Lab Summary Week 4: Ritual Fire

John Hoffman, Helen Murphy, and Sam Wege


This week, our lab had two separate but related focuses. We experimentally analyzed the methodology of two rituals: 1. Greek sacrifice of bovine femurs and tails and 2. Anglo-Saxon cremation rites. The goal of the study of Greek sacrifice was to recreate an ancient practice based off of only the snippets of description we have surviving in order to attempt answers to questions otherwise nearly impossible to examine with the extant historical and documentary record. Especially important among these was an attempt to develop an understanding of the timing of these rituals, with a special focus on how tension and theater would have contributed to the spiritual significance of the event. The cremation study sought to determine some of the specifics of where a body would have been placed on a pyre in order to achieve full cremation, as well as to what degree the fire would have had to be monitored and tended during the rite. We also wanted to gain a better experiential sense of what it would have been like to be a part of such an event.

Greek Sacrifice

Sacrifice Background

Ritual religious practice was an integral part of ancient Greek life. This practice ranged from small scale, domestic rites to expensive, public, and extravagant ceremonies. At the heart of many religious ceremonies was an act of sacrifice operating as a bridge between earth and the heavens which allowed for communication between human and divine (Parker 2011: 124). Shared cultural memory and values from the 7th-5th centuries BC, as expressed in myth, epic, or theatre, demonstrate that animal sacrifice was a valued and commonly practiced part of religious tradition which continued from an older, idealized Bronze Age (Homer Il. 1.38-49; 1.545-566; Homer Od. 3.455-460; Aristophanes Peace 1016-1059). Images, especially those decorating ceramics, of these ceremonies provide further documentation of their existence and give a glimpse into their practice. It has been argued by contemporary scholars of Greek religion and religious practice that these ceremonies of sacrifice were essentially paired with communal feasting (Parker 2011: 127). Indeed, Theophrastus provides commentary in the 3rd and 4th centuries BC on the Greek practice of religious sacrifice, arguing that the Greeks only sacrifice those animals which are specifically pleasurable to them, claiming that “we persist in sacrificing to the Gods, for the sake of our own pleasure, and not for the sake of gratifying the Gods” (Theophrastus ap. Porph. Abst. 2.25-26, from Parker 2011: 127). Ancient sacrifice, religious ceremony, communication with the gods, and communal feasting were all fundamentally tied up in one another. As such, it is no great leap for us to argue that the combination of religious ceremony and communal feasting would result in a form of community-building integrated into the religious ceremony itself; the religious ceremony essentially must also become a social and communal ceremony. Firm belief in this argument underlies our discussion throughout the following report.

Today, it is more challenging to examine this communal aspect of ceremony specifically than it is to investigate methods or empirical results of sacrifice; there is no archaeological record of shared group emotional reaction, nor is there an extant historical record about the feelings shared by a community during sacrifice. As such, we turn to experimental archaeology as a method of engaging with these ceremonies in a way that allows us to try to get at the heart of what was happening during these ceremonies that is omitted from or uncertain in the historical and material record. What might have happened during the ritual itself beyond the simple fact of sacrifice? How might the sacrifice have demonstrated its success or failure? What emotions might a community have felt during different stages of the sacrifice, especially in the face of potential failure – and the social consequences associated? How might these emotions have shaped and brought together a community?

Groups A and D, therefore, were assigned to the large fire ring, where they constructed a miniature funeral pyre made of logs laid perpendicularly across one another. This pyre was stuffed with brush and covered by sticks, but not before two pork shoulders had been laid within it — one on top of the pyre, and the other beneath it. The goal was to ascertain how a corpse might have been laid on a pyre during a cremation rite, as the archaeological record cannot provide an answer to this question. We hoped that burning the pyre and observing how the pork shoulders had fared after the burn would provide some illumination in this respect, as well as giving us more experiential insight into what being. apart of such a ceremony might have been like.

Sacrifice Research Questions

To address the challenges posed by the archeological and textual evidence, we developed a series of research questions that focused our study on the logistical and emotional aspects of the religious rituals.

First regarding the sacrificial altar itself: Is the structure depicted on Greek vases a viable construction for sacrifice? If so, how long does the stacked wood maintain its structural integrity? What tending might the fire require? Given these, how long a time could the ceremony potentially have lasted? How short?

Relating to the articles being sacrifices: What is the timing of the burning of both the wrapped femur and bovine tail? Do the climaxes of these events overlap? Is tension built into the rituals, that is, does it look like it might fail (are the Gods listening)? Furthermore, can these rituals be contained in a metal brazier (similar in size to recovered metal artifacts that have been hypothesized to serve as sacrificial braziers)?

Sacrifice Methods

Archeological and textual evidence show that sacrifice was an essential part of Mediterranean spirituality for the better part of the 1st millennia BC up through 300 AD. Our experimental methodology was developed from images on vases and textual evidence which depict and describe the various aspects of these rituals, as seen in the images below.

Preparation of Fire

Our experiments were conducted in round fire pits, with wood laid to construct a log cabin structure that could serve as a stable “altar” for the sacrificial elements. In each of the two fire pits 12 logs were placed in rows of 3, each row perpendicular to the one below. Some small twigs were placed in the center to aid in ignition. 

The sacrifice fire prior to lighting (Photo: Sam Wege)

Three essential objects of sacrifice have been identified from historical and archaeological records: a femur wrapped in fat (i.e. Homer Od. 3.455-460); tails (Van Straten 1995); and splanchna, the heart, liver, and kidneys of the animal being sacrificed (Morton 2015: 66-67; Arist. Part. an. 665 a28 – 672 b8). These objects of sacrifice were burned by being placed flat on the altar itself. The ritual concluded with the dousing of the bone in wine (Homer Od. 3.459) and its removal from the fire for photographing and interpretation. 

Preparation of Objects of Sacrifice

Wrapped femurs were covered in omentum fat such that no part of the bone was showing. The goal was to wrap each femur in two full sheets of omentum (as seen in below) in accordance with the ritual’s description in Homer. Two femurs were left unwrapped to serve as controls to compare the post burning damage and discoloration of the bone. See below images for the bone before and after wrapping, as well as for the process of wrapping.

We also roasted splanchna over the fire on spits. For this, we used two different styles of spit to investigate the effect of various utensils for cooking. One was an L shaped spit and the other a thin, wide, flat spit. We were testing to see if this change would have an effect on the ability to cook as well as quality of flavor for the splanchna.

Roasting splanchna (Photo: Sam Wege)

Hypotheses were formed for the interaction between fire and offering for both wrapped femurs and tails. We hypothesized that the femurs would sputter for a long time with little change to fire but then there would be a sudden, tall (1-2m), and hot burst of flame somewhere between 5 and 10 minutes into the process. This hypothesis was derived from Morton 2015’s observations of the interactions between wrapped lamb femurs and fire. We adjusted his findings to be slower but more intense with the bovine because there is so much more fat and overall mass to burn off first. We hypothesized that the tails would follow the pattern established by Van Straten 1995, curling upward after a prolonged period of continuous drooping into the fire (see image below for scenes of this from ancient images).

Curling tails (from Van Straten: 1995)

Sacrifice Results

We were able to produce significant experimental results at both firepits, with multiple runs emulating the evidence we have for these rituals. We found that tails curl and wrapped femurs do indeed ignite, and these events occur at approximately the same time under the same altar conditions, though this time varies considerably between individual runs.

When the curl or ignition was delayed, all workers reported feelings of anticipation, with some disappointment at the failure. We imagine that these feelings would have been multiplied under the spiritually significant conditions of the ritual in the classical era.

ItemBone weight (g)Fat weight (g)Time to fat ignitionTime to start tail curlTime to tail fully curledTime to removal
1st wrapped bovine femur without tail2,7608243 min 25 secN/AN/A30 min 26 sec
2nd wrapped bovine femur without tail3,3061,0222 min 20 secN/AN/A12 min 20 sec
Lamb femur177N/A6 min 03 secN/AN/A9 min 00 sec
1st tail aloneN/AN/AN/A11 min 14 sec14 min 34 secLeft in brazier
2nd tail, with unwrapped bovine femurSee belowN/ANo additional flame12 min 04 sec30 min 08 sec32 min 00 sec
3rd tail, with unwrapped bovine femurSee belowN/ANo additional flame8 min 49 sec12 min 23 sec*12 min 23 sec*
Unwrapped bovine femur3,503N/ANo additional flameN/AN/A32 min 00 sec
Firepit 1 Immolation Results

*experiment concluded before tail was able to complete the curling process.

A bovine tail burning (Photo: Sam Wege)
ItemBone weight (g)Fat weight (g)Time to fat ignition Time to start tail curlTime to tail fully curledTime to removal
1st wrapped bovine femur with tail23822834 minutes2 minutes**Inconclusive19 minutes
Lamb femur1501504 minutes*N/AN/A6 minutes
Tail aloneN/AN/AN/A2 minutes6-10 minutes**Left in brazier
Unwrapped bovine femur1777N/ANo additional flameN/AN/A18 minutes
2nd wrapped bovine femur with tail13447505 minutes2 minutes5 minutes21 minutes
Firepit 2 Immolation Results

**Our first two runs with tails appeared to have curls in the wrong direction; however, for the second, the appearance was discovered to be produced by a structural failure of the altar platform. It is possible this was the case for the first run as well. During the first run, the tail was flipped over after 6 minutes due to the wrong direction of the curl, after which, the tail seemed to straighten. This effect was not replicated in any other run.

Of significance, the bones burned with and without fat wrapping appear similar post sacrifice. Intuitively, bones left for longer in the fire are more structurally damaged. The bone wrapped in extra fat incurred significantly more structural damage in a similar amount of time as the bone wrapped in minimal fat, indicating that the fat induced a higher heat on the bone while burning.

Sacrifice Discussion

The results of our sacrifice experiment show an exciting similarity to the archeological, textual, and pictographic evidence we have for the ancient rituals; however, there are a number of important considerations to account for before making sweeping conclusions. Our results were mutually inconsistent, with the timing of the tail curling and fat igniting ranging from three to fifteen minutes after placing the items onto the altar. Groups C and E at firepit 2 ran the experiment under different altar conditions every time, which inhibits reasonable comparison of the runs within the group. These different conditions are not wholly problematic, as they illuminate altar construction and integrity as possibly the most important controlling variable in the ritual. The final run for firepit 2 best replicated the ritual depicted in the evidence, but from there, we should have replicated that set-up for multiple additional runs to better understand the consistency or inconsistency of the ritual. Our experience of tension occurred during the many runs with imperfect altars and may thus be an inaccurate understanding of realistic tension during the event. 

Although firepit 2 was able to coordinate the timing of the burst of flame with the curling of the tail to create an especially dramatic spectacle, firepit 1 found that tails consistently took far longer to curl than femurs to release the great bursts of flame. It is necessary to note that at firepit 2 the tail and femurs were put on concurrently, while at firepit 1 they were put on separately. An essential question of focus for future experimental research should be to conduct more tests to see if placing them on together directly lends itself to their concurrent spectacle. 

For more future experiments, we suggest replicating the fire depicted on the vases and placing the sacrificial articles perpendicular to the top wood layer (the set-up of the final altar). We found that the altars were less successful at maintaining structural integrity and support for the offering when the object of sacrifice was placed parallel to the wood. Successive runs should be conducted with a new top layer, but ideally, an entirely new altar would be constructed for each run to replicate the depiction of individual sacrifices in the texts. 

Finally, we would like to address some of the variables that impacted our ability to conduct a replicable experiment. Many of these stemmed from the fact that we were working outside. Foremost was our inability to control the wind. It was challenging at times to judge the size of the fire and the amount of growth by comparison. Additionally, as mentioned above the fire was not perfectly controlled. As such, there were surely fluctuations in temperature. Future experimental design should either better control for changes in fire temperature, or systematically record temperature to include differences in discussion. 

Sacrifice Conclusions

From the sacrifices we conducted, we may conclude the following:

  1. The event would have been a theatrical spectacle and based on our timings, the sacrifice of the femur and the tail could have been conducted simultaneously for dramatic effect as both the burst of flame and curl of the tail occur between 3 and 5 minutes in a well-built fire.
  2. The woodpile depicted on the vases is workable and is easily maintained to enable multiple visible sacrifices.
  3. In a well-built fire without extraneous circumstances, there is not much time for tension before the flames burst and the tail begins to curl (only a minute and a half to two minutes), though in less optimal circumstances, the ritual takes more time and occasionally appears to putter out before reigniting, producing the theorized tension; however, this is likely a product of methodological error and requires further experimentation.

Anglo-Saxon Cremation Rites

Cremation Background

“Anglo-Saxons” in early medieval England practiced multiple different death rites, including both inhumation and cremation ceremonies. Given the lack of textual records of these ceremonies, apart from a passage in Beowulf that describes a cremation, evidence for their proceedings comes largely from the archaeological record, which is biased toward materials that actually survived for over a thousand years. Additionally, most cremated remains have been moved from their original place, either by people collecting bone remnants to inter after the ceremony itself or by various other occurrences throughout time. Thus, it’s difficult to know anything specific about how the ceremony was performed. How was the body laid on the pyre? How was the pyre itself built? Did more wood need to be added throughout the cremation? How did it feel to be a part of this event?

Cremation Research Questions

What kinds of temperatures are achieved during the cremation process, and how do those vary based on location, time, and method of burning? Are they high enough (800 degrees Celsius) to achieve cremation?

How does the placement of the body (pork shoulder) in the pyre affect its cremation?

How long does the cremation process take, and does the fire require more wood or tending?

Cremation Methods

Textual evidence from this period of Anglo-Saxon history is basically nonexistent, and so the majority of our knowledge comes from archaeological evidence, mainly cremated remains that were subsequently buried and thus preserved. Jacqueline McKinley finds that the hands and other extremities of the body frequently are not fully burnt in cremated remains, and suggests that this indicates corpses were laid on top of pyres, where hands and feet further from the heart of the fire and low in flammable fat would not have burned as well (1994, from Lucy: 2000). Calvin Wells, on the other hand, believes that corpses were placed beneath the pyre in order to ensure that heat would be maintained throughout the cremation process (1960, from Lucy: 2000). In order to shed some light on which of these theories, if either, might be closer to the past reality, we procured two pork shoulders and placed them at different locations in a pyre, intending to observe their appearance after a simulation cremation.

The building of the pyre was based upon McKinley’s theory of construction (Lucy: 2000: 105). To build the pyre, we first dug out a trough in the fire ring roughly an inch deep. Six logs were laid across the trough, and then four were placed perpendicular to these, with two across each set of three logs. This layering was repeated once more. One of the bottom logs was removed to allow for the placement of Group A’s pork shoulder, while Group D’s pork shoulder was placed on the opposite side of the pyre, on top of the layers of logs. Both pork shoulders were wrapped in linen napkins and secured with a brooch.

The logs of the pyre were then packed with brush obtained from the Carleton Arboretum brush pile, and a series of thinner sticks were leaned against the logs. The space between these sticks and the pyre itself was also packed with brush, for a final height of 115 centimeters. It was 140 centimeters wide.

We attempted to light the pyre from top to bottom, using a burning log from a different fire, but the brush proved slow to catch and eventually a piece of paper was crammed into an opening in the sticks and set alight. The temperature of the fire was measured every five minutes using a radar thermometer, as well as the temperatures of both pork shoulders (as best as we could). The following day, we reconvened at the fire pit to gather the remains of both pork shoulders, with the intention to place them in an urn.

Cremation Results

After the fire had been lit, it caught quickly, going from smoke to full blaze in five minutes. The sticks surrounding the pyre were edged in white ash by 35 minutes into the burning, and at 40 minutes they had collapsed onto the pyre. The pyre itself, being made from thicker logs, burned slower, but by 95 minutes had ceased to produce visible flame. At its hottest, around 15 minutes, the fire measured 930 degrees Celsius, and maintained a temperature of 91 degrees Celsius even just before we left at the end of our lab period. It should be noted that technically we were attempting to measure the temperature of the pork shoulder under the pyre, but given that it was surrounded by the fire and we were using a radar gun, it seems reasonable to assume that the temperature readings were more reflective of the fire itself than the pork shoulder.

Pork shoulder A reached 930 degrees Celsius at its hottest, although it should be noted that its location beneath a considerable amount of wood and ash made measurement via radar gun somewhat suspect. However, assuming this is accurate, it is well within the range hot enough to successfully cremate a body, although that temperature was certainly not sustained for the time required to finish the process (at least in a modern crematorium). Throughout the cremation process, collapsing wood from the pyre fell on top of pork shoulder A, obscuring it from view. When it was retrieved the following day, however, the top of the pork shoulder had been largely burnt to ash, exposing bone with coloration suggesting it had achieved a temperature of 900-1000 degree Celsius. However, the bottom half of pork shoulder A, which had been in the trough dug in the fire pit the entire time, was still entirely intact. The green of the linen napkin it had been wrapped in was still visible, with the cloth fully present, and when peeled away the linen revealed cooked pork skin and flesh, still pink and moist.

Pork shoulder A (Photo: Austin Mason)

Pork shoulder D achieved a maximum temperature reading of 656 degrees Celsius, but, again, since a significant amount of flame obscured the shoulder for much of the time, these readings may not be the most accurate. However, this temperature is not considered great enough to successfully cremate a body. Shoulder D remained on top of the pyre until minute 28, at which point logs collapsing beneath it caused it to fall out of the fire. Although we had not intended to “tend” the fire in any way during the cremation, Austin used two sticks to return pork shoulder D to the pyre in order to achieve more realistic results. Unlike pork shoulder A, pork shoulder D did burn to some degree on all sides — none of the linen napkin remained when we returned the next day to retrieve the remains. However, it also failed to fully cremate. The coloring of the exposed bone revealed a wide range of temperatures (900-1000 degrees Celsius on the top, but closer to 500-700 Celsius beneath), and when we cut into the remaining flesh at the lab, it was revealed to still be pink and wet inside, although apparently fully cooked (we did not consume the pork). Group D notes that “the body smelled like burnt flesh and a bittersweet mixture of caramelized meat and carbonized matter” (Ritual Fire Lab Data Group D).

More detailed data for both groups A and D can be found in the lab data reports on this site. Group A here and Group D here.

No sign of the brooches could be found upon sifting through the ash the following day. Given that the experiment was performed on public grounds and the fire showed signs of having been moved the next day (pork shoulder D had been relocated to the other side of the fire pit to allow for another fire to be burned), it is possible that someone else found and took the brooches, but given their small size it is also possible that they simply melted in the fire or we unidentifiable among the considerable ash buildup.

Cremation Discussion

Our cremation results suggest that neither Wells nor McKinley was fully correct. While the lack of exposure to the fire did indeed keep the bottom of pork shoulder A from cremating at all and, although not complete in this experiment, one could imagine that with a hotter, longer-lasting fire pork shoulder D might have fully cremated, practical matters still exclude McKinley’s theory from full belief. While this experiment worked with pork shoulders that none of the experimenters had any sentimental attachment to, it’s easy to imagine that seeing a loved one fall off of a pyre with their flesh half-burnt could be a very traumatic experience, and definitely not one respectful to the dead. As such, it seems likely that some degree of “tending” of the fire did occur. Clearly, the body needed to be laid on top of the pyre to start with, in order to allow all sides of the body exposure to adequate heat. However, in order to maintain that heat, more logs would likely have needed to be added to feed the fire past the initial 105 minutes of burning that the fuel produced. Additionally, covering the body after the first minutes of the rite would hopefully prevent an traumatic event like the partially-burned body falling from the pyre and ensure that the remains fell into the pyre instead. We also theorize that watching the process of a loved one burning might have been disturbing, and so covering the body would have allowed for a more comforting send-off, as well as giving more of a sense of transformation: the body is set on the pyre and covered, and then, the next day, turned to ash. Obviously, this inference comes from the assumption that we, as modern humans, have similar sensibilities to those of ancient Anglo-Saxons, so it cannot be a certainty.

Experientially, the main sense of awe we felt was in the sheer size of the fire. Our pyre was considerably smaller than one that would have been used by Anglo-Saxons, and yet it still produced flames taller than some of our heads, with a heat that kept us ten or more feet away at all times. I, at least, had never been near a fire this size, and it was fairly awe-inspiring. We could imagine that, at night, surrounded by people singing and drinking mead, with a much larger fire, the experience would have been fairly magical. The building of the pyre itself was also a strangely communal event, with all eight people in the two groups helping to dig a trench, stack logs, and break and place brush — it was easy to picture such an event as a community-building event, a time for those who had known the dead to come together.

In future experiments, a more accurate measure of the temperature of the fire would be incredibly useful. Additionally, adding wood to the top of the pyre after the initial burn in order to maintain temperatures sufficient to fully cremate the pork shoulder should be investigated. Our professor, Austin Mason, ran an unofficial experiment of this in his backyard after our class experiment, and reported positive results, as seen below.

Cremation Conclusions

The cremation results offer the following conclusions:

  1. The body would likely not have been placed underneath the pyre, touching the ground.
  2. Some degree of tending the fire would have been required to achieve full cremation.
  3. The cremation itself would certainly have been a spectacle with mystic undertones.
  4. The building of the pyre itself could well have been a community event, as well as the celebration during cremation, allowing for a greater sense of connection between all those at the cremation.

Further Reading/Bibliography

Aristophanes, Peace. Translated by Henderson, J. (1998). Loeb, Cambridge. 

Beowulf. Translated by Heaney, S. (2000). Farrar, Straus, and Diroux, New York.

Ekroth, G. (2013). What we would like the bones to tell us: a sacrificial wish list. In Ekroth, G., and Wallensten, J. (eds.), Bones, Behaviour, and Belief: The Zooarchaeological Evidence as a Source for Ritual Practice in Ancient Greece and Beyond. Swedish Institute at Athens, Athens, pp. 15-30.

Forstenpointer, G., Galik, A., and Weissengruber, G. (2013). The zooarchaeology of cult: Perspectives and pitfalls of an experimental approach. In Ekroth, G., and Wallensten, J. (eds.), Bones, Behaviour, and Belief: The Zooarchaeological Evidence as a Source for Ritual Practice in Ancient Greece and Beyond. Swedish Institute at Athens, Athens, pp. 234-242.

Homer, Iliad. Translated by Fagles, R. (1990). Penguin Books, Hammondsworth, Middlesex, England.

Homer, Odyssey. Translated by Murray, A.T. (1995). Loeb, Cambridge. 

Lucy, S. (2000). Cremation Burial Practice. In The Anglo-Saxon way of death: Burial rites in early England, Sutton Publishing, U.K., pp. 104-122.

McKinley, Jacqueline I., Robert Bond, David Wicks, and Julie M. Bond. The Anglo-Saxon Cemetery at Spong Hill, North Elmham Part VIII: The Cremations. East Anglian Archaeology Report 69. Dereham: Field Archaeology Division  Norfolk Museums Service, 1994, pp 82-86.

Morton, J. (2015). The experience of Greek sacrifice: Investigating fat-wrapped thigh bones. In Miles, M. (ed.), Autopsy in Athens: Recent Archaeological Research on Athens and Attica, Oxbow Books, Oxford, pp. 66-75.

Parker, R. (2011). Killing, dining, communicating. In Parker, R., On Greek Religion, Cornell University Press, Ithaca, pp. 124-170.

Porphyry. De Abstinentia. Translated by Taylor, T., in Pearse, R. (ed.), Early Church Fathers – Additional Texts, The Tertullian Project. Web URL:

Silver, A. (1969). The ageing of domestic animals. In Brothwell, D., and Higgs, E. (eds.), Science in Archaeology: A Comprehensive Survey of Progress and Research, Thames, London, pp. 284-302.

Wells, Calvin. “A Study of Cremation.” Antiquity 34, no. 133 (1960): 29–37.

Williams, H. (2010). At the funeral. In Carver, M., Sanmark, A., and Semple, S. (eds.), Signals of Belief in Early England: Anglo-Saxon Paganism Revisited, Oxbow Books, London, pp. 67-82.

Van Straten, F. T. (1995). Hiera Kala: Images of Animal Sacrifice in Archaic and Classical Greece. Brill, Leiden.

Group Data Reports

0 thoughts on “Lab Summary Week 4: Ritual Fire

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.