The Period Transport of Liquids

The transport and the storage of liquids are one of the biggest problems in the reenactment of any time period. Archaeological finds are only a few and making a keg or flask needs skill. For a person living in 21st century, it is much easier and cheaper to load a barrel of beer and some bottles of water to a car and after that hide everything in a tent. On historical events, there are principles of hiding modern bottles, however we would be lying, if we said that it is a generally valid and strictly followed convention.

If we move from a camp to a march, there is a necessity to have a field bottle, because in our luggage there is a limited space for equipment. In such a case, we are going to plan our way close to the springs and streams. Scandinavian streams (Old Norse lœkr) and mountain rivers have stayed drinkable even up till now, so if the Old Norse people made a good journey plan, they had no thirst. In the corpus of Old Norse dictionary, there is a term rǫst (“mile”), which literally means “distance between two halts”. Literary sources show existence of route with some fixed halts, which were located near the water streams.

Reconstruction of the farmstead Stöng, Iceland.

Even buildings and farmsteads were built near to the water streams. Water is necessary for a household, and people settled there not only because of water, but also because of fish. In some sources, the connection of a farm and a stream more than obvious:

Next to Ásólf’s hall, there was a river. Winter started and the river was full of fish. Þorgeir claimed that they settled on his fishing grounds, so Ásólf moved and built the second hall on west near to another river.
(The book of settlement, chap. 21, Hauskbók version)

The same situation was during the settlement of Iceland. Settlers often took up land, surrounded by two water streams. In addition, there was the law that the settler could take more land than she or he could walk around in one day. The farmstead Stöng, which was built in 11th century and covered by volcano ash in 1104, follows the same logic – it was built on a hill approximately one kilometer above the Fossá river. In densely built-up areas, water drained from wells. The most of farms did not need wells, because they had access to water streams (Short 2003: 74).

The containers for a water transportation can be divided to big volume containers and small volume containers. Among the big volume containers belong barrels, buckets and bigger ceramic vessels. Their volume can be between ones and hundreds of litres and they served for crowds, e.g. farm residents, merchants or soldiers on war expeditions. However, the dimension limits mobility, as can be shown by the quote from the Eyrbyggja saga (chap. 39):

Then too was it the custom of all the shipmen to have their drink in common, and a bucket should stand by the mast with the drink therein, and a locked lid was over it. But some of the drink was in barrels, and was added to the bucket thence as soon as it was drunk out.

The transport of barrels at the Bayeux tapestry.

The small volume containers were using for needs of individuals and they were parts of personal equipment. We are talking about different kinds of flasks, bags and bottles, which had limited volume – only up to several litres, but it was not difficult to carry them. It is necessary to add, that there are almost no preserved containers from Scandinavian area, so we have to use the written sources or look for the analogic finds from the period Europe.

The barrel from Haithabu.

The biggest container from the Viking period is a barrel (Old Norse: tunni, verpill). The barrels are well preserved in archeological, written and iconographic sources. In the previous written example, we can see the barrels were used for long-term storage of water on ships. Barrels also served for fermentation and storage of beverages in the halls. A big barrel with the volume of approximately 800 litres was found in Haithabu, Germany. Similar finds are known also from the Rome Empire period. Barrels of this kind are also depicted in the Bayeux Tapestry, where they are loaded on both carts and shoulders and carried to the ships. The Tapestry comments this depiction with these words: “These men carry arms to the ships and here they drag a cart laden with wine and arms.

A slightly smaller container is represented by a bucket, a tub and a vat (Old Norse: ker). The main advantage is a handle for the easier transport. It could be the most frequent big volume container of the period. A bucket was not provided with a permanent lid, because the liquid was meant for an immediate consumption. If it was necessary, the bucket could be covered by a removable lid (Old Norse: hlemmr or lok, see the quote from Eyrbyggja saga). The finds of buckets are well preserved in Oseberg and Haithabu. In Haithabu, they found imported big volume ceramics (so called Reliefbandamphoren) as well, which could be used for similar purpose thanks to transportation eyelets.

Opening of a bottle.  Made by Jakub Zbránek and Zdeněk Kubík.

We know only a few finds of flasks and bottles (Old Norse: flaska) made of leather, ceramics, wood, metal and glass in Early medieval Europe. Absence of local anorganic bottles in Scandinavia is a sign of the fact that organic materials were mainly used. From the following list, it is evident that ceramic, metal and glass bottles were imported to Scandinavia.

There are only a few written mentions about bottles from Scandinavia and they all are of the late date. It is interesting that some mentions are connected with bynames of people living in the Viking Age. We can find Þorsteinn flǫskuskegg (“bottle beard“) and Þorgeirr flǫskubak (“bottle back“) among the Icelandic settlers.

  • Leather bottle, made by Petr Ospálek.

    Leather bottles – it is the only kind mentioned in Old Norse sources. In Grettis saga (chap. 11), there is a funny story of Þorgeirr flǫskubak who is attacked by an assassin to his back, but he manages to survive, because the axe of the assassin hits a leather flask:

“That morning, Þorgeirr got ready to row out to sea, and two men with him, one called Hámundr, the other Brandr. Þorgeirr went first, and had on his back a leather bottle [leðrflaska] and drink therein. It was very dark, and as he walked down from the boat-stand Þorfinn ran at him, and smote him with an axe betwixt the shoulders, and the axe sank in, and the bottle squeaked, but he let go the axe, for he deemed that there would be little need of binding up, and would save himself as swiftly as might be. [Now it is to be said of Þorgeirr, that he turned from the blow as the axe smote the bottle, nor had he any wound. [Thereat folk made much mocking, and called Þorgeirr Bottleback, and that was his by-name ever after.”

This part continues with a stanza with this meaning: “Earlier the famous men cut their swords into enemies’ bodies, but now a coward hit a flask with whey by an axe. Even though it is a nice example of an Old Norse perception of society decline, but we can notice the mention about whey (Old Norse sýra). The whey was mixed with water in a ratio 1:11 and created a popular Icelandic drink, the so-called blanda (for the exactl mixture, see here, page 26). The saga suggests that Þorgeirr has got such a drink in his flask.

The leather flasks are mentioned in Anglo-Saxon sources and are archaeologically documented in Ireland, where were found some decorated pieces from 12th century. They are lightweight and ideal for long hikes. They are resistant against damage too. But sometimes water is running through, whis is a disadvantage. Summary, I recommend to reconstruct of leather variants.

A replica of a wooden bottle, made by CEA.


  • Ceramics bottles – ceramics bottles were popular for the whole Early medieval period. They were used in the the Roman times (Roman ceramics amphoras for a wine transporting are known from Rhineland), in the Migration period, as well as in the period of 9th to 11th century. One piece was found in Winchester, England (11th century, photos here, here, here), another one in Gnezdovo, Russia (10th century, photo here) and yet another in Great Moravian Staré Město (9th century, photos here and here). In Belgian Ertvelde-Zelzate (9th century, here), a painted flask was found. Analogies of this bottle were found in Dorestad and in Norwegian Kaupang too. The find from Kaupang is represented by nine orange painted shards – the only proof of ceramics flasks in Scandinavia (Skre 2011: 293). The similar shape to Roman amphoras remained popular in the Rhineland, and it devepoled into so-called Reliefbandamphoren that are up to 70 cm high. Some pieces were found in Haithabu as well. Ceramic bottles seem to be popular in Eastern Europe as well.

    The pottery industry of Viking Age Scandinavia was not very developed, so we can presume that all the ceramic bottles in Scandinavia were imported. Me and my colleagues were using this type for years and it proved to be very practical. On the other hand, the use is very questionable in Scandinavia.

  • Bronze bottle from Aska.

    Metal bottles – an unique copper-alloy bottle was found in the woman’s grave in Aska, Sweden. According to works, which I found on the internet (here and here), the grave dated to 10th century and the container is considered a Persian import, because of the inscription. The origin limits the usage in reenactment. A similar bottle was found in FölhagenGotland, and it is dated to the of 10th century (the picture on demand).

  • Glass bottles – I am aware of two Scandinavian bottle necks made of glass, they are very rare finds. The first one was found in Haithabu and is dated to the 9th century (Schiezel 1998: 62, Taf. 13:1–2). The second one was found in a rich female grave from Trå, Norway, dated to the 10th century. Pictures on demand.

All the mentioned bottles except the glass and metal examples do have the eyelets. So, we can suppose that they had got a strap for a hanging. To my knowledge, stoppers are never preserved, so they probably were made of wood. The experiments showed that oaken lathed or hand-made mushroom or cylinder-shaped stoppers are functional. While a simple wooden stopper works for wooden and leather bottles, in case of other materials, it is useful if the stopper is a bit smaller and wrapped in a textile, so the neck is not destroyed by the harder material of the stopper. 

I believe that the article provided a brief summary of Early medieval liquid containers. For reenactment purposes, I recommend to use the barrels and buckets for camp life and the bottles for a march. This can also lead to reconstructing proper banquet tools, like spoons, scoop and ladles, that are present in the sources. If needed, write your feedback into the comments, the problem of a liquid transportation is still opened. Many thanks to Roman Král, Zdeněk Kubík, Jan Zajíc and Jakub Zbránek, who helped me with this article and answered my questions. 


The book of settlement – Landnamabók I-III: Hauksbók, Sturlubók, Melabók. Ed. Finnur Jónsson, København 1900.

Grettis saga – Saga o Grettim. Přel. Ladislav Heger, Praha 1957. Originál online.

Eyrbyggja saga – Sága o lidech z Eyru. Přel. Ladislav Heger. In: Staroislandské ságy, Praha 1965: 35–131.

Cleasby, Richard  Vigfússon, Gudbrand (1874). An Icelandic-English dictionary, Toronto.

Schietzel, Kurt (1998). Die Glasfunde von Haithabu, Berichte über die Ausgrabungen in Haithabu 32, Neumünster.

Short, William R. (2010). Icelanders in the Viking Age: The People of the Sagas, Jefferson.

Skre, Dagfinn (ed.) (2011). Things from the Town. Artefacts and Inhabitants in Viking-age Kaupang. Kaupang Excavation Project Publication series, vol. 3., Århus.


For those interested in wooden barrels, buckets and ceramic vessels, I recommend these books:

Hübener, Wolfgang (1959). Die Keramik von Haithabu, Neumünster.

Janssen, Walter (1987). Die Importkeramik von Haithabu, Neumünster.

Wesphal, Florian (2006). Die Holzfunde von Haithabu, Neumünster.

The Length of Early Medieval Belts

There are some “truths” in reenactment that are not questioned even though they should be. These are called “reenactorism” and engaged by both newbies and veterans. In this article we will show one of these, the myth of a long belt in Early medieval Europe, following the work done by German reenactor Christopher Kunz.

It is fully evident from the preserved material that there was a number of approaches to belt wearing in the Early Middle ages. These approaches originated alongside cultural environment and local development, social ranking, gender and usage method. The assumption of using a uniform belt type with the same width and length is wrong. On the initiative of beginning reenactors who often raise questions about belt length, in this article we will try to map the legth of men’s leather belts according to iconography and finds in burial complexes.

Fig. 1: Grave no. 59 from the Haithabu-Flachgräberfeld burial site
Arents – Eisenschmidt 2010b: 308, Taf. 10.

Simple belt with a short end (up to approx. 20 cm)

This form best resembles present belts, which are manufactured approximately 15 cm longer than the waistline. In seven graves from Birka, Sweden (488, 750, 761, 918, 949, 1030, 1076) the buckles are no more than 10 cm far from each other (Arbman 1943) and similar positions could be found throughout Europe – we can mention Great Moravian (i.e. Kalousek 1973: 33, Fig. 13) or Danish graves (Arents – Eisenschmidt 2010b: 301, Taf. 3). There are no belts with hanging strap-ends in Early medieval iconography, which is rather schematic than detailed. Belts are scarcely visible in painted iconography as they usually seem to be overlapped by pleated upper tunics, which can be interpreted as an element of fashion. As a result the belt looks like a narrow horizontal line.

There is a certain contradiction between some burial positions and strap-end decor, where some of Early medieval belts had strap-ends that hung down when threaded through the buckle. The most graphic evidence comes from depictions of people and animals which can be seen on the strap-ends and placed lengthwise. In some cases, there are figures of naked men depicted on the strap-ends, which could imply that the hanging end could reach down to the genitals and symbolically represent or emphasize them (Thomas 2000: Fig. 3.16, 3.27). In the listing below we will attempt to suggest several manners of tying these belts.

Fig. 2: A selection of painted iconography of 9-11th century depicting a belt hidden in tunic pleats.
From the left: British Lib. MS Arundel 60, 4r, 11th century; BNF Lat. 1, 423r, 9th century; British Lib. MS Stowe 944, 6r, 11th century; XIV.A.13, 29v, 11th century

Fig. 3: Strap-ends depicting a naked man.
Thomas 2000: Fig. 3.16, 3.27.

Fig. 4: A rare depiction of hanging strap-end in Western Europe iconography. Manuscript: Latin 1141, Fol. 14, 9th century.

  • Loose end
    The simplest form is represented by a belt worn in its nearly maximal length. The end is then short enough not to obstruct manual labour and because it copies the belt, it can be hidden in a pleated tunic. Depictions of loose belt ends can be quite typically observed in 13th and 14th century. Moreover, we know a belt from Early medieval Latvia which had a metal ring at its end, used to grapple on a buckle tongue. The very same method was is also known from Čingul mound, Ukraine, from 13th century (Отрощенко – Рассамакин 1983: 78).

Fig. 5: Reconstruction of belts from 400-700 AD in Zollernalb region, Germany.
Schmitt 2005: Abb. 15.

Fig. 6: Reconstruction of Haithabu type belts.
Arents – Eisenschmidt 2010b: 140, Abb. 61.

  • Tucked behind the belt
    Another simple way of wearing a belt is tucking its end behind the already fastened part of the belt. We have at least one piece of evidence of this wearing from Anglo-Saxon England, where a belt passed through the buckle, flipped back and end tucked behind itself was documented (Watson 2006: 6-8). This forms a perpendicular line on the belt and keeps the face side of strap-end exposed. In case of pleated tunic covering the belt it can be easily adjusted to form a line.

Fig. 7: Strap-end being flipped back after going through the buckle and tucked behind the already fastened belt. Shrublands Quarry, Watson 2006: Fig. 6.

  • Tucked in a slider
    Metal belt sliders are very scarce in terms of archeological material. One of this kind was found within Gokstad Barrow (C10439) and adjusted to fit a strap-end from the same grave (Nicolaysen 1882: 49, Pl: X:11). Another slider was presumably found in Birka grave no. 478 (Abrman 1943: 138) and three more made of sheet bronze were apparently found in Kopparvik, Gotland (Toplak 2016: 126). According to sliders usually appearing in relation to spurs or garters where they are 2-3 centimeters wide (i.e. Andersen 1993: 48, 69; Thomas 2000: 268; Skre 2011: 72-74), we can assume that if the sliders were used with belts more, we would be able to detect them more easily. It is possible that they corroded over time, that organic sliders were used too or that they will be found during a more detailed research. Generally we can assume that the sliders were used in cases where the buckles did not include holding plates – in opposite cases the holding plates would not be visible after using the slider.

Fig. 8: Reconstruction of the belt from grave no. 478 at Birka.
Abrman 1943: 138, Abb. 83.

Fig. 9: Attempt for a reconstruction of the belt from Birka grave no. 949 applying a leather slider.
Author: Sippe Guntursson.

  • Puncturing two holes
    A relatively elegant reenactor’s solution is to puncture two consecutive holes and tuck the belt behind its buckle. All the belt’s components therefore remain visible. This solution was documented in case of at least two archeological finds from Britain and Belgium, 6th-7th century. (De Smaele et al. in pressWatson 2002: 3). The same system is known from Early medieval Latvia. In case of pleated tunic covering the belt it can be easily adjusted to form a line.

Fig. 10: Puncturing two holes that enables threading the strap-end behind a buckle.
Author: Erik Panknin.

  • Attaching by a thong
    Another aesthetical, yet undocumented manner of attaching a belt is adding a thong which holds the buckle’s tongue while the strap-end continues further behind the buckle. We have no evidence for this manner.

Fig. 11: Fixing the buckle with a thong attached to the belt. An unfounded hypothesis.
Author: L’Atelier de Micky.

  • Tucking into a buckle slot
    Buckles having a rectangular slot aside from the typical loop are very common in Eastern-European regions. After fastening the belt using the loop’s tongue, the strap-end could be tucked into this slot and hanged downwards. In case of pleated tunic covering the belt it can be easily adjusted to form a line.

Fig. 12: Reconstruction of the belt from Berezovec barrow.
Степанова 2009: 250, рис. 18.

  • Knot on a belt
    The most frequent solution among reenactors is undoubtedly a knot performed like this: after going through the buckle, the strap-end is tucked behind the belt from below and then passed through the resulting loop. This means achieving a perpendicular line on the belt and keeping the strap-end’s face side visible. This knot-tying, although with much shorter belt than standardly used in today’s reenactment, could be found in France during the Merovingian age (France-Lanlord 1961). With a high probability, the same solution was found in a grave from St Michael’s Church graveyard in Workington, England. Knots were often worn in 13th and 14th century.

Fig. 13: Reconstruction of a Merovingian belt from St. Quentin.
France-Lanlord 1961.

Composite belt with a long end

Some of the Eastern-European Early medieval decorated belts are manufactured in a more complex way, having one or more longer ends. In case of a belt constructed to have more ends, one of these ends – usually the shorter one – is designed to be fixed by the buckle, while the others are either tagged on or formed by the outer layer of two-layered belt. Long ends of these costly belts are designed for double wrapping, tucking into a slider or behind the belt. The length of the ends is not standardized, however we are unable to find any belt that would reach below its owner’s crotch when completely tied. While looking for parallels, we can notice that a belt compounded this way has many similarities to tassels on horse harnesses. Apparently, the belts were worn by riders or emerged from such a tradition, then maintained the position of wealthy status even after being adopted by neighbouring non-nomadic cultures. At last we can state that longer belts were designed mainly to hold more decoration and to allow the owner to handle the length more flexibly, whether for practical or aesthetical reasons.

Fig. 14: Composite belts with long ends.
A, b – belts from Gnezdovo (Мурашева 2000: рис. 109, 113), c – belts from Nové Zámky (Čilinská 1966: Abb. 19), d – belt from Hemse (Thunmark-Nylén 2006: Abb. III:9:3), e – reconstruction of belt tying from Káros, Hungary (Petkes – Sudár 2014).


The topic of belt lenght in reenactment is definitely a controversional one as it touches every male reenactor. Belts are sometimes costly and even a hint, originally meant as constructive critic, can easily cause negative emotions. There is no need for them though, as there is probably no reenactor who has never worn a long belt. We suppose that this reenactorism, used in practice for more than 30 years over the whole world, is caused by these factors:

  • unwillingness to perform one’s own research leading to imitation of a generally accepted model
  • bad access to sources or their misintepretation
  • easily obtainable and cheap, yet historically inaccurate belts sold on the internet in standard length of about 160 cm
  • unwillingness to talk about the problem by both organizers and attendants

In this article, we demonstrated that historical belts often did not have any hanging ends and that the maximum length where the end would reach was the crotch, which could have a symbolic meaning. Any of the aforementioned manners of attaching should not be incompatible with the sources we have at our disposal, however as we already mentioned, both the length and style of wearing followed local traditions. Western Europe therefore preferred delicately hidden belts while in Eastern Europe, the richly decorated belts were worn on public display.


Andersen, A. W. (1993). Lejre-skibssættinger, vikingegrave, Gridehøj. Aarbøger for Nordisk Oldkyndighed og Historie 1993: 7–142.

Arbman, Holger (1943). Birka I. Die Gräber. Text, Stockholm.

Arents, Ute – Eisenschmidt, Silke (2010a). Die Gräber von Haithabu, Band 1: Text, Literatur, Die Ausgrabungen in Haithabu 15, Neumünster.

Arents, Ute – Eisenschmidt, Silke (2010b). Die Gräber von Haithabu, Band 2: Katalog, Listen, Tafeln, Beilagen, Die Ausgrabungen in Haithabu 15, Neumünster.

Čilinská, Zlata (1966). Slawisch-awarisches Gräberfeld in Nové Zámky. Archaeologia slovacca, Fontes, t. 7, Bratislava.

De Smaele, B. – Delaruelle, S. – Hertogs, S. – Scheltjens, S. – Thijs, C. – Van Doninck, J. – Verdegem, S. (in print). Merovingische begraving en middeleeuwse bewoning bij een bronstijd grafveld aan de Krommenhof in Beerse, AdAK rapport 17, Turnhout.

France-Lanlord, Albert (1961). Die Gürtelgarnitur von Saint-Quentin. In: Germania 39, 412-420;

Kalousek, František (1971). Břeclav-Pohansko. 1, Velkomoravské pohřebiště u kostela : archeologické prameny z pohřebiště, Brno.

Мурашева, В.В. (2000). Древнерусские ременные наборные украшения (Х-XIII вв.), М.: Эдиториал УРСС.

Nicolaysen, Nicolay (1882). Langskibet fra Gokstad ved Sandefjord = The Viking-ship discovered at Gokstad in Norway, Kristiania.

Отрощенко, B. – Рассамакин, Ю. (1983). История Чингульского кургана // «Наука и жизнь», 1983/07, 78-83.

Petkes, Zsolt – Sudár, Balázs (2014). A honfoglalók viselete – Magyar Őstörténet 1, Budapest.

Schmitt, Georg (2005). Die Alamannen im Zollernalbkreis, Materialhefte zur Archäologie in Baden-Württemberg Band 80, Pirna. Available:

Skre, Dagfinn (2011). Things from the Town. Artefacts and Inhabitants in Viking-age Kaupang, Aarhus & Oslo.

Степанова, Ю.В. (2009). Древнерусский погребальный костюм Верхневолжья, Тверь, Тверской государственный университет.

Thomas, Gabor (2000). A Survey of Late Anglo-Saxon and Viking-Age Strap-Ends from Britain, University of London.

Thunmark-Nylén, Lena(2006). Die Wikingerzeit Gotlands III: 1–2 : Text, Stockholm.

Toplak, Matthias (2016). Das wikingerzeitliche Gräberfeld von Kopparsvik auf Gotland : Studien zu neuen Konzepten sozialer Identitäten am Übergang zum christlichen Mittelalter, Tübingen : Eberhard Karls Universität Tübingen.

Watson, Jacqui (2002). Mineral Preserved Organic Material from St Stephen’s Lane and Buttermarket, Ipswich, Portsmouth : English Heritage, Centre for Archaeology.

Watson, Jacqui (2006). The Identification of Organic Material Associated with Metalwork from the Anglo-Saxon Cemetery at Smythes Corner (Shrublands Quarry), Coddenham, Suffolk, Portsmouth : English Heritage, Centre for Archaeology.