Final Report Summary - AMPT (Ancient maritime pitch and tar: a multidisciplinary study of sources, technology and preservation)
The AMPT project has applied and developed experimental simulation and analytical chemistry techniques to research the sources, technology and preservation of tars and pitches from maritime archaeological contexts. Tars and pitches are black sticky substances obtained by the destructive distillation either of wood, bark, or tapped resin from soft and hardwood trees such as pine, spruce and birch. These products were widely used in the past as waterproofing agents and timber preservatives, especially in maritime contexts. In the medieval period, their role in shipbuilding and maintenance led them to acquire crucial strategic and political importance for the developing European seafaring economies and naval fleets.
Tars and pitches from the Newport Ship provided a focal archaeological assemblage for the research. The Newport Ship (discovered in 2002 on the bank of the river Usk in Newport, Wales) is the most complete example of a fifteenth century clinker-built vessel ever found in the United Kingdom (UK). The dating evidence indicates that it was constructed after AD 1445 and came to rest in Newport soon after AD 1468, while associated finds suggest contacts with the Iberian Peninsula. This unique assemblage presented an opportunity to investigate ship building techniques and maintenance strategies alongside examination of differential preservation due to burial environment. Interpretation of the Newport material was supported by the analysis of comparative material from other archaeological sites enabling the impact of both intra and intersite variability in burial environment to be compared. Interpretation of the archaeological data was further supported by analysis of experimentally produced and artificially aged tars.
The key aims of the project were:
(a) to understand the range and sequence of timber waterproofing / preservative used on the Newport Ship, both in its original construction and subsequent repairs;
(b) to model degradation pathways of tar and pitch, from the Newport Ship and the comparative archaeological material and using simulated degradation in the laboratory.
From the Newport Ship, 88 samples were analysed by gas chromatography-mass spectrometry (GC-MS). These were selected to provide:
(i) transects across the ship (port-starboard and bow-stern) so that chemical variability in tars could be mapped over the vessel with respect to building sequence and burial conditions;
(ii) multiple samples from a single strake to investigate inherent variability in tar from a single vessel / burial locale;
(iii) samples from areas of repair on the vessel (different tarring events).
Comparative GC-MS analysis was carried out on:
(a) 19 tarry samples of caulking material from 6 out of 8 wrecks (11th and 15th century) excavated between 1996 and 1997 in the Roskilde harbour in Denmark;
(b) 20 tarry samples of caulking material from the 15th century clinker-built wreck Mönchgut 92 excavated in 2010, at the entrance of the Bay of Greifswald, Germany;
(c) 19 samples of archaeological tar from the wreck Arles-Rhône 3, a Gallo-Roman (1st century AD) river barge discovered in 2004 on the bank of the river Rhône, France;
(d) 10 samples from the Doel cogs, two Hanseatic (14th century) flat-bottomed clinker vessels discovered in 2000 and in 2002 in the banks of the river Scheldt in Belgium;
(e) 3 tar samples from the 'pitch' cargo of a Gallo-Roman trading vessel discovered in 1982 off the coast of Guernsey;
(f) samples of black tarry materials from a basket excavated in Oslo harbour and from two barrels excavated respectively in Skarvøy, Norway and in Østersjøen, Sweden, all now in the collection of the Norwegian Maritime Museum, Oslo;
(g) samples of black / brown interior surface deposits on various Greek trade amphorae from the Eastern Mediterranean in the collections of the British Museum;
(h) samples of lumps of tarry-looking material from the 7th century AD ship burial at Sutton Hoo (Suffolk, UK).
A further 24 samples from the Newport Ship were analysed for bulk light stable isotopic ratios (dD and d13C) comparing ratios for original and repair tars. Isotopic analysis was also carried out on samples from the Roskilde, Mönchgut, Guernsey, Doel and Arles-Rhône wrecks and the Sutton Hoo ship burial.
Samples were collected from traditional tar making in Norway and also manufactured at small scale in the laboratory. The Norwegian tar was used in artificial degradation experiments which aimed to simulate different archaeological environments in the laboratory in order to examine how the chemistry of the tar is altered after deposition.
All of the samples from the Newport Ship and the comparative wreck sites were found to be conifer derived and manufactured at temperatures of 300 - 350 degrees Celsius. Amongst the Newport Ship samples no pattern of use related to the building of the ship could be identified from the tar chemistry. The high variation in the composition of tars from a single strake demonstrates that such mapping is not possible. The repairs, however, were distinctive in composition and statistical analysis of the data allows the samples to be identified as three groups, two distinct repair groups plus the original construction group. Isotopic analysis has shown that the repair tars are made from material originating at a different latitude to the tars from the body of the vessel. Distinctive degradation markers seen in the Newport Ship tars were also observed in tars from other wrecks examined and are probably attributable to specific environmental conditions prevailing in the burial environment.
The results of the study set new parameters for the potential and limitations of tar and pitch analysis as a means to understand past technology, behaviours and object histories. The results show that after archaeological degradation, tar and pitch chemistries can be highly variable, even within a single production / use event, so that precise chemical indicators for production conditions or use related chronologies are not viable. Nevertheless, statistical analysis is shown to be a helpful tool in making broad observations given a large enough dataset and this should encourage future studies to incorporate larger numbers of samples than has been usual for previous archaeological studies of these materials. Some distinctive molecular characteristics have been linked to particular burial conditions and this will have implications for interpretation of material from similar assemblages in future. The isotopic analysis of the tars yielded significant results for the Newport Ship material that lend an important new dimension to the interpretation of the repairs on the ship and moreover demonstrate the utility of this approach to tracing mobility of these materials in medieval Europe.
Pauline Burger (AMPT project fellow) (burgerpauline@gmail.com via e-mail)
Rebecca Stacey (British Museum) (rstacey@thebritishmuseum.ac.uk via e-mail)
Nigel Nayling (Newport Ship) (n.nayling@tsd.ac.uk via e-mail)
Project website: http://www.britishmuseum.org/research/research_projects/all_current_projects/ancient_maritime_pitch_and_tar.aspx
Tars and pitches from the Newport Ship provided a focal archaeological assemblage for the research. The Newport Ship (discovered in 2002 on the bank of the river Usk in Newport, Wales) is the most complete example of a fifteenth century clinker-built vessel ever found in the United Kingdom (UK). The dating evidence indicates that it was constructed after AD 1445 and came to rest in Newport soon after AD 1468, while associated finds suggest contacts with the Iberian Peninsula. This unique assemblage presented an opportunity to investigate ship building techniques and maintenance strategies alongside examination of differential preservation due to burial environment. Interpretation of the Newport material was supported by the analysis of comparative material from other archaeological sites enabling the impact of both intra and intersite variability in burial environment to be compared. Interpretation of the archaeological data was further supported by analysis of experimentally produced and artificially aged tars.
The key aims of the project were:
(a) to understand the range and sequence of timber waterproofing / preservative used on the Newport Ship, both in its original construction and subsequent repairs;
(b) to model degradation pathways of tar and pitch, from the Newport Ship and the comparative archaeological material and using simulated degradation in the laboratory.
From the Newport Ship, 88 samples were analysed by gas chromatography-mass spectrometry (GC-MS). These were selected to provide:
(i) transects across the ship (port-starboard and bow-stern) so that chemical variability in tars could be mapped over the vessel with respect to building sequence and burial conditions;
(ii) multiple samples from a single strake to investigate inherent variability in tar from a single vessel / burial locale;
(iii) samples from areas of repair on the vessel (different tarring events).
Comparative GC-MS analysis was carried out on:
(a) 19 tarry samples of caulking material from 6 out of 8 wrecks (11th and 15th century) excavated between 1996 and 1997 in the Roskilde harbour in Denmark;
(b) 20 tarry samples of caulking material from the 15th century clinker-built wreck Mönchgut 92 excavated in 2010, at the entrance of the Bay of Greifswald, Germany;
(c) 19 samples of archaeological tar from the wreck Arles-Rhône 3, a Gallo-Roman (1st century AD) river barge discovered in 2004 on the bank of the river Rhône, France;
(d) 10 samples from the Doel cogs, two Hanseatic (14th century) flat-bottomed clinker vessels discovered in 2000 and in 2002 in the banks of the river Scheldt in Belgium;
(e) 3 tar samples from the 'pitch' cargo of a Gallo-Roman trading vessel discovered in 1982 off the coast of Guernsey;
(f) samples of black tarry materials from a basket excavated in Oslo harbour and from two barrels excavated respectively in Skarvøy, Norway and in Østersjøen, Sweden, all now in the collection of the Norwegian Maritime Museum, Oslo;
(g) samples of black / brown interior surface deposits on various Greek trade amphorae from the Eastern Mediterranean in the collections of the British Museum;
(h) samples of lumps of tarry-looking material from the 7th century AD ship burial at Sutton Hoo (Suffolk, UK).
A further 24 samples from the Newport Ship were analysed for bulk light stable isotopic ratios (dD and d13C) comparing ratios for original and repair tars. Isotopic analysis was also carried out on samples from the Roskilde, Mönchgut, Guernsey, Doel and Arles-Rhône wrecks and the Sutton Hoo ship burial.
Samples were collected from traditional tar making in Norway and also manufactured at small scale in the laboratory. The Norwegian tar was used in artificial degradation experiments which aimed to simulate different archaeological environments in the laboratory in order to examine how the chemistry of the tar is altered after deposition.
All of the samples from the Newport Ship and the comparative wreck sites were found to be conifer derived and manufactured at temperatures of 300 - 350 degrees Celsius. Amongst the Newport Ship samples no pattern of use related to the building of the ship could be identified from the tar chemistry. The high variation in the composition of tars from a single strake demonstrates that such mapping is not possible. The repairs, however, were distinctive in composition and statistical analysis of the data allows the samples to be identified as three groups, two distinct repair groups plus the original construction group. Isotopic analysis has shown that the repair tars are made from material originating at a different latitude to the tars from the body of the vessel. Distinctive degradation markers seen in the Newport Ship tars were also observed in tars from other wrecks examined and are probably attributable to specific environmental conditions prevailing in the burial environment.
The results of the study set new parameters for the potential and limitations of tar and pitch analysis as a means to understand past technology, behaviours and object histories. The results show that after archaeological degradation, tar and pitch chemistries can be highly variable, even within a single production / use event, so that precise chemical indicators for production conditions or use related chronologies are not viable. Nevertheless, statistical analysis is shown to be a helpful tool in making broad observations given a large enough dataset and this should encourage future studies to incorporate larger numbers of samples than has been usual for previous archaeological studies of these materials. Some distinctive molecular characteristics have been linked to particular burial conditions and this will have implications for interpretation of material from similar assemblages in future. The isotopic analysis of the tars yielded significant results for the Newport Ship material that lend an important new dimension to the interpretation of the repairs on the ship and moreover demonstrate the utility of this approach to tracing mobility of these materials in medieval Europe.
Pauline Burger (AMPT project fellow) (burgerpauline@gmail.com via e-mail)
Rebecca Stacey (British Museum) (rstacey@thebritishmuseum.ac.uk via e-mail)
Nigel Nayling (Newport Ship) (n.nayling@tsd.ac.uk via e-mail)
Project website: http://www.britishmuseum.org/research/research_projects/all_current_projects/ancient_maritime_pitch_and_tar.aspx