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

The helmet from River Somme, France


While writing the Rethinking the helmet from Yarm, we mentioned a helmet found in the River Somme near the town of Abbeville. With almost a year’s gap, the helmet was republished in print. From that reason we feel the time is ripe for a separate post on this helmet, which may hold the key to understanding the larger group of helmets scattered around the North Sea.

Circumstances of the find, storage and publication

The helmet was allegedly found in the Somme River near the French town of Abbeville, apparently in the 19th century, certainly before 1875. After its discovery, it was given to the Museum of Artillery (Musée d’artillerie) by the archaeologist Boucher de Perthes (1788 – 1868) along with a number of other objects in Paris, which today corresponds to the Army Museum (Musée de l’Armée). The object is still there today under the inventory number H3. The inventory number was given to the item in 1890 at the latest.

Location of the town of Abbeville on the map of Europe.

To date, the helmet has only been published in a limited way. The earliest known publication of the helmet was provided by Viollet-le-Duc (1875: 104-5), who included a fairly faithful drawing. The helmet also appears in Robert’s inventory (Robert 1890: 167). As far as we know, the helmet was not published in the 20th century, apparently because it was considered a forgery. Recently, the French researcher Baptiste managed to examine the object and published it for the first time with photos in 2022 (Baptiste 2022: 88-9). The photos modify some of the details known from the earlier drawing.

In the following text, we are inclined to the opinion that, despite its bizarre appearance, the Somme river helmet could be original and not a modern forgery. While it is true that the helmet is part of a highly dubious collection, there were no known helmets in the 19th century with similar features that could be copied – the helmet from Vendel XIV was found in the 1880s-1890s, the Bremen helmet in 1924, the Groningen helmet in 1941, the Yarm helmet in the 1950s and the Coppergate helmet in 1982. Combining the then unknown dome accessories would be a work of a brilliant mind with good typological knowledge of the parallels at a time when they had not yet been discovered. We cannot fully rule out that the Somme helmet is a forgery based on period iconography, but in that case we would have to ask which early medieval scenes were widely known in 1870. It is equally possible that the helmet was a compilation of several pieces or an early medieval or even recent modification of an older helmet. At the same time, the question arises that the Yarm helmet could be a forgery based on the Somme helmet, but as we have shown above, the helmet is almost unknown in the literature even among experts. Despite a number of similarities, both helmets show differences. However, authenticity needs to be verified by archaeometallurgical analysis and a detailed publication.

The River Somme helmet. Source: Viollet-le-Duc 1875: 104.

Description of construction

The basis of the helmet is a perforated rectangular strip of relatively uniform width, forming the future circumferential band, which has been wound into an oval shape copying the head, folded at the nape and riveted. The riveting is done with at least two rivets and is slightly offset from the center of the helmet’s width. The lower edge of the band is curved at a right angle outside the intended center of the helmet and forms a brim. The height of the circumferential band in the bent state is at least 4 cm. In the bend of the band there is an unknown number of holes, close to 12, which seem to have been punched towards the inside of the dome after the bending the band and which were probably used to fasten the lining. The spacing of these holes is roughly 3-4 cm. The bvrim shows unusual and unaesthetic wrinkles that were created by hammering. All the riveting found on the helmet is done with rivets with unaesthetically hammered flat heads.

The helmet from River Somme. Source: Baptiste 2022: 89, edited.

Another rectangular band with a width of at least 5 cm is riveted to the inside the oval in a line from the forehead to the nape of the neck using pairs of rivets. The band slightly widens towards the circumferential band. The rivet used to fixate the ends of the circumferential band on the nape is apparently used to fasten the vertical band. The vertical band was perforated along its longer sides in advance, but the openings do not appear to have been made perfectly symmetrical and the number of holes on each side varies between 4-6. Perpendicular to the vertical (in a line from ear to ear) another band is riveted that touches the underside of the aforementioned vertical band at the top of the helmet. In terms of the number of rivets and holes, both bands are identical. The crossing of the bands is not, at least according to the available materials, riveted. By connecting these bands, a basic skeleton is created, the advantage of which lies in the possibility of control of the result and comparison with the size of the head of the future wearer. The top of the dome is rounded, not decorated with a raised ridge or projection.

Four triangular perforated fillings are subsequently inserted into the frame an their attachment completed the dome of the helmet. The fillings are attached to the circumferential band with pairs of rivets, to the vertical bands with pairs or triplets of rivets. The corners of the vertical bands are probably attached to the circumferential band and to the fillings at the same time, thus tripling the strength of the material. The fillings are flat, without reinforcements, with significant defects today. According to Viollet-le-Duc, the dome has a length of 21 cm and a height of 17.5 cm including the brim.

The helmet from River Somme. Source: Baptiste 2022: 89, edited.

The dome is then fitted with three remarkable components. The first of these is a long, hourglass-shaped nasal that reaches up to the vertical band and widens towards the lower end in a spatula-like manner. It is fixed with a total of three rivets – two are fixed to the circumferential band, one to the vertical band. The bottom half of the nosepiece is convex, a feature used to minimize bending and damaging the wearer’s nose. The bottom edge of the nasal is slightly curved; overall, the nasal seems to extend beyond the wearer’s nose. The brim appears to have been intentionally removed at the point of deposit. The second component is the low triangular cheek-sguards that do not reach the level of the lower end of the nasal and that create apertures for the eyes. The upper edge of the cheek-guards is bent at right angles and is firmly riveted to the lower side of the dome’s circumferential band with at least three rivets. The edge of the cheek-guards that was in contact with the face appears to be slightly bent and adapted for more comfortable wear. The last part to mention is the neck guard. This is formed from a single triangular sheet metal, the point of which is hammered down to a firm point, inserted into a hole in the brim of the dome and curled into an arch. The lower edge of the guard is rolled up.

The helmet from River Somme. Source: Baptiste 2022: 89, edited.

Analogies and dating

Comparison of the find with known parallels and proposed dating in the available literature is not satisfactorily resolved. Due to its dome construction, the helmet undeniably fits into the so-called Bandhelme (banded helmets), as evidenced by the overall shape, width of the bands and the number of used rivets. Variants operating with a four-segment dome and vertical bands are common in 5th-8th century (Miks 2009: Abb. 4), and the vertical bands used in the forehead to nape line occurs in the Continent, Middle East (Vogt 2006: 283-293) and Anglo-Saxon England (Hood et al. 2012Meadows 2019Tweddle 1992), with a subgroup of helmets that have solid bands in both directions that cross at the top (Bretzenheim and Voivoda, see Vagalinski 1998Vogt 2006: 283-6). As for the dome, the helmet simply cannot be dated to the post-800 horizon – as shown by the Coppergate helmet, which is dated to the 3rd quarter of 8th century (Tweddle 1992: 1082).

Construction scheme of so-called Bandhelme. Source: Miks 2009: Abb. 4.

The bending of the edge of the dome outwards and the creation of a brim are known for a total of four helmets of this construction: the helmet from the Novakovo, Bulgaria, which is dated to the 5th-6th century (Biernacki 2012: 99-100; Parušev 2002), the helmet from Bremen, Germany (Grohne 1929: 73-5; Lonke 1925; 1946: 37-41; Vogt 2006: Taf. 56), the helmet from Groningen, the Netherlands (van Griffen 1954; Kubik 2016: Fig. 16) and the helmet from Yarm, UK (Caple 2020). The last three named, especially the helmet from Yarm, which also has holes on the brim and similar rivets, form the closest parallels to the find. Based on the pointed rivets, the helmets from Bremen and Groningen are approximately dated to the period 650-800 (Kubik 2016: 99; 2017: 158), which was also confirmed by the revision of the stratigraphy of the Bremen find (Bischop 2014: 418-9; Rech 2004: 62-4 ). The Yarm helmet is dated analogically to the same period (see Rethinking the helmet from Yarm).

Photographs of the helmet from Yarm, taken by Matt Bunker.

Helmets from Groningen (left) and Bremen (right). Source: Kubik 2016: Fig. 16.

The convex nasal shows a great similarity to the nosepiece of the mask of the Yarm helmet. Other close parallels in the 6th-8th century horizon are represented by nasals of the helmet from Coppergate (Tweddle 1992), the helmet from Ultuna (Tweddle 1992: Fig. 544), the helmet from Vendel XIV (Tweddle 1992: Figs. 537, 539) an the helmet from Endre backe, Gotland (Nerman 1969: Taf. 65). In terms of length, the nasal is close to the nasal of the Coppergate helmet. Generally speaking, the nasal itself does not contribute to the narrowing of the proposed dating.

The neck guard is more promising as it is not a common feature. Analogous protectors, but attached by hinges, are found on helmets from Ultuna, Vendel XIV, Valsgärde 5 (Lindqvist 1931) and Broa, Gotland (Nerman 1969: Taf. 66). All four helmets can be dated between 520-720 in absolute dating (Kyhlberg 2013: 111; Tweddle 1992: 1109, 1115). As for cheekpieces, after the 6th century they seem to have receded on the Continent and survive in Scandinavia (Tweddle 1992: Figs. 537, 539) and Anglo-Saxon Britain (Fern 2019: 80-9; Meadows 2019: 25; Tweedle 1992: 989-997 ), while the youngest cheeks are recorded for the helmet from Coppergate in the 3rd quarter of the 8th century.

When evaluating the helmet, we face a lack of analogous pieces that would allow us to determine whether the helmets from Bremen, Groningen, Somme river and Yarm form one group of approximately the same age, or whether they are different stages of development with different production dates. As far as we can tell, all of these helmets move on a chronologic axis, the beginning of which is marked by the helmet from Vendel XIV (520-625; Tweddle 1992: 1109) and the end of which is given by the Coppergate helmet (3rd quarter of the 8th century; Tweddle 1992). The Somme River and Yarm helmets are so similar to each other that one can speculate about a similar age. If we accept this, the period from the construction of the helmet from Valsgärde 6 (620-720; Kyhlberg 2013: 111) to the end of the use of band-shape neckguards (1st quarter of the 8th century) becomes our dating limits. If we were to consider all four helmets as products of the same generation, the dating would be limited on the one hand by the use of pointed rivets to the middle of the 7th century, on the other hand by the use of a neck protectors to the 1st quarter 8th century. Previous publications date the helmet to the 9th-10th century (Baptiste 2022: 88) or the 12th century (Viollet-le-Duc 1875: 104), which does not seem to us to be a correct conclusion.

If the Somme helmet is genuine and completely manufactured at once, it is a valuable artifact most likely from the period 620-775, expanding the limited corpus of 7th-8th century helmets around the North Sea. Its rediscovery in the literature can shed new light not only on the helmet from Yarm, but also on the development of banded helmets, which were still produced plain and without face or neck protection in the 6th century (e.g. Hood et al. 2012). However, modern review and proper publication in the press is needed before general acceptance by the academic community.


This article would not have been possible without the help of several people. First of all, we need to thank Adam Kubik, who brought the helmet to our attention. Our big thanks goes to researcher and reenactor N. P. Baptiste, who personally examined the helmet and subsequently published it. We also express our gratitude to Michal Havelka (baba_jaga_atelier), who helped us edit the illustrations.

I hope you liked reading this article. If you have any question or remark, please contact me or leave a comment below. If you want to learn more and support my work, please, fund my project on Patreon or Paypal.


Baptiste, N. P. (2022). Armes & Armures. Tome 1, VIe-XIIe siècles, Bayeux.

Biernacki, A. B. (2012). Early Byzantine Iron Helmets from Novae (the Diocese of Thrace). In: Böhlendorf-Arslan, B. – Ricci, A. (eds.). BYZAS 15, Byzantine Small Finds in Archaeological Contexts, Istanbul, 91-104.

Bischop, Dieter (2014). Der Bistumssitz Bremen im späten 8. und 9. Jahrhundert. In: Weiss, R.-M. – Klammt, A. (eds.). Mythos Hammaburg. Archäologische Entdeckungen zu den Anfängen Hamburgs, Hamburg, 417–433.

Caple, Chris (2020). The Yarm Helmet. In: Medieval Archaeology 64:1, 31-64.

Fern, Chris (2019). Characterising the Objects. In: Fern, Ch. – Dickinson, T. – Webster, L. (2019). The Staffordshire Hoard. An Anglo-Saxon Treasure, London, 29-122.

Grohne, Ernst (1929). Bremische Boden- und Baggerfunde. In: Jahresschrift des Focke-Museums 1929, 44–102.

Hood, Jamie et al. (2012). Investigating and Interpreting an Early-to-mid Sixth-century Frankish Style Helmet. In: British Museum Technical Research Bulletin 6, 83–95.

Kubik, A. L. (2016). Introduction to studies on late Sasanian protective armour. The Yarysh-Mardy helmet. In: Historia I Świat 5, Siedlce, 77-105.

Kubik, A. L. (2017). Hełmy Azji Południowo-Zachodniej pomiędzy VI – VIII w.n.e. Zarys problematyki, Siedlce.

Kyhlberg, Ola (2013). Leadership cult and burial practise. In: Munktell, Ing-Marie (ed.). Valsgärde 1, 2 & 4, Uppsala, 109-144.

Lindqvist, Sune (1931). En hjälm från Valsgärde, Uppsala.

Lonke, Alwin (1925). Ein Spangenhelm aus Bremen. In: Praehistorische Zeitschrift 16, 196-197.

Lonke, Alwin (1946). Römisch-Germanisches. Sechs Untersuchungen, Gießen.

Meadows, Ian (2019). The Pioneer Burial: a High-Status Anglian Warrior Burial from Wollaston Northamptonshire, Oxford.

Miks, Christian (2009). Relikte eines frühmittelalterlichen Oberschichtgrabes? Überlegungen zu einem Konvolut bemerkenswerter Objekte aus dem Kunsthandel. In: Jahrbuch des Römisch-Germanischen Zentralmuseums 56, 395-538.

Nerman, Birger (1969). Die Vendelzeit Gotlands II. Tafeln, Stockholm.

Parušev 2002 = Парушев, Веселин (2002). Средновековен шлем от Варненско // Оръжие и снаряжение през късната античност и средновековието IV – XV в. Отв. ред. Baлepu Йomoв et al., Варна, 191-194.

Rech, Manfred (2004). Gefundene Vergangenheit – Archäologie des Mittelalters in Bremen, Bonn.

Robert, Léon (1890). Catalogue des collections composant le Musée d’artillerie en 1889, T. 2, Paris.

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

Vagalinski, Lyudmil (1998). Ein neuer spätantiker Segmenthelm aus Voivoda, Schumengebeit (Nordostbulgarien). In: Archaeologia Bulgarica 1, 96-106.

van Griffen, A. E. (1954). Zwei merkwürdige Funde. In: Palaeohistoria 3, 225-244.

Viollet-le-Duc, Eugéne (1875). Dictionnaire raisonné du mobilier français de l’époque carlovingienne à la Renaissance, T. 6, Paris.

Vogt, Mahand (2006). Spangenhelme: Baldenheim Und Verwandte Typen, Mainz.

2 Responses

  1. There was a paper on helmets from Vyatka that had an example that looked somewhat similar, a domed cap with a nasal and a single hinged, narrow back protector, from the Pyanobor period. Unfortunately, I don’t know Russian and there’s only so far I can get with Google translate, so I haven’t been able to get more details on it though. I imagine that a helmet from Vyatka and one from Abbeville wouldn’t be connected, but its still interesting to see

Leave a Reply

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