'Look at all the BONES...' 

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

Smelt Team:
Neil Peterson, Rey Cogswell
Smelt Master : Darrell Markewitz


    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.


The intent of this experiment was to build a standard ’Short Shaft’ furnace, on a prepared clean base, then undertake a careful examination of the final debris field.

Figure 1 : underlay of concrete slabs - view to east
Figure 2 : sand pad

    As this was intended to be a short duration experiment examined in detail, a hard flat base was created by laying a set of 24 x 24 inch (60 x 60 cm) concrete patio slabs down over the natural grass surface of the main smelting area. A square frame was created from 6 foot long pieces nominal 2 x 2 inch lumber (so about 4 x 4 cm by 1.8 m). This was used to contain a fill of course sand, establishing a flat surface with a depth again at 4 cm. This combination would allow for easy recording of any created debris, and also in clearing of the area afterwards.

    As the function of the furnace itself was not under investigation, a long proven design and building methods were used. (1)
The furnace shaft was raised on a plinth of stone blocks, filled with charcoal fines. This allows for both better control of slag levels, and assists in the extraction process. There were a selection of natural granite blocks on hand. (acquired for the Icelandic Stone Block series, also ongoing at the same time). These were irregular sizes, but all roughly 10 cm tall, and varied in width and length. These were laid as one full circle, roughly 30 cm internal diameter (ID), with a second layer positioned with a gap to the front (extraction) side.


Figure 3 (2) : Stone plinth as established

    The mix used for the construction of the furnace shaft was powdered potters clay (EPK), locally dug course sand and dry shredded horse manure, combined in rough thirds by volume. This was hand blended with water to what is considered an ideal consistency for building.
A metal form at 28 cm wide, wrapped in paper to prevent clay sticking, was used to establish and control the internal diameter.


Figure 4 : At the completion of the lower layer, the metal form in place.

    Figure 4 also shows the establishment of a ‘Beardsley Break’, here seen with a coating of charcoal fines. (3) As is the normal practice, the body of the furnace was built up of hand sized ‘bricks’ of the clay mix, pressed together and around the metal form. As a full batch height was finished, the form was pulled out and clear, with the interior filled with a mix of sand and wood ash to stabilize and assist drying.  As the structure got higher, it proved necessary to spiral wrap rope around the exterior to prevent slumping (impressions of the rope can be seen in figure 5).

Figure 5 : Finished furnace, extraction and tapping arches cut,
tuyere installed.
Figure 6 : Overall measurements

    The extraction arch was cut to almost the height of the tuyere, total 20 x 20 cm, with a smaller tapping arch on the lower edge. The forged copper tuyere was set to 22.5 degrees down angle, 5 cm proud of inner wall, and the base depth adjusted via charcoal fines to 20 cm. With the total height of the clay walls at 66 cm above the stone plinth, the net result was stack height (above tuyere) of 48 cm.
A number of concrete blocks were used to support the tuyere and air system. The placement of these would remain clear within the debris field during later examination after the completed smelt


Figure 7 : Overall working area, at the start of the smelt, view to south east


The combination of gentle drying fire the day before, plus the pre-heat sequence also using fine twigs, served to place a layer of wood ash over the charcoal fines base. Some of the fines had also been consumed, so the functional base distance was closer to 29 cm (below centre of tuyere).

ash base

Figure 8 : Ash over charcoal fines, at the start of the experiment.

    The sequence undertaken for the smelt itself was again fairly standard :
    There was considerable variation in the burn rates, with outliers at 11 and 23 minutes / bucket.

Figure 8 : Working area, the following morning

ADDITIONAL : Photo series over the progression of the smelt.

Adding BONES :

    A quantity of animal bones of various types had been gathered for addition at various points during the progress of this iron smelt. Some were retained after household meals, the pieces with meat on were donated by the local grocery store meat department (Foodlands, Dundalk ON). 


Figure 9 : Bone pieces as collected.
    Although the pieces were photographed against a scale ‘as fresh’, no other specific measurements were made of sizes or weights. Because of this, the results should be considered ‘qualitative’. The initial concept was to make two additions, one at the start of the smelt process and one half way through ore additions, hence the grouping seen in figure 9.

    At the end of the preheat phase, the extraction arch was opened and a set of bones was placed over the hot ash and still burning charcoal fragments :

bones 1

Figure 10 : Bones at base level, showing individual placement. Tuyere is to the left side this image.
    The first pieces of charcoal were carefully added through the extraction arch, to ensure the few remaining hot pieces of charcoal would serve to ignite the upper layers. Once it was clear this had happened, the arch was replaced, and the furnace filled with rough charcoal. From this point the smelt sequence was carried out as normal.

    At roughly two hours into the sequence, a second group of bones was placed on the top of the filled furnace. At this point (13:40) there had been a total of 5 kg of combined iron slag and ore added to the furnace (so still early in the iron reduction process).

bones 2

Figure 11 : Bones placed part way through the smelt sequence
    With the decision to make three separate bone additions, some of the larger pieces (Pork ribs and Turkey breast, both with meat) were cut in two. The bone pieces added at this time were placed to one side of the furnace, away from the tuyere. This was done to echo the normal placement of ore, as the hottest part of the furnace is directly above the tuyere, typically resulting in faster consumption to that side (and hence faster dropping down the stack in this area). The temperature at the top surface of a working furnace has been measured previously to the range of 400 - 500 C, so even in the short time it took to grab the camera and record this image, heat effects are visible (meat cooking, exposed fats starting to carbonize) (5)
    These pieces have been intentionally placed away from the side of the furnace above the tuyere. The area close around, especially just above, the tuyere is the hottest part of the furnace, so charcoal above this is consumed faster. Normal practice is to place ore in the same location seen with the bone placement, to ensure the best possible progressing through the lower reduction zone.

    At the end of the ore additions (estimated at about three hours since the initial charcoal filling), a last set of bone pieces was placed :

bones 3

Figure 12 : Bones placed at the end of ore additions.
    One extra last full bucket of charcoal was added to cover these bones, then the charcoal in the stack allowed to burn down in preparation for extraction.

    Another aspect visible in figure 13 are the short bright orange streaks. These are likely small particles of ore that have been reduced to fragments of metallic iron, but under the effect of the air blast are being re-oxidized (burned). This may even be a sign that the combination of too high a slag bowl and internally draining liquid slag is resulting in the bloom being eroded by the blast almost as fast as new metal is being deposited. (One possible cause of the smaller than expected final bloom size, which was found to be a fairly flat ‘cake’, rather than the more typical half bowl shape as it was extracted.)


Notes :

1) See : ‘If you don’t get any IRON…

2) Images taken during the build suffered in quality because of the contrast between the bright sunlight and the shadow created by the temporary overhead cover used at that point.

3) This feature named after George Beardsley, who had demonstrated it’s use during a cast iron session at the Scottish Sculpture Workshop in 2014. The purpose is to prevent any cracking of the clay shaft from extending upwards, caused when the clay shrinks while formed over an irregular stone base.

4) Air was controlled by the sliding plate blast gate, set to marks for 800 / 900 / 100 LpM. More precise measurements undertaken during experiment # 90 (October 2021) have been applied here to correct the original numbers.

5) It is worth noting that despite the high temperatures over the top of the furnace, the upper charcoal is not showing much effect. It is usual to see contained water vapourizing (but hard to photograph). Inside the walls of the furnace at this level there is no actual free oxygen to support combustion. Air / oxygen injected at tuyere level has combined with carbon from the charcoal, with further reaction with the iron oxide ore creating a carbon dioxide atmosphere at the top of the furnace.

unless otherwise credited - Text and photography © Darrell Markewitz