Apr 152011

If you would like to re-create Queen Elizabeth’s ‚Phoenix‘ gown from the painting, there’s a reprint of the black fabric with the pearls available here on Spoonflower.
The very same design, without the pearls (in case you’d like to add your own…), is available as a reprint here; and also, there’s a cream / gold version of that fabric with the pearls available here on Spoonflower.
Note that while reprints of this design aren’t exactly historically accurate (after all, the original was decorated with cording embroidery and real pearls!); but they’re a good possibility to ‚get the look‘ without having to invest too much work.


Examining the Phoenix gown

In March 2007 I had the chance and pleasure to be able to examine the ‚Phoenix‘ gown reproduction Jean Hunnisett made for Glenda Jackson and ‚Elizabeth R‚ at the Museum of London.
Suzi Clarke, a costumer and former student of Jean Hunnisett as well as Janet Arnold, took part in the examination and told me some more things about the gown, for which I’m incredibly thankful.

I would love to describe what I felt when the first item – the overskirt – was unpacked and laid out on a table for me to examine, but no matter how hard I think, I can’t find the words. I once wrote elsewhere that this gown is holy to me, and it really is.
So rather than desperately trying to find any more words, I will just say that I had to take a *deep* breath before taking the following pictures.
My hands were more shaky than usual (I do have a medical condition which renders me unable to hold my fingers still, but this was really, really ridiculous and probably made me look as if I was suffering from a bad hangover…), but somehow I managed, even without a flash, to come forth with not too badly blurred images…

Without further trying to write something that won’t escape my keyboard anyway, here’s the result of my examination. I didn’t try to take the pattern, because this is given in Jean Hunnisett’s book „Period costumes for stage and screen, 1500-1800“. But I guess my pictures will help you understand the construction even more, as well as give you a chance to marvel at the craftsmanship involved in this reproduction.
I have to say that the photos on this page have been reduced to a size of 50% of the original pictures I took, for space- and traffic reasons. Also, I’ve not published all my pictures on this page; just the ones that are really needed to make you understand how the gown was made.


The gown is made from black cotton velvet.

The leaf embroideries, according to Suzi Clarke, have been carried out with a Cornely machine – note that this type of machine is indeed called ‚Cornely‘, a French brand, and not „Cornelly“ – many websites misspell this (so did I before I properly researched those machines). Basically it’s a chain stitch embroidery machine type which is capable of doing various cording works as well.
The brand’s logo looks like this, which should clarify how it’s spelled:

I suspect that for this particular embroidery, a Cornely A3 or L3 might have been used.
Some websites on the Cornely machines:
History of the Cornely brand (French)
What a Cornely machine is capable of embroidering (English)
Another nice website, explaining and showing what a Cornely machine can do (French)
French website, offering various models of Cornely machines for sale
Interesting enough, this particular type of embroidery machine has lately also been used to embroider the designs on Padmé’s ‚Peacock‘ gown from Star Wars, Episode 3 – see a comprehensive site on that costume at Padawansguide, and also note my own research for it here. It has apparently also been used for the embroideries on the Star Wars Episode 2 Senate gown.
The original Cornely machines have been copied, though – the Singer 114W and 114E as well as the Consew 104 are copies of Cornely embroidery machines. Here is a website on the copies.

The embroidery thread used is golden metallic.

Each leaf is 6.2cm wide and 7.7cm high. The pattern repeat (including the couched cording) is 14.8 x 8.2cm (height x width). In the following picture I’ve marked this pattern repeat with red lines as well as cut-out pieces of the measurement tape:

I must admit that I was a tiny bit disappointed by the embroidery. No! Don’t get me wrong! It’s beautiful. It’s really beautiful, and very impressive in person.
However, it’s different from the original in the portrait, and it could have been better. I’ll explain.
I do own an embroidery machine, too – a Brother one, though. It comes with a software that allows you to make your own embroideries. I took the chance to do just this from the above shown measurements picture. The resulting embroidery would look like this in my embroidery software’s preview:

By the way, this single motif – with the lines around and the three lines in between other motifs – would result in 5163 stitches (the Brother software can tell exactly, though, of course, it has to be said that this is not a Cornely machine and therefore the amount of stitches might slightly vary for this machine).
I repeat, over 5000 stitches for one motif.
If I can trust Jean Hunnisett’s information about the Phoenix gown reproduction being embroidered with 600 of these motifs, this would result in over three million embroidery stitches for all the motives.
However – now compare this motif to the embroidery on the original painting; it’s best to do this side by side with the above shown ‚measurements‘ picture:

As you can probably see, they’re much likely – but there are some differences, the biggest ones probably being the ‚rose‘ in the Hunnisett motif and the fact that the original leaf tips are almost pointed, not rounded. Also, the outline on the Hunnisett gown is zigzagged, not running stitched. Also, the original motif fills the cording lattice more than the one in the reproduction.
To have an exact comparison of how a more original motif would look like, I took the task and programmed that, too, in my software:

This motif would only result in 2178 stitches per motif, resulting in „just“ a bit over 1.3 million stitches for the whole 600 motifs. Feel free to compare this to the above shown original as well as my computerized embroidery version of the Hunnisett gown and to the photograph of the Hunnisett gown’s embroidery.
Just to have a side by side comparison again, here’s one of all the versions:

Hunniset’s original embroidery My digitized version of that embroidery The original embroidery in the painting My digitized version of the painting embroidery

In an earlier version of this page the following information has not been included (mainly because I forgot to mention the invention date of half-way affordable digital embroidery machines for the public, which I should have and which was around the beginning of the 1990s), but Suzi Clarke pointed out that the following information should be added:
There were no computerized embroidery machines available in the early 1970’s, and Mrs. Thorold stitched every one of the 600 motifs individually, without the aid of computers. If that kind of technology had been available, it would have probably been used on the costumes.
I personally would like to add that a Cornely machine works a bit different from a computerized embroidery machine and indeed the operator of such a machine has to do a „bit“ more than just programming the stitching, threading the machine, putting the fabric onto a special embroidery hoop and watch the work being done, occasionally rethreading the machine and moving the hoop to a different place, as it’s the case on modern embroidery machines.
Instead the operator of the Cornely machine has to guide the embroidery machine along the path of the embroidery in each and every motif, which is a lot more work to do.
Nevertheless, if I would get the chance to buy a Cornely machine at an affordable price, I’d take it without much thinking because the embroidery achieved by this machine is a „bit“ different from what a modern, computerized embroidery machine can achieve. The result looks more ‚handmade‘, which it in fact is.

The glued-on half pearls are 1cm each, some of them came off the gown, as you can see in the first picture above. The pearls are different on bodice and skirt, which is why I will get back to them later.

The couched cording has been carried out by hand, with a metallic cord that, when frayed, is quite rough to the touch – a tiny frayed end on which we tested this can be seen in this:

There was a funny moment during the overskirt examination when Suzi exclaimed (and I think those were the exact words):
„Hunnisett, you dirty cheat!“
This was when she saw that Jean Hunnisett obviously had machined straight lines in the places where she would later do the hand couched cording. This can be seen in some of the pictures, but here’s one where it is very well visible:

After thinking about this for some time, I think that it might well be possible that this either was done by Jean Hunnisett before the gown was embroidered (to indicate exact placements for the leaf motifs) or maybe even was done by the embroiderer, Phyllis Thorold.
The cording is just 2mm wide. The three ’stripes‘ between, over and under the repeating motifs are embroidered (together with the leaves, as it seems, as the embroidery is carried out in the same metallic thread); the later added cording goes over, not under them.


The overskirt was made from three panels of fabric. The hem circumference is 323 centimeters, the front has turned over edges of one inch, which can be felt under the black cotton lining.
Suzi told me that originally it wasn’t planned to border the bottom end of the skirt with straight lines of soutache cording, but the embroidery shrank the velvet so that the skirt was too short in the end, and this is why this was done.

The ouches with which the center front edges are decorated are not metal buttons, but were made like this:
A square of velvet, three layers, folded to 1.2 x 1.2 cm, was sewn or glued as a base to the velvet.
Over this a cross of stamped, metallic, sticky foil has been laid, the size approximately 2x2cm (square) with the crosses‘ edges turning inwards for 3mm. The edges of the cross have been bent down over the edges of the velvet square beneath. Last not least, a 1x1cm square acrylic cabochon stone, amethyst colored all over the gown, has been glued to the center.
I hope this can be seen in the four following pictures, which each show different ‚ouches‘. On some of them, the sides of the metal ‚cross‘ have come loose, showing the velvet square beneath.

The skirt is cartridge pleated to a black twill waistband:

It has been altered at the front, where the fabric is laid over the twill waistband – according to Suzi, this might have been done to shorten it for a different production. It’s a rather crude alteration, though; some of the cording and embroidery came loose during this process:


Once I temporarily got over the sheer respect I had for the gown, the bodice – in shape – didn’t look too spectacular.

However, the sheer genius of the construction of this bodice – already discussed in Janet Arnold’s „Patterns of Fashion“ – can only be understood if the bodice is opened at the front and spread out with the sleeves facing upwards:

Yes, that’s all one piece of fabric. The motives that are angled at the front do so because it is cut on the bias. This cut would still allow for moderate stretch because of this bias cut front.
Interesting enough, the cream cotton lining of this bodice is cut in three pieces:

of which the two front pieces are reinforced from the inside with Rigilene, as you can probably see as the Rigilene is puckering against the cotton fabric.

By the way, opening the bodice… there is a velcro-equipped fake chemise banding under the opening.

Also, there is a modesty panel under the left front side. Suzi noted that the hooks and eyes with which the bodice is closing all point in one direction instead of alternating. This was done because of this modesty panel.

The slashes with white fabric puffing out of them are also fake – those are bias-cut strips of some chiffon fabric (too thick for silk chiffon; so I assume it’s polyester).
Yes, strips, not turned tunnels (as Hunnisett herself suggests in her book) or slashes in the fabric with other fabric puffing out from them (as it would historically have been). You can clearly see this in the following second picture, where I have turned the fraying edges of one of the strips to the outside.

The sleeves are equipped with downwards turnable cuffs, of which the black velvet side is facing inwards when turned up. Suzi suggested that this might have been done because of the the ruffs Glenda Jackson wore in the movie; with the ruffs, the cuffs would be turned upwards, and in case the gown was ever worn without the ruffs, they could be turned downwards, revealing the velvet side and lenghtening the sleeve.

As I have already said, the pearls are different on overskirt and bodice. While the overskirt has glued on half pearls, the pearls on the bodice are full pearls, stitched on. You can see this very well in the following picture. Nevertheless, all the pearls – overskirt and bodice – seem to have come from the same manufacturer, as they all have the same color and luster. If you have ever bought artificial pearls and tried to find the exactly same color elsewhere, you will know how impossible this is if you are not lucky enough to find some that have been produced by the same manufacturer, or sometimes even from the same batch.

This picture also shows the sleeve head (left side of picture), cartridge pleated to the shoulder strap. It’s a bit difficult to see, but if you look carefully, you can see some black stitching running over the cartridge pleats.

Along the tabs at the waistline there’s a row of pearls, it’s puckering up as soon as the bodice is laid out flat, but smoothes when it’s worn.

The inside hem of that bodice has been neatly hand stitched.

Along the neckline, a black twill tape has been sewn to the mini-tabs. This is, as Suzi suggested, to prevent the wearer from being scratched by possibly coming lose cording, which is hard to the touch (see description of cording above) and would therefore have scratched the wearer along the neckline. At least one of the cordings did come lose, so that thought is not farfetched.


This is a standard farthingale by Jean Hunnisett; the pattern is given in her book. Surprising is that the bottom casing doesn’t have a hoop, and never seemed to have one, as it is sewn close.

To me, in person, it seemed a bit small for the Phoenix gown, considering that it looks like this:

when worn. Back at home it dawned me that my assumption that the farthingale could have belonged to anything else than this gown is probably wrong for the following reason:
I had in mind just what you see above – Glenda Jackson as Queen Elizabeth I. while *sitting* – when I thought this, which is no surprise – she’s sitting most of the time while wearing this gown, which makes her skirt look much wider.
However, I do have one screenshot of her standing with the gown:

and this picture should make it clear that it’s indeed the correct farthingale used with this skirt.


The petticoat held a surprise for me. When I first saw it, the first thing I could say was „Oh – a Rococo jupe!“, because that’s what it, in fact, looks like.

The measurements of the pleating can probably be estimated from this picture:

The portion under the pleats has been reinforced with something that I would call ’stiff crinoline fabric‘ (and other people would perhaps call it horsehair or stiff linen). It’s been sewn to the bottom of the petticoat with zigzagged lines.

It seems, though, as if the petticoat didn’t originally belong with this gown. It doesn’t have the „Glenda Jackson“ tag in it, as bodice and overskirt do. Instead, it is inscribed „Lady Rochford“:

And from this picture you can probably guess the rather tiny waist circumference…:

That’s it, you’re through. I *could* have examined the forepart, but this was really, really just a triangular piece of cream-yellow decorator fabric, so I didn’t.
I hope you had fun following me through the examination (well, if you just had a tiny portion of the fun *I* had when doing it, then I would bet you had a blast!). If you should feel that I have forgotten something or would like to comment on my examination, feel free to drop me an email – the address is given at the bottom of each page in this web.

  3 Responses to “Examining the Phoenix gown”

Comments (2) Pingbacks (1)
  1. hello , please i need used sewing machine cornely , thanks and best regards

    • I do not have a Cornely machine, and therefore certainly do not have one for sale.
      Your best bet would be to search the internet for a used one. Be advised though that even used ones aren’t cheap – I have never seen one for under US-$ 2,500, and even THAT was the lowest price I ever saw! – so good luck 🙂

Kommentar verfassen