E r i c  P e t t i f o r

Cooking Clan of the Cave Bear Style
blank space - 0.1 K

Cooking Clan of the Cave Bear Style!
Boiling water in a skin pot over a fire (or not...)

by Eric Pettifor

Ayla was slicing pieces of yam to put into a skin pot that was boiling over a cooking fire.
      Clan of the Cave Bear by Jean M. Auel (page 140)

While looking for a project to explore for an archaeology course at Simon Fraser University (Arch 372, taught by Michael Wilson) I recalled that I had read somewhere that cooking could be done in a hide container over a fire. When I called my sister in Hope, British Columbia, to inquire if I could use her land to build a fire on, she informed me that I had probably read this in Clan of the Cave Bear. I bought a copy and fortuitously flipped directly to the above quote on page 140. Inquiries to knowledgeable persons failed to yield any knowledge of this method, and a query to Jean Auel to date remains unanswered [April 26, 1995: received letter from Jean Auel, but no specific references as they are buried in her research notes, and she is hard at work ~ good news Earth's Children fans! ~ on book 5.]

Looking for information on construction of such a pot, I came across the book Plains Indian and Mountain Man Arts and Crafts: An Illustrated Guide, by Charles W. Overstreet (1993). Herein I found information on the difference between rawhide and tanned leather, and several projects, none of which were a pot, but interesting for construction methods.

I ordered two half hides of rawhide from Buckskin Fur and Leather Company in Calgary, Alberta. Overstreet's book had not prepared me for what I picked up at the Greyhound parcel depot ~ two rolls of a stiff material not unlike plastic! I had thought that I would essentially be making a couple of leather bags, and that rawhide was simply a more robust form of leather.


Two pots were simply constructed from a single piece of skin each. The first step was cutting out circular pieces which would eventually be attached to metal rings purchased from Tandy Leather in Vancouver. Had I chosen to go all the way with an accurate reconstruction of a hypothetical Neanderthal pot, the rings could possibly have been fashioned from green willow (Overstreet, 1993). However, my main concern was to test whether this method of cooking was indeed possible.

Due to the difference in size of the half-hides and their irregularities, one pot would have a diameter of 14", the other of 18". There is a mathematical correspondence between the diameter of the circle cut from the hide, and the diameter of the ring it is to be attached to to form the mouth of the pot. Unfortunately, not being a mathematician, I used trial and error, starting with roughly twice the diameter of the ring and cut back from there, winding up with approximately 24" for the 14" ring and 28" inches for the 18" ring. The 14" inch was the first made and the diameter I finally arrived at made it a bit deeper than what I'd envisioned, and it had a 'scrunchier' (more pleats) rim than the 18" pot I created subsequently. Thus I called it 'the ugly pot', and the 18" one 'the good pot'.

To construct them I first drew circles of the incorrect diameter (2X) on the rawhide, also taking care to mark points 1/3 of the circumference apart around the edges for support thongs to be attached later, and then cut the circles out using Wiss Metal Master tin snips ("Made from special molybdenum steel, the non-slip serrated jaws cut up to 18-gauge low carbon cold rolled steel."). I marveled that anyone could have worked with this material using only stone tools, but I also suspect that necessary cutting was either done prior to it drying, or else there was water enough to spare nearby to soak it in. That's the next step. For my purposes the bath tub served well enough for soaking the rawhide circles.


I made rawhide strips for attaching the hide to the rings. These are easily made from scrap, starting from the edge of a piece and cutting around in a spiral pattern towards the center. This then needs to be soaked as well to make it pliable enough to work with.

spiral cut to make rawhide
thongs from a piece of scrap


The next step was to make a dual set of holes around the circumference, as well as three extra sets of two for the support thongs at the points previously marked.

Then I pressed the circle into the ring and began attaching it.

The long rawhide 'string' went through one hole, under the ring, then through the corresponding hole in towards the centre. From thence it went over to the next hole, through and under the ring to the corresponding hole and through towards the outside. And so on, stopping only a moment at the support holes to attach a thong ring, until the pot was complete and ready to hang-dry by the three thong rings.

Unfortunately, I lacked the wit to take clear pictures of the pots prior to their exposure to fire, but I did make a small (5" diameter) trial pot, shown here.

This has roughly the proportions of the good pot, perhaps a little shallower. I wanted the pots to be wider than tall on the premise that cooking would be more efficient if a wider area was exposed to the flames.


The first trials took place on Sunday the 10th of March, 1996 at Rosewood Gardens (a garden centre) in Hope, British Columbia, the proprietor of which is my sister Betty.

The pots were suspended over a fire from a tripod constructed of bamboo. While Auel's Neandertals wouldn't have had access to this material, it is strong and readily available when scavenging around a garden centre, since such poles are used as supports for plants. The circular thongs on the pot were attached to leather thongs which in turn were attached to a wire ring with a hook on it (another material Neandertals wouldn't have had access to, but since the home of my sister and her partner is on the site of the garden centre, coat hangers were easy to scavenge as well). The pot could be hung from a ring attached to another long thong, this also of leather, which passed through a metal ring secured to the top of the tripod. The tripod itself was held together with a strip of rawhide tied wet so that when it dried it bound everything quite securely. On one leg of the tripod was a metal loop (again coat hanger) to which the long thong could be attached, and the height of the pot could thereby be adjusted.


The first trial was with the ugly pot. The objective in all trials was to get water to boil. I filled it half full (4 litres). After 38 minutes the water was steaming hot, but not boiling. There was massive shrinkage of the pot, until the level of the water was near the brim.


The good pot was filled initially with seven litres of water. As it shrank I bailed out water to keep it from spilling over onto the fire. The water became steaming hot after about 12 minutes, but failed to boil over a trial time of one hour, thirty-seven minutes. Shrinkage left it seriously misshapen as one side moved up close to the plane of the ring.

Lopsided shrinkage in the 'good' pot

This may have been due to un-uniform thickness of the rawhide circle the pot was made from. Hide varies in thickness, and on the half hide this pot was made from it was impossible to find a circle of this area that was uniform in thickness throughout. Also thickness varies from hide to hide. Since I didn't specify a thickness grade when ordering, Buckskin Leather sent me the two best they had available. The ugly pot was made from the thicker of the two, and that, coupled with its greater vertical proportion, may have contributed to its greater uniformity in shrinkage.


The ugly pot again. 3.5 litres of water. Again, water became steaming hot in short order, but failed to boil. We (myself, Betty, and friend Val) decided that it was hot enough to brew tea in, and created a tea bag from cheese cloth and chamomile. Popped it in at one hour, five minutes.

Terminated trial at one hour, twenty-five minutes, still no boiling. The tea tasted distinctly of rawhide. Not terribly appetizing .cheese cloth chamomile tea bag in 'ugly' pot


The method of cooking by putting hot rocks into a container (leather, or mud slaked basketry) has been popular amongst diverse groups of hunter-gatherers. The container can be as simple as a skin pressed into a hole dug in the ground, or draped over a circle of rocks. The latter is the method I chose to use for comparison purposes with the pot-over-fire method.

Sedimentary rocks should be avoided, since trapped moisture can cause the rocks to explode when heated (Wilson, 1996; McParland, 1977). The best rocks to use are igneous (though porous igneous can explode since it contains gas pockets which expand when heated (Wilson, 1996)). Then there are the rocks which I used. A couple of them were fine, but most were nice looking cobbles of the variety not shown below.

This is one of them after being heated and submerged. Unless one fancies a high mineral content in one's diet, this type of rock should be avoided. It didn't explode or shatter (I wore safety goggles just in case), it crumbled. With transportation and handling it has crumbled even more and seems to have totally given up on continued existence as an integrated entity. It is the metamorphic equivalent of granodiorite, though only slightly metamorphosed so that one almost has to use one's imagination to detect oriented bands of hornblende (the dark mineral) visible after cracking, and it would be (was) a very tough call for a novice rock seeker without first cracking the cobble. The white portion is sodium feldspar, and the off-white/grey mineral is quartz (Wilson, 1996). Further adding to its fragility is the fact that it is weathered, and one can see the weathering rind on a broken section. They, like so much of the material in this experiment, were scavenged from around Rosewood Gardens, some from the surface of the ground, others from a dirt pile produced as a result of recent digging. Better would have been cobbles protected from weathering by a stream, though these are susceptible to tumbling which can create incipient fractures (small cones of percussion going into the material) in the surface, which might come apart if stressed (Wilson, 1996). Still, a cone fragment or two in the stew would be better than a significant portion of crumbled metamorphic equivalent of granodiorite.

In his experimenting with heating and cooling rocks, Pat McParland (1977) discovered that "in general a rock could be heated and submerged as many as seven times" before breaking, and his experiment involved cold water each time. He notes that the figure could be higher if the water in the trials had been allowed to remain hot. This is an important factor in the consideration of efficiency (how many trips to the stream bed) of the method. The rocks I was using did not conform to his general rule.

The 'container' held 16 litres of water. I removed hot rocks from the fire one at a time and deposited them into the container using a shovel. After about twenty five minutes the water was steaming hot, but not boiling. It was clear that my fire was too small, that I didn't have enough cobbles to boil that much water, and that the number I did have was diminishing due to crumbling.

We decided to halt for dinner. We had hoped to cook it in the manner described in Clan of the Cave Bear (Auel, 1980), but, alas, it was not to be. The stew was cooked in the kitchen (recipe at end of this document), and desert was marshmallows roasted on sticks of maple.

Trials continued on March 23, 1996, focusing primarily on the hot rock method. Betty had the fire going prior to my arrival in the morning and reported hearing a loud bang. One of the rocks that formed the circle, a sedimentary, had split suddenly. Thus we learned that rock selection is important for ring stones as well, not just rocks directly in the fire.


Hot rock method using the ugly pot containing 3.5 litres of water. Again, I transferred the rock with a shovel, taking care not to bring along too much ash or coals. The ugly pot attachment thongs were connected to the ring with a single long thong wound through and amongst them and the ring with the intention of making the horizontal level of the pot more adjustable. While this worked for the most part, when I dropped in the rock, the whole pot shifted spilling much of the water. However, the water that was left boiled throughout (as opposed to bubbling in the immediate vicinity of the rock) with one rock in about a minute. This would be a fast method for making a quick cup of tea.


I decided to give the good pot, now ugly and misshapen, another trial with 4 litres of water. It took 2 rocks to get boiling throughout in six minutes, boiling petering out for four minutes after that to a complete non-boil.


The ugly pot, 3.5 litres of water. Three rocks to boiling throughout in four and half minutes. Four minutes after that to a non-boil.


I decided I was going to get the over-the-fire method to work even if I had to destroy the pot in the process. Showing it no mercy, I lowered it directly into the flames. Twelve minutes later one of the support thongs burned through and the contents spilled onto the fire without it having achieved a boil.

In The Clan of the Cave Bear, Auel's Neandertals make sure the level of the water remains above the flames (Auel, 1980, page 81). Clearly I had not been as careful as they.

I considered repairing it and trying again, this time keeping a closer eye on the flames, perhaps pouring water on any area in danger of burning, but even in so considering it was already clear that, in comparison with the hot rock method, the leather pot-over-fire approach is a real non-method, or, at best, a grossly inefficient one. I could have boiled water twice in the time it had taken just to lose a support thong.


Cooking using the hot rock method would be more efficient than using leather pots over a fire.


If this experiment were to be done again, the first thing I would require would be more time. Ideal would be about a week in the summer when it could reasonably be expected not to rain in this part of the world. Two of three days of the first weekend at Rosewood Gardens in Hope were lost to rain. The second weekend I could only spend the Saturday (thankfully weather cooperated) due to the demands of school and end of semester crunch.

With more time, I would explore the optimum method of hot rock cooking. I find it interesting that in trial six two rocks got the water to boiling in six minutes, while in trial seven, three rocks had it boiling in four and a half. It raises the question of what is the optimum time to number of rocks to volume relationship. I successfully demonstrated that hot rock cooking is faster than cooking in a leather pot over the fire, but have yet to determine just how fast the hot rock method can be. Should people with fire pits put away their microwaves?

[Feb 21, 1997: Received an email from Bonnie (Dakona) Farner which calls into question my assumption that cooking by this method is 'quick'. While she is sceptical of the skin-pot-over-fire method, she notes from experience that cooking a stew with the hot rock method takes hours. She also includes the ingredients for a tried and tested hot rock stew! Full text of her message is included here.]

It would also be a good idea to make a special trip to a stream to collect unweathered igneous cobbles which could be reused over many trials without disintegrating. Better yet, send a geologist or someone experienced with construction and operation of sweat lodges, since exploding sedimentary rock is unwelcome in that situation as well.

In presentation of this project in tutorial, Michael Wilson observed that perhaps the reason the pots failed to boil over the flames is that the energy being released by steam was equivalent to that coming in from the flames, and that if the diameter of pot the were less this might be somewhat mitigated. TA Bob Muir wryly noted that perhaps it would be simpler to just put a lid on it, a point which was made again by Lloyd Bogart (and Bert Kamphuis) by email, along with other insightful comments [Aug 30, 1996, personal correspondence ~ Jean Auel as well thinks a lid would be a good idea].

If I were to construct another leather pot, I would fold the edge of the leather over the rim towards the inside of the pot, so that they would be less exposed to flame. Overall the pots took the direct heat well, with the exception of the final trial where not only edge damage was a concern, but the flames charred off a thin bit off the bottom of the pot. Further trials with a better designed pot, carefully watched (never mind the proverb), could also determine how well the pot could stand up to direct flame before it was destroyed. Given the damage to the bottom of the pot from one trial of limited duration, I suspect that the pot overall would not last many trials.

(left) Damage to bottom of ugly pot: charred layer flaking off

(below) Edge damage to ugly pot: Had the ring been made of willow rather than metal it is likely that the structural integrity would have been compromised (though once shrinkage has occurred the rawhide, while becoming softer, does keep its shape even when containing boiling water)


Rawhide is an amazing material. I even created a blade from it with a serrated edge and used it to cut a tomato. It can be used in the construction of shields, armor, boats, masks, drums, and all manner of things (not forgetting that it can also be tanned and made into leather), and as a binding is very strong and useful in both construction and repair.

However, most fascinating from an archaeological perspective is that it is virtually invisible in the archaeological record. Not only time and decay play a role here in consideration of human artifacts. Rawhide is so irresistible to dogs that anything discarded (and some things no doubt even before they were willingly discarded) would have been made short work of by camp or wild dogs.

But even if rawhide artifacts shone forth from the archaeological record as brightly as stone, I doubt, based on the results of this experiment, that we would find skin pots that had been used for cooking over a fire. Neandertals and other hunter-gatherers had better things to do with their time.


Auel, Jean M. (1980) The Clan of the Cave Bear, Bantam Books

McParland, Pat (1977) Experiments in the firing and breaking of rocks. The Calgary Archaeologist v.5, pages 31-33

Overstreet, Charles (1993) Plains Indian and Mountain Man Arts and Crafts, Eagle's View Publishing.

Wilson, Michael (1996) in conversation

NEANDERTAL STEW (a la The Clan of the Cave Bear, pages 81,82)
        [ Back to trial 4 ]


bison (we substituted cow)
wild onion (substituted domestic)
unspecified herbs (chose marjoram, cloves, garlic, and bay leaf)
thistle stalks (we omitted)
watercress (substituted bamboo shoots)
small immature yams (substituted grocery store yam)
cranberries (substituted canned cranberry sauce with whole cranberries)
"wilted flowers from previous days growth of day lilies for thickening" (substituted potatoes)

Note on other substitutions from Elizabeth B. Naime (Feb, 1998)

Add meat first, then potatoes (if used), then softer ingredients which don't take as long to cook. You might want to have salt and pepper on hand as well. Auel used the adjective "salty" for coltsfoot (omitted, see below), and this may be what that ingredient was intended for [August 30, 1996, personal correspondence ~ Auel notes that the saltiness may actually be from a source where the coltsfoot was flavoured with ash.]

Some ingredients were omitted and not substituted for as my sister Betty, wise in plant-lore herself, expressed skepticism about some of them and looked them up. The roots of milkweed should be avoided, and the other parts "must be boiled in three or four different batches of water to remove toxic substances and make them safe to eat." The cooking water should be discarded (Magic and Medicine of Plants, senior editors James Dwyre & David Rattray, 1986, published by Readers Digest Association Canada Ltd.). The tome also mentioned that "coltsfoot may cause cancer if taken in large doses or repeated small doses." Betty also expressed some reservations about lilies, noting that some varieties are poisonous. I should add that the mushrooms we used were from the grocery store, and that anyone using wild mushrooms they've picked themselves does so at their own risk. [August 30, 1996 ~ further info from Jean Auel.]

THANKS are especially due to Betty and Kari of Rosewood Gardens in Hope, and to Val Little for reading through the entire text of Clan of the Cave Bear and noting for me the pages with references to cooking and food.

© 1996 Eric Pettifor