Navigation – Plan du site

AccueilNuméros21Studies on WoodThe Bremen-Cog: environmental issues

Studies on Wood

The Bremen-Cog: environmental issues

Mike Belasus, Aoife Daly et Adolfo Miguel Martins
p. 257-261

Résumés

La cogue de Brême avait été datée par dendrochronologie dans les années 1960, alors que la discipline en était encore à ses balbutiements. On avait pu alors établir une date d’abattage des arbres pendant l’hiver 1378/1379 dans les Weser Uplands, ce qui a été confirmé en 2007. En 2016, une nouvelle analyse complète du navire a été réalisée. Tous les échantillons analysés proviennent d’arbres probablement abattus dans le dernier quart du xive siècle dans l’arrière-pays proche de Brême. En outre, une tentative de reconstruction de la morphologie des arbres et une enquête détaillée sur leurs caractéristiques de croissance ont révélé la mauvaise qualité du bois, qui ne semblait pas destiné, à l’origine, à la construction navale.

Haut de page

Texte intégral

  • 1  “Bremen-Type” refers to the shipbuilding method defined on the basis of the Bremen-Cog. It is used (...)

1After its recovery from the Weser River in 1962 just 4 km west of Bremen’s medieval centre, the Bremen-Cog, a late 14th-century ship built according to the “Bremen-Type” shipbuilding method,1 was primarily seen as a technical monument from the heydays of the Hanseatic League. Even though it was recons­tructed within the German Maritime Museum in Bremerhaven, which opened its gates in 1975, it took many years before the construction of the ship was analysed following an archaeo­logical approach.

2During the project entitled “From the North Sea to the Norwegian Sea – Interdisciplinary research on the Hanse” funded by the Leibniz Association at the German Maritime Museum, the ship became the focus of archaeological investigation, when the question about the technical basis of Low German shipwrights for the building of seagoing ships was raised. A generous contribution from the support society of the museum (Förderverein Deutsches Schiffahrtsmuseum e.V.) enabled a comprehensive dendrochronological analysis of the ship’s timbers. This analysis was linked to the ERC-funded project “TIMBER”, within which research on the ship is currently continuing, and was extended with the analyses of the timber quality and an attempted reconstruction of tree morphology patterns based on the outboard planking.

3Apart from still being the best preserved late medieval cargo vessel in northwest Europe more than 50 years after its discovery, and the basis for the technical definition of a particular shipbuilding method, the Bremen ship also played a role in the development of a reliable oak chronology in this part of the continent. While dendrochronology here was in its infancy during the 1960s, the Bremen-Cog turned out to be the first archaeological find of the area whose timber’s provenance could be established on the base of dendrochronology (Haneca et al. 2009, p. 6).

1. Early dendrochronology

4At that point in time there were no reliable chronologies for the North German Plain. While in southern Germany climate signals were distinctly visible in the ring growth pattern of oak trees, the situation in the north proved to be more difficult and at the time of the ship’s discovery there was no reliable ­cover­age of master chronologies.

5Nevertheless, three samples from the ship could be dated on the basis of the, at the time, recently constructed south German master chronology, which was made using more than 1000 analysed dendrochronology samples from upper and lower Franconia and Hesse (fig. 1). From these samples, one provided a bark edge dated to winter 1378/79. The comparison with a site chronology from Ziegenhain/Kassel (Bauch 1969, p. 125), and a historical investigation by J. Delfs (1952), which gives evidence for timber rafting on the Weser to Bremen already during the 13th century, narrowed the origin of the dated construction components down to the Weser Uplands about 200 km upstream from Bremen (Liese, Bauch 1965; Bauch 1969). In 2007 the published data on this early analysis was reanalysed on the basis of a widely extended base of chronologies and confirmed dating and provenance for two of the three early dated samples (Daly 2007, p. 115–116).

Fig. 1: Map of Germany with the Weser passing through Bremen and Bremerhaven in the North German Plain

Fig. 1: Map of Germany with the Weser passing through Bremen and Bremerhaven in the North German Plain

The headstreams, Fulda and Werra, unite in the southern part of the Weser Uplands (D). The regions producing the samples for the south German main curve are Upper Franconia (A), Lower Franconia (B) and Hesse (C) with Ziegenhain/Kassel.

(Drawing M. Belasus using a ginkgomaps.com base map)

2. New dendrochronology and investigations into timber morphology

6During archaeological investigations between 2015 and 2018 the opportunity was taken to carry out new comprehensive dendrochronological analyses of the ship’s timbers that benefitted from an extensive coverage of oak chronologies for the North German Plain. This had been developed over a period of more than 50 years since the original dating of the ship (Daly 2017). The aim of these new analyses was to discover information on timber procurement for shipbuilding purposes in Bremen.

7The strategy for the new analysis of the Bremen ship was to sample as much as possible a wide variety of the ship’s ­construction components. Sampling was carried out by coring, using a specially designed borer that bores a 16 mm diameter hole in the timber, to a maximum depth of 30 cm. The core extracted is 6 mm wide. The position of the coring is aimed at maximising the number of rings retained in the core, from the outermost towards the original centre of the tree. The timbers are assembled in the reconstructed ship, so sampling took place during April 19–22, 2016, on in situ timbers chiefly inboard of the hull: a few were also taken from outboard hull planks. Care was taken to spread the sampling around the hull so that a representative sample of timbers was achieved.

8Of 78 analysed timbers, 55 were successfully dated (including the three from Bauch’s analysis). Only heartwood was observed in most timbers, and the estimated felling date of the trees from which the samples came was in agreement with Bauch’s estimated building date of the Bremen ship at around AD 1380. In other words, no evidence of later additions to the ship has emerged in this analysis. This underlines the interpretation that this ship was not an old ship when it sank.

9Of all the dated samples, just three have sapwood preserved. One of the timbers sampled by Bauch had complete sapwood to bark edge, but this was not the case for any timbers examined in this analysis.

10Using the correlation between every dated tree-ring curve with every other, the material can be divided into a small number of groups. Averages for each group have been calculated. The groups are a mixture of different components from the ship. No apparent differentiation can be made between the outer planks of the hull and its inner timbers.

11Group 1 is the largest group and it is made from the tree-ring curves of 34 timbers. It is 243 years in full length, but a version of the average version is made with a minimum of five trees in depth, and this is 164 years in length, and covers the period AD 1186 to 1349. The correlation between the average for this group and a wide range of site and master chronologies is mapped in fig. 2A. The group achieves the highest correlation with datasets around the Bremen hinterland. This seems to indicate that the bulk of the timber for building the ship was harvested in the locality of the shipbuilding site.

12Group 2 is made from just ten trees: most are framing timbers, but a pair of ceiling planks are also assigned to this group. The full length of the average for Group 2 is 262 years, but again a reduced version, only including the period where at least five trees are represented, is made and this is 156 years long, covering the period AD 1163 to 1318. As the correlation between this group and site and master chronologies indicates, a source area within the Bremen hinterland is again most likely (fig. 2B). The remaining timbers that have not been assigned to a group are all dated with chronologies for Lower Saxony. To conclude, the provenance analysis of the Bremen ship presents no evidence of long distance transport of timber for its construction, apart from the two earlier analysed timbers by Bauch. Rather, we are seeing that the Bremen ship is chiefly built of timber from the Bremen hinterland.

13The results of the dendrochronological analyses prompted questions about the available timber quality and tree morph­ology in the direct hinterland of a town like Bremen.

Fig. 2: Maps showing the correlation between the two dendrochronological groupings of timbers from the Bremen Ship (A: Group 1; B: group 2)

Fig. 2: Maps showing the correlation between the two dendrochronological groupings of timbers from the Bremen Ship (A: Group 1; B: group 2)

The correlation statistic, Student’s t-test, is plotted using circles. The higher the t-value, the larger is the circle. The average for each group is compared with an extensive tree-ring dataset for oak in Northern Europe. The turquoise circles represent large regional master chronologies while the brown circles represent local site chronologies (See Daly 2007 for a detailed explanation of the method). The river data is from Lehner & Grill (2013, www.hydrosheds.org, accessed 3 March 2020).

(Drawing A. Daly)

14With the aid of structure-from-motion technology (a reflex camera and AGISOFT software), a portable coordinate measuring machine (Faro-Arm) and 3D modelling software (Rhinoceros 3D), three selected planks from the ship were documented and their morphological features (grain direction, tree rings, position of the pith and traces of branches) used to recons­truct a general morphology of the original trees and to allow us to describe their original nature (fig. 3) (Martins 2017; Colson 2017). The results of this attempt at reconstructing trees from planks gave enough information to characterise the trees’ morphology and shed light on their growing conditions. The main characteristic is the number and position of branches, which are distributed all over the trunk. A tree creates branches to have its leaves catch sunlight for photosynthesis. In a dense population of trees the branches would be concentrated in the upper part of the trunk. The lower part would be for the most part branchless, as leaves here would not get enough light to thrive. When a tree grows in a rather open landscape and catches light from all directions it will produce branches all over its trunk, as is the case with the trees used for building the Bremen ship. It is therefore assumed that the trees present in the ship’s outer planking were growing in a rather open landscape. However, the trees were not short but rather tall and wide, offering enough material for the production of planks with an average width of 0.506 m and up to 9.52 m long.

Fig. 3: An example of the reconstruction of a tree’s morphology from a plank

Fig. 3: An example of the reconstruction of a tree’s morphology from a plank

In this example, a portable coordinate-measuring machine (Faro-Arm) was applied to document all the important features where they were accessible (A). The positions of branches and their growing directions are established (B). The shape and circumference of the original tree trunk is estimated (C) and combined with the assumed size of the evident branches at the time of the trees felling (D). From this information, the position of the plank in the original tree is established (E).

(Drawing A. M. Martins)

15Further detailed investigation of the outer planking revealed a poor timber quality. From the detailed descriptions of all surviving timbers by Werner Lahn (1992), it had already become clear that over 50% of the outer hull planking was repaired during the building process. In some cases dead branches (knots) were cut out and a rectangular patch was placed to fill the hole, waterproofed with the aid of moss caulking and secured with clamps. Other repairs were needed on cracks that appeared in the planks. These were repaired using moss caulking and iron clamps. The investigation showed that not all dead branches were patched up and these fell out at a later point. They are evidence of ingrown deadwood, which occurs when it takes a tree too long to incorporate a broken or cut off branch. In other areas, the trees produced clusters of branches, giving evidence of the repeated removal of branches from the same spot. Apart from creating a weakness in the planks’ strength, all branches divert the tree’s grain in a straight trunk. When the tree is converted into about 0.05 m thick planks by means of tangential sawing, the fibres are cut regardless of their direction. The cut fibres create weaknesses, which often cause the planks to crack, and, if they are built into a ship’s construction while the timber is still green (unseasoned), tension is created as they shrink during the drying process. The cracks follow the direction of the fibre.

16That cracks occurring in the outer planking have been repaired is not an unusual observation on tangentially produced planks and this can also be seen in other ship finds, e.g. in Doel 1 or Poel 11. While Doel 1 is built from tangentially sawn oak planks, Poel 11’s planks were produced from tangentially cleft pine trunks (Vermeersch, Haneca 2015; Belasus 2017b). In most cases these cracks appear longitudinal and rather straight and can easily be repaired without greater consequences. In the case of the Bremen ship longitudinal cracks appear to be a minority (fig. 4 A). Other repaired cracks appear curvy, indicating a crooked trunk, or they are straight until they curve away at one end revealing the extension of the planks by cutting well into an adjacent main branch (fig. 4 B). However, the most numerous cracks in the ship’s construction are the most disturbing. These cracks travel diagonally through the planks, in the worst cases cutting the plank in two from the overlapping plank above to the overlapped plank below, reducing the longitudinal strength the strake should have given to the hull structure and remaining inaccessible for proper repair. These cracks are the result of the twisted growth of the oak it was cut from (fig. 4 C). The twisted growth of trees is a phenomenon that can be genetic, however, it is often the result of severe injury to the trees phloem, a thin layer of cells just under the bark, the purpose of which is to ­conduct water and nutrients vertically from the roots up to the branches. When this phloem is cut in one area, the tree is able to prevent the parts of the tree directly above the injury from being deprived of nutrients by creating alternative diagonal cell ­connections to bridge the injury. This causes a ‘re­programming’ of the phloem’s cell structure. Each new layer in the following years will circle diagonally around the trunk. The phloem becomes sapwood and, in the years to follow, hardwood with twisted grain (Kubler 1991). Spiral grain does not necessarily weaken the timber: as long as it is used as a full block and the angle of the twist does not exceed 37°, it can even add strength to the timber (Leelavanichkul, Cherkaev 2004, p. 134). This is, for example, the case with the recovered capstan of the vessel.

Fig. 4: Distribution of different grain courses in the tangentially sawn planks of the Bremen ship and their origins in the tree

Fig. 4: Distribution of different grain courses in the tangentially sawn planks of the Bremen ship and their origins in the tree

A: Planks sawn from the trunk of a straight-grained tree. B: Planks sawn from a crooked tree or a cut that was extended into a main trunk for the production of longer planks. C: Planks cut from trees with spiral grain.

(Graphic: M. Belasus using the 1980 University of Hannover’s/German Maritime Museum’s photogrammetry recording of the ship)

3. Conclusions

17The recently carried out extensive dendrochronological analy­ses of the building timbers of the Bremen-Cog did not change the previous date of a felling in winter 1378/79. Neither did the ­components of the ship, which were successfully analysed before, experience a change in their provenance in the Weser Uplands during the re-examination of their published curves. Nevertheless, now that we have a massively extended dataset of regional oak chronologies for the North German Plain, we can state that all trees in the new analysis have their origin in the hinterland of Bremen, an area that was directly influenced by the town’s interests.

18The trees were far from being ideal for shipbuilding. The reconstruction of their basic morphology with branches all over their trunks from below to above places them in a rather well-lit park-like landscape. Their considerable size, indicated by the lengths and widths within the planks, indicate their location at the edge of a forest or in small patches of woods in the countryside. Ingrown deadwood, indicated by the dead, patched up or fallen out dead branches, confirm that they received no maintenance to ensure quality timber. However, the trees were most likely used during their growth. Clusters of knots confirm repeated removal of branches or twigs, and the evidence of twisted grain in many of the outer planks suggest severe damage to the trunks already early in their life.

19The evidence speaks for woods whose main purpose was not as a source of quality timber but for other purposes, e.g. providing acorns to feed pig herds, as a hunting ground, or for collecting firewood. Why they were finally felled and converted into planks, and if these planks were originally intended for the purpose of shipbuilding is still unknown.

Acknowledgements

20
The research presented in this paper is a result of activities within three research projects. Daly and Belasus are currently funded by the project entitled “Northern Europe’s timber resource – chronology, origin and exploitation (TIMBER)” (2016–2021), which has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No. 677152). Belasus was previously funded by the project “From the North Sea to the Norwegian Sea – Interdisciplinary research on the Hanse” (2015–2018), which received financing through the Leibniz Competition 2015, funding line 4: “Promoting women for academic leadership positions” by the Leibniz Foundation, Germany, during which the dendrochronological analysis was generously financed by the support association of the German Maritime Museum (Förderverein Deutsches Schiffahrtsmuseum e.V.). Martins was funded through the project “ForSeaDiscovery”, supported by the People Programme (Marie Curie Actions) of the European Union’s Seventh Framework Programme (FP72007–2013) under REA grant agreement no. PITN-GA 2013‑607545. The authors are very grateful to the German Maritime Museum for their generosity in allowing us to examine the Bremen ship.

Haut de page

Bibliographie

Bauch W.
1969 “Die Bauzeit der Bremer Kogge”, in Die Wittheit zu Bremen (eds), Die Bremer Hanse-Kogge: ein Schlüssel zur Schiffahrtsgeschichte; Fund, Konservierung, Forschung, Bremen, Röver, p. 123–126.

Bauch J., Liese W., Eckstein D.
1967 “Über die Altersbestimmung von Eichenholz in Norddeutschland mit Hilfe der Dendrochronologie,” Holz als Roh- und Werkstoff 25, 8, p. 285–291.

Belasus M.
2017a “Historical Ship Archaeology in the Shadow of Historism and Nationalism. A German Perspective,” in A. Brooks, N. Mehler (eds), The Country where my Heart is: Historical Archaeologies of Nationalism and National Identity, Gainesville, University Press of Florida, p. 222–241.

Belasus M.
2017b “Connecting maritime landscapes. Or early modern news from two former ‘Baltic Cogs’ (Mecklenburg-West Pomerania, Germany),” in J. Gawronski, A. van Holk, J. Schokkenbroek (eds), Ships and Maritime Landscapes. Proceedings of the Thirteenth International Symposium on Boat and Ship Archaeology (ISBSA 13), Amsterdam 2012, Eelde, Barhuis Publishing, p. 179–184.

Colson A.
2017 “Digital Documentation of Ships in Cultural Heritage: A European Review,” International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XLII-2/W5, p. 129–134.

Daly A.
2007 Timber, Trade and Tree-rings. A dendrochronological analysis of structural oak timber in Northern Europe, c. AD 1000 to c. AD 1650, PhD thesis, University of Southern Denmark.

Daly A.
2017 “Dendrochronological analysis of timbers from the Bremen Ship, Germany,” Dendro.dk Report 46, Copenhagen (on line: at the German Maritime Museum, Bremerhaven, Germany).

Delfs J.
1952 Die Floesserei im Stromgebiet der Weser, Bremen, Dorn, Schriften der wirtschaftswissenschaftlichen Gesellschaft zum Studium Niedersachsens e.V. 34.

Haneca K., Čufar K., Beeckman H.
2009 “Oaks, tree-rings and wooden cultural heritage: a review of the main characteristics and applications of oak dendrochronology in Europe,” Journal of Archaeological Science 36, p. 1–11.

Kubler H.
1991 “Function of spiral grain in trees,” Trees 5, p. 125–135.

Lahn W.
1992 Die Bremer Kogge I, Hamburg, Ernst Kabel Verlag, Schriften des Deutschen Schiffahrtsmuseums 30.

Leelavanichkul S., Cherkaev A.
2004 “Why the grain in tree trunks spirals: a mechanical perspective,” Structural and Multidisciplinary Optimization 28, p. 127–135.

Lehner B., Grill G.
2013 “Global river hydrography and network routing: baseline data and new approaches to study the world’s large river systems,” Hydrological Processes 27.15, p. 2171–2186. On line: www.hydrosheds.org.

Liese W., Bauch J.
1965 “Das Alter der Bremer-Kogge,” Bremisches Jahrbuch, p. 38–42.

Martins A. M.
2017 “Reconstructing Trees from Ship Timbers: Data analysis and schematics,” in R. Valera Gomes,K. Trápaga Monch (eds), Árvores, Barcos e Homes na Península Ibérica (Séculos XVI–XVIII), Zaragoza, Instituto de Arqueologia e Paleociêcias da U.N.L, p. 63–76.

Vermeersch J., Haneca K.
2015 “Construction Features of Doel 1, a 14th-Century Cog found in Flanders,” IJNA 44.1, p. 111–131.

Haut de page

Notes

1  “Bremen-Type” refers to the shipbuilding method defined on the basis of the Bremen-Cog. It is used here to avoid the term “Cog/Kogge” as an archaeological/technical definition, which can lead to contradictions and misunderstandings (Belasus 2017a).

Haut de page

Table des illustrations

Titre Fig. 1: Map of Germany with the Weser passing through Bremen and Bremerhaven in the North German Plain
Légende The headstreams, Fulda and Werra, unite in the southern part of the Weser Uplands (D). The regions producing the samples for the south German main curve are Upper Franconia (A), Lower Franconia (B) and Hesse (C) with Ziegenhain/Kassel.
Crédits (Drawing M. Belasus using a ginkgomaps.com base map)
URL http://journals.openedition.org/archaeonautica/docannexe/image/1860/img-1.jpg
Fichier image/jpeg, 2,9M
Titre Fig. 2: Maps showing the correlation between the two dendrochronological groupings of timbers from the Bremen Ship (A: Group 1; B: group 2)
Légende The correlation statistic, Student’s t-test, is plotted using circles. The higher the t-value, the larger is the circle. The average for each group is compared with an extensive tree-ring dataset for oak in Northern Europe. The turquoise circles represent large regional master chronologies while the brown circles represent local site chronologies (See Daly 2007 for a detailed explanation of the method). The river data is from Lehner & Grill (2013, www.hydrosheds.org, accessed 3 March 2020).
Crédits (Drawing A. Daly)
URL http://journals.openedition.org/archaeonautica/docannexe/image/1860/img-2.jpg
Fichier image/jpeg, 3,1M
Titre Fig. 3: An example of the reconstruction of a tree’s morphology from a plank
Légende In this example, a portable coordinate-measuring machine (Faro-Arm) was applied to document all the important features where they were accessible (A). The positions of branches and their growing directions are established (B). The shape and circumference of the original tree trunk is estimated (C) and combined with the assumed size of the evident branches at the time of the trees felling (D). From this information, the position of the plank in the original tree is established (E).
Crédits (Drawing A. M. Martins)
URL http://journals.openedition.org/archaeonautica/docannexe/image/1860/img-3.jpg
Fichier image/jpeg, 336k
Titre Fig. 4: Distribution of different grain courses in the tangentially sawn planks of the Bremen ship and their origins in the tree
Légende A: Planks sawn from the trunk of a straight-grained tree. B: Planks sawn from a crooked tree or a cut that was extended into a main trunk for the production of longer planks. C: Planks cut from trees with spiral grain.
Crédits (Graphic: M. Belasus using the 1980 University of Hannover’s/German Maritime Museum’s photogrammetry recording of the ship)
URL http://journals.openedition.org/archaeonautica/docannexe/image/1860/img-4.jpg
Fichier image/jpeg, 1,2M
Haut de page

Pour citer cet article

Référence papier

Mike Belasus, Aoife Daly et Adolfo Miguel Martins, « The Bremen-Cog: environmental issues »Archaeonautica, 21 | 2021, 257-261.

Référence électronique

Mike Belasus, Aoife Daly et Adolfo Miguel Martins, « The Bremen-Cog: environmental issues »Archaeonautica [En ligne], 21 | 2021, mis en ligne le 01 juillet 2022, consulté le 09 décembre 2023. URL : http://journals.openedition.org/archaeonautica/1860 ; DOI : https://doi.org/10.4000/archaeonautica.1860

Haut de page

Auteurs

Mike Belasus

Saxo-Institute, University of Copenhagen

Articles du même auteur

Aoife Daly

Associate Professor, Saxo Institute, University of Copenhagen

Articles du même auteur

Adolfo Miguel Martins

University of Wales Trinity Saint David

Haut de page

Droits d’auteur

CC-BY-4.0

Le texte seul est utilisable sous licence CC BY 4.0. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.

Haut de page
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search