OITIS Field Repor t no. 1 A Report on Preliminary Work on Papa Stour, Shetland 2010 Jill Campbell, Endre Elvestad, Mark Gardiner and Natascha Mehler OITIS (The Operation of International Trade in Iceland and Shetland) is a joint project between the RömischGermanische Kommission of the German Archaeological Institute and Queen’s University Belfast to examine trade and ishing in Iceland and Shetland during the period from the 15th to early 18th centuries. This interdisciplinary study draws primarily upon evidence from archaeological and historical work. The present report series makes available the results of ield survey and investigation in advance of the synthesis and publication of the complete project. They are interim reports, contain provisional results, and are therefore subject to revision and modiication. Front cover illustration: Surveying at Hamnavoe Jill Campbell, School of Geography, Archaeology, and Palaeoecology, Queen´s University Belfast (jcampbell66@qub.ac.uk) Endre Elvestad, Stavanger Sjøfartsmuseet (endre.elvestad@stavanger.museum.no) Dr Mark Gardiner, School of Geography, Archaeology, and Palaeoecology, Queen´s University Belfast (m.gardiner@qub.ac.uk) Dr Natascha Mehler, Institut für Ur- und Frühgeschichte, Universität Wien (natascha.mehler@univie.ac.at) 2 Contents Introduction 4 Aims and methods 4 Historical evidence 4 Place-names 5 Oral tradition 5 Hamna Voe Archaeology Maritime background: Hamna Voe Maritime background: Culla Voe Summary of maritime archaeology Site Survey Hamnavoe Description of building Interpretation of the building sequence Tulloch´s Böd, Culla Voe Description of building Interpretation of building sequence 5 6 6 8 8 9 9 9 11 13 13 14 Conclusion 15 Acknowledgements 16 References 16 3 Introduction This is a report on preliminary visits to Papa Stour undertaken in May 2010 as part of a wider ield survey of Hanseatic trading sites in Shetland. The visits were intended to evaluate the evidence for the survival of remains relating to Hanseatic trade on Papa Stour. The importance of the island as a base for German traders and as a site of imported goods was indicated by the only major excavation to have taken place on the island, at The Biggings, where substantial quantities of imported pottery were found. Those include Paffrath Ware of the 12th and early 13th centuries, north German and South Scandinavia redwares and stonewares from Lower Saxony (Crawford and Ballin Smith 1999, 156-67). Pottery was not a main item of trade, but provides some indication of trading activity on the island. While The Biggings relects the pottery at a consumer site, no trading sites had been excavated in Shetland before the start of work of the OITIS project. Papa Stour lies west of mainland Shetland and would have been passed by ships travelling along that coast of the island group. It has long served as both as a base for ishing and as a centre for agriculture. It has a number of voes (or bays) which have been used as ishing bases at various times, but in the late 19th century the main activity was concentrated around West Voe where William Adie constructed a pier. The island is sharply divided into two halves. The better land lies to the east of Hill Dyke, while the land to the west is extremely poor, in places lacking any soil at all. The sites surveyed in the preliminary work both lie on the western half. Aims and methods The visits undertaken in May 2010 were to evaluate the potential of the island for future study of Hanseatic trade. The initial survey looked particularly at the buildings which, though it was apparent from irst inspection were not evidently 16th or 17th century in date, might include fabric of buildings from that period. The survey also considered the surroundings of the buildings where other evidence of Hanseatic trade might survive, and a more general survey was undertaken of the inlets as anchorages and for landing boats. The work on the maritime archaeology was undertaken by Endre Elvestad. Two buildings were recorded photogrammetrically and an analytical study was made of the building fabric. The buildings were photographed with targets which surveyed using a Leica total station. The photographs were corrected to create orthophotos using PhoToPlan and the fabric of selected 4 elevations drawn up in AutoCAD by Jill Campbell. The immediate surroundings of the buildings were also surveyed using the total station, drawn up in AutoCAD by Mark Gardiner and these plans were correlated with satellite photographs which were brought into AutoCAD. This allowed a record of wider area to be generated. The plans for this report have been drawn up by Libby Mulqueeney from the CAD surveys. Historical evidence This report considers two areas which were examined in the preliminary survey: Hamnavoe farm in Hamna Voe and Tulloch’s Böd in Culla Voe. Both settlements lie on the western half of the island and beyond the Town Dike which encloses the better land of the island. Crawford and Ballin Smith (1999, 40) suggest that a settlement was established at Hamnavoe for commerce rather than for agriculture because the soils are fairly poor, although the anchorage is the safest in the island. The historical evidence for the early presence of German merchants on Papa Stour is better than for most other places in Shetland. A deed of 1452 records the sale of land on the island by Henrik Soost who was evidently a Lübeck merchant resident in Bergen in the mid-15th century (SD 1, no. 22). Crawford and Ballin Smith (1999, 39) have speculated that he might have acquired the land in default of payment of debts by a Shetlander, or because it was in some other way connected with trade. Rolf Hammel-Kiesow at the Archiv der Hansestadt Lübeck has provided the following information: during 1441 and 1473 the Oberstadtbücher of Lübeck mention several times a Hinricus Soost / Hinricus Susato / Hinricus Soest, but it is uncertain if this is all the same person. He was most likely a Bergenfarer and Soost is reported to have bought and sold some houses in Lübeck. Unfortunately, the sources do not mention any journeys made by him to Shetland. A list of trading sites in Shetland of 1563 mentions the harbour of “Sunte Mangens Eilandt“ and the harbour in “Papposunt” (SD 1, no. 140). The identiication of these poses some problems. The irst, St Magnus Island, is the older name for Papa Stour. The second is Papa Sound which might be the sound to the south of Papa Stour and therefore the harbour might be on the mainland, perhaps near Sandness as Brian Smith has suggested (Crawford and Ballin Smith 1999, 39n). Alternatively, it may be a reference to West Voe on which Papil is situated and where there were certainly “Dutchmen” in the 1650s (Smith 1984, 14). The reference to “Dutch- men” almost certainly refers to Germans. It is clear from written evidence that Shetlanders could or did not distinguish a Dutch from a German trader, confusingly calling German merchants “Dutch” (e. g. Mehler 2009, 95f.). We know little more about the presence of German merchants on Papa Stour until 1602 when Orne Mair, a Hamburg merchant at Gluss in Northmavine, sought the removal of Simon “Harratstay”, a Hamburg merchant from Gunnister Voe, Northmavine, to Papa Stour on the grounds that there was too much competition for the supplies of ish (Court Book of Shetland 1602-1604, 16-17, 9394). Simon Harriestede, elder and younger, can be traced in the Hamburg records sailing to Shetland from 1588 when the former was paying for goods taken in the ship of Hans Meier (possibly the same person as Orne Mair mentioned above). Simon Harriestede – father and son – appear to have sailed most years to Shetland until 1626. Whether they continued to be based at Gunnister for long after 1602 is uncertain. However, by 1626 the younger Simon Harriestede seems to have transferred his activities to Papa Stour (see also Gardiner and Mehler 2010). It is recorded exceptionally that year that the Jacob Surman’s ship, in which he had an interest, had gone to “Papoien” (Staatsarchiv Hamburg 1988; see also Smith 1984, 16). In 1603 there is a record of an affray at the Dutch booth and “frie coupsta” of Papa when two men attacked another (Court Book of Shetland 1602-1604, 77). Papa Stour continued to be used as a trading base by Germans until the later 17th century (Smith 1984, 14). Hibbert writing in 1822 commented that The kirk, a neat structure, and not very old, is situated near the centre of the island. A merchant from Holland, gave to it, about a century ago, a bell, a silver cup for the administration of the sacrament, and a curious copper bason for holding water in baptisms, in which appeared several religious emblems (Hibbert 1822, 552-53). Place-names George Peterson drew our attention to the placename of the rock, “Herrit’s Baa” near the mouth of Culla Voe, which he tentatively connects with Simon Harriestede who certainly was present in Papa Stour. The other place-name of note is “Dutch Loch” which lies to the north of Hamna Voe. There is a further place-name of interest, not discussed by Peterson, “da Dutch Loch” which lies between West Voe and Housa Voe and may be related to German merchants operating from one or both places. Oral tradition Hamna Voe George Peterson discussed the places associated with German or “Dutch” traders on our visit. We have augmented our record of his comments with the account he gives in his book, The Coastal Placenames of Papa Stour (Peterson 1993) which incorporates both oral tradition and historical research. He suggested to us on our visits that the Fore Wick and Housa Voe on the south and east of the island were poor harbours because the sand in the bays provides insecure hold for the anchors of ships. The tradition is that the main harbours used by the German merchants were Hamna Voe in the south and Culla Voe in the north. The site on Hamna Voe may have been the farm at Hamnavoe which was occupied by Gabrielson in 1671 when a list of Mowat tenants was drawn up. It was refurbished in the 1850s, but may have been the site of a Hanse building which had been improved. Peterson (1993, 13) has traced the history of the farm in his book. From about the 1720s it was occupied by the Henry family – Thomas, his son James and his son Robert. It was later used by Laurence Fraser who had served in the Royal Navy. After returning to Shetland in about 1800 he set up as a general merchant at Hamnavoe. The Fraser family were still occupying the house in 1851, but before 1856 it had been bought by the irm of Thomas M. Adie who used the house as a shop and ishing station. Their agent on Papa Stour was Hugh Hughson who was succeeded in 1871 by his son, also called Hugh. The shop was closed in 1898, but the building continued to be occupied until 1904 by William Johnston and later by his daughter Helen Johnston and her husband James Nelson. Hamna Voe is not an ideal harbour. It has a depth of water of four fathoms in the harbour, but only two fathoms at the entrance at low tide (see further below). Nevertheless, it was used by German traders and it is said that the Dutch Loch was used to supply water for the ships anchored in the voe (Illus. 1) Culla Voe The other voe said to be used by German traders was Culla Voe in the north of the island. The site occupied by them is said to have been at Tulloch’s Böd. The name comes from Gilbert Tulloch who came from Northmavine and used the site in the 1830s as a shop. His trade was undercut by the Adies’ operation at Hamna Voe established in the early 1850s and Tulloch subsequently moved to Scalloway (Peterson 1993, 40-41). 5 Illus. 1 Hamna Voe looking south-west to the farm (centre middle distance) and towards the mouth of voe. Tiptans Skerry is just visible in the middle of the voe near the mouth. It has already been noted that the rock off Point Lodge near the entrance of the voe is known as Herrit’s Baa (Peterson 1993, 40). Archaeology Maritime background: Hamna Voe Blaeu’s map of Shetland dated to 1654 gives the place-names, “Hana Vo” [Hamna Voe], “Midgairth”, “Goird” and “Housa”. The same names are repeated by Johannes van Keulen in his map published in Amsterdam in 1730. The map by Herman Moll of 1745 simply repeats the irst and third of these names. The anchorages at Papa Stour were not marked on nautical maps before the later 18th century. Thomas Preston’s New Hydrographic Survey of the Island of Shetland published in 1781 shows Hamna Voe as the only anchorage and noted “8 feetwater going in”. It recorded that “Olis Voe”, Ollie’s Voe or West Voe was for small vessels. The map of the Depot General de la Marine of 1803, evidently based on Preston’s survey, also marks Hamna Voe by an anchor, indicating a safe anchorage. Another map from 1807, the Admiralty Chart of Shetland similarly marks Hamna Voe as an anchorage. The North Sea Pilot from 1827 gives the following description of Papa Stour: off the former lies Papa Stour, a large rocky island, steep-to in every part; several small Voes of 7 and 8 feet water are on its eastern side, and one called Hamna Voe on its western side; there is anchorage within the latter in 8 feet water. The North Sea Pilot from 1847 adds that In Housa Voe, Papa Stour, ships of moderate burden may ind occasional anchorage, or wait for the tide to the southward. According to a later North Sea Pilot the entrance to Hamna Voe was dificult and the anchorage was mainly used by small sloops and boats. 6 The place-name Hamna Voe derives from the Old Norse “Höfn” denoting a harbour and “Vágr” denoting a bay. An interesting aspect of bays named “Hamna Voe” in Shetland, is that they seem to have similar topographical features. The North Sea Pilot (1847, 33) states that the place-name is often found in basin-formed bays with a narrow entrance. Hamna Voe on Papa Stour has these characteristics. The inlet is obstructed by Tiptans Skerry, a shallow bank that divides the inlet into two possible entrances. There are several cairns surrounding the bay. On the highest point of the Hill of Feilie facing the ocean towards west and south, there is a burial mound with a newer cairn in the middle (Illus. 2). In the eastern end of the bay there are two burial mounds with newer cairns in the middle and several other stone structures nearby. The protected bay surrounded by burial mounds intended to be seen from the waterways, indicates that the anchorage was frequented in prehistoric periods. The connection between burial mounds and the sea routes is exempliied in the following quote from Beowulf: The Geats built a mound then, in ten days, high and broad on the hill, a beacon for the warrior widely seen by sailors. Many meads are recorded in Shetland. A mead is a term derived from Old Norse, “mið”, that is a navigational device made by a straight sightline through two ixed points. The meads were used to ind an exact position on ishing grounds or a safe line into anchorages or through fairways free from sunken rocks or skerries (Melchers 2005). It is interesting that the cairns surrounding Hamna Voe seem to have been used in a similar navigational system. The cairn at the top of Hill of Feilie might have been used as a landmark, signifying the location of the harbour from a certain distance at sea. One of the cairns on the eastern ridge was used Illus. 2 The cairn at Hill of Feilie. Illus. 3 The cairn that might serve as a mead. in a mead, “Cairn over Yellow Hammer”, meaning a straight line from the cairn through Yellow Hammer, that is a cliff close to Tiptans Skerry. The mead provides a safe entrance to Hamna Voe between the Tiptans Skerry and the southern shoreline. The meads used in navigation have very long traditions, and the mead at Hamna Voe is dificult to date (Illus. 3). The larger collection of noosts (the Shetland term for boat shelters) on the southern shore of the bay indicates that the Voe was used until recently (Illus. 4, 5), and the mead might have been used by the locals. On the other hand it is likely that the need for a precise navigational system must be greater for foreign seafarers that did not have the precise knowledge of the local topography. It is therefore possible that the mead also was used by Hanseatic merchants in navigation towards Hamna Voe. In addition to the noosts and cairns, another recorded maritime structure was a pier close to the settlement of Hamnavoe (see also below; Illus. 6). The pier is quite eroded and it was dificult to determine the original extent. Since there was no other structures related to it, such as noosts or cleared landing places, the pier might have been used for loading and unloading goods, not necessarily by ships, but by smaller vessels, because the water depths outside the pier seems to shallow for larger vessels. The position of the pier is not accidental. The place is well protected against swell from the sea and there is a short distance to the deeper part of the bay. A position further towards north-west would not pro- Illus. 4 Noosts in Hamna Voe. 7 vide suficient sailing depth, even for smaller boats, and the shore south of the settlement is exposed to wind and waves from south-west. The place-name Dutch Loch to the north of Hamna Voe is a good indication of Hanseatic maritime activity. Supplies of fresh water were necessary for offshore sailing ships, and an important attribute to a high quality anchorage. The naming of the lake after the German merchants indicates that the lake was an important source of water for trading ships, and perhaps not only for those trading in Hamna Voe. Illus. 5 Noosts shown on the east side of Hamna Voe on the Ordnance Survey 25-inch map surveyed 1877. Illus. 6. The pier in Hamna Voe. Illus. 7 Noosts in Culla Voe. 8 Maritime background: Culla Voe In Culla Voe there is another site with a possible booth. Culla Voe is not described in the pilots or marked on older sea maps. The reason is probably a dificult entrance, shallow water in the protected part and the possibility of swell with northerly winds. Despite that, there are many remains of old noosts on the south-east shore in the bottom of the bay (Illus. 7). The noosts seem to be of different generations with newer and smaller ones inside older and larger ones. The noosts might belong to different farms from those in Hamna Voe, but it is also possible that the same farm had noosts or landing places in alternative bays that made it easier to land a boat under different weather condition. There were no visible structures on the shore close to the Tulloch’s Böd in Culla Voe (Illus. 8). The shore close to the houses is steep and stony, but some 100 metres to the south there is a lat beach with gravel that is excellent for landing boats. There were no remains of seamarks or cairns in the entrance to Culla Voe. Summary of maritime archaeology As the pilots and the remains of noosts indicate, both Hamna Voe and Culla Voe are good harbours for smaller and rowed boats. The inner part of the bays, where we ind the noosts, are well protected from waves from any direction. They are probably the part of a tradition going back to the prehistoric period of placing the noosts at the same place. For larger vessels these anchorages are dificult. The narrow and exposed inlets make navigation into the bays problematic, especially during contrary winds. Leaving the anchorage might be dificult for the same reasons. As stated in the pilots the bays are quite shallow with a water depth around 8 feet or 2.6 metres, that is somewhat shallow for larger trading vessels. This aspect might indicate that the ships belonging to the Hanseatic merchants were not of the largest. Illus. 8 The settlement in Culla Voe. The possible landing places are to the right in the picture. Site Survey Hamnavoe The preliminary survey concentrated on the two areas drawn to our attention by George Peterson – Hamna Voe and Culla Voe. The rooless and now ruinous farmhouse at Hamnavoe is said to have been on the site of a booth occupied by “Dutch” merchants. The house was last occupied in 1909 (Davidson and Carter 1998, 831) (Illus. 9 and 10). Description of building The farmhouse at Hamna Voe is largely of 19thcentury date: the window and door lintels have drilled quarrying-holes made for splitting the stone. If we take it that the building is aligned north-south, then the east (or front) wall is largely of single-period construction with a central doorway and two windows set symmetrically either side (Illus. 11). The wall is built of coursed stone roughly squared. The windows both have internally splayed jambs Illus. 9 Hamnavoe looking south-east at the rear of the farm and towards the voe. The farmhouse is on the left and an byre in the centre. 9 on the sides adjoining the gable walls. The window jambs incorporate brick in their construction. The front wall is bonded to the north and south gables. The south gable wall is similarly built of squared, coursed stone with large side-alternate quoins. On the inner face the south gable incorporates hearths at ground and irst loor level. The jambs of the lower hearth are of stone, but those of the upper hearth are entirely built of brick. The north gable is partly ruinous and the lower part is buried internally under rubble. There is some evidence that it too incorporated a hearth, certainly at ground-loor level. There are traces on the external face of the line of a shallow-pitch roof of adjoining building, probably of recent date. The west wall has a single doorway which has in recent years been adapted to take a feeding trough. The wall is bonded at the ends to the gables, but there are clear breaks in the coursing which suggest it pre-dates the 19th-century building activity. The break at the south end of the west wall is marked by a crack which has been patched at some date (Illus. 12). A similar break in coursing, though not marked by a structural failure, is evident close to the quoins at the north end of the same wall. There is also clear evidence of patching around the doorway, suggesting that this was a later opening. It seems to have been cut through the site of a former window, the south jamb of which can be traced on the internal face, though not on the external face. In summary, the west wall clearly belongs to an earlier structure which was evidently ruinous or was knocked down when the building was reconstructed and re-ordered in the 19th century. Illus. 10 Hamna Voe from the irst-edition six-inch Ordnance Survey map, surveyed 1878. Hamnavoe farm and Dutch Loch are also shown. The upper part of the side walls of the building above the heads of the windows and door is marked by a clear change in fabric when the walls were raised and joists inserted to form an upper storey. It must have been at this time that the upper ireplace was inserted (Illus. 13). Illus. 11 Front (east) wall of Hamnavoe farmhouse looking west. 10 Illus. 12 South end of exterior of west wall showing break in fabric between the wall and gable end. Illus. 13 Inside of west wall looking west. There is a clear distinction between the larger stones below the joist holes and the stones above. Interpretation of the building sequence There are three clear phases marked by the building work. The west wall of the house is from the earliest phase. The character of this building is uncertain, except that it seems to have had a window which survives in part in the later doorway. At some point in the 19th century the building was largely reconstructed to form a substantial house with large windows divided into two by a central passageway. The height of the building was raised in the third phase when an upper storey was added to occupy the roof space. We can try to tie the structural development of the building into the history of the farm. The substantial investment in re-building the farmhouse with quarried lintels in Phase 2 seems most likely to belong to the possession of the property by Thomas M. Adie who would have had the capital to invest in such work. This would agree with the oral tradition given to us by George Peterson. A mid-19th century date is not out of keeping with the building. Furthermore, the 1846 map of Papa Stour (SRO RHP 83385, reproduced in Crawford 1999, endpapers) appears to show the farmhouse orientated eastwest, in contrast to the present north-south alignment, suggesting that the re-building took place after that time. A building with north-south alignment is shown on the irst-edition six-inch Ordnance Survey map surveyed 1878. The earlier wall incorporated into this house is of uncertain date. It could belong to the occupation by the Fraser family in the early 19th century or might be earlier. Equally, we cannot date the third phase when the height of the building was raised and an extra storey inserted. The house is surrounded by a number of other ruinous buildings, which all appear to be farm buildings with the possible exception of the building to the north-east which adjoins the beach (Illus. 14). This may have been constructed to store ish. Behind the house is an enclosed ield which is higher than the surrounding land and was evidently reserved for growing fodder or perhaps vegetables. The rectangular ield is enclosed by a slight wall, but there are traces of an earlier curved boundary which partly coincides with the later, but encloses a larger area. This boundary bank passes beneath the building on the foreshore and is therefore earlier than it. There are earthwork traces of a further building on the shingle bank to the east of the house. This was probably constructed in connection with the use of the foreshore as an ayre to dry ish. Traces of a stone pier were recorded beyond this. Illus. 14 Survey of buildings and earthworks at Hamnavoe. 11 Illus. 15 Culla Voe looking north. Tulloch’s Böd is in the mid-distance with a sheepfold adjoining it to the right. There is a further sheepfold in the foreground. Illus. 16 Culla Voe from the irst-edition six-inch Ordnance Survey map, surveyed 1877, with Tulloch’s Böd. 12 Illus. 17 Survey of buildings and surroundings of Tulloch’s Böd. Illus. 18 Photogrammetric model of the walls of Tulloch´s Böd (in metres). Tulloch’s Böd, Culla Voe Peterson (1993, 40) notes that there are three beaches in Culla Voe – East Beach, the Mid Beach and the Beach of Tulloch’s Böd. Around these cluster ruined buildings which served as ishermen’s lodges in the past (Illus. 15, 16 and 17). Tulloch’s Böd lies at the north end of the beach to which it gives its name and adjoins a sheepfold which lies immediately to the north-east. The building is a single-storey croft of at least two phases. It is currently rooless and the east wall has been almost entirely removed. It was used latterly as a sheepfold. A number of trees have been planted in the interior in recent years. Description of building The front (south) wall has a single unsplayed window, 0.48m wide, to the west of the doorway (Illus. 18, 19, 20 and 21). It is now blocked. When clearing the window for photography fragments of glass were discovered which are discussed further below. There is a possible window to the east of the doorway, though the evidence for it is uncertain. The doorway was initially constructed 1.4m wide, but was later reduced to 0.75m by adding masonry to the west side. The blocking runs through and is integral with the thickening of the wall on the interior (see below). The lower part of the masonry of the front wall on the west of the doorway is divided into lifts (see Illus. 22). This is not apparent on the upper part of the wall above the height of the window head, which appears to have been built with less care, or to the east of the doorway. The interior face of this wall has been added to the outer face at a later stage to create a wall about 1.3m wide in contrast to the north wall which is only 0.7m wide. However, the surviving top surface of the wall does not relect this, suggesting that the upper part of both the exterior and interior elements of the wall have been rebuilt at that height. There are traces of lifts low in the exterior face of the wall to the east of the doorway. 13 Illus. 19 South (front) wall of Tulloch’s Böd. Illus. 20 North (rear) wall of Tulloch’s Böd. The gable wall on the west of the building contains a hearth. The lower part of the wall is marked by the use of large water-rounded stones and a construction in lifts at the same level of those on the south wall (Illus. 20). No lifts are evident in the upper part of the wall. There is a hearth in the interior face of the wall, which was originally set centrally, but is now closer to the interior of the south wall because of the additional band of masonry added on that side. The survival of the north wall is poor, particularly towards the east end (see Illus. 20). There are no openings in this wall face. There are traces of lifts in the masonry which, at least on the west side, match those in the west gable wall. Only a few stones survive to mark the position of the east wall. The present crosswall is related to a late use of the building as a sheepfold. The building must have either become ruinous or had structural problems as the south face certainly and possibly the west gable end were reduced in height. In Phase 2 an inner wall was added to the south almost doubling its width, the doorway reduced in width, the window blocked and the wall rebuilt (Illus. 22). Interpretation of building sequence The lifts suggest that the lower part of the south, west and north walls of the building, and also the window on the south wall and wider doorway belong to a single phase of construction. The building was originally constructed about 11m long with a doorway set towards the centre and hearth in the west wall. There were no certain internal divisions. 14 Illus. 21 West (gable) wall of Tulloch’s Böd. In its irst phase the building clearly had a domestic use marked by the presence of a hearth, but may also have served a commercial function which is implied by the unusual width of the doorway. The best evidence for the date of this phase is given by fragments of glass found in the window opening. These must precede the second phase when the window was blocked. Three types of glass are clearly distinguishable from the fragments recovered: i) three pieces; heavily weathered; two of similar widths and one slightly thinner; one piece with grozed edges ii) two conjoining piece with green tinge produced by the use of potash from kelp; one has grozed edges iii) three small, clear pieces Dr Hugh Wilmott (Shefield) who examined these suggests that types i) and ii) are 18th-century or earlier, and type iii) is 19th century. The presence of glass of this period is notable because even in the mid-19th century many small houses in Shetland would have had no window openings in the wall. When Tulloch’s Böd was renovated, the window was thought unnecessary and blocked. The glass therefore indicates that the irst phase of the building is 18th century or earlier, and also implies that it is unlikely to have been a simple croft. The different dates and types must relect breakages and replacement of individual panes, and also continuing purchase of glass. If it was a merchant’s store, something implied by the width of the doorway, then the owner may well have had access to glass from trading activities, allowing the import from England or Germany. (We owe most of the discussion in this paragraph to Ian Tait, Shetland Museum). The building had become ruinous by the time it was renovated in the second phase of construction work. The work involved reducing the walls in height and then rebuilding them, as well as constructing an inner wall to support the front (south) wall which must have been deemed to be structurally unsound. It may be signiicant that the south wall was not totally taken down to ground level and rebuilt, although such work would hardly have been more labour than the construction of an inner face to the wall which doubled its width and reduced the internal loor area. The roof must have been removed during this work as the upper part of the walls, at least on the south side and probably in the west gable were rebuilt. The reduction in the width Illus. 22 South wall of Tulloch’s Böd showing the contrast in masonry between the coursed west side (background) and the reduced door and east jamb (foreground). of the door implies that the restored building served a different function, while the closing of the window suggests that its status may have changed. This second phase cannot date to before the 19th century, as glass of that period was found in the irstphase window. It is not simple to relate the building history of Tulloch’s Böd to the historical evidence. It is quite possible that in the 1830s Gilbert Tulloch took over an already existing trading building dating to the 18th century or earlier. After he left, the building may have fallen into disrepair and was subsequently renovated to serve as a ishing booth. This interpretation is not the only one possible of the evidence. There are no further buildings immediately adjoining Tulloch’s Böd, although there are a number of other ishing booths around Culla Voe. A stoneedged platform to the west of the building may be the base of a fuel stack Conclusions None of the surveyed standing structures certainly belongs to the period of German trade, but evidence of earlier structures have been found at both Hamnavoe and Tulloch’s Böd. The wide doorway and the presence of 18th-century or earlier glass at Tulloch’s Böd has suggested that the building may be of some signiicance for the study of trade in Papa Stour. 15 Acknowledgements The work in Shetland was undertaken in conjunction with Stavanger Museum. The 2010 season was funded by grants from the Deutsche Forschungsgemeinschaft and the Römisch-Germanische Kommission of the German Archaeological Institute. Judith Benedix, Ronny Weßling (both Vienna), Dr Nigel Melton (Bradford) and Lindsay Poole (Belfast) were helping us in the ield and we are most grateful to them. We are indebted to Dr Barbara Crawford of the University of St Andrews who helped us in our visits to Papa Stour and supplied us with a copy of The Coastal Place-names of Papa Stour, and to Jane Puckey and George Peterson who met us on the island and gave us the beneit of their considerable local knowledge. Dr Ian Tait of Shetland Museum, Brian Smith of the Shetland Archives, Dr Rolf Hammel-Kiesow of the Archiv der Hansestadt Lübeck and Dr Hugh Wilmott of the University of Shefield all provided valuable information. References Primary Sources Court Book of Shetland 1602-1604, ed. G. Donaldson (1958). Edinburgh. SD1: Shetland Documents 1580-1611, ed. J. H. Ballantyne and B. Smith (1994). Lerwick, Shetland. Staatsarchiv Hamburg (1988) Verzeichis der Hamburger Shetland-(Hitland-)Fahrer 1547-1646 (Typescript). Hamburg: Staatsarchiv Hamburg. Gardiner, M. and Mehler, N (2010), The Hanseatic trading site at Gunnister Voe, Shetland, Post-Medieval Archaeology 44/2 (in press). Hibbert, S. (1822) A Description of the Shetland Islands, Comprising an Account of their Geology, Scenery, Antiquities, and Superstitions. Edinburgh. Mehler, N. (2009), The Perception and Interpretation of Hanseatic Material Culture in the North Atlantic: Problems and Suggestions, Journal of the North Atlantic Special Volume 1, 89-108. Secondary literature Crawford, B. E. and Ballin Smith, B. (1999) The Biggings, Papa Stour, Shetland: The History and Excavation of a Royal Norwegian Farm. Society of Antiquaries of Scotland Monograph Series 15, Edinburgh. Davidson, D. A. and Carter, S. P. (1998) Micromorphological evidence of past agricultural practices in cultivated soils: the impact of a traditional agricultural system on soils in Papa Stour, Shetland, Journal of Archaeological Science 25, 827-38. 16 Melchers, G. (2005) Shetland mead names – some notes on their structure and character, in P. Gammeltoft, C. Hough and D. Waugh (eds), Cultural Contacts in the North Atlantic Region: The Evidence of Names, 157-72. Lerwick, Shetland. Peterson, G. P. S. (1993) The Coastal Place-Names of Papa Stour. Brae, Shetland. Smith, H. D. (1984) Shetland Life and Trade 15501914. Edinburgh.