[This is an article from Cariadoc's Miscellany. The Miscellany is Copyright (c) by David Friedman and Elizabeth Cook, 1988, 1990, 1992. For copying details, see the Miscellany Introduction.]

Building a Conjecturally Period Pavilion

Some years ago, we decided to try to build a pavilion that would be authentic, reasonably portable, and provide adequate space and shelter for long camping events such as Pennsic. After looking at all of the period pictures of pavilions that we could readily find and failing to find any period descriptions of how they were made, we concluded that the best we could manage would be a conjecturally period pavilion-one consistent with what we knew about period pavilions and period materials. We have so far built three, starting with a half size model, going on to one that we thought was large enough for us, and ending with one that is large enough for us; in building each we attempted to correct problems discovered in its predecessor. This is a description of our third pavilion.

Basic Design

One common period pavilion design is roughly conical, with a center pole. It seems clear from pictures that at least some such pavilions did not have side poles. One of our objectives was portability (we wanted something that could, if necessary, be taken to Pennsic by air; our second pavilion was, twice), so we designed our pavilion with a center pole but no side poles.

Pavilions of a generally conical shape seem to come in two varieties. In one, both the roof and the walls slope, although the roof is steeper than the wall. In the other, the walls are roughly vertical (see picture above). We chose the latter design because we thought it would be easier to build.

Without side poles, one needs something to keep the shoulder of the pavilion from collapsing inwards. It appeared that in at least some pavilions this was done with a hoop; what it was made of we do not know. In our (ten sided) pavilion the hoop consists of a wooden decagon made of ten dowels, each the length of a side, connected by leather corner pieces. A taut rope connects each pair of adjacent corner pieces, pulling them towards each other in order to keep the dowels from pulling out of the corner pieces. The ropes keep the frame together, the dowels keep it apart.

The entire weight of the pavilion hangs from one center pole. This is a problem when one wishes to pitch the thing. Getting a long center pole lifted to vertical while the weight of an entire pavilion is hanging from the top end is not easy. We could barely manage it with our second pavilion (an 8' diameter pavilion with a 12' center pole); the third (a 14' diameter pavilion with a 15' center pole) would require someone considerably stronger-and longer-than either of us.

We found a solution to this problem in pictures of period pavilions that show two ropes running from the top of the center pole down to the ground outside the pavilion. We concluded that there were actually three ropes (the third would be hidden behind the pavilion and pole). Their function was to hold the pole upright. After the pavilion had been pitched its ropes would hold up the pole, which suggested that perhaps the function of the extra ropes was to pitch the pole without the pavilion.

The system, as we worked it out, goes as follows. First you pitch the center pole, using its own three ropes. Near the top of the center pole is a pulley with a rope running through it. Once the pole is pitched, you use the rope and pulley to pull the pavilion up the pole. I do not know if this is a correct reconstruction of how period pavilions were pitched, but it works.

The walls of the pavilion consist of rectangles of fabric, hung from the dowels just as a curtain hangs from a curtain rod. Each rectangle is sewn to the adjacent rectangles, except at the door. The bottom corners of each rectangle have cloth loops sewn to them, which attach to stakes holding the bottom edge of the wall down and a little out. Since the edge of each rectangle is under the corresponding corner block, the resulting tension is at the ends of the dowels where they fit into the corner pieces, not at the middle. This is important; a strain in the middle of a dowel would tend to bend or break it.

The roof of the pavilion consists of ten isosceles triangles; their tips are cut off to make a hole for the center pole. The triangles are sewn together along the long edges, from the tip almost to the base, making them into one giant cone of cloth. Cloth loops are sewn to the roof at the seams that join the triangles, near the bottom of the cone (see picture below). Ropes run from these loops through the corner pieces out and down to a stake in the ground. The combination of the tension of these ropes, the tension down of the walls, and the tension of the roof pulling in on the corner pieces holds the frame (dowels and corner pieces) to shape.

One important feature of this design is that the structure is maintained almost entirely by tension and compression. A dowel can take much more force lengthwise-compression-than crosswise. A rope can take far more tension without breaking than a wooden dowel of the same weight can take either compression or bending force. The only crosswise force (bending force) on the structure is the weight of the wall sections where they are suspended from the dowels.

Materials

The pavilion and the groundcloth are made of light canvas. The canvas is waterproofed with beeswax. This is done by rubbing beeswax onto the canvas, ironing it in, then rubbing more beeswax on while the canvas is still hot and ironing that in. We do not know if this is a period technique for waterproofing cloth; we got it out of an early twentieth century camping manual. The manual recommended paraffin, which is a nineteenth century synthetic substitute for beeswax, so we used beeswax instead. Judging by our experiments, the waxed cloth is not noticeably more inflammable than the unwaxed cloth. We have had somewhat mixed success with this technique; how well it works seems to depend in part on how closely woven the cloth is. Test your tent before Pennsic.

The frame is made of ten hardwood dowels 3/4" in diameter and about 4'6" long. They fit into the leather corner pieces. The center pole consists of three lengths of hardwood about 1.75" in diameter, each about five feet long. They are joined by copper tubes about 8" long, slightly smaller in diameter than the pole; the ends of the pole pieces are filed down so that they can go into the tubes when we wish to assemble the center pole. The ends of the pole pieces are lubricated with beeswax so that they can be gotten out of the tubes when we wish to disassemble the center pole. The top section of the pole has a pulley set into it, as described below. A second small pulley is attached to the top of the pavilion. Both must be large enough to allow a 1/4" rope to go through them (see below).

Most of the rope consists of 1/8" braided cotton line; probably there are better period materials available. The rope by which the pavilion is suspended from the center pole is 1/4" sisal. The tent stakes should be hardwood dowels, 1" in diameter and 12" long. Their tops should be painted a light color to make them easier to find and harder to trip over at night.

In total, you will need:

19 yards of roof cloth 54" wide or 18 yards 60" wide

21 yards of wall cloth 54" wide

12 yards of ground cloth 60" wide or 13 yards 54" wide

about 6 lbs of beeswax to waterproof all the cloth

10 3/4" dowels 4' 6" long

2 square feet of ~ 8 oz vegetable tanned leather for the corner pieces

10 tension ropes 5' 3" long

10 stake ropes 14' long

20 ropes 1' long for attaching the walls and the groundcloth to the wall stakes

23 stakes (10 for stake ropes, 10 for the walls, 3 for the center pole ropes)

1 center pole rope 40' long

3 24' long ropes for pitching the center pole

3 center pole sections 1.75" x 5' hardwood or somewhat larger diameter softwood

2 metal tubes about 8" long sized to just fit around your center pole sections

2 pulleys, one ordinary and one designed to inset in the center pole

1 cleat

1 metal ring (or materials to make one) for the top of the roof.

patience

Construction

Hoop and Center Pole

The picture below shows one corner piece and the two dowels that it joins. The corner piece is a leather tube, with tabs at the end for the tension ropes to attach to and an additional leather piece around it for the rope going to the tent stake to pass through and tie to.

The corner piece is made of 8 oz vegetable tanned leather, cut out as shown on the figure. The small holes are for rivets, the large holes in the tabs are where the tension ropes attach. Roll the piece into a tube (4" long, not counting the tabs, and about 3/4"in diameter), rivet it closed with four rivets, then soak it in water. When thoroughly wet, vegetable tanned leather becomes somewhat flexible and stretchy. Put a 3/4" dowel in from each end, meeting at the middle of the tube. Arrange the dowels at the angle they will have in the completed frame (144deg. for a decagonal pavilion like ours) and leave them there until the leather has dried. You now have a bent tube.

Next cut the additional leather piece. Make it into a loop with the ends overlapping, as shown in the end view, and rivet the ends together with two rivets. Slide it onto the tube and rivet them together with a rivet through the remaining rivet holes (marked a on the drawing of the pieces). Your corner piece is done. Make nine more of them (assuming you do not think you need any spares).

The picture above shows the top section of the center pole, with the inset pulley. It also shows the top of the roof, with its own pulley. A rope runs from the center pole (it is tied above the pulley) down through the roof pulley, up through the pole pulley, then down again. This gives a mechanical advantage of 2 for one; a pound of pull on the rope produces two pounds of lift on the roof. That is enough so that my ten year old son can get the pavilion about half way up.

The design of the inset pulley is based on similar designs in the masts of ships; the hardware we used was intended for that purpose. We have been told that it was in use before 1600, but have not checked it. The use of pulleys and the physical principle of mechanical advantage have been known since classical antiquity. If you cannot locate a pulley to inset, a smoothly polished hole through the center pole works pretty well even without a wheel.

The center pole has a rain hat, made of leather or metal, set above the pulley. This is intended to keep rain from coming through the center hole. The cleat is attached to the bottom section of the center pole near its top end. The center pole rope ties to it when the whole thing is pitched.

The Roof

The roof is made out of ten triangles of cloth; the details of how to convert a piece of cloth into ten triangles of appropriate size and shape are explained later. The picture below shows how the triangles are assembled. They are sewn together down to about 16" from the base; beyond that they are hemmed. The seams are done as shown in the picture; this gives a stronger seam than simply sewing the two pieces of cloth together. A loop of cloth about three inches long is sewn securely to the seam at its end, as shown; it is attached under, not over, the roof. The loop is made from a 6" strip of folded cloth, made by folding and sewing together a rectangle of heavy cloth (preferably some of the scrap left over from making the roof) about 6" long by 2" wide. It must be strong and attached securely, since it is going to bear considerable strain. The bottom corners of the triangles have ties attached to them so that they can be tied together under the stake ropes to keep the skirt from flapping. Similar ties are also attached at the bottom of the base of each triangle, as shown; they will tie to the wall. We use braided cord for the ties.

The tips of the triangles are cut off. A strip of folded cloth like those used for the loops is sewn to the truncated ends around a metal ring as shown. The lower pulley attaches to this ring by a rope that runs from one side of the ring to the pulley (where it is tied to the metal loop on the outside of the pulley-this is not the rope that runs through the pulley) and back to the other side. The ring should be large enough so that it will not get stuck on the cleat when you slide it up the center pole.

The Ailettes

At the bottom edge of the roof, there are small gaps where the ropes run from the cloth loops out to the tent stakes. In order to keep the rain from getting in through them, we made ten little shield shaped pieces of cloth. The top edge of each was sewed to the roof just above the gap. Near the point at the bottom we attached two ties, which tie around the stake rope, as shown in the picture below.

Both the shape and the position of these pieces suggest ailettes-small pieces, often shield shaped, which attached to armor at the shoulders. The purpose of ailettes is not entirely clear-possibly they were intended as shoulder protection, possibly as a place to display one's arms when not carrying a shield. So far as I know, nobody has suggested that they were intended to keep out the rain.

The Walls

The walls consist of ten rectangles of cloth, sewn together using the same kind of seam used for the roof. We use the full width of the cloth and make the length about 4" more than we want the height of the wall to be, to allow for a hem at the bottom and a tubular hem (through which the dowels will run) at the top. In our pavilion, the wall is slightly over five feet high (so are we), so we started with rectangles 5' 6" long. Each section of wall has a strip of cloth, made in the same way as the loops described above, sewn to it as shown. This is for attaching the ties in the middle of the roof sections, in order to tie down the skirt (the part of the roof that projects over the frame) in bad weather. The seam joining each two sections of wall has a loop sewn into it at the bottom for staking down the wall.

To make it possible to get in and out of the pavilion, we simply neglected to sew together one pair of wall panels. To make the gap larger, slide the panels apart along their dowels, just as you would slide a curtain along its rod. For bad weather, slide them back together. To keep rain from getting in, we made a door 2' wide and about the same height as the wall. Its top has three ties which attach to cloth strips sewed to the bottom of the roof near its edge, just outside the hoop. Each edge of the door has an additional three ties, which attach to similar cloth strips sewed to the wall panels on either side of the door opening, about a foot in from the edge. Normally, the door is rolled up above the door opening; for bad weather or privacy it can be unrolled and tied to the wall.

The Groundcloth

The groundcloth is a decagon just small enough so that none of it projects past the walls-it is not intended as a collector of rainwater. The seams should be done like those of the roof and wall. It has loops at the corners so that it can be staked out. There should be a hole at the center for the pole to go through. It should be waterproofed with beeswax.

Measuring String

The picture below shows a measuring string, used to figure out where the stakes go. It also shows the arrangement of stakes you are trying to produce, seen from above, with the stakes numbered from 1 to 13.

Pitching The Pavilion

Decide where you want the center and put a stake in there (call it the central stake). Put loop A of the measuring string over the stake. Stretch the string out, and put a stake (call it stake 1) in the ground (do not hammer it all the way in yet) at loop B (i.e. a distance AB from the central stake). (This is step one in the picture below.) Put loop C over that stake. Pull the string out as shown in the figure, and put in a stake (call it stake 2) where loop B is; it will be a distance AB from the center and BC from the previous stake. (This is step two in the picture below.) Keeping loops A and C over their respective stakes, move B to the other side of the stake that C is over and put in another stake (3). Put loop C on stake 3, and repeat the process to put in stake 4. Put loop C on stake 2, and put in stake 5. You are spacing stakes evenly around the central stake, with a distance BC between each pair of stakes.

You could keep going around until you had all ten stakes in. I prefer to sight from stake 1 across the central stake, measure a distance AB from the central stake with my measuring string, and put stake 6 there (as shown). I then put in stakes 7-10 in the same way I put in 2-5, starting with 6 instead of 1. That procedure reduces somewhat the tendency of small errors in the measuring process to add up as you go around the circle.

If you have done everything right, stakes 1-10 are now evenly spaced around a circle as shown; if they are not quite right, adjust the positions before hammering them in. Then put in stakes 11-13 as shown (by the x's in the picture of the arrangement of the stakes); they are for the ropes that hold up the center pole. They should be a little farther from the center than the other stakes and evenly distributed around the circle.

The next stage is to spread out the groundcloth (the hole goes over the center stake), then the roof of the pavilion (on top of the groundcloth). Next take a stake rope and tie one end around a loop on the roof. Run the rope through the stake rope hole on the corner piece, and tie it, so the corner piece cannost slide on the rope. Tie a taut line hitch to the other end of the rope (if you do not know how ask a Boy Scout to show you) and put it loosely over the corresponding stake. Repeat for the other nine loops.

Next assemble the frame. Each dowel goes into its two corner pieces-after first being run through the tubular seam at the top edge of its wall section. The tension ropes should already be in the corner pieces, one rope running from each corner piece to each of the two adjacent corner pieces. Pull the tension ropes tight and tie them. The frame is now held together by the tension ropes and attached to the loops in the roof. Note: The frame assembly could probably be done after the pavilion roof was raised. That would make it easier to raise the pavillion, since you would not be lifting the weight of the walls. On the other hand, putting in the dowels might be a little harder at that point.

Assemble the center pole. Attach the center pole rope and pulley as shown in the picture (several pages back) of the pulleys for raising the pavilion. One end of the rope attaches to the center pole. The rope then runs down through the pulley attached to the roof, back up through the pulley at the top end of the center pole, and down again through the hole in the roof of the pavilion. The three center pole-stake ropes should be attached to the top of the center pole above the rain cap.

Raise the center pole; its bottom end will now be in the hole of the roof, sitting next to the center stake. As the pole goes up let the rope which attaches the top of the center pole to the roof of the pavilion run through the pulleys, so that the roof stays on the ground as the center pole goes up. Attach the pole's three stake ropes to stakes 11-13 with taut line hitches. Tighten them until the center pole stands vertical. Now crawl under the pavilion, being careful not to pull any of the dowels out of their corner pieces while doing so. Grab the center pole rope and start pulling. The pavilion will rise up the center pole until the metal ring reaches the cleat. Ease the ring over the cleat (you did make it big enough, didn't you?). Keep hoisting. Have someone on the outside checking that none of the stake ropes are taut and, if any are, loosening them. When the pavilion gets all the way up the center pole, tie the rope to the cleat. Have someone outside go around tightening the stake ropes until everything stands straight and looks proper.

If you have done this right, the center pole is sitting in the hole in the center of the ground cloth, next to the center stake. Take out the center stake. Stake down the walls of the pavilion and tie the loops on the ground cloth to the same stakes. Spread out your bedding in the pavilion and take a nap; you are done.

Sizes

So far, I have assumed that you want to make your pavilion exactly the same size as ours and have exactly the same width of cloth available. This is unlikely. In order to modify our design for your pavilion, you will want to know two things: how big you want it to be, and how wide the cloth available is.

It is convenient if the width of a wall section is the width of the cloth (minus 2" seam allowance). Start by assuming you are doing it this way, and figure out how many sides your pavilion should have in order to be about the size you want; you may find graph paper (or trigonometry) useful in doing so.

Next figure out how tall you want the pavilion wall to be. Add 4" and you have the length of the pieces you will cut for your wall sections. Multiply by the number of walls, add an extra piece for the door flap plus necessary pieces of scrap for making loops and things, and you know how much wall material you need.

The roof is made up of isosceles triangles with two long sides and one short side. The length of the short side of the triangle is the width of one side of the wall plus about 8" for seam allowance and for the extra needed for the overhang. The length of the long sides is about twice the length of the short side for a decagonal pavilion or about 1.75 times the short side for an octagon. Triangles of this size do not quite fit on cloth of the width you are using for the walls; your choice is to buy slightly wider cloth for the top or to cut out the triangles with one little corner left off, cut out a small triangle of that size plus seam allowance, and sew it on as shown above. The amount of cloth you will need is 5 times the length of the long side of the triangle for a decagonal pavilion (or 4 times for an octagon) plus about 3'; if you are using cloth the same width as you used for the sides, add another 4' for a decagon or 3' for an octagon. Before cutting (or even buying) your cloth, try cutting the pieces out of graph paper and taping them together to make a scale model; this will show you what the proportions of the pavilion will look like and if your measurements are off.

All of these figures assume that you want your pavilion to have about the same proportions as ours-in particular, that you want your roof to slope at about 45deg. . For a steeper roof, make the triangles longer relative to their width; for a shallower roof, make them shorter.

The sizes we give for the measuring string, the length of the dowels, and the length of the various ropes are all for our pavilion. If yours is going to be bigger or smaller than ours, scale them up or down accordingly.

The Next Pavilion

The pavillion described in this article is our third attempt at a conjecturally period pavillion. If we ever do a fourth try, we will probably make the following changes:

1: Eliminate the Ailettes

The ailettes are intended to keep rain from coming in through the holes at the corner where the ropes come out. We have not yet seen any evidence of them in period pictures. That suggests that we may have solved a problem that does not exist for a correctly designed medieval pavilion.

One could eliminate the problem by sewing the roof panels together all the way to the edge, replacing the hole where the rope now goes through with two cloth loops, one on the outside of the roof, one on the inside. The two loops would be sewed together through the cloth of the roof. The corner piece would attach to the inside loop, the rope from the tent stake to the outside loop. Thus the tension of the rope would pass through the roof, but without a hole. The portion of the roof beyond the loop would hang down, making a skirt.

2. Eliminate the waterproofing

Our present system of waterproofing is a lot of work, only conjecturally period, and works only moderately well-the roof mildewed one Pennsic and has leaked since, so we currently have a double roof, not the single roof described in the article. I am told that tightly woven canvas is effectively waterproof. When it starts raining a little gets through at first, but once the canvas gets wet the threads swell and the leaks stop. If I made another tent I would try to get suitable canvas and experiment to see whether the tent provided adequate rain protection without waterproofing.

3. A thicker center pole

It has not yet broken, and probably never will, but seeing it bend visibly makes me nervous. If I did it again I would probably use 2 1/4" hardwood or 3" softwood.

4. Add Crow's feet

In addition to these, there is one more change that we might try-eliminating the hoop entirely. Adding crow's feet should give a more uniform tension around the edge of the roof. It is possible that this would make our system of a leather and dowel hoop superfluous. One could try redesigning the pavilion by sewing the walls directly to the roof and relying on the tension of the external ropes to hold the whole system to shape.

Final Words

We have tried to describe our pavilion as it now exists. We have described how the pavilion should have been built, not how it was. The first time we pitched the current version was at a Pennsic War-and we discovered that we had made the roof triangles about a foot too short. We spent that Pennsic with a temporary skirt pinned around the edge of the roof. It is still there, except that now it is sewed on. It is easier to make the roof the right size to start with, so that is how we have described it.

The current design works and is consistent with what we know about how period pavilions could have been built, but it is still only a conjecture about how they were built. If you think of ways in which the design could be improved while remaining consistent with medieval technology, by all means try them and let us know how they work.

Some years ago, after experimenting with the use of beeswax to waterproof a pavilion, we reported on our technique to a good friend in the SCA, explaining that it could have been used in the Middle Ages although we had no evidence that it had been. At the next Pennsic, we were told that there was another pavilion there which had been waterproofed in the same way as ours. When we spoke to the owners, they assured us that it was a period technique. After a little further conversation, we discovered that their source for that fact was our friend and her source was us. In two steps, via someone who is very much more careful about issues of authenticity than most people we know in the Society, conjecture had been converted into fact. We would prefer, with your help, to convert conjecture into fact by a slower but more reliable process-experimenting with what works and checking the result against the evidence of period sources.

[by Cariadoc and Elizabeth]


Webbed by Gregory Blount of Isenfir