Research by Charles Mellor, may be reproduced if properly credited.
This diagram is based upon an extant garment shown in ‘Lost Treasures
Persia’. The garment is shown from the back, laid out flat. Front details
are conjectural, based upon garments painted in miniature painting of that
period and locale. This particular caftan is dated at last quarter of
sixteenth century, though this type of garment was in style from about 1300
The original garment is composed of “dragonslayer” brocade, which shows a
scene of a Persian man/warrior slaying a Chinese style dragon in a landscape
setting. The picture is in black & white and the color is described as
light blue. The brocade seems to be quite detailed with a repeat of
approximately 18 inches vertical. As is typical for brocades from this
period in Persia, the motif is repeated with colors alternated or varied
and also with the scene reversed from left to right on alternate repeats. As
is also typical for brocades of this period, it is most unlike anything made
nowadays. This particular type is really more like you photocopied a
painting and then pasted multiple copies together edge-to-edge, and then
covered up the edges with overlapping leaves from the trees in the scene.
No attempt is made to match the patterns on the back of the garment. Based
on extant Turkish garments I have seen, I would guess that the patterns
might have matched at center front, but probably not anywhere else. There
are even places were tiny slivers of fabric were pieced together, most
likely as an economy because this fabric would have been very expensive. In
this pattern I have only shown the seams that I believe are “functional”
rather than economical.
Both men and women wore this style of garment as an outer covering. At
this length it almost sweeps the floor on me, but comes up to mid calf if
belted. I have seen paintings of both men and women wearing this type of
garment in every length from hip length to floor length. Women more
typically would wear it long, but the split seems to be about 80 – 20, with
both sexes wearing longer garments for formal or indoor activities and
shorter garments for outdoors or on horseback.
This type of garment is usually worn over other layers of a similar cut of
garment, perhaps of a lighter weight. Another garment worn under this is
the “Mongol” or crossover coat. Some people believe that women never wore
the Mongol coat, but the pictures clearly show that they did. Pants similar
to the Turkish “salwar” were always worn by both sexes under all the other
layers. I have seen very few paintings were a short garment is worn over a
longer one (not counting a short coat over long pants), but I have seen many
pictures with longer coats worn over another coat that is short enough to
tuck into the pants. It is very hard to guess what the most inner layer of
garments would have looked like. There are paintings that show men wearing
knee-length white pants that seem to be cut about the same as salwar. It is
hard to tell from the picture, but they are made out of fabric that is
either crinkled like gauze or has a fine vertical stripe. In some paintings
the women are wearing what seems to be a gauzy undershirt with sleeves about
10″ longer than their fingertips that sometimes fall down over their hands.
I saw one painting with a man wearing an undershirt of crinkled
gauze/stripes that seemed to fasten at the side of the neck like a Cossack
I can’t tell from the photo of garment what the weight of the fabric would
be, but based on paintings of the period, a very drapey, softer fabric was
most common. I have seen no evidence in paintings that the super-starched
Turkish look was ever used in Persia. Based on paintings, other appropriate
fabrics would be: brocades with small, over-all geometric patterns, brocades
with small over-all patterns of Chinese motifs (most typically cloud-scroll
patterns), brocades with arabesque patterns, brocades with foliate or
acanthus in either over-all or serpentine runners, solids with gold
embroidery or stamping randomly applied or applied to hem and wrist, Turkish
style brocades, Venetian style brocades, cut velvet in Turkish patterns.
Even though everybody thinks stripes are typical of Middle Eastern design,
stripes are never shown for use in garments in Persian art, except for
peasant/labourer class clothing. Based on paintings, this particular type
of garment is frequently lined with fur or silk shag in improbable colours.
The garment in the photo seems to be lined with a lightweight broadcloth of
yellowish green with a repeat pattern. It is possible that garments of this
type may have been lined with Ikat silks in period (as they were in the 19th
– 20th cent) but this is entirely conjectural. I know they had that fabric,
but I have seen no pictorial evidence to support it’s use in lining.
These garments most typically buttoned from neck to waist, though they were
not always worn buttoned and sometimes show no evidence of any fastening in
the paintings. As far as I can tell from all the evidence I have seen,
buttonholes were not used. Instead, loops or tabs were used. Buttons were
often grouped in clusters of two, and from the evidence I have seen were
most typically either toggle or ball buttons. Sometimes horizontal stripes
of braid where used to accent the closings. Based on the paintings, it is
either appropriate or even essential for the garment to gap at the closing
and show the next layer of fabric underneath. It is hard to tell for sure,
but some of the paintings seem to show the women wearing several layers of
these buttoned at the neck and belted at the waist and open to show a
vertical sliver of skin on their torso. Some people believe this was for
nursing or in the harem. I have examined several paintings showing this
with a magnifying glass in person, and I cannot say whether they are showing
skin or undertunic.
- This could be made without the slit in the arm that allowed the
sleeve to be used as a hanging sleeve. In that case, the sleeve would be
shortened to be about 10 – 12 inches longer than the fingertips and worn
pushed upon the wrist.
- This could be made with short sleeves, slightly above elbow.
- When made with short sleeves, sometimes a full-length sleeve would
also be made. This would be worn over the short sleeve and attached with a
button at the shoulder for warmth.
- Sometimes a full-length sleeve would be attached with buttons
right at the edge of the short sleeve.
- Either variation 4 or 5 can be worn attached by the buttons, but
with the arm out of it, as a hanging sleeve.
- Sometimes for riding this garment would be slit up the back to the
seat, or the side slits would be extended higher than shown.
- Prior to about 1400, the taper of the sleeve would not be as
great; making a sleeve that was more open at the wrist. Since this would
not stay up when pushed up at the wrist, the sleeves were shortened
somewhat, to approximately knuckle length.