Iron from Zealandic bog iron ore – more than a theoretical possibility?

Tekst: H. Lyngstrøm (publiceres som del af Sjællands Jernalder, Nordiske Fortidsminder 2011)

The frequent application by archaeologists of Werner Christensen’s distribution map for the occurrence of bog iron ore in Denmark (1966) is queried, and the argument is made for the use, relative to Zealand, of Kristian Rørdam’s mapping data in combination with actual observations. Subsequently, against the background of the archaeological and geological record, metallurgical analyses and experimental archaeological research, the contours are sketched of iron production based on bog iron ore from Zealand.

In 1985 and 1986, the Danish National Museum, to the great surprise of the archaeological profession, excavated two iron-smelting sites in Western Zealand near Korsør (Andersen et al. 1987, 176f). The furnaces were found in connection with reconnaissance prior to the laying of a national natural gas pipeline network at Espevej and Skydebjerggård. In and around the furnace at Espevej, two smelting slag blocks were found, weighing respectively 4.8 and 8 kg, as well as 8 kg of slag in the form of many smaller pieces (Voss 1991, figs. 6-7; Buchwald 2005, 193), whereas only very few slag pieces were found by the furnace at Skydebjerggård. The slag from both furnaces was subjected to metallurgical analysis (Buchwald 2005, table 8.2), and there was no doubt that it was formed by the reduction, i.e. smelting, of bog iron ore from Zealand. Here lay the first evidence for Iron Age farmers having produced iron from Zealandic bog iron ore – and the sources of information relative to an understanding of how this happened.

Christensen’s map and bog iron ore on Zealand

In several respects, the archaeological discoveries at Espevej in Boeslunde and Skydebjerggård in Eggerslevmagle fractured archaeological preconceptions. Iron smelting? That wasn’t something that happened on Zealand. And furthermore, the furnaces in Western Zealand were so well preserved that the discoveries opened the eyes of archaeologists relative to a completely different and earlier form of iron extraction than that known previously. Even so, this was apparently not a phenomenon which was discussed at the symposium Sjællands Jernalder (Iron Age Zealand) held in 1990 (Lund Hansen & Nielsen 1992). And the reason for this was perhaps that archaeology was, at that time, still very much influenced by Werner Christensen’s model popular-scientific article Myremalm (Bog iron ore) and, in particular, the distribution map which accompanied the article (Christensen 1966, 47).

The map is frequently referred to in Danish archaeology (Hedeager 1988, 258; Lund 1991, 165; Jensen 2003, 48) so Werner Christensen’s work – and thereby the background for his distribution map – is definitely worth discussing.

As a starting point, it is important to know that a national mapping programme relative to occurrences of bog iron ore in Denmark has never been carried out. This is presumably because the aim of geological mapping is most often to record occurrences of economically important raw material resources such as gravel, sand or chalk – and that bog iron ore has only represented a modest potential in this respect. In 1966, when Werner Christensen as section geologist in the Danish Geological Survey drew up the map, he had studied bog iron ore over a longer period. He was, as the press wrote at the time, our leading expert in the field, who in 1950 had written a weighty memorandum concerning the extraction and exploitation of bog iron ore in Jutland (Skive Folkeblad 21^st July, 1956). It was this memorandum which, in 1951, made Christensen the obvious candidate to lead the practical work in dealing with around 1300 export applications which, in the post-war period, were submitted to the Danish state in connection with commercial exploitation of the bog iron ore deposits. And up until 1956, Christensen, in his headquarters, respectively at Hammerum and Bolderslev, took innumerable core samples and carried out almost 4000 analyses of Danish bog iron ore. Here then was a man with a detailed and balanced knowledge of bog iron ore in Denmark who, in 1966, passed on his knowledge to the general public.

On his map, Christensen (1966, 61) shows the distribution of bog iron ore as it is presumed to have been prior to the intensification of extraction in the years leading up to the Second World War. Using a key comprised more or less solid circles, he marked the areas where he himself had systematically prospected for bog iron ore. Each circle symbol corresponded to a topographical map on the scale of 1:20,000; an ordnance map (around 71 km²). A solid circle showed that there was, in practice, bog iron ore everywhere that hydrological conditions permitted, whereas an open circle denoted very scattered occurrences of bog iron ore. Using a cross, Christensen marked where information from the literature, archives or collections suggested there was bog iron ore in the area. This symbol was not used by Christensen to indicate the quantity of bog iron ore.

Accordingly, the distribution map shows primarily Christensen’s study area – and secondarily where Christensen knew that there was bog iron ore, without necessarily having seen it personally. Relative to the possible production of iron on Zealand, it would be interesting to know from where Christensen obtained his information.

Many of the places marked with a cross are where Christensen knew that bog iron ore had been included in the building materials for the churches on Zealand. In Frederiksborg county, this applies to the churches in Helsinge, Annisse, Ramløse, Asminderød, Kregme, Skævinge, Strø and Egebjerg, whereas in Odsherred, bog iron ore is found as part of the masonry of Vig and Højby churches. But most of the points marked are due to the fact that Christensen built on Rørdam’s mapping of the soil conditions in Frederiksborg, Copenhagen and Roskilde counties. Geologist Kristian Rørdam’s mapping was carried out at the end of the 19^th century and is involved all too rarely in the archaeological discussion of possible iron extraction on Zealand. But in Frederiksborg county, Rørdam observed that: Bog iron ore layers are found in many bog hollows, and furthermore: The thickness of the layers varies as a rule between ½ and 1 foot and nowhere seems to exceed 2 feet. Their extent is very variable, but in not a few cases is it possible to follow the same bog iron ore layer over several acres (Rørdam 1893, 88f). Among the c. 50 areas to the north of Copenhagen where Rørdam found bog iron ore (Rørdam 1894, 240), he described a deposit in the bog Niverød Mose, 300 m to the south of Langstrup, as being the most extensive. Here, the bog iron ore lay in a 35 cm thick layer covering 10 ha. Rørdam estimated that the total weight of the ore deposit was around 84,000 tonnes, and his analyses revealed that the ore had relatively high iron content (74.75% Fe₂O₃).

Rørdam analysed a total of 11 samples of bog iron ore distributed across ten localities in Frederiksborg county. The richest ores he found, apart from that in Niverød bog, were in Handskemagerrenden near Øverupgaard in Søborg parish (68% Fe₂O₃), SE of the farm of Avleholm in Ramløse parish (65.25% Fe₂O₃), at Skovløberhuset in Knurrenborgvang in Asminderød parish (63.06% Fe₂O₃) and by the river 1 km east of Lønholt in Grønholt parish (60.55% Fe₂O₃).

In Copenhagen and Roskilde counties, Rørdam (1899, 97) found not in a few places layers of bog iron ore. The largest ore layers in this area were recorded by him in a bog NE of Torkildstrup in Saaby parish, where the deposit extended over 5 ha, and in the Værebro river valley, SE of the lake Løged Sø (Rørdam 1899, 98).

For an archaeologist, Rørdam has several valuable pieces of information concerning the bog iron ore on Zealand. Among other things, he draws attention to the fact that the bog iron ore in North Zealand is so accessible that it requires less work to dig up a cubic metre of ore than to dig up a corresponding quantity of peat because the layers are as a rule very accessible as they are only covered by 1-2’ of peat-like soil which frequently is reddish-brown from the precipitated iron ochre (Rørdam 1893, 89). Rørdam notes at the same time that ore extraction has the added advantage that it can be carried out at all times of the year and therefore can take place at the farmers’ time and opportunity, when other more important tasks are not imminent (Rørdam 1894, 241). Furthermore, extraction does not require any previous knowledge and quarrying the ore will in certain cases actually improve the soil, so the sale of the ore, as he writes, can be considered as net profit (Rørdam 1894, 241).

Rørdam’s early observations can probably be confirmed by the museums in Frederiksborg county. At Holbo District Cultural-Historical Centre natural occurrences of bog iron ore have frequently been encountered in connection with the museum’s excavation activities. And according to museum curator Liv Appel, there is a tendency for the bog iron ore to occur in particular areas within the museum’s area of responsibility. She describes the ore as dark brown and solid, forming c. 30 cm thick layers, and as often being mixed with occasional small and medium-sized stones. At the Folk Museum in Hillerød, natural occurrences of bog iron ore have similarly frequently been observed. Museum curator Esben Aasleff estimates that bog iron ore occurs everywhere within the museum’s area of responsibility: Hillerød, Allerød and in Halsnæs. It is dark brown or orange-brown, is often solid and forms 20-30 cm thick layers which can be broken up in large cohesive slabs. He describes the ore as being homogeneous without many stones or organic components. Also at Hørsholm Museum, natural occurrences of bog iron ore have been frequently met with, and these appear to occur throughout the whole of the museum’s area of responsibility: in Asminderød, Langstrup, Farum, Isterød and in Rude Skov. Museum curator Mette Palm describes the bog iron ore as reddish-brown, very dark brown or black, and as being concentrated within delimited, small areas. It can often be broken up as 5-10 cm thick slabs. From Frederiksborg county, more recent analytical data are available for bog iron ore from Store Dyrehave (75.6% Fe₂O₃) (Lyngstrøm 2008a, table 2), Hillerød (70.4% Fe₂O₃), Ramløse (76.1% Fe₂O₃), Pederstrup (73.4% Fe₂O₃), and two ores from Søborg (respectively 82.6% and 73% Fe₂O₃) (Buchwald 2008, 121).

At Odsherred Museum of Cultural History, natural occurrences of bog iron ore have been encountered on repeated occasions. Here, too, there is a tendency for the bog iron ore to occur in particular parts of the museum’s area of responsibility (especially in the Egebjerg area and the area around Borren and Højby, but also in Asnæs). Museum curator Arne Hedegaard Andersen describes it as dark brown or orange-brown and as being frequently mixed with organic material. A single analysis has been published of bog iron ore from Odsherred. The sample was taken at Stenbækgård and contained 62.6% Fe₂O₃ (Buchwald 1998, table 2).

However, to the SW of Odsherred, at Holbæk Museum, natural occurrences of bog iron ore have only rarely been met with. And when this has occurred, it has been in particular regions (especially in Svinninge and towards the SW). Museum curator Niels Wickmann describes the material as being reddish-brown and has seen solid bog iron ore, which can be broken up into cohesive slabs, and also both crumbling ore and ochre ore. It is most often homogeneous without many stones or organic components. A single analysis has been published from the museum’s area of responsibility; a sample of ochre ore from Nøkkentved near Mørkøv. It contained 91.7% Fe₂O₃ (Buchwald 1998, table 2).

The distribution map that has repeatedly formed the basis for an archaeological line of argument reflects, accordingly, Werner Christensen’s work which was focussed on the bog iron ore of Jutland at a time when it was being exploited industrially. On Zealand, Frederiksborg county, as well as large parts of Roskilde and Copenhagen counties, were mapped by Kristian Rørdam in the 1890s. Valuable information concerning the distribution and quality of the bog iron ore can also be obtained from museums having archaeological responsibilities. Similarly, metallurgical analyses of the bog iron ore can establish whether its content of iron oxides (more than 60% Fe₂O₃) made it suitable for the extraction of iron during the Danish Iron Age.

Iron smelting on Zealand

There is, accordingly, the theoretical potential for the Iron Age farmers to have manufactured iron from Zealandic bog iron ore at settlements other than those at Espevej and Skydebjerggård. And since 1985, the archaeological record relating to ironworking on Zealand has also been expanded. This is rarely in the form of actual in situ furnaces but as refuse pits with extraction slag, shaft fragments and tuyeres for furnaces.

One of the best finds comes from the extensive excavations at Lysehøj near Korsør which were carried out by the Museum of Southwest Zealand in 2006. The furnaces here were not only an integrated part of a settlement dating from around the birth of Christ but also in context with pits which apparently contained exclusively discarded material from iron smelting: slag, parts of furnace walls and tuyeres. But also at Stenhusager, only 1 km distant from Espevej, a pit was found in 1998 which contained pottery, slag and fragments of tuyeres from iron smelting furnaces.

In another part of Zealand, at Tystrup I near Fakse, Sydsjællands Museum has located the remains of two furnaces, a refining forge and slag from smelting and refining. This took place in association with a settlement from the Roman Iron Age (Hansen & Staal 1996, 61). Three pieces of smelting slag were analysed from here (Buchwald 2005, table 9.4).  The material for one of the analyses was obtained from a piece of slag weighing 4.2 kg which was fused on to the furnace wall. It had almost the same metallurgical composition as a piece of slag from Gørlev, near Holbæk (Buchwald 2005, table 9.4). And not far from Gørlev, at Bjergene, an archaeological excavation in 2003 uncovered a stone heap containing fragments of vitrified furnace shaft and 22 kg of slag from ironworking. Around 75 m from the heap, a pit was found containing bog iron ore. Both of these features are dated to the time shortly after the birth of Christ (Borby Hansen 2006, 37), and the slag fragments included two intact examples. They weighed, respectively, 6.3 kg and 5.4 kg, were planoconvex, with a diameter of around 25 cm and a thickness of between 10 and 12 cm. This is more than double the size of the planoconvex slag bodies which were formed in the refining processes at, for example, Æbelholt Kloster. So, solely on the basis of the morphological form of the slag bodies from Bjergene, there is reason to assume that these were formed during iron smelting. This was confirmed by a metallurgical analysis (Jouttijärvi 2006, figs. 4, 5 and 6) which, furthermore, underlined the likelihood that the smelting took place using local bog iron ore. It is possible that the planoconvex bodies of smelting slag were formed at the base of a furnace with a rounded floor. Perhaps a furnace which could be used several times? From Koppedal Museum’s area of responsibility in Eastern Zealand there is a corresponding planoconvex body of smelting slag from Store Holmegård (SØL 390). This was found in a pit containing pottery dated to the Early Pre-Roman Iron Age. The slag has been analysed, but the results of the analysis are not published. The same applies to a body of smelting slag found in a secondary context at Sneglehøj (TAK 1269).

In Frederiksborg county there are several recorded finds relating to iron smelting. These include the site of Brolandsgård II where, as early as 1985, smelting slag was found in a pit dated to around the birth of Christ, and Sindshvile, where were excavated in 1991 and 1992 pits which contained 26 kg of slag resulting from both smelting and forging. These latter pits are dated to the Germanic Iron Age. Subsequently, material from Lyngebækgård III (HØM 240), Rævemose (HØM 400), Korsbjerg Have and Vassingerød (NFHA 2765) has been added.

On Zealand, archaeological evidence of iron-smelting furnaces has, accordingly, been identified in situ at Espevej, Skydebjerggård, Lysehøj and perhaps Tyrstrup I. While the complete bodies of base slag from Bjergene and Store Holmegård enable us to evaluate the basal diameter of these furnaces.

Additionally, numerous fragments of slag from ironworking have been found at settlement sites – or scattered in the landscape. With regard to quantity of slag, there is most commonly less than 10 kg at each locality. These slag bodies are often heavy, in some areas slightly magnetic, and they have an uneven surface. On occasional pieces of slag it can be seen how they have solidified in a channel or depression. A number of these slag bodies have been dated to the Pre-Roman Iron Age or to the time just around the birth of Christ – but there is also slag dated to the Roman and Germanic Iron Ages.

It is certain that the farmers used large quantities of wood and charcoal for smelting and forging – and calcium oxide (CaO) and potassium oxide (K₂O) are the main components of charcoal ash. Systematic experimental work (HAF 17/08) has clarified how the slag formed in all stages of ironworking reacts with the ash formed when the wood burns, and that the slag from iron smelting has a lower concentration of calcium and potassium oxide than that formed during refining and forging. In this way, the slag from Gørlev, Bjergene, Tystrup I, Store Holmegård, Sneglehøj, Korsbjerg Have, Brolandsgård II, Lyngebækgård III, Rævemose and Vassingerød is identified as resulting from iron smelting. And the argument that smelting took place on the basis of local bog iron ore is supported, not only by the ore deposits present on Zealand and the prevailing exceptionally close relationship between raw material and production (Giles 2007, 397ff), but also by the scientific analyses of the slag and iron.

Zealandic iron

Identification of the provenance of slag and iron with the aid of analysis of slag inclusions has been discussed in innumerable contexts (most recently Blakelock et al. 2009). There is a great deal to suggest that the slag inclusions in the iron following its smelting from bog iron ore from Zealand – as well as slag resulting from the subsequent refining and forging of the iron – will contain high values of phosphorous and calcium oxide (P₂O₅ and CaO), and that aluminium (A₂O₃) and potassium oxide (K₂O) are also of significance for the identification. In three knives from Viking times, iron has been found which could originate from smelting of Zealandic bog iron ore. One of the pieces was used to forge a knife found in a pit-house at Birkely in Northern Zealand (FRM S7x87; Bodilsen 1993, 121f). This was found together with three other iron knives as well as various antler and bone waste. It is a single-edged, straight iron knife, 8.1 cm in length, of which the tang comprises 1.1 cm. The maximum blade width is 1 cm, maximum back thickness 0.3 cm, and it was manufactured as technological type I. [note 1] The knife had clearly been forged from a recycled item which originally had been welded together using iron from three different smelting operations. The bog iron ore which formed the basis for one of these smelting operations was probably dug up in Eastern Denmark. The welds between the individual pieces of iron were fairly good and the iron in the lamella, which perhaps was made from local bog iron ore, had a content of 0.2% carbon and 0.5% phosphorous – the iron had a high slag content and was judged to be AH 5 [note 2] (Lyngstrøm 2008b, no. 5).

Two other pieces that could originate from Zealandic bog iron ore were identified in a knife found in one of the graves at Nordre Grødbygård on Bornholm (BMR 1399x1368; Wagnkilde 1999). The body was that of an adult individual and in the grave – in addition to the knife – there were only a couple of iron rivets. This is a single-edged, straight iron knife, which is 10.2 cm long. It was forged as technological type IV [note 1], which is a traditional Danish knife type in the 10^th century (Lyngstrøm 2009, ryc. 1): two layers of iron almost lacking in carbon are laid about a middle lamella containing about 0.8% carbon. It was the carbon-free iron (0-0.3% carbon and 0% phosphorous) which could have been extracted from Zealandic bog iron ore. The knife has been hardened and tempered. The welds are competently executed, and the slag content of the pieces of iron is moderate: AH 2 and AH 3 [note 2] (Lyngstrøm 2008b, no. 33).

The final example of a piece of iron which possibly was smelted from Zealandic bog iron ore is a knife from Kaagaarden on Langeland (LMR 11563x394; Grøn et al. 1994, 96 and 193). This lay, as the only artefact in the grave, at the hip of an adult male. It is a double-edged, straight iron knife, 14.4 cm in length, of which the tang comprises 3.9 cm. The maximum blade width is 2 cm. It was forged from a single piece of iron as technological type I [note 1], and the iron contains 0% carbon and 0.3% phosphorous. It contained a large number of slag inclusions: DH 4 [note 2] (Lyngstrøm 2008b, no. 48).

These three examples of iron produced from bog iron ore were included as reference material in a series of experimental-archaeological smeltings of bog iron ore from various localities on Zealand. The most important of these, in this respect, are the trials using bog iron ore from Store Dyrehave in Hillerød (HAF 09/02; HAF 17/08) because the smelting was followed up by trials involving refining of the iron blooms and forging of iron tools.

The bog iron ore was obtained from an open drainage ditch, where it occurred in palm-sized dark brown clumps. Prior to smelting, five pieces were subjected to metallurgical analysis (Lyngstrøm 2002, table 2). The first smelting (2001/II) lasted 9 hours and 55 minutes, during which the bellows were used four times for a total period of 2 hours and 25 minutes, and where use was made of 20 kg roasted bog iron ore and 31.5 kg charcoal. An iron bloom weighing 2.4 kg was produced, of which 835 g could be refined as iron ingots. The second smelting (2001/III) lasted 7 hours and 30 minutes, during which the bellows were used for a total of 2 hours, and use was made of 23.5 kg roasted bog iron ore and 39 kg charcoal. The smelting produced a bloom of 3.1 kg, of which only 67 g could be refined to iron ingots.

The refined iron was used in experimental forging of copies of pins and knives from the Pre-Roman Iron Age. Of the four pins, two were copies of pins with a coiled head and one, 2002/6, had a high swan’s neck (copy of Grarup; Becker 1961, Pl. 119/3). It was forged starting with 44 g of refined iron from iron bloom 2001/II. After being heated for the ninth time, the piece fractured at a large slag inclusion and 17 g was severed. After 45 minutes and 33 heatings, a pin weighing 7 g lay on the anvil. The severed iron was subsequently used to make 2002/7, which was a copy of a pin with a coiled head, and a low swan’s neck (copy of Ris; Becker 1961, Pl. 116/4). After 27 heatings, the smith concluded his work with a pin weighing 5 g. In forging the small needle with an eye, 2002/9 (Müller 1895, no. 133), the smith exchanged his 1 kg hammer for a smaller hammer, on which the head only weighed 300 g. Subsequently, he forged 4 g ingot iron from 2001/II into a pin weighing 1 g, using 17 heatings. The smith was not satisfied with the progress of the forging using iron from 2001/II and believed that there were pockets with more than 0.8% carbon in the iron. This was confirmed by subsequent metallurgical analysis.

The smith also used 17 g of iron from 2001/III to make a pin with a ring-shaped head, 2002/10 (copy of Vejle; Jensen 1997, fig. 88). The pin was forged in 50 minutes and with 42 heatings – it weighed 9 g. Iron ingots from 2001/II were also used in experiments with making copies of knives from the cemetery at Lønhøjgårdsvej (SKJ 175). The knives are from the Late Pre-Roman Iron Age and were all forged as technological type 1. [Note 1] For knife 2008/1, use was made of a piece of ingot iron of 38 g and, by 18 heatings, a small corner-handled knife (SKJ 175x75) was copied. During forging, sand was scattered on the piece at the 4^th, 8^th and 12^th heating. A total of 9 g of hammer scale and welding balls was collected on and around the anvil. Knife 2008/2 was forged from 30 g of ingot iron, making a corner-handled knife of 24 g, and 2008/6 (SKJ 175x513) was forged from a piece of ingot iron weighing 56 g with 37 heatings in 41 minutes and 418 hammer strokes, producing a knife weighing 32 g. A piece of cut-off iron, weighing 8 g, was picked up, as well as 10 g hammer scale and welding balls around the anvil. After five heatings, the piece was dipped in sand and after a further four heatings it was brushed. It was not welded.

The experiments involving smelting of Zealandic bog iron ore and forging of the iron which was produced show that Zealandic bog iron ore does not differ much from that from Jutland. Of course, it takes much longer to copy exact processes and forms than it took an Iron Age smith to forge a knife or a pin. But the systematic trials show that it is possible to forge artefacts from iron extracted from Zealandic bog iron ore which, in appearance and weight, correspond closely to the artefacts that were forged in the Iron Age – at the same time as the forging technique and the iron quality were the same.

It was a minority of archaeologists who, prior to the discovery of the furnaces at Espevej and Skydebjerggård, would have believed that, in the Iron Age, iron was produced from bog iron ore from Zealand. But the discovery made it necessary to re-evaluate our perception of the distribution and applicability of the bog iron ore. Since the first symposium on the Iron Age of Zealand, the archaeological record has been expanded, analytical methods have developed and experiments have been carried out, all of which suggest that iron production could have been much more than a theoretical possibility for the Iron Age farmers on Zealand.

Notes

1. The various forging techniques which were used for iron knives during the Danish Iron Age are explained in Lyngstrøm 2008b, fig. 77.

2. The slag inclusions are described according to the Swedish Jernkontoret’s standard SS 11 11 16, where a letter denotes the appearance of the slag inclusions and a figure their quantity. This is depicted in Lyngstrøm 2008b, fig. 12.

Literature

Andersen, A.H., F. Kaul & O. Voss 1987. Danmarks ældste jernudvindingsovne. I: Rigsantikvarens Arkæologiske Sekretariat (red.) Danmarks længste udgravning. Arkæologi på naturgassens vej 1979-1986, s.176-180.

Becker, C.J. 1961. Førromersk jernalder I Syd- og Midtjylland. Nationalmuseets Skrifter, Større beretninger VI. København.

Blakelock, E., M. Martinón-Torres, H.A. Veldhuijzen & T. Young 2009. Slag inclusions in iron objects and the quest for provenance: an experiment and a case study. Journal of Archaeological Science 36, s. 1745-1757.

Bodilsen, I. 1993. Kregme i vikingetiden. En undersøgelse ved museumsgådern ”Birkely”. Arkæologi i Frederiksborg amt 1983-1993, s. 115-140.

Borby Hansen, B. 2006 Langerød og Tuse – 3000 år før motorvejen. I: Wickman, N. & B. Juhl Andersen (red.) Vejens skjulte skatte – Holbækmotorvejens forlængelse, s. 21-45.

Buchwald, V.F. 1998. Myremalm. Geologisk Tidsskrift 1998:1, s. 1-26.

Buchwald, V.F. 2005. Iron and steel in ancient times. Historisk-filosofiske Skrifter 29. Det Kongelige Danske Videnskabernes Selskab. København.

Buchwald, V.F. 2008. Iron, steel and cast iron before Bessemer. Historisk-filosofiske Skrifter 32. Det Kongelige Danske Videnskabernes Selskab. København.

Christensen, W. 1943. Nogle Analyser af Myremalm fra Danmark. Naturhistorisk Tidende 7. Aargang Nr. 5, s. 70-71.

Christensen, W. 1966. Myremalm. Dansk Natur – Dansk Skole årsskrift 1966, s. 41-67.

Christensen, W. 1968. Myremalmens historie i Danmark. Hedeselskabets Tidsskrift.

Giles, M. 2007. Making metal and forging relations: ironworking in the British Iron Age. Oxford Journal of Archaeology 26(4), s. 395-413.

Grøn, O., A.H. Krag & P. Bennike 1994. Vikingetidsgravpladser på Langeland. Rudkøbing.

Hansen, K.M. & B. Staal 1996. Enkeltgård og landsby fra Jernalderen – de arkæologiske undersøgelser på Fakse Losseplads. Kulturhistoriske Studier, s. 60-66.

Hedeager, L. 1988. Bind 2. Danernes land, fra ca. år 200 f.Kr. – ca. 700 e.Kr. Gyldendals og Politikens Danmarks Historie. København.

Jensen, J. 1997. Fra Bronze- til Jernalder. Nordiske Fortidsminder serie B, bind 15. København.

Jensen, J. 2003. Danmarks Oldtid, Ældre Jernalder 500 f.Kr. – 400 e.Kr. Gyldendal

Jouttijärvi, A. 2006. Smed eller smeltemester – analyse af slagger fra Bjergene. I: Wickman, N. & B. Juhl Andersen (red.) Vejens skjulte skatte – Holbækmotorvejens forlængelse, s. 47-56.

Lund, J. 1991. Jernproduktion i Danmark i romersk jernalder I: Fabech, C & J. Ringtved (red.) Samfundsorganisation og Regional Variation. Norden i Romersk Jernalder og Folkevandringstid Jysk Arkæologisk Selskabs Skrifter XXVII, 1991, 163-169.

Lund Hansen, U. & S. Nielsen 1992. Sjællands Jernalder. Beretning fra et symposium 24.IV.1990 i København. Arkæologiske Skrifter 6. København.

Lyngstrøm, H. 2002. Thomsen, slaggerne og den eksperimentel-arkæologiske metode. Drik og du vil leve skønt. Publications from The National Museum, Studies in Archaeology & History vol. 7, s.132-138

Lyngstrøm, H. 2008a. Kasserede Kilder. I: Lyngstrøm, H, O.T. Kastholm Hansen, E. Brinch Petersen og B. Storgaard (red.): Fra Vandkunsten til Vestpynten af Amager – 20 arkæologiske essays, s. 55-60.

Lyngstrøm, H. 2008b. Dansk Jern – en kulturhistorisk analyse af fremstilling, fordeling og forbrug. Nordiske Fortidsminder serie C, bind 5. København.

Lyngstrøm, H. 2009. Technologia produkcji zelaza i wyrób nozy zwlaznych na terenie danii od 500 r. przed chr. Do 1000 r. po chr. Wiadomosci Archeologiczne Tom LX, s. 189-195.

Müller, S. 1895. Ordning af Danmarks Oldsager. II Jernalderen. Kjøbenhavn.

Rørdam, K. 1893. De geologiske Forhold i det nordostlige Sjælland. Beskrivelse til Kaartbladene ”Helsingør” og ”Hillerød”. Danmarks geologiske Undersøgelse 3. Kjøbenhavn.

Rørdam, K. 1894. Om Betingelserne for dansk Staalproduktion. Nogle Tilføjelser fra den praktiske Geologis Side. Ingeniøren 3:48, s. 239-241.

Rørdam, K. 1899. Beskrivelse til Kaartbladene Kjøbenhavn og Roskilde i 1:100,000. Danmarks geologiske Undersøgelse 1:6. Kjøbenhavn.

Voss, O. 1991. Jernproduktionen i Danmark i perioden 0-550 e.Kr. I: Fabech, C & J. Ringtved (red.) Samfundsorganisation og Regional Variation. Norden i Romersk Jernalder og Folkevandringstid Jysk Arkæologisk Selskabs Skrifter XXVII, 1991, s. 171-184.

Wagnkilde, H. 1999. Slaviske træk i bornholmske grave fra tiden omkring kristendommens indførelse. Meta 1999:2, s. 3-20.

Upubliceret

HAF 07/01 Teknisk rapport og beretning for Dansk jern. Historisk-Arkæologisk Forsøgscenter, Lejre.

HAF 09/02 Teknisk rapport og beretning for Forsøg med dansk myremalmsjern. Historisk-Arkæologisk Forsøgscenter, Lejre.

HAF 17/08 Teknisk rapport og beretning for Herkomstbestemmelse af myremalmsjern. Historisk-Arkæologisk Forsøgscenter, Lejre.

Industrirådet. 1918. Myremalmsundersøgelser 1918, rapport fra ”Myremalmsudvalget”.

Skive Folkeblad 1957 Malmen i den jydske jord – et moderne eventyr (21.7.1957).