Recent Advancements in the Riddarholmen Ship Puzzle: A New Interpretation of the Architecture of an Early 16th-Century Clinker-Built Gun-Carrying Warship

ABSTRACT The 16th-century Riddarholmen Ship was discovered in the middle of Stockholm in 1930. Despite being exhibited since 1947, the efforts to reconstruct the ship have been limited. A substantial portion of the recovered parts has never been put on display. This paper aims to shed new light on the ship´s architecture using the original material from the excavation together with an inventory of the timbers in the collections of the Medieval Museum. From this, it is argued that the ship originally had a full deck, three masts, and a forecastle and that it resembles an early purpose-built, gun-carrying warship.


Introduction: At the Medieval Museum
A visitor who enters the subterranean exhibition hall of the Stockholm Medieval Museum is met by the well-preserved forward part of the Riddarholmen Ship (Figure 1).The ship was discovered and excavated in the town centre in 1930.Two wrought iron breech-loaded guns are also on display.These are fairly well-known as they are early examples of ship artillery and were found loaded together with a large number of projectiles.The ship and its guns provide a snapshot of the conflicts and strife in Stockholm some 500 years ago, at the end of the Kalmar Union and during the early rule of King Gustav I (r.1523-60).
While the exhibited ship is genuinely impressive, it remains an empty shell.It is difficult to envisage or imagine how people were able to dwell, stand, or move aboard a vessel like this, let alone fire cannons.The irregularly placed floor timbers form a veritable obstacle course for those who want to move around and through the hull (Figure 2).Obviously, some vital parts of the ship, such as decks and bulkheads, are missing.Few people know that solid archaeological information and evidence regarding these structures were available at the time of the excavation.However, for various reasons, the fragments of the ship's interior ended up in a storage magazine of the museum's collections rather than the exhibition.
As part of an ongoing research project, this author visited the Medieval Museum magazine for a closer look at the material that is not on display.After a brief inventory, it was clear that some fascinating and diagnostic vessel parts were preserved but have not yet received sufficient scholarly attention.These ship parts are neatly lined up on pallet racks and are easy to access.
This paper aims to summarize some conclusions regarding the ship's architecture drawn from examining ship timbers in the magazine, the original plans and notes from the excavation, and the exhibited ship.The focus will be on piecing together the space-defining structures aboard the ship, such as deck levels, bulkheads, and the scarce fragments of a forecastle.I will assess and interpret these different parts through comparisons with contemporary images and other archaeological finds.The results presented here do not claim to be the final reconstruction of the Riddarholmen Ship.I realize that extracting more information from the recovered and preserved parts would likely be possible if they were digitally recorded and processed, but that would require a much larger project with a completely different budget.Hopefully, the results presented here might stimulate the initiation of such a project and raise the necessary funding.
I have chosen to present the result in text and handdrawn two-dimensional plans, as the original data from the excavation come in this form and I am used to working in this way.The result will be comprehensive enough to argue that the Riddarholmen Ship is an early example of a purpose-built gun-carrying warship.The Riddarholmen Ship may thus add to our understanding of the transition from the medieval combined merchant and warships and the early modern purpose-built men-of-war.

The Discovery
Stockholm is a town built on several islands, with its medieval centre on the island Gamla Stan -The Old Town.Today the smaller, adjacent island of Riddarholmen is separated from Gamla Stan by a subway shaft, but less than 100 years ago, this was a strait that formed sheltered water and a lively and frequently used harbour (Figure 3).In the 1920s, car traffic increased significantly, requiring the construction of several new roads.During dredging for two new bridges between Riddarholmen and Gamla Stan, two old cannons were discovered on the seabed.The discovery was made in early 1930, and the amanuensis Tord Olsson Norberg was commissioned to examine the location in more detail.A diver went down in the turbid water and concluded that an old ship lay deeply embedded in the seabed where the cannons were found (Fischer, 1983, p. 819;Hansson, 1947, p. 10;1960, p. 42;Nordberg, 1930, p. 263;1931, p. 194;1950, p. 19).The diver also made a surprisingly detailed site plan.There is reason to return to this plan below as it contains some important information for reconstructing the ship (Figure 4).
As the wreck was regarded as well-preserved, ancient, and very interesting, the city of Stockholm decided to finance an archaeological excavation.In the 1930s, Swedish archaeologists mainly worked on prehistoric sites and, to a lesser extent, were concerned with medieval material, except for castles and similar manifest remains.The situation was similar when it came to ships.Experts excavated prehistoric ships, whereas historical shipwrecks were salvaged commercially for souvenirs and waterlogged black oak, the latter used for the fabrication of furniture (Arnshav, in press;Cederlund, 1983;Eriksson, 2014, pp. 21-36).
The archaeological excavation of the Riddarholmen Ship was thus a pioneer work that was soon to be followed by several others, for instance, the wrecks from Kalmar and the ship Elefanten sunk in 1564 (Adams, 2013;Åkerlund, 1951;Ekman, 1942).In the 1930s, archaeological work under water was still in its infancy.Even though Elefanten was partly excavated under water, it would still be long before archaeologists began to dive.Instead, a cofferdam of boards was built around the Riddarholmen Ship, and water was pumped out.Photographs from the excavation that was carried out between June and August 1930 reveal that it was very muddy and slippery work (Figure 5a and b).
As excavation is a process in which contexts and spatial relationships between artefacts are removed and destroyed, recording and documentation must be made continuously as the soil is removed.The amount and quality of the information gathered during fieldwork depend on the skill of the involved archaeologists.Recording features that appear incomprehensible at first sight is necessary as one never know what questions researchers will address to the material in the future.
Archaeologist Hans Hansson conducted the excavation of the Riddarholmen Ship together with   Thanks to this plan, we know that the missing deck beam, when the ship was excavated, measured 4.5 m.In the bow, there appear to be several levels of crossbeams (SSM Archive 10090100_SR195_SRARKRITN, public domain).
engineer John Söderberg under the supervision of the above-mentioned Tord Olsson Norberg.These were experienced professionals who knew what they were doing and could adapt and adjust their methods to the special conditions in the trench.
The different layers of loose timbers and artefacts were recorded separately, meaning many different plans were made (Figure 6a and b).As there will be reason further on to return to these, these different layers have not been combined and compiled into a synthesis.Hence, there has been no real interpretation of the archaeological contexts and ship parts discovered loose aboard the ship.The present study relies on the material from the excavation.It was possible to carry out the study only because of the high quality of the material and the skills of the archaeologists who excavated the site.
When the hull had been emptied and only frames and planking remained, 11 hull cross-sections were drawn.Before the ship was taken apart, all parts were marked and named individually.The strakes of planking were numbered from the keel with even numbers on the port side and uneven numbers on the starboard side.The frames were labelled with letters starting in the bow (Fischer, 1983, pp. 5-14, describes the excavation in detail, as does Hjulhammar, 2010, pp. 119-123).

Interpretations
The discovery and the excavation spurred discussions concerning the ship's age and identity, as well as speculation on why and when it sank, which became a recurring topic in the press during the summer of 1930.The cannons and, not least, the relatively large number of projectiles and chambers that were discovered aboard the ship underlined that it was a warship.The stock of one of the cannons turned out to be marked with the symbol for Stockholm, and it appeared likely that the cannon and, consequently, the ship were intended to defend the city (Figure 7).Archaeologist Hans Hansson suggested that the Riddarholmen Ship could be the city's Bardsa mentioned in written sources.According to a document from 1473, that ship was equipped with 13 cannons and 64 chambers, which could be compared with the four cannons and 43 chambers found on board (Hansson, unpublished Promemoria in the MM archive; Hansson, 1960, pp. 43-47;also Fischer, 1983, p. 47;Hjulhammar, 2010, pp. 119-122).Hans Hansson's assumption must be considered well-founded, based on the information available at the time.However, it turned out he was wrong.
During the 20th century, the methods for dendrochronology developed radically.As a part of the work for an unusually ambitious thesis for a bachelor's degree in archaeology at Stockholm University, Arnd Fischer took samples for a dendrochronological analysis of wood from the Riddarholmen Ship.The analysis was carried out by Thomas Bartholin at Lund University and showed that the trees were felled during the 1520s.Fischer did not reveal any candidate for the ship's identity but suggested that the ship may have belonged to Gustav I (Vasa) and that it had been intended to protect the city (Fischer, 1983, pp. 46-54).
Dendrochronological methods are continuously developed and refined.Around 35 years later, historian Ingvar Sjöblom had the samples re-analysed.The new analysis was carried out by Hans Linderson.In general, the analysis confirmed the previous results but as the statistics for sapwood has developed since the 1980s it was concluded that the wood for the ship may had been cut between 1516 and 1524 (Linderson, 2015;Sjöblom, pers. comm.).
The written source material from the early 16th century is fragmentary, scarce, and far from comprehensive.The few ships that are mentioned in the preserved written accounts represent an unsystematic selection of ships that mainly belonged to the very top strata of society.As pointed out by the late Jan Glete, a history professor, a warship may have sunk in the middle of Stockholm without the event being mentioned in preserved material (letter from Glete cited in Fischer, 1983, p. 47).It may thus be difficult to find the identity of the Riddarholmen Ship with certainty in written sources.

Assemblage and Reconstruction
After their excavation, the parts of the ship were dried and treated with linseed oil.The hull remained dismantled until 1947, when the foremost part of the hull was pieced together in the newly opened Stockholm City Museum.During the assembly work, it was noted that the wood had shrunk to a varying degree during the 17 years that had passed since the excavation.To fit the pieces together, iron bolts of a diameter similar to that of the original dowels were used to attach the planking to the frames.As the ship is clinker-built, the shrinkage is not visible thanks to the overlap of the strakes (Fischer, 1983, pp. 19-20;Hansson, 1947, pp. 10-12;1957, p. 1).It should be noted, however, that the wood from the ship, which is almost exclusively made of oak, is still in surprisingly good condition.
In the first rebuild in 1947, slightly more than the forward part of the ship was included.After a major rebuild at the museum during the 1960s, the ship was moved and refitted again on the bottom floor of the museum, and now the aftermost part was included in the reconstruction.Some sketches of the ship in its reconstructed form were made during its assembly in 1947, but no real reconstruction plans or drawings were made.
The reconstructed and exhibited parts of the ship were limited to the hull, such as the keel, stem, frames, keelson, and planking, whereas the ship's interior was left out in the magazine.I find it a bit surprising that there seem to have been no real efforts to reconstruct the ship in its entirety through drawings.The only traces of such activities that I have found are some drawings made by John Söderberg, preserved and stored in the archive of the Stockholm City Museum.The drawings were made in 1941 before the ship was assembled but have not been published except for the cross-sections that were included in Fischer's thesis (1983, Appendix 3).It is possible that Söderberg's reconstruction drawings reflect the ideas of the original appearance of the vessel that flourished at that time (SSM, archive inv.10090100_SR195_SRARKRITN.).
Arnd Fischer proceeded thoroughly with his bachelor thesis, a work that included visits to the museum's magazine to look at the parts of the ship that were not on display.He was thus well aware that some of the timbers had not been included in the reconstruction, among these the wales, but also some occasional remains of a deck structure.Fischer was of the opinion that the ship was decked in the bow and possibly also in the stern (1983, p. 29, see also Hansson, 1960, p. 45).When he wrote his thesis, Stockholm's new Medieval Museum was under construction, and the Riddarholmen Ship was soon to be moved to this new location.Fischer ends his text with a wish that the missing parts would be included when the ship was exhibited again.That did not come true.
The work to move the Riddarholmen Ship to the new museum was entrusted to Axel 'Acke' Lindberg, a retired boat builder with much experience in clinker-built boats.He summarized his observations and ideas in connection with the assembly in a report where he argued that the Riddarholmen Ship originally had a transom stern, something that is also hinted at in the welded steel frame that supports the ship in the exhibition and provides an idea of where the hull ended at the stern (Lindberg, 1985, pp. 21-30).
Figure 7.The two wrought iron guns that were found prior to the discovery of the wreck.A: The first gun that was discovered.The barrel is conical with a calibre of 10 and 14 cm at the muzzle.It was discovered without a chamber.The marking on top of the stock, aft of the barrel, is the symbol for Stockholm.B: The second cannon, which has a calibre 3.6 cm.The house mark on the side probably belongs to the gunsmith (Fischer, 1983, pp. 33-38).Sketches based on available measurements and the exhibited guns.(Niklas Eriksson).
Unfortunately, Lindberg completely overlooked not only the material from the ship stored in the museum's magazine but also the exemplary archaeological documentation from the excavation kept in the museum's archives.If he had included that material, the exhibited ship would have looked very different.For instance, Lindberg argues that the ship has 'no traces of a shelf clamp' (Lindberg, 1985, pp. 7, 9, my translation), and he thus concludes that the Riddarholmen Ship 'obviously had no deck' (Lindberg, 1985, p. 31, my translation).Moreover, he argues that the ship had only one mast (Lindberg, 1985, p. 31).The present study of the ship proves all of these statements to be false.

A New Reconstruction
After an inventory of the timbers and ship parts in the magazine of the Medieval Museum, it was clear that some interesting and diagnostic parts of the vessel had never achieved sufficient scholarly attention.As their existence will affect the interpretation of the ship's architecture, I found it necessary to make a new reconstruction.The work has consisted of a review of the drawings and notes from the excavation, which are kept in the archives as well as an inventory of the parts of the ship that are stored in the warehouse.Some parts of particular interest in interpreting the ship's architecture have been recorded through measurements, photos, and sketches.The reconstruction is presented on hand-drawn two-dimensional drawings and focuses on the space-defining elements of the ship's architecturesuch as deck levels and bulkheadsrather than defining the original hull form.
A first step in the reconstruction has been to produce a site plan (Figure 8) that includes the hull components necessary for the reconstruction.The plan was drawn by combining information from the plans and sketches made by John Söderberg during the excavation.Two plans, in particular, were used: one that shows the ceiling planks and other timbers from the ship's interior and one that shows the planking and the frames (Figure 6a and b).The archaeologists who excavated the Riddarholmen Ship were skilled in their profession but had little ship experience.Some errors in the plans, for instance, around the keelson, have been adjusted and corrected on the basis and comparison with photos from the excavation and recording of the preserved timbers in the magazine.
When John Söderberg made his reconstruction in 1941, he drew several sections based on the sections drawn during the excavation.His reconstruction looks like a Viking ship with a very wide and low hull.Part of the explanation for this result is that Söderberg did not use the timbers that provide information about the hull's cross-section, like deck beams, shelf clamp, etc., which resulted in a hull that was far too wide.According to his reconstruction, the hull originally was 6.25 m wide (Fischer, 1983, p. 21;Lindberg, 1985, p. 6).
I have redrawn Söderberg's cross-sections and compared them with the original field drawings.They are very similar, but when they diverge, I have relied on the field drawings.I have added ceiling planks, keelson, and shelf clamp to these drawings as these parts are important for the reconstruction (Figure 9).
When the Riddarholmen Ship was found, it was listing to starboard.The preserved hull had reshaped as the upper part of the starboard side had fallen out.To largely reconstruct the cross-sectional hull shape is a matter of determining how much the upper part of the starboard side has separated from the coherent bottom.One way to assess the original shape could be to try to fit the frames and the planking together.Another way is to base the reconstruction of the original hull shape on the deck beams.
As already mentioned, Acke Lindberg argued that the Riddarholmen Ship did not have a deck as there were no traces of a shelf clamp (Lindberg, 1985, pp. 7, 9).However, the ship did have a shelf clamp which was recorded in situ during fieldwork, in accordance with good archaeological practice.However, none of the plans that show the shelf clamp have been published.In the notes from the excavation, it is never mentioned as a shelf clamp; it is registered as a ceiling plank.The timber is still preserved in the magazine, and there is no doubt that it is a shelf clamp (Figure 10a).As the timber was properly recorded in situ and is also visible in several photos, it is also possible to determine its original position (compare Figure 8a).
The shelf clamp is important in reconstructing the ship in cross-section as it has notches that reveal where the beams that supported the deck were attached.Unfortunately, only one deck beam was preserved in its entire length during the excavation (Figure 8b).After examining the beam, the shelf clamp, and the material from the excavation, it can be concluded that the original location of the beam was just aft of section F, which makes it possible to reconstruct this section with reliable precision (Figure 11).
Further aft, the reconstruction becomes slightly more erratic as no deck beams were recovered during excavation.However, when the Riddarholmen Ship first was discovered, several of the crossbeams were still in place, but unfortunately, the starboard side was damaged amidships due to the construction of a bridge pillar before the site was emptied from water, which explains the splintered aft end of the shelf clamp.This is why diver Erik Hellström's site plan from February 1930 is important, as it was made before the damage (Figure 4).On the plan, Hellström has drawn the above-described preserved deck beam in situ at section F, which is recognizable, thanks to the two notches for carlings.Around 2.5 m aft of the beam at section F, he has drawn another beam.Judging from the site plan, it looks as if he has cleared the beam from mud towards the port side in order to assess and measure the breadth of the hull.He measured the beam to be 4.5 m long.Judging from Hellström's site plan, the now-lost deck beam was placed at frame M, making reconstructing another section possible (Figure 11).

Hull Dimensions and Displacement
According to the available material from the excavation the preserved length of the ship, measured from the outside of the stem to the aftermost preserved end of the keel, is 19.7 m (different numbers appear in Fischer, 1983, p. 21;Lindberg, 1985, p. 6).The sternpost is missing but the notch into which it was attached has an angle of 8 degrees and provides an impression of an upright sternpost, with only a slight incline (Lindberg, 1985, p. 18).The overall length of the ship would have been just above 21 m.
At the reconstructed section M the hull becomes 5.45 m wide, measured at the inside of the planking.This shape corresponds well with the shape of the exhibited ship.Section M is spaced ahead of amidships, which means that greatest width would have been 5.5 m, or just above that inside the planking.
A noticeable feature, when the hull is seen from the side, is that the stem is relatively upright.Several large late medieval clinker-built ships reveal long curved stems with a considerable rake (see Auer & Maarleveld, 2013, p. 35 with references).Shipbuilding manuscripts of the late 16th and early 17th centuries, such as Matthew Baker's, the Newton or the Harriot manuscripts, describe the construction of carvel-built warships and indicate that the rake of the stem should be more or less equal to the width of the hull (cf.Bellamy, 2006, p. 10;Eriksson, 2021, p. 122).
The rake of the stem of the Riddarholmen Ship is just above 2 m from the forward end of the keel, which is less than half the ship's breadth.The upright stem and the relatively long keel in relation to the  overall length have provided additional lateral plane on the relatively shallow, narrow, and probably quite fast hull.
Compared to wrecks of both of merchant vessels and warships dated to the decades around the turn of the 16th century, the Riddarholmen Ship appears relatively long, low and slender (compare Auer & Maarleveld, 2013, p. 41, figure 40, or Eriksson, 2021, p. 124, Table 1 with references).The keel-to-beam ratio of the Riddarholmen Ship is 3.2:1, which may be compared to 2.8:1 of Mary Rose, 2.5-2.59:1 of the carracks in the Timbotta manuscript from 1445, or 2.67:1 of Lomelinna (see for instance Friel, 1994, pp. 81-83;Guerot et al., 2023).
Besides being long and narrow, the rediscovered shelf clamp reveal that the Riddarholmen Ship had the deck placed relatively low with a limited hold below (compare Figure 11).The capacity of late medieval and early modern ships was usually established through the cubical numbers achieved from multiplying keel length x beam x depth in the hold (see Ferreiro, 2007, pp. 191-194 for an overview).Depth in the hold aboard the Riddarholmen Shipfrom the deck beam down to the keelis 1.45 m.
In order to get an idea of the ship´s displacement, a solid model of the bow back to section M was made, using a 3D CAD program (Figure 12).The analysis was made by Fred Hocker at the Vasa Museum and revealed that the displacement of the part of the ship with the deepest possible draught of about 2.1 m, with the lower wale totally submerged amidships, is about 28 tons.The rest of the ship should displace at least this much and probably slightly more, something that would give a total displacement of a little over 60 tons, perhaps as much as 70 tons.

Wales, Standards and Railing
Other parts of the construction that are not included on the exhibited ship are two wales that ran in parallel on the outside of the planking.The wales consist of several planks that are scarf-joined together and attached to the hull with treenails.In contrast to the rest of the ship, which is made of oak, the wales are made of pine.These are mentioned in the notes from the excavation, and some of these wales have been relocated in the magazine (Figure 14c).
Among the interesting timbers in the magazine, there are also vertically oriented reinforcements that were attached to the outside of the planking in the bow.Another way to describe them is as top-timberframes placed on the outside of the planking.Timbers with similar locations, for instance, on Mary Rose, which is contemporary with the Riddarholmen Ship, are mentioned as standards (Marsden, 2009, pp. 220-241), and this term will be used below.There are two preserved standards from the Riddarholmen Ship; both are made of pine (Figure 13).One of these was discovered loose outside the starboard side in the vicinity of its original position (Figure 14a).
The standard has a hole in it, likely for a rope and possibly for the main tack of the main sail.The timber stood on top of the upper wale and was attached to the wale using iron nails.Above the wale, it was attached with dowels.The inside of the standard reveals that there were four strakes of planking above the wale.
The other standard is longer, but as the upper end is very eroded, it is not possible to determine its original length.The erosion indicates that it has been sticking up above the bottom sediments for several hundred years.The position where it was found is not known in detail, and it is registered as part of a frame.It is clear from the shape of the timber that it was placed at a similar level on the outside of the hull as the other standard, standing on top of the upper wale.The lower end is shaped to follow the sheer of the planking, which indicates that it was located before the other wale (Figure 14b).Traces of the wales and the standards are visible on the exhibited ship through a number of treenail holes.
An image that corresponds quite well with the arrangement seen on the outside of the bow of the Riddarholmen Ship's planking is preserved in the British Library (Figure 15).It is assumed to show a battle that took place in 1416 between the French and English in the mouth of the Seine, but the drawing was made towards the end of the same century, which explains how the English ship can be armed with the kind of artillery that was modern at the end of the same century and when the Riddarholmen Ship sank.The French ship on the right is clinker-built and equipped with wales in a similar way as the Riddarholmen Ship.In the bow, standing on top of the upper wale, we find standards that support the forecastle.A French soldier in the forecastle throws a lance against the enemy, whereas an archer aims from the corresponding position on the English ship.
It is likely that the Riddarholmen Ship had a similar forecastle in the bow, which is supported by other fragments in the magazine.These consist of thin boards with semi-circular cutaways.During the excavation, these were found loose in the bow, and I would like to suggest they are the remains of a bulwark planking for a forecastle.These boards are made of carefully planed and shaped oak, only 17 mm thick.One of the fragments is bevelled towards one end in order to be scarf joined (Figure 16).A well-known example of a forecastle with a similar railing is the    so-called Mataró ship model from around 1450, which originates from Catalonia but is now in the collections of the Maritime Museum of Rotterdam (see, for instance, De Meer, 2009, p. 46).The semi-circular cutaways may be decorative, but they may, all the same, be argued to be functional as the shape would allow longbows and cross-bows to be fired through the railing.
Like the forecastles, tops in the masts are built up platformsused for battle as well as vantage points.Preserved images reveal that railings on fighting tops, forecastles, and sterncastles were built in similar ways and they are sometimes referred to as topcastles (see for instance, Ellmers, 1994, p. 43;Runyan, 1994, p. 53).This is hardly surprising as their function was the same and they should be of lightweight construction in order not to jeopardize the ship's stability.Whereas the number of preserved forecastles from ships contemporary with the Riddarholmen Ship are limited to some timbers from Gribshunden (sunk 1495) or Mary Rose (sunk 1545) (Eriksson, 2014(Eriksson, , pp. 25-28, 2020, pp. 262-276;, pp. 262-276;Marsden, 2009), there is actually a nearly complete top preserved in Gamleby church in Sweden.Unfortunately, nobody remembers where it came from or how and why it ended up in the church (Hall & Halldin, 1963, p. 99;Hofrén, 1945, pp. 105-110).
The top from Gamleby is built up around a steady framework, but in between these, the construction is very light, with battens thinner than 10 mm with semi-circular cutaways that are scarf-joined in the same way as the fragments found on the Riddarholmen Ship (compare Figure 17).It is thus not too farfetched to see the fragments from the Riddarholmen Ship as part of a rail.Judging from where they were found, this would be the rail in the bow, quite possibly from the forecastle, one that rested on reinforcing the standards mentioned above.Some images of late medieval ships reveal forecastle railings with circular openings (see, for instance, Figure 15), and the fragments could derive from such a railing as well.

Guns and Gun Deck
Four guns were found on the Riddarholmen Ship.One of these was a very small cast bronze handgun, whereas the other three were wrought iron breechloaded guns, a kind of weapon common during the Late Middle Ages but replaced by more efficient and reliable cast bronze muzzle-loaders during the 16th century (Figure 7).A disadvantage with the wrought iron guns was that they were relatively fragile and could not be loaded as hard, shot as long, or thrown away as heavy projectiles as later cast iron or bronze muzzle-loaders.On the other hand, they were relatively lightweight and did not require the same kind of supporting sturdy hull structures around them.They worked well onboard a relatively light clinkerbuilt ship like the Riddarholmen Ship.Later cast bronze or iron muzzle-loaders were much heavier and rested in wheeled gun carriages that stood on a massive gun deck.In order not to jeopardize the ship's stability, such heavy artillery was placed low in the hull and fired through gunports in the sides of the hull (for a discussion on the transition from clinker to carvel and the introduction of shipborne artillery, see, for instance, Adams, 2013;Marsden, 2009, or Fontana & Hildred, 2011).
When the Riddarholmen Ship was excavated, several of the chambers for the guns were still loaded, which has been interpreted as the ship being ready for action.The chambers, ammunition, the handgun, and one of the wrought iron guns were found below deck in the ship's bow.Several short boards reveal that these artefacts stored in a space separated by a bulkhead (Figure 18).Exactly where the two complete guns (Figure 7) were located in the wreck is not known as these were recovered before the ship was discovered.The drawing from the British Library provides an idea of what it might have looked like (Figure 15).The first two wrought iron guns that were discovered from the Riddarholmen Ship were placed in stocks hung in swivels (Figure 7).During action, the majority of the ship's guns would have been placed amidships with the swivels in holes in a now-lost gunwale.
When the Riddarholmen Ship was excavated, the uppermost parts of the hull, including the gunwale, were missing.But as shown above, the deck's substantial remains were preservedthe floor on which the artillerists stood when they fired the wrought iron cannons.In the earliest texts, it is concluded that the ship was decked in the bow and possibly also in the stern.Support for such a conclusion was observed already in the diver Erik Hellströms sketch of the wreck in February 1930.He reveals several deck beams and some deck planks along the starboard side (Figure 4), corresponding to the nail holes from the only preserved deck beam (compare Figure 10b).
However, several other elements from a now-lost deck have not received attention despite being included in the plans, sketches, and photographs from the excavation, the already mentioned shelf clamp being perhaps the most important.With its row of square or rectangular notches, it reveals where the cross beams originally were located and enables a reconstruction of the deck's orientation (Figures 10a and 18).Only the forward part of the shelf clamp is preserved, which reveal that the forward half of the Riddarholmen Ship was decked.It is likely that this deck continued aft for the entire length of the ship.
Notches in the preserved deck beam (Figure 10a) indicate that the deck was built up of two long lengthwise carlings.In the material in the magazine, fragments of shorter half-beams ran in between the shelf clamp and the carlings (Figure 10c).Below the shelf clamp is another thick plank with a square notch (Figures 8b and 18).The function of this notch is unknown, but it is likely that it reveals the original location of another now-missing crossbeam.The plank has not been relocated in the magazine.The railing and the standards found in the bow indicate that the Riddarholmen Ship had a forecastle.Some of the crossbeams shown in Erik Hellström's sketch that were missing when the ship was excavated may have been the remains of such a structure (Figure 4).
There is a plethora of contemporary images of ships of forecastles, which have sterncastles as well (for instance Figure 20) and is possible that the Riddarholmen Ship also had a sterncastle, even if no remains of this structure are preserved.

How Many Masts?
Several details from the rigging were found during the excavation.The blocks, deadeyes, and other details deserve a closer study, but this is beyond the scope of this article, but it is important to discuss the number of masts.According to Acke Lindberg, the ship was rigged with only one mast (Lindberg, 1985, p. 31), but he was not the first or the last to express this opinion (see, for instance, Hansson, 1960, p. 45;Hjulhammar, 2010, p. 215).The square notch in the keelson reveals the location of this mast.However, another mast step was encountered during the inventory of the ship parts in the museum's magazine (Figure 19).Judging from the shape of the timber and the location where it was found it was placed on top of frame C (Figures 11a  and 18c).It is noticeable that the square notch for the foremast is not cantered but pushed towards and along the starboard side, likely to make room for a bowsprit.The ship would likely have had an additional small mizzen mast in the stern.The timber was recorded and drawn in situ during the excavation, but it obviously was not recognized as a mast step and has not been mentioned in any publication concerning the ship (Figures 6b and 8c).

Conclusion
In 1514-15, the anonymous 'Edebo Master' painted the interior of Edebo church in Sweden with scenes from the Old and the New Testaments.On the western wall, he painted a scene of Saint Ursula and her maidens travelling on the sea.Christ is depicted as a fellow passenger, which is regarded as an allegoric expression of the ship of the church (Cornell & Wallin, 1965, pp. 34-35) (Figure 20).
Material culture depicted in church paintings from this time has a tendency to be contemporary with the artist rather than with the illustrated scenes.This applies not least to depicted ships (see, for instance, Eriksson, 2020, pp. 266-270), and that is why Jesus, Ursula, and the maidens travel aboard an early 16th-century ship.According to a recent dendrochronological analysis, the Riddarholmen Ship was built some years after the anonymous master painted the ship in Edebo church.
In contrast to the existing opinion that the Riddarholmen Ship was an open one-masted vessel, the present study has argued that the Riddarholmen Ship reveals many similarities with the ship depicted in Edebo; it was clinker-built, with a high rising curved stem with a small mast placed near the stem, a much larger mainmast and a small mizzen mast that balances the pressure from the other two in the stern.The image also depicts fore-and sterncastles with railing reminiscent of the fragments found on the Riddarholmen Ship.
During the Late Middle Ages, warships were usually big merchant vessels drafted for war, equipped with lighter weapons, and manned by soldiers (e.g.Bill, 2002;Runyan, 1994).I would suggest that the long and slender Riddarholmen Ship is a purposebuilt warship rather than a ship that served such a dual purpose.The room below deck is very low and allows for very limited cargo space.In contrast, the deck forms a perfect platform for artillerists to operate pivot-hung wrought iron ordnance.At the same time the limited space below deck offers little room for provisions necessary for a large number of soldiers during longer campaigns.The wrought iron cannon with the Stockholm coat-of-arms suggests the cannonand possibly also the shipbelonged to the city.It is likely in this context we should see the Riddarholmen Ship as a vessel aimed to patrol and defend Stockholm.
The Riddarholmen Ship thus provides a unique opportunity to understand how a clinker-built guncarrying warship actually looked.Dated to the early 16th century, it is a species that was soon to be extinguished and replaced by much larger, carvel-built vessels with cast muzzle-loaded guns in ports along the hull side (Eriksson, 2019).
The deck, the bulkhead that defines the room where the ammunition and the chambers were found below deck in the bow, as well as the forecastle, forms a link between the clinker-built hull and the well-known guns that were found aboard.These rooms are crucial for understanding how the Riddarholmen Ship functioned as a warship.
Surprisingly, these vital components of the Riddarholmen Shipthe shelf clamp, standards, mast step, and other pieces of the puzzlehave been forgotten in the warehouse of the Medieval Museum.
That a substantial part of the Riddarholmen Ship has never been refitted is likely due to other more prominent focus areas.It has not been possible to focus on the find all at once.The ship's architectonical analysis has been overshadowed by other more established research perspectives and questions, for example, the construction of the hull, the armament found on board, and striving and ambitions to find and reveal the original identity of the ship.
An important result of this study of the Riddarholmen Ship is how important it is to address relevant archaeological questions about the material.It is also never too late to address these questions if the archaeological information has been properly excavated and recorded.

Figure 1 .
Figure 1.The bow of the Riddarholmen Ship as exhibited in the Stockholm Medieval Museum.Unfortunately, the wales and the standards from the forecastle are not included on the exhibited ship.(Photo: Jens Ullenius, https://commons.wikimedia.org/w/index.php?curid=23224398).

Figure 3 .
Figure 3. Location map drawn by John Söderberg (translated and edited by Niklas Eriksson).

Figure 4 .
Figure 4. Diver Erik Hellström made the site plan in February 1930 before construction work that damaged parts of the wreck.Thanks to this plan, we know that the missing deck beam, when the ship was excavated, measured 4.5 m.In the bow, there appear to be several levels of crossbeams (SSM Archive 10090100_SR195_SRARKRITN, public domain).

Figure 5 .
Figure 5. A. The bow of the Riddarholmen Ship during excavation in 1930; B. The photo is taken amidship with the camera pointing towards the bow.Note the keelson on the left, the shelf clamp along the starboard side on the right, and the deck beam with the two notches towards the bow (SSM graark_10064944 and graark_10064942, with permission, BY-NC-SA).

Figure 6 .
Figure 6.Two examples of site plans made by John Söderberg.A. The first plan reveals the empty hull is the one that appear in several publications.B. The second plan reveals the ceiling, shelf clamp and other elements of the hull that have never been published.Bow to the right (SSM10090100_SR195_SRARKRITN, public domain).

Figure 8 .
Figure 8.A revised site plan that includes components crucial for reconstruction, with bow to the left.A: shelf clamp, B: ceiling plank with notch for crossbeam, C: foremast step, D: keelson, E: deck beam, F: cask with canister shots, G: location of railing fragments, compare Figure 16.(Niklas Eriksson).

Figure 9 .
Figure 9.Sections seen from the bow drawn from John Söderberg's originals, supplemented with shelf clamp, remains of ceiling and keelson.The positions of the sections are indicated on Figure 8 (Niklas Eriksson).

Figure 11 .
Figure11.Three reconstructed cross-sections seen from the bow.From the left is section C (compare Figure9) with the foremast step (compare Figure19).In the middle is section F, with the preserved deck beam and the two standards standing on top of the upper wale (compare Figures 10: b, 12 and 13).To the right is section M with the deck beam recorded by Erik Hellström in February 1930 (compare Figure4).(Niklas Eriksson).

Figure 10 .
Figure 10.Some of the parts recorded in the museum's magazine that have been included in the reconstruction, A: shelf clamp.B: deck beam, C: half-beam.(Niklas Eriksson).

Figure 12 .
Figure 12.Solid model of the bow of the RIddarholmen Ship based on the reconstructed cross-sections (compare Figure 11) created to calculate displacement (Fred Hocker, The Vasa Museum).

Figure 13 .
Figure 13.Four views of the two standards that were attached to the outside of the planking in the ship's bow.Both are made of pine and were resting on top of and nailed to the upper wale.A: Standard with a hole through it, possibly for a bowline.It is preserved at its full length.B: Standard with an eroded upper end.The angle of its lower end indicates it was placed closer to the bow, where the sheer is more pronounced.(Niklas Eriksson).

Figure 14 .
Figure 14.Side elevation of the Riddarholmen Ship.A: Standard with hole for possible bowline, B: The approximate location of a longer standard, C: The double rows of pine wales, D: The approximate location of the rail fragments.(Niklas Eriksson).

Figure 15 .
Figure 15.The drawing probably depicts a battle that took place in the Seine estuary in 1416 but was made towards the end of the same century.The clinker-built French ship to the right reveals the same arrangement of wales and standards as the Riddarholmen Ship.Note the pivot-hung, wrought iron cannons amidship on the English ship to the left.(Beauchamp Pageant MS, Cotton Julius E. IV Art.6f.18v, reproduced with permission).

Figure 16 .
Figure 16.Two fragments of a railing with semi-circular cutaways were found loose in the ship's bow.(Niklas Eriksson).

Figure 17 .
Figure 17.In Gamleby church in Sweden, there is nearly complete late medieval top or crow's nest.It is made of oak and measures around 160 cm in diameter.Around the outside, run two rows of thin oak battens with semi-circular cutaways similar to the fragments found on the Riddarholmen Ship.(Niklas Eriksson).

Figure 18 .
Figure 18.Lengthwise cut-through section of the Riddarholmen Ship.A: shelf clamp, B: ceiling plank with notch for cross beam, C: foremast step, D: fragments of railing, E: space defined by collapsed bulkhead in which ammunition and spare rigging details were discovered.(Niklas Eriksson).

Figure 19 .
Figure 19.The foremast step that was recorded in situ during the excavation; compare Figure 8c.Note that the notch for the mast is not centred in order to allow space for a bowsprit.(Niklas Eriksson).

Figure 20 .
Figure20.In 1514-15, two years before some of the trees were felled for the Riddarholmen Ship, the anonymous 'Edebo Master' decorated the western wall of the Edebo church in Sweden with the legend of the Saint Ursula and her maidens.They travel aboard a ship that reveals many similarities with the Riddarholmen Ship: clinker-built, a curved and very upright stem, three masts, forecastle, railing, and so on.(Ulrika Södrén).