept in the Archaeological Museum in Herne, inv. no. 1930:32, Grab 143, F 1. The grave is dated to 730-770 (Hernö 2007: 53; Lehmann 2015: 387).

Fig. 6: the sword from Bocholt-Lankern. Source: Lehmann 2015: Abb. 125, Taf. 35.

The second specimen comes from grave 133 from the West German site of Dorsten-Lembeck, belongs to the Mannheim type and is dated to 730-800 (Lehmann 2015: 393-5, Taf. 47-49). The upper guard and the pommel cap are interspersed with an oval plate that has a beaded edge that projects outward. The crossguard was provided with decorative rivets at the ends, which are now absent. The hilt components are decorated with 0.5 cm wide gilded foil with punched decoration, the remains of which have been preserved on the pommel cap. On the guards, the foils are organized into grids with seven vertical and one central horizontal band; horizontal bands are absent on the pommel cap. In the free square areas between the bands, there are pits filled with tubes. The tubes have an inner diameter of 0.15-0.2 cm, extend about 0.15 cm deep into the components and are lined with gold foils on which the almandine fillings rest. The find is now kept in the Archaeological Museum in Herne, inv. no. MKZ 4208,17c Grab 133, F 1.

Fig. 7: 3D scan of the sword from Dorsten-Lembeck.

The third German find is a hitherto poorly published sword, found in the Eggenstein-Leopoldshafen gravel site in 1982 (Geibig 1991: Cat. no. 73). The sword had long resided in a private collection and was recently purchased by the curator Astrid Wenzel of the Baden State Museum in Karlsruhe from the heirs of the original finder (Landesarchiv Baden-Württemberg 2024). The sword can be classified under the Altjührden or Immenstedt types. The upper guard and the pommel cap are interspersed with a sheet of copper alloy, which protrudes outside the pommel with a beaded edge. The guards are faceted, with each facet decorated with tubes placed in a cross grid. On the pommel cap, the tubes are organized in rows and the number of tubes decreases upwards. The nature of the tubes is currently unclear, but from the available photos it appears that solid and not hollow copper alloy tubes have been inserted into the components. The upper parts of the tubes are apparently modified in such a way that they form a motif of concentric circles. However, other possibilities cannot be ruled out. The object is now kept in the Baden State Museum in Karlsruhe, inv. no. 2017/96.

Fig. 8: the sword from the Baden State Museum in Karlsruhe.
Source: Landesarchiv Baden-Württemberg 2024.

The fourth and northernmost sword comes from the grave 410 of the site of Schortens (Rötting 1999: 238; Westphal 1999; 2002: 68-70). Typologically, this weapon can be classified as the Altjührden type, and the entire grave is dated to the period 730-770 (Westphal 1999: 70). The upper guard and the pommel cap are not interspersed with a copper alloy plate. The guards are faceted and angular at the ends, each facet being decorated with tubes placed in a cross grid. The central tubes are larger and have Y-shaped compartments that divide the ring space into thirds. The location of the tubes on the pommel cap is not exactly known, a combination of large and small tubes can be expected. Larger rings have a diameter of 0.7 cm, smaller ones have a diameter of 0.3-0.4 cm. The material of the tubes was analyzed with the result that the copper content is 64.4-83.3%, zinc content 11-17.1%, tin content 1.4-2.6% (Westphal 1999: 251). The fillings of the tubes have not been preserved. The sword is kept in the State Museum of Natural Sciences and Early History in Oldenburg.

Fig. 9: a proposed reconstruction of the sword hilt from Schortens.
Source: Westphal 2002: 68.

The fifth sword with analogical decoration comes from the Netherlands and is reported to be from two different locations – the IJssel River in Rheden (Bjørn – Shetelig 1940: 119-120; Ypey 1962-1963: 164) and the Meuse River near s’Hertogenbosch (Maczek 2021 : 202). Typologically, this weapon can be classified as the Altjührden type and is dated to the second half of the 8th and the first half of the 9th century based on analogies (Maczek 2021: 202). The upper guard and the pommel cap are interspersed with a sheet of copper alloy, which protrudes outside the pommel with a beaded edge. The guards are faceted, with each facet decorated with tubes placed in a cross grid. The tubes of each facet are interspersed with three larger rings, in which there are cross-shaped compartments. A similar combination of large and small tubes are found on the pommel cap. The larger rings are 0.8-0.95 cm in diameter and extend to a depth of 0.15-0.2 cm, the smaller ones are 0.35 cm in diameter and extend to a depth of 0.15 cm (Ypey 1962-1963: 166). In some of the smaller tubes, the lining with gold foil, on which the almandine fillings rest, has been preserved. The sword is kept in the National Museum in Leiden, inv. no. RSH 1.

Fig. 10: the sword with tubular decoration from the Netherlands.
Source: Rijksmuseum van Oudheden 2024.

The sixth sword is a find from a riverbed in the town of Souffelweyersheim, Alsace, France (Geibig 1991: Cat. no. 98; Menghin 1980: 232; Westphal 2002: 149-150). The sword belongs to the Mannheim-Speyer type (Müller-Wille 1982: 153) and is datable to the late 8th or early 9th century (Westphal 2002: 150). The pommel is divided into three fields, which are divided by pairs of copper alloy beaded wires. The same beaded band separates the pommel cap from the upper guard. On the underside of the base and on both sides of the crossguard we see decorative rivets with a beaded edge that hold the decorative copper alloy plates. The edges of the partitions are decorated with edging lines made of inlaid copper alloy wire. The filling decor of the hilt components is made from the same material and using the same method, depicting wavy lines ending in animal heads or flowers. Copper alloy tubes with almandine fillings are placed in these ends. The sword is kept in the Mainz State Museum, inventory no. 3339.

Fig. 11: the sword from Souffelweyersheim.
Source: Menghin 1980: Abb. 6.1; Westphal 2002: 149.

The last and largely overlooked sword was found during a flood in the Thielle River, Switzerland in 1887 (Boissonnas 1914: 18, Pl. XXV, Cat. No. 107), making it the southernmost find. In the 1950s, it was subsequently sold by the Fischer Gallery to the collection of E. A. Christensen in Copenhagen (Bruhn Hoffmeyer 1968: Cat. no. 29, Fig. 9, 13), to be sold again in 2012 at a Bonhams auction (Bonhams 2024). It is currently offered for sale by David Aaron Gallery (Aaron 2024). The sword belongs to the Mannheim-Speyer type and is datable to the late 8th or early 9th century. It is a specimen very similar to the find from Souffelweyersheim – we see the same division of the pommel and the same methods of decoration of the hilt components. The motif of inlay is not readable. The tubular decoration filled with red material is only visible on the pommel; each lobe of the pommel cap is equipped with one copper alloy tube.

Another sword typologically close to Petersen’s type B is known from a British private collection and should be mentioned (see Petty – Hill 2023). The division of the hilt decoration is similar to that of the Dorsten-Lembeck sword, with the difference that the space is divided by vertical bands of silver inlay, between which are pairs of tubes. There is a legitimate suspicion that this is a well-made forgery, which is why we are not working with this specimen in the article.

Fig. 12: the sword of the Thielle River. Source: Bonhams 2024.

Coloured stones in sword hilts are also known from the 7th century (e.g. Lindqvist 1931: 375; Paulsen 1967: 88-9), but the described group of swords seems to be the very first to apply a decoration consisting of pits organized into a cross grids. The tradition of pit decorated hilts is currently dated to the period from the 1st half of the 9th century to the 12th century and includes an evolution line with a trajectory from Petersen’s type E swords to Petersen’s type T variants, which marginally influenced other Petersen’s types that usually they lack decoration – types H/I, X, W (Kainov – Novikov 2024: 185-198; Shchedrina 2018; 2021; Tomsons 2018; Zozulja – Kainov 2022). In the light of our comparison, however, it appears that the tradition has its roots at least a generation earlier, in the 2nd half of the 8th century in the Frankish Empire.

It is not a completely new idea. Already Geibig (1992-1993: 216-7) and Menghin (1980: 251) connected the swords of Petersen’s types D and E with the Souffelweyersheim sword, but rather because of the shape and division of the pommel and the presence of decorative rivets at the guard ends. A conclusive link between the two groups was hindered by the insufficient examination of the Petersen’s type E swords, as well as the revision of swords with tubular decoration. An important starting point for further research was recently brought by Kainov in his book, which represents the best current study of E type swords: the author defined the evolutionary stages of Petersen’s type E and proved that the oldest type E swords (the so-called early E3 group) can be dated back to the 1st half 9th century, they are based on Petersen’s types A and special type 1 and are more or less contemporary with Petersen’s type D swords (Kainov – Novikov 2024: 191-193).

The study of swords with tubular decoration yields the following findings, which indicate that this group may indeed have served as one of the ideological bases for the production of the earliest swords of types D and E in the first half of the 9th century:

  • The tube decorated swords chronologically ideally precede Petersen’s D and E types. The two tube decorated swords stand close to Petersen’s special type 1, which is believed to be the shape inspiration of the E type.

  • The pit decoration is dominantly linked to the pommels created according to Geibig’s construction I, meaning a two-part pommel pierced by a tang that is peened on top. Almost all swords decorated with tubular decoration meet this requirement.

  • The earliest type E swords lack the inlay between the pits, which corresponds to most swords decorated with tubular decoration. One tube-decorated sword (Souffelweyersheim) has inlaid border lines at the guard edges, which are known from the younger Petersen’s type E variants.

  • The rectangular grid division and the massive copper alloy tubes correspond well to the decoration of Petersen’s type D hilts (see especially Kainov – Zuzulja 2014; Schoknecht 1988). The cross grid of tubes is essentially very similar to Petersen’s type E pit decoration, at the same time it resembles the inlaid decor of Mannheim-Speyer type swords, arranged into a chessboard structure (see Menghin 1980: 228). The Mannheim-Speyer is also considered to be the source of inspiration for type D and E swords (Menghin 1976: 12). One Norwegian type D sword (specifically S3962) has an almost identical decoration organisation as the sword from the Netherlands.

  • In three cases (Dorsten-Lembeck, Souffelweyersheim, Thielle) decorative rivets holding decorative plates appear at the guard ends. These rivets are a common phenomenon of D and E types swords as well.

  • Petersen’s type D swords and some type E variants have wide faceted guard with angular ends, which corresponds well to Petersen’s type B swords and its continental variants including the Altjührden type.

  • Swords decorated with tubular decoration come from a region that we assume played a key innovator and distributor role in the development of European swords (see also Müller-Wille 1982: 149).

  • Two nearly identical examples of the earliest type E swords come from continental Europe (see Müller-Wille 1977: Cat. no. 83), which appears to be a relatively high number in a set of five pieces. It is theoretically possible that swords with tubular decoration co-stimulated the creation of the earliest Petersen type E swords on the continent, not in Scandinavia.

Fig. 13: a proposal for the evolution of 8th-12th century swords with tubular and pit decoration.
Source: Kainov – Novikov 2024: 193, 198, edited.
Higher resolution here.


The author team would like to take this opportunity to thank the heirs of Mr. Eugène Thirion, who agreed with the publication. The photos of an approximate reproduction were taken by Mr. Henri Lehance, to whom we also thank. We are grateful to the following researchers and institutions for providing unpublished data: André Van Bossche (Viking Museum, Hamme), Eddy Bruyère (Museum of Ancient Arquebusiers, Visé), Jean-Luc Schutz (Grand Curtius Museum, Liège) and Astrid Wenzel (Baden State Museum, Karlsruhe). We must also mention Diego Flores Cartes, who is the author of fantastic diagrams.

We hope you liked reading this article. If you have any question or remark, please contact us or leave a comment below. If you want to learn more and support our work, please, fund our project on PatreonBuymeacoffee or Paypal.


Aaron, David (2024). Viking Sword of Petersen Type D. In: [online]. [2024-03-01]. Available here.

Aksdal, Jostein (2017). Dei anglosaksiske sverda : L-typesverd i England og Skandinavia. In: VIKING – Norsk Arkeologisk Årbok LXXX, 59–88.

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

Arbman, Holger (1937). Schweden und das karolingische Reich. Studien zu den Handelsverbindungen des 9. Jahrhunderts, Stockholm.

Behmer, E. G. (1939). Das zweischneidige Schwert der germanischen Völkerwanderungszeit, Stockholm.

Bilogrivić, Goran (2009). Karolinški mačevi tipa K. In: Opusc. archaeol. 33, 125–182.

Bjørn, Anathon – Shetelig, Haakon (1940). Viking Antiquities in Great Britain and Ireland. Part IV: Viking Antiquities in England with a Supplement of Viking Antiquities in the Continent of Western Europe, Oslo.

Bogaert, A. J. van (1962). Drie ‘Viking zwaarden’, noormannentijd, gevonden in de “Wijde Dorma” en “Luijsenbos” te Hamme, Hamme.

Boissonnas, Jean (1914). Collection Charles Boissonnas. Armes anciennes de la Suisse, Paris.

Bonhams (2024). A Rare Viking Sword Of Petersen Type D. In: [online]. [2024-03-01]. Available here.

Bruhn Hoffmeyer, Ada (1954). Middelalderens tveæggede sværd II, København.

Bruhn Hoffmeyer, Ada (1968). Gammelt jern. E. A. Christensens våbensamling, København.

De Baere, Sven (2019). Archeometallurgisch onderzoek van Vikingzwaarden, Gent.

De Prelle de la Nieppe, Edgar (1907). Une épée carlovingienne. In: Annales de la Société Royale d’Archéologie de Bruxelles 21, 81-96.

Dunning, G. C. – Evison, V. I. (1961). The Palace of Westminster Sword. In: Archaeologia 98, 123–158.

Evison, V. I(1967). A Sword From the Thames at Wallingford Bridge. In: The Archaeological Journal 124, 160–189.

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.

Geibig, Alfred (1992-1993). Der Hort eines Edelmetallschmiedes aus der frühslawischen Siedlung Rostock-Dierkow. Die Schwertgefāßteile. In: Offa 49/50, 215-227.

Ghellinck, Joseph de (1968). Catalogus van de tentoonstelling gewijd aan wapens en wapenrustingen behorend tot Belgische privé verzamelingen Kasteel Laarne, Brussel.

Hasse, Georges (1935). Les Vikings en Belgique. In: Revue belge d’archéologie et d’histoire de l’art 5, 199-211.

Hasse, Georges (1939). A propos de Vikings. In: Bulletin de la Société Royale Belge d’Anthropologie et de Préhistoire 54, 102-105.

Hernö, Anke (2007). Studien zu frühmittelalterlichen Gräberfeldern im westlichen Westfalen (Bocholt-Lankern, Wulfen-Deuten, Dorsten-Lembeck), Münster.

Hjardar, Kim – Vike, Vegard (2016). Vikings at War, Oxford – Philadelphia.

Hošek, Jiří et al. (2019). Ninth to mid-sixteenth century swords from the Czech Republic in their European context. Part I, Praha – Brno.

Hošek, Jiří et al. (2022). Ninth to mid-sixteenth century swords from the Czech Republic in their European context. Part II, Praha – Brno.

Jones, L. A. (2002). Overview of Hilt & Blade Classifications. In: Oakeshott, Ewart – Peirce, I. G. (eds). Swords of the Viking Age, Woodbridge, 15–24.

Jørgensen, A. N. (1999). Waffen & Gräber. Typologische und chronologische Studien zu skandinavischen Waffengräbern 520/30 bis 900 n.Chr., København.

Kainov – Novikov 2024 = Каинов, С. Ю. – Новиков, В. В. (2024). Новые находки мечей в Гнёздове (2017-2020 гг.), Москва.

Kainov – Zozulja 2014 = Каинов, С. Ю. – Зoзуля, С. С. (2014). Меч из раскопок кургана 1/1902 Михайловского некрополя // Военная археология 3, 29-46.

Kivikoski, Ella (1963). Kvarnbacken. Ein Gräberfeld der jüngeren Eisenzeit auf Åland, Helsinki.

Košta, Jiří (2021). Early Medieval Swords from the Czech Republic (9th–10th Centuries), Praha.

Košta, Jiří – Hošek, Jiří (2014). Early Medieval swords from Mikulčice, Brno.

Landesarchiv Baden-Württemberg (2024). Schwert. In: Landesarchiv Baden-Württemberg [online]. [2024-03-01]. Available here.

Lehmann, Ulrich (2015). Wurmbunte Klingen. Studien zu Konstruktion, Herstellung und Wertigkeit der frühmittelalterlichen Spatha in Westfalen, Münster.

Lindqvist, Sune (1931). Båtgravarna vid Valsgärde. In: Fornvännen 26, 372-377.

Maczek, Dušan (2016). It Is to the Water That It Must Return. An Inquiry into the Water Deposition of Swords in the Carolingian Period (ca. 750 – 1000 AD) in North-western Europe, Leiden.

Maczek, Dušan (2021). Non-funerary sword depositions in Carolingian Europe. In: Willemsen, Annemarieke – Kik, Hanneke (eds). Dorestad and its Networks. Communities, Contact and Conflict in Early Medieval Europe, Leiden, 199-206.

Menghin, Wilfried (1976). Ein karolingisches Prunkschwert aus dem Altrhein bei Mannheim. In: Anzeiger des Germanischen Nationalmuseums 1976, 7-13.

Menghin, Wilfried (1980). Neue Inschriftenschwerter aus Süddeutschland und die Chronologie karolingischer Spathen auf dem Kontinent. In: Spindler, Konrad (ed.). Vorzeit zwischen Main und Donau. Neue archäologische Forschungen und Funde, Erlangen, 227-272.

Müller, Heinrich – Kölling, Hartmut (1990). Europäische Hieb- und Stichwaffen aus der Sammlung des Museums für Deutsche Geschichte, Berlin.

Müller-Wille, Michael (1970). Ein neues Ulfberht-Schwert aus Hamburg. In: Offa 27, 65-91.

Müller-Wille, Michael (1977). Krieger und Reiter im Spiegel früh- und hochmittelalterlicher Funde Schleswig-Holsteins. In: Offa 34, 40-74.

Müller-Wille, Michael (1978). Das Schiffsgrab von der Île de Groix (Bretagne): ein Exkurs zum Bootkammergrab von Haithabu. In: Berichte über die Ausgrabungen in Haithabu 12, 48-84.

Müller-Wille, Michael (1982). Zwei karolingische Schwerter aus Mittelnorwegen. In: Studien zur Sachsenforschung 3, 101-154.

Paulsen, Peter (1967). Alamannische Adelsgräber von Niederstotzingen (Kreis Heidenheim), Stuttgart.

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

Petty, David – Hill, Robert (2023). Viking Era Swords, Farnham Common.

Rahir, Edmond (1928). Vingt-cinq anneés de Recherches, de Restaurations et de Reconstitutions, Bruxelles.

Rijksmuseum van Oudheden (2024). Tweesnijdig ijzeren zwaard. In: Rijksmuseum van Oudheden [online]. [2024-03-01]. Available here.

Rötting, Hartmut (1999). Zu Bestattung und Repräsentation im friesischen und fränkischen Stil im Spiegel herausgehobener Grabanlagen von Schortens, Landkreis Friesland. In: Both, Frank – Aouni, Heike (eds.). Über allen Fronten. Nordwestdeutschland zwischen Augustus und Karl dem Großen, Oldenburg, 231-248.

Shchedrina 2018 = Щедрина, А. Ю. (2018). Новый взгляд на деталь меча из коллекции Харьковского исторического музея // Харьковский историко-археологический сборник 23, 18-32.

Shchedrina, A. Yu. (2021). Sword from Krasn͡ianka. Complex of traditions and technologies. In: Acta Militaria Mediaevalia XVI, 205-224.

Schoknecht, Ulrich (1988). Vier neue Wikingerschwerter aus dem Bezirk Neubrandenburg. In: Ausgrabungen und Funde 33, 142-146.

Stalsberg, Anne (2008). Herstellung und Verbreitung der VLFBERHT-Schwertklingen. Eine Neubewertung. In: Zeitschrift für Archäologie des Mittelalters 36, 89-118.

Stein, Frauke (1967). Adelsgräber des achten Jahrhunderts in Deutschland – Text, Berlin.

Tackenberg, Kurt (1960). Die Bedeutung der Schwertsignierungen auf dem Abdinghofer Tragaltar. In: Westfälische Forschungen 13, 13-20.

Tomsons, Artūrs (2018). Zobeni Latvijas teritorijā no 7. līdz 16. gadsimtam, Rīga.

Vierck, Hayo (1980). Ein westfälisches “Adelsgrab” des 8. Jahrhunderts n. Chr. Zum archäologischen Nachweis der frühkarolingischen und altsächsischen Oberschichten. In: Studien zur Sachsenforschung 2, 457-488.

Vlasatý, Tomáš (2019). Petersen type M sword. In: Project Forlǫg – Reenactment and science [online]. [2024-04-12]. Available at:

Vlasatý, Tomáš (2020a). Number of European swords of 9th-11th century. In: Project Forlǫg – Reenactment and science [online]. [2024-04-12]. Available at:

Vlasatý, Tomáš (2020b). Wood species used for sword grips. In: Project Forlǫg – Reenactment and science [online]. [2024-04-12]. Available at:

Vlasatý, Tomáš (2021). Construction of wooden sword grips of 9th-12th century. In: Project Forlǫg – Reenactment and science [online]. [2024-04-12]. Available at:

Westphal, Herbert (1999). Herstellungstechnische Merkmale ausgewählter Schwerter des Gräberfeldes Schortens. In: Both, Frank – Aouni, Heike (eds.). Über allen Fronten. Nordwestdeutschland zwischen Augustus und Karl dem Großen, Oldenburg, 249-253.

Westphal, Herbert (2002). Franken oder Sachsen?: Untersuchungen an frühmittelalterlichen Waffen, Oldenburg.

Willems, W. J. H. – Ypey, Jaap (1985). Ein angelsächsisches Schwert aus der Maas bei Wessem, Provinz Limburg (Niederlande). In: Archäologisches Korrespondenzblatt 15, 103-114.

Ypey, Jaap (1962–1963). Vroeg-middeleeuwse wapens uit Nederlandse verzamelingen. In: Berichten van de Rijksdienst voor het Oudheidkundig Bodemonderzoek 12–13, 153-176.

Zozulja – Kainov 2022 = Зозуля, С. С. – Каинов, С. Ю. (2022). Мечи типа «W» на территории Древней Руси // Археологические вести 36, 107-129.

One Response

  1. Very interesting, as usual. You may relate the discovery of a sword at the Musée de Denain, in France. Denain is a french town near the boundery with Belgium. This the title of the study : DELATTRE Bernard, LEMAN Pierre & HANTUTE Gaston. Une épée de l’époque Viking au musée de Denain – 1987.

Leave a Reply

Your email address will not be published. Required fields are marked *