Sunday, 30 November 2008

The old measures

It's a funny thing to me that there was a person locally who after 3 months still knew me as a dancer and dance teacher, and not that I'm a mad keen costumer. Myself I identify primarily as costumer and crafter, and many people reading this probably don't even know I dance. But I do spend 2-4 hours a week doing or practising medieval dancing for much of the year. I dance because it's fun, but I also devote extra hours to this as a way to slow down my descent into very unfit. And like most of my other pursuits I find myself seeking authenticity in this too, and end up teaching because I want to pass the fruits of my knowledge on (rather than for any skill I possess at teaching).

Anyway, enough prelude, today's musings concern a particular genre of Dance, the Old Measures, danced in London Law Courts (Inns of Court) in the late 16th and 17th Century.

These dances range from extremely simple to approaching moderate in difficulty. That makes them excellent dances to teach beginners and also to warm up with at the beginning of a session. Quadran Pavanne is my favourite warm up dance - it's little more than walking, but introduces the concept of singles and doubles, and gives the participants a chance to chat to each other in greeting (it may be best to use non-hopped doubles if this is done before stretching). The basic number of moves needed for the Old Measures are small - it's easy to teach a new alman if the person knows a couple of others. And it is easy to arrange to teach the dances in an order which introduces one new concept per dance. (eg doubles & singles, then turns, then set & turn, etc)

Another attractive prospect of these dances is that they teach many of the basic moves needed for English country dance - singles, doubles and reverances, 4 beat doubles, set & turn, chorus & verse structure, etc. And I consider it likely that these moves were performed in the same manner for both aalmans and English country dance, as both were being danced in the same place by the same people. Think of English country dance as what you dance at the free dance after the formal set of the debutante ball, or maybe it was more the modern dances you danced at the after party.

Anyway, the very simplicity of the old measures may be one reason they are not danced very often. One possible solution lies in more period presentation of the dances. the old measures weren't the hip dances of the day, they were the compulsory dances at the start of the night. make them the dances that everyone knows, everyone (including the older and less fit members of your group) dance. Or dance a more period number of repeats - we generally dance 2-4 times the period preferred number of repeats, which makes it nicer for beginners, but the period 1-2 repeats avoids boredom by people who know the dance. Make things more interesting by dancing the old measures as a set of 6-10 dances (as was done in period), each danced back to back - make the experienced dancers work to remember the dances correct because they don't have 3 more repeats to get the dance right. (a nice dancemaster will however provide cheatsheets). Anyway, those are a few options to try to create more interest in the old measures.

But I think the most likely reasons some of the old measures are danced less than others is that while music and instructions are widely available for some (like the ubiquitous black alman), they are more difficult to find for most of the old measures. There are actually two good publications which cover the near complete collection of old measures. Practise for Dauncinge and Durham's "Dances from the Inn's of Court", and del's dance book also has a few allemande. Any of these are good starting points, Dances from the Inns of Court is my favourite, giving a good historical overview, comparisons of original sources, and a carefully thought out reconstruction accompanied by sheet music for a melody line and an optional CD. However it is not free, nor is it instantly available (order from here). Del's dance book provides well spelt out reconstructions, and I've always found del's a great source for showing beginners the dance they've just danced written down, and for helping intermediate dancers to teach their first dance from, however the historical background is very scanty, and the range of old measures supplied is limited. Practise for Dauncinge is also available online, with a large range of old measures, and good historical background, and gives original instructions as well as reconstructions. It's not quite as easy to follow as Del's, but it isn't difficult to interpret.

Here's a chart to show the historical trend of what was being danced (or at least recorded) when:

MS Rawl Poet. 108 SRO DD/WO 55/7 Harleian 367 Douce 280 Rawlinson D.864 Inner Temple, vol 27 RCM 1119 RCM 1119 Fol. 2
year c. 1570 1594 1575-1625 ? 1607 1630-1633 ca. 1640-1675 after 1640 ? after 1640 ?
The Quadran Pavan yes yes yes yes yes yes yes yes
Turkelone yes yes yes yes yes yes yes
The Earl of Essex Measure yes yes yes yes yes yes yes yes
Tinternell yes yes yes yes yes yes yes
The Old Alman yes yes yes yes yes yes yes
The Queens Alman yes yes yes yes yes yes yes
Madam Sosilia Alman yes yes yes yes yes yes yes
The Black Alman
yes yes yes yes yes yes yes
Other old measures 8


Interestingly, there is little variance between the sources. To me this reflects what Durham and Casazza have written about - this is a set of formal dances that everyone learned, and were quite old fashioned for most of the above manuscripts. The proliferation of other old measures in the first manuscript may reflect that it still lies in an era when the old measures are popular dances of the day, with what will become the classics still being chosen, and new dances created - or the author may just be a more avid dancer than his successors.

Here's a chart showing which reconstructions of old measures are available in each of the books I have mentioned:

Practise for Dauncing Del's dance book Inns of Court
The Quadran Pavan yes
Turkelone yes yes yes
The Earl of Essex Measure yes yes yes
Tinternell yes yes yes
The Old Alman yes
The Queens Alman yes yes yes
Madam Sosilia Alman yes yes yes
The Black Alman yes yes yes
lorayne Allemayne yes yes yes
Brownswycke yes

The newe allemayne, yes

When it comes to sheet music for these dances, a much wider range of music is available, and yet little of it is easy to come by. Eric's music search engine can help greatly with finding music (both sheet and recorded, with a good collection of the quadran pavanne), but is not complete, and some dances are hard to fins due to the variety of spellings used in their names. Here is a chart linking to versions of sheet music available for the old measures, sorted by most easily available source:
(The names listed are the arranger/transcriber/composer of the version, where multiple names are listed, click on each for multiple versions. "Inns of court" is not available online.)

Practise for Dauncing Del's dance book Inns of Court other
The Quadran Pavan Casazza
yes Avatar,
Turkelone Casazza Del yes
The Earl of Essex Measure Casazza Hendricks yes Hendricks
Tinternell Casazza Casazza yes
The Old Alman Casazza
The Queens Alman Casazza Smith yes Hendricks,
Madam Sosilia Alman Casazza Phaedria yes Avatar,
The Black Alman Casazza Yardley,
yes Exeter
lorayne Allemayne Casazza Phaedria,
yes Smith,
Brownswycke Casazza

The newe allemayne, Casazza

yes Smith

I'd love to make a later post describing which sheet music is better for which purposes (eg key suited to trumpets, this one good for a group playing, this one better for a virtuoso player, etc), and am handing music out to local musicians (let me know if you are local to me and haven't recieved any) but so far the only comment i've recieved is that 1 page versions are better than 2 page ones. So dear blog reading musicians, your comments are sought.

Tuesday, 30 September 2008

Aachen Cathedral Treasury - Part 3: Shrine with nice pictures (tents!)

Following on from my previous posts concerning the Aachen Cathedral Treasury, here is another selected treasure.


Karlsschrein (Charlemange shrine)- 1200-1215 Aachen , gilded silver and enameled copper.

This is a lovely piece of artistry. I read somewhere that 12th C german metalsmiths were copied by the manuscript artists because they were were considered the superior artists of the time. For a colour pictures of the shrine, a glimpse of colour scheme, search for 'Aachen' at this link, and look for the gold and silver metalwork. (sorry direct links not working) It's only a single colour image of the shrine, but it shows well how the colours are and how a few more details can be seen in colour (eg broaches clasping cloaks) than can be seen in the high resolution black and white photos linked to below. Ithink it might have been all gold originally, but the gilding wore off.

Now to talk about specific aspects of the shrine's artwork and how it relates to concrete objects of the era....

The clothing depicted on the shrine is mainly same as 1180, possibly with less droopy sleeves, although with so few women depicted, and then mainly only Mary, it is hard to tell with regards to sleeves.. Are silversmiths behind or conservative, or does fashion not feature the sudden changes seen in France at around 1200? I think it may be rather the characters on the shrine - saints are very often dressed conservatively in fashions a little out of date and lacking the excesses of fashion. But it's also possible German fashion doesn't have a big change from late 12th C to 13th C clothing -this would be a lovely project for someone else to follow up.

There are some really lovely depictions of braids on garments, and the ample stones reflect a period practice we mostly only read about, since extant garments generally have had these stolen, and in manuscripts we can't reliably distinguish these from braid with a blobby pattern.

The material and method used enables us to see a number of features not normally seen - for example crisp lines of applied braid on many of these shoes (almost in the same pattern as the extant shoes of Philip of Swabia of around this time). Manuscripts cannot show the difference between an applied raised braid and a flatter painted on sewn on decoration in this situation, while much sculpture has been weathered, obliterating the crisp lines that make such a distinction possible. The precious and religious nature of this shrine has resulted in excellent preservation.

About half the shoes are in this style with
Many others saints on this shrine have one piece of braid (less high, often wider), running up the vamp of the shoes and far up the calf - these are long boots, and remind me of this pair of hose/buskins.

The crowns shown match well with the extant crowns of this era (mostly german) that I have seen, especially the fleur-de-lis on top.

There are a couple of nice pictures of armour, including details of Charlemagne putting on (or taking off?) his armour. The lines of where the chainmail stops (especially on the legs) are clearer than on many other pictures I've seen. And nice details of surcoats, haulberks without a concealing surcoat, and haulberks with built in hoods (I think).

There are a lovely picture of tents from a fairly close perspective , and a second depiction with similar tents with an interesting feature I'll discuss at the end. I'm quite excited because tents are relatively rare in artwork of this era, and pictures with this level of detail even more so. They appear to be single bell tents, held up by a central pole and a network of guy ropes. The Central Pole is not visible, but I believe it is implied by the use of a ball and decorative caps on top of the centre of the tent. Charlemangne's tent is topped with the eagle (symbol of German royals), while the other tents are topped with crosses.

The section of tent immediately under the ball appears to be a separate piece from the tent roof below it, as clear seams (including textural differences) show on several tents but is probably still a fabric (cloth or leather) as it falls in a curve with gravity, rather than stiffly. This corresponds with stiffener panels used in recreations.

The roof of the tents appears to have been decorated with patterned fabric. Other tents show a pattern of radiating stripes that could be stripes (more accurately arcs), or could simply be tension lines from the guy ropes.

From the edge of the roof, several guy ropes are seen anchored to the ground. They attach to the roof of the tent via a triangular or y-shaped feature. The spacing between guy ropes is quite small, although guy ropes on the front of hte tent have been omitted. This may be to create a doorway, but more likely is for artistic reasons, to not block the view of the protagonists. (other guy ropes end at the edge of the tent to prevent overlapping other characters).

The walls of the tent descend from somewhere slightly inside of the edge of the roof, and descend fairly vertical until about halfway down, and then flare out to the ground. The way the walls flare is consistent with being pegged out (approx one peg per guy rope, so also at fairly regular intervals). I believe I can see a one of these pegs, but it is simply a tapered blob.

It is not clear if any additional structures such as internal wall poles are present or absent, but if the lines on the roof are stress lines, I would guess not.

A doorway opens from the front of the tent in a gentle curve, but the way this nicely frames the scene inside makes me believe this is likely to be an artistic cutaway, rather than a realistic doorway shape. Then again, it might be both a good artistic cutaway shape and a good door way shape.

On the second picture, one tent is depicted with a strange window in the roof. Maybe this is a skylight? Interestingly, the lines I previously identified as tension lines in the roof run through the "window" in the roof from centre pole to guy ropes. Perhaps this is a way the roof was reinforced?

This reminds me a great deal of the tents used by my sca neighbours.

other items
Also depicted on the reliquary are some nice lamps and pennons, which I've added to my forthcoming next post on the topic.

Tuesday, 23 September 2008

Peasant Shoes I

My friend Asfridr has doccumentation for this type of really simple shoe, I hope she will upload it soon. In the meanwhile, this journal post of hers may be of interest.

These are very very simple shoes to make. They use only one piece of leather and put seams at the toe and heel to turn it into a rough bag shape. Gathering is then used to shape bag into a shoe.

Asfidr showed us how to make mockups out of felt. She described two main ways of making toes and two ways of making heels. I tried both on two mockups:

This first style has a slightly stubbed toe

The second a pointy toe

I decided I liked the second version of both toe and heel better, so made them up out of leather

I used thinnish alum tawed leather because of the simple expedient - cheap and available. Thicker leather would probably be better.

These were very easy to make, much easier and quicker than turnshoes. Only trouble was my code - I wanted to use a nice medieval waxed shoe thread (in preparation for more turnshoes), and I haven't got the right mix of wax yet. (that might be a blog post of it's own). If you want easy shoes, use a commercial pre-waxed thread.

I also would be happier if my stitches were tighter (again the wax probably isn't helping here), but this doesn't affect the usability of the shoes.

Here are 3 views of the shoe



detail of the stitching at heel and toe:

And what you've been waiting for - the shoes on:

I field tested them at Rowany Festival. Here's a report:
they wet through slowly- but combined with wool hose wasn't unpleasant, slowed down moisture getting through and the hose were only damp, not dirty of muddy. Thicker leather might mean these shoe lasted hours before getting wet, and they dry relatively quickly for leather. Drying overnight alongside the fire (away from the direct heat) might work in a medieval hut.
Also I took my shoes off and went barefooted when it was dampest, which the canny medieval peasant probably did (the farmers sowing are sometimes shown barefooted), at least in more temperate weather. Having dry shoes to put my cold feet into on the cooler festival days after a while barefooted worked well enough for me.

I'd also like to try tallow on them - I think that might work wonders. But then I'd have to be more careful about transporting them As I'd rather not have tallow on my clothes. Shoes lined with dry straw or fleece might last most of the day if you didn't cross rivers in them. Or overshoes - I'm not sure if peasants would wear pattens, but some kind of cheap overshoe could have been used.

I was comfortable. this leather is a bit thin, gravel paths, and sharp edges hurt. Thankfully these were really rare on the Glenworth site. I didn't have any trouble with my slight tendancy to roll my feet inwards. Seams did not rub. Shoes could be worn barefoot.
The lacings were the main adjustment. Too tight and walking normally was difficult, too loose and they fell off. The was a good spot in between. I was using cotton lacings (I've replaced them since with leather which shouldn't get/look a grotty), and over a long day they loosened off. Extra long lacings that looked around the ankle and tied at the front of the ankle might work better - one of my lacings was longer and I did this, and it stayed on a bit better, and Ii think the thong dragged in the mud less.

I replaced this with leather thonging (shoelaces I think - thanks op shop), and this looks cleaner, but knots in it do not stay done up as well as braid (yes i do always have some fingerloop braid handy). Also the ends dangle about, but I need them to tie the knots. I could move the knot tying place from the back to the side - but then they wouldn't fit either foot. I could sew them closed (I don't really need to loosen the lace to get in and out of the shoes), but I like the adjustability, for when my foot changes size, the lace swells when wet, stretches from use etc, and also it is a little easier to get off If I can untie it. Finally, bringing the laces over to the front and tying a second knot at the front of the ankle seems a better option, but my leather thongs are just a little too short for this unfortunately.

Dancing was no good in these - they fell off mid bransle- but maybe if I had had the lacing just right it might have been ok. They were also slippery on the dance floor, just like my turnshoes. I didn't break into a full run, but I suspect this would have been a problem, while a few jogged steps were not. I guess barefoot might have been an option for the frolicking peasant.

They lasted more than 3 days worth of wear. And another 6 months of at least once a month. You can see one scrape on the back in the close up shot of the seam after 3 days, and they are only a bit dirtier after 6 months. Sturdier heavier leather would probably be sensible.

Monday, 22 September 2008

egg tempera

What else have I been doing? I've playing with egg tempera at a workshop day. It was fun, and one day when I magically have more time I'll try again with a better idea of what technique to use. (sorry the second shot is so dark that you can't see detail easily - it reflects too much with the flash).

Timeline of construction of a pair of Hose

Months ago in February, (I admit, I've been slack about posting) I decided I wanted lots of pairs of hose for Rowany festival, since that part of my wardrobe was lacking (only 2 pairs for 4 days! and one of those thick wool.) I'm getting pretty good at hose by now, given that I've been using the same pattern each time. So I decided to time how long it took to construct a pair. Or to be more precise I constructed one first to check the fit in the particular fabric (it changes slightly each time), and the exact way I was doing the hose, and then timed the second one of the pair.

Construction, Pattern, stitching technique
I've been making copies of the 14th Century Hose from MOL because I think they are the same pattern as 12thC ones for reasons I'll describe in a post soon.

I decide to experiment with a couple of things on this pair. Firstly, I've found the rear seam to be quite weak (running stitch as described on the London hose snapped on the second wear in linen on linen on another pair), and the transition around the seam at the heel difficult. I wondered why the seam used on the foot didn't simply extend up the back of the leg. I guess it would be inelegant in wool with raw edges showing, but in linen all my edges were contained anyway. I tried this method (see pictures to left), and in linen it works just as well, maybe slightly better than the other method. The transition at the heel is slightly tricky, but it's even trickier in linen with double folds changing from this seam to flat felled apart seams.

I chose to make a decorative hem using herringbone stitch on this pair of hose. The pictures to the right show the inside (decorative so it will look prettiest when turned over) and outside (plain) top hems. I have no evidence for this type of hem for this period, or on hose. It just is used on hems elsewhere and when and was pretty.

The last is a shot of how I handle the top of the vamp (top of the arch of the foot). This is a slight adaption for the linen versus a wool that doesn't need hemming. The picture is mostly there to show how I do this.

These hose were constructed on my way to work, sitting at the bus stop, on the train, at the train station, and so there was a lot of stopping and starting and a little unpicking. I probably could have made these slightly quicker under optimum conditions.

So here is my construction progress in pictorial form:

I think the decorative hem took longer than a simple line of backstitching as seem in the London hose would have. But the white linen thread on purple linen fabric is pretty.

Saturday, 31 May 2008

An apron

Just for a change, I've made something not 12th C based.
But to be honest it's mainly because I have no evidence of what was used in the 12th Century as aprons, and I was a little inspired by some early 15th Century pictures I'd seen.

Here's a butcher and another trader (baker?) from the picture. The manuscript is a few added pages to a copy of Romance of Alexander, it's actually "Marco Polo, Li Livres du Graunt Caam, in French prose, with miniatures by Johannes and his school."England c. 1400 (MS. Bodl. 264). To see more (but not higher resolution unfortunately - these are really small): from the bodlein library index page select view all, and scroll down to folio 218 recto. Zoom in to the marketplace (big red blob on the left) and see the aprons worn by the traders in context, with the 3rd trader in an apron (partly concealed) too.

On reflection (I didn't have the pictures with me when constructing this) I think mine is much too large compared to the illustrations, but it does work quite nicely. There's a better reconstruction on the medcos forum. It's also possible these aprons don't have ties on them, that they simple sit under the belt. Somethign about the way the fabric gathers up a little suggests this to me.

If you are interested in apron designs in general, may I recommend karen larsdatter's aprons links page. You'll notice that nearly all the aprons are styles that are tied around the waist with nothing above the waist, but if you check out her blacksmith's aprons page, you'll find a few aprons like this, namely:

  • Nature at her forge, Roman de la Rose (Bibl. Sainte-Geneviève, MS 1126, fol. 115), c. 1350-1360
You can however wear this apron in that manner by simply tucking the small triangle above the ties in behind the rest of the apron like gwynfor did with this apron at festival. you'll also notice that nearly all the aprons are white - which I'm assuming means linen, easy to bleach. I'm substituting ramie which has a similar stiffness to linen, for a fraction of the price.

So making this up out of nowhere, here's the experimental cutting plan I tried:

The pattern is quite simple - square (as in equal length all sides) of fabric and some ties. The ties could be strips of fabric or something firmer (that needs no sewing) like twill tape. My ties are about 1m long each - mainly because this is the width of the fabric I had available. They are more than long enough for this job, but not so long they tangle or drag on the ground.

I've used letters rather than numbers for the measurements because I think of how I made this apron as a method, rather than a pattern - like T-tunics, I prefer to craft things to the wearers measurements. So measurement A is the width of material I had handy, in this case 105cm, the width I cut my square to. I then turned my square around 45 degrees until it resembled a diamond, and pinned one of the corners high on my chest (just below where I pin my keyhole necklines closed) using a pennanular broach. I marked out where just above my waist sat on this rectangle (B - 40cm for me). This is the point where I attached the ties, simply sewing them to the edges of the fabric at this point. The last step was to trim off the bottom point of the diamond. I chose to try my apron on, and get a friend to mark out my desired apron length (slightly longer than the sides, but shorter than my ankle length dress) with a pin in the centre (length C). I then turned this into a pleasing curve to cut off.

Because this was experimental and made the night before leaving for rowany festival (our big camping event), I didn't bother to neaten the edges, hem or in any way finish this garment, and hardly anyone noticed. The lack of finishing didn't detract from the garment's usefullness. Now that it has been washed, it is starting to fray, and i must choose to finish the edges or make a new one (maybe even to annother design). I say this because others may wish to quickly make an apron at some stage, or to make enough aprons for helpers at an event, and this seems a fairly quick way to make disposable aprons. Of course I'm sure that that isn't the medieval approach - they would have carefully finished edges to ensure the durability of the garment.

Tuesday, 1 January 2008

What have I been doing lately? beading

Well I've discovered a fairly good way to do beading on the train. Couching pre-strung beads. And using fake parchment (real vellum donations welcome) provides a stiff base so you don't need a frame to keep a taut background.

I've discovered this also makes a good introductory or beginner project, as long as the person has basic stitching skills (anti tangling skills especially). I've done this at a demo now, and several people had fun with it - quick and colourful.

historical background:
Grizel's website has some great links to medieval beading. She unearthed the pieces that got me started on beading on parchment.
This set of pieces in particular, and this one are on parchment.

By contrast this cap is on a linen base.

Most pieces I do not know if they are on parchment, linen or some other background. Some pieces that might work well in parchment, but could equally well be on fabric are for example beaded chests, or boxes, or cups.

my first piece:

I chose the form of a decorative roundel as it is a shape that is used quite small, and one we see in period tapestries and embroideries particularly. The pattern is a simple geometric, in a style I'm sure I've seen in Romanesque art, but I can't point you to a picture today.

It sure looks impressive from here, and is a faster technique than many methods of embroidery. I think I've already said I did this on the train, but I'll just confirm, that once beads were strung, this was one of the most portable projects I've ever done. (The buttons being the only slightly more portable thing).

I read somewhere (probably on Grizel's site or mailing list) that beads were generally couched in period, and that is why long strings of them fall off so easily. They also fall off when helped by humans, generally in order of their resale value, but that still would be unlikely to happen in the same way if each bead was individually sewn on in the way many of us would think to sew on beads otherwise.

I haven't he expertise to disagree with this statement (and reputable books that truly cover medieval beading are few and far between), and indeed what small evidence I've seen suggests a lot of the large scale beading was done this way. (I'm sorry if this sounds like doubting people, but I try to doubt everything including my own assumptions, so I'm noting this really to say I need to provide better evidence, not that others whose evidence I haven't seen isn't correct.)

fine points of technique
When I first started trying this I couched the beads down every 7th bead or so. And I was having trouble controlling the exact placement of the beads. Then I saw this fragment of a beading.
This piece has some beads removed -probably the more precious blue lapis based on the handful of remaining beads. I've reconstructed where beads might have sat based on the size of the lapis beads elsewhere on the piece. Based on this size there are generally 4, sometimes 3 beads between tie downs. But if this section wasn't lapis beads, and instead something as big as the coral beads, there might only be 2-3 beads per tie down.

Tieing down the beads every couple of beads in the tricky bits was much easier.

When I came to a corner in the pattern my instinct was to try and use smaller beads to better fill the gap. This meant that I was threading on only a few beads at a time before couching them down. But in the longer parts I was threading up long strings of beads before couching them and I could have used pre strung beads or a necklace as my string. I wonder how medieval beaders handled this? Did they pre-string all their beads (you loose less that way)? Did they individually thread beads as they couched? Did their beads have as much size variation as the cheap glass beads I was using? I doubt I can answer that question with certainty as it is likely some beaders worked one way and some another, but can I see any evidence for any of these questions in medieval beading works?

  • This piece is unravelling square by square, rather than by colour. But the inticate geometric pattern of the squares would make it difficult to fill in single beads of different colours, quite different style from the outlined large pieces that are more common. But it does argue for some pieces that are individually strung.
  • On this piece Jen/Grizel notes that "and the technique is more like cross stitch or Victorian needlepoint in the design NOT like the flowing couchwork of the Germans. Straight lines, not contoured ones." I think this is similar to the above piece, and both probably stitch down all colours with the same thread.
  • This piece has a diversity of bead sizes (within each colour) similar to my cheap seed beads. It also has one spot (pictured right) where a smaller blue bead has been squeezed into a corner, but nearly none of the other spots do this, so it could be a coincidence.
Finally the prevalence of small metal plaques - note the small gap with thread showing in the middle. I could have avoided it but it would have been difficult. But if I'd had a metal plaque to put in the middle, I could have been a fairly mediocre beader and you wouldn't have a gap in the middle.