'Look at all the BONES...' 

Addition of Animal bones into a Bloomery Iron Smelt
June 20, 2020
Wareham, Ontario


    Several recent papers have suggested the presence of small fragments of bone sometimes found within the debris fields related to bloomery iron smelting point to a possible 'ritual' practice, even so far as proposing a functional impact on iron bloom quality. How might the physical process within a complete iron making sequence effect the ability of bone to endure, and thus remain to be recovered archaeologically? A typical 'short shaft' furnace will be constructed and operated through to bloom extraction on a clean working surface. Both bone pieces and meat containing bone of several animal types will be added, before, during, at at the final stage of the smelting process. Afterwards, the debris field will be examined in detail to determine what remains of the bones.



Figure 1B : Smelting area the following morning

    A fast look the day after the smelt did not immediately reveal any traces of the added bone pieces. At this point, the interior of the furnace was still too hot to allow for detailed examination. The smelting area was left un-modified, with overhead cover to protect from rain, until July 15.

Figure 2B : Area as examination starts - view to east Figure 3B : Area as examination starts - view to north Figure 4B : Laying out the grid rods

Explanation of the numbering below :
grid 5.1
grid 5.2
grid 5.3
grid 5.4
grid 5.5
5.1 empty 5.2 empty 5.3 furnace rear north 5.4 furnace rear south 5.5 large plate of tap slag placed to cool
grid 4.1
grid 4.2
grid 4.3
grid 4.4

4.1 support blocks / piece of furnace wall / some cleared tap slag 4.2 support blocks with tuyere above 4.3 furnace north 4.4 furnace south (image for 4.5 missing) fragments of extraction arch cover / broken tap slag
grid 3.1
grid 3.2
grid 3.3
grid 3.4
grid 3.5
3.1 tap slag cleared to east side 3.2 debris / stones pulled from furnace base 3.3 broken furnace wall, slag bowl / main path for extraction 3.4 stones pulled from furnace base / slag bowl pieces 3.5 stones pulled from furnace base / tap slag pieces
grid 2.1
grid 2.2
grid 2.3
grid 2.4
grid 2.4
2.1 empty 2.2 to east is edge of extraction debris
arrow marks visible bone fragment
2.3 large piece of slag bowl (to west) / east is edge of extraction debris 2.4 largely empty 2.5 discarded tap slag to south edge
grid A1.1
grid A1.2
grd A1.3
grid A1.4
grid A1.5
1.1 magnetic group pieces
(placed after extraction)
1.2 piece of slag bowl (placed during extraction) 1.3 large piece of slag bowl (to east) 1.4 empty 1.5 empty

Figure 5B : Interior of furnace, view from top of shaft Figure 6B : Detail of furnace interior Figure 7B : Interior of furnace, view from the extraction arch

    Although not part of the main purpose of this smelt, the interior of the furnace clearly shows effects that can be attributed to lower internal temperatures than have been observed in the past :
    Lower internal temperatures over this smelting sequence were also predicted by the longer than normal charcoal burn rates recorded.

    As the grid photographs were taken, care was taken to look for any bone fragments that were exposed on the surface. One one piece was observed, located at grid +50 cm east and +60 cm south. This was clearly a section of one of the chicken leg bones, deposited with larger pieces of unburned charcoal.


Figure 8B : Bone piece recovered on debris surface (detail to 2.2 above)

    The working area was cleaned through the fingers, each square cleared individually. Fragments down to 5 mm were examined, and any pieces judged to be bone pieces had their positions noted and were bagged individually for later scaled photography. It should be noted that with one individual undertaking this work, no specific images of recovery in place were made. A total of 15 pieces were recovered,
The list below is taken from field notes, locations are +east / +south in cm. Sizes maximums in mm

50 / 60       
4 small pieces, hollow porous tubes (chicken thigh)
found near top of loose charcoal materials pulled out during extraction
20 x 13 (triangular)
20 x 8
32 x 8
7 x 6

80 / 50
1 small piece, slightly cylindrical, likely chicken thigh, but heavily carbonized
5 x 20

80 / 85   
1 piece of flat porous bone (turkey breast bone?)
42 x 26 x 6
3 pieces of flat very solid bone (beef leg section)
found on top of fines layer raked out
45 x 26 x 12
30 x 15 x 5
(one since broken)

70 / 100
1 large piece, with slag attached to one side (clearly a piece of turkey breast bone)
found in same orientation as originally placed
90 x 45 x 40
4 smaller fragments of porous bone (may be fragments of above?)
25 x 12 x 7
22 x 12
15 x 7
15 x 5

    The images of the bone pieces recovered are photographed against a 10 mm grid.

Figure 9B : Largest bone recovered, section of turkey breast bone from 70 / 100
shown with slag attached side down
Figure 10B : Fragments of chicken leg bone, primarily recovered from 50 / 60
Figure 11B : Pieces of solid bone, beef leg bone segments, recovered from 80 / 85

Conclusions :

A)    The collection of fragments recovered at grid 80 / 85 were contained within a mass of charcoal fines, obviously material that had formed the initial base of the furnace. These fines would have been raked clear out of the furnace at the beginning of the extraction process to create more working space. The most obvious origin of these pieces would have been the bones placed on top of the fines and fine ash layer before the addition of charcoal. From previous experience, it is known that the air blast does not allow for much actual combustion in this area. Heat is generated from the slag bowl, which over the course of the smelt will slag down, usually coming to rest on the top of the fines layer.

B)    The large turkey breast bone recovered at 70 / 100 is most likely the meat covered piece placed at the top of the furnace at the end of the smelt sequence. It had been placed initially with the central ‘fin’ upwards. If it stayed in this orientation as it dropped, it would have presented the opposite side to the fluid slag bowl, which could have resulted in the same encrusting slag as found. This last set of pieces were exposed to the internal heat of the furnace for the shortest time, suggesting the water contained in the encasing meat had helped to protect this specific piece during the decent. That no other fragments (especially of the larger the beef bone) were found suggests that without this protective meat layer, the internal heat was enough to consume the other pieces from this layer.

C)     The small fragments of chicken leg bone recovered from 50 / 60 are more uncertain as to their initial placement. They are furthest from the extraction arch, suggesting they were pulled clear early in the emptying of the furnace. They show no surface carbon marks, which suggests they were exposed to heat, but in an oxygen and carbon reduced area. As the bones placed in the bottom of the furnace were also deposited on an ash layer (about 2 cm thick) this suggests that these, quite thin and fragile bones, were initially placed at that location.

D)     There was only a single, very small fragment that was recovered that showed extensive carbonization, the piece of chicken thigh bone at 80 / 50. The exact conditions that this bone was exposed to are uncertain. It’s find location, similar to group at A may suggest a similar origin?

    I had participated in an experimental test by Emma Harrison of bone added to a bloomery iron smelt, an addition to the first phase of the ‘Turf to Tools’ project at the Scottish Sculpture Workshop (SSW) in 2014. (1B) That experiment would form part of Harrison’s 2015 initial report ‘Iron, Turf and Bone : An explorative study into Viking Age Icelandic bloomery smelting and the effect of bone in iron smelting’. (2B) At SSW, a fairly standard short shaft furnace was used for several smelting operations. Harrison added a total of 5.2 kg of prepared (cleaned) beef leg bones broken to small pieces, as a final layer after the main sequence of ore charges, with a final charcoal cover. On final examination of the debris after the bloom had been extracted and the furnace allowed to fully cool, only 7 small fragments of bone were recovered, the largest (approximately) 40 x 23 mm. The majority of the bone added was completely consumed within the furnace, with no discernible effect to the already existing bloom itself.

Figure 12B : Only Bone fragments recovered from Harrison’s 2014 test
Image by E. Harrison, pg 35 (poor quality scan from printed copy)

    Taking note of Harrison’s earlier work, and comparing it to the results recovered during this experiment, a number of elements become clear :

    It is important that this experiment makes no attempts to suggest ‘why’ Norse smelt masters may have chosen to, or even if they ever did, add bone pieces into a working furnace. Extreme care must be taken in any attempt to apply ritual practice from other cultures, remote in time, geography, working methods, and especially radically differing world views.
    The fairly consistent destruction of any bones added during charcoal charging does suggest that it is unlikely to recover archaeologically anything beyond the uncommon and smallest traces of added bones. The only likely situations where bone has much chance of leaving traces after being added to a functioning bloomery iron furnace :

Notes :

1B) The sequence data for this specific smelt.

2B) The final submitted version, which included a number of further experiments and considerable analytical detail :
 ‘ Smelting the Ancestors? An experimental investigation into the effect of bone on slag systems and extractive metallurgy ‘ (2016)
Available as a PDF

3B) Standard modern cremation practice (for a complete human body) is temperatures into the range of 1000 C, over 2 - 3 hours. Typically anything placed on the top of this type of furnace would take about 30 - 45 minutes to drop to the slag bowl level, subjected internal temperatures at 1250 C plus.

unless otherwise credited - Text and photography © Darrell Markewitz