Mamluk Cards

Home · History · Topkapi deck · Inscriptions · Reconstruction · Shop

A reconstruction of the Topkapi deck of Mamluk cards

How it started

when I started this project late in 2013 the only deck of Mamluk cards that had been available was the Facsilime by Jan Bauwens, published by Aurelia Books in 1972. This deck as well as the second print run from 1977 have long been out of print and are now very expensive and hard to find. The more pleased was I when I managed to get a copy of this deck. Not only are the Mamluk cards an important part of playing card history, the Topkapi deck in particular also is plain and utterly beautiful. The facsilime includes hand drawn recreations of the missing cards that very neatly pick up the look and feel of the original cards.

However it seems inevitable that a facsilime together with cards recreated from scratch will not look entirely consistent. Also the colors of the original cards are mostly faded, often bits of the color have entirely dissapeared and in some cases the colors are inconsitently reproduced with some of the original cards looking more greenish and others looking more reddish. But it's the blue patches on the cards that is most distracting about the original cards. These patches had been added to distinguish court and pip cards (see the section about the structure of the deck), but they were just painted over the existing designs and now sort of spoil the balance and geometry of the cards.

All this let me wish for a consistent and neat looking reproduction of the entire deck. Since there was no such thing at the time, I started to try and make one myself.

The underlying idea

Traditional Islamic art is a lot about symmetries, regular geometric patterns and circles. From the faksilime cards it is obvious that the creators made heavy use of rulers and dividers. The "royal" pattern in the king cards is a nice example for a distinctively symmetric structure:

It is believed that the cards from the Topkapi deck, unlike other remaining examples of the Mamluk pattern, were mainly used for representative purposes. We may assume strong saturated colors, most of which have faded away in the cards that still exist. We also may assume that the cards were intended to look very neat and ornamental. So one might guess that the people back then would have enjoyed using a tool that easily provides perfectly symmetric layouts, perfect circles and strictly regular patterns.

The idea was to create a deck that preserves this very spirit of the Topkapi deck, and that follows the ideas and principles that can be found in the deck rather than redrawing every bit exactly as it looks in the original cards. Since 15 out of 56 cards are either missing or no photos of them are available, these cards would have to be reconstructed anyway.

So I used a vector graphics program to recreate the whole deck from scratch. At the same time I tried to stay as close to the original cards as possible. Here are two examples of what the results look like in comparison to the original cards:

The complete layout of the deck

As described in the Topkapi deck section it appears to be very likely that the original type-1-deck from the Topkapi museum with 43 remaining cards used to have four suits containing 10 pip cards and four court cards each, making a total of 56 cards. The well-known blue patches on the court cards are later additions and have not been reproduced.

Line Drawings

The images below show the use of design principles, like symmetries, size ratios and geometric patterns starting with the cards that still exist.

Nine of Coins:

Take the card frame, fit the card image into it, locate the suit symbols, define a idealized drawing frame, define a grid inside of it that would make the distances between the centers of the symbols equal.

Move the suit symbols to fit into the grid. Apply this approach for other cards where images of the origial cards don't exist. Make some compromise where other patterns change the balance of the card design. Draw identical suit symbols into the new locations.

Six of Scimitars:

Take the card an find all the circular bits that define the suit symbols. Amazingly the cards are really done from parts of (almost) perfect circles whereever you look (almost). The distances between the circles tend not to be quite as regular. Finding a regular pattern for the lower half of the scimitars that would apply to all the cards of the suit was not too difficult.

Finding a pattern for the upper half that applies to all scimitar cards was a bit more tricky. A set of circles that are not quite concentric define the areas for each scimitar including black and white bits together. The lines for the black bits are designed as to make areas that contiuously become thinner further down the area for the complete part and stay in the middle of that complete part. This is approximated by more circular parts. As a result a curve for the left half of the card won't make a complete circle together with the corresponding curve for the right part any more. More circles then define the remaining bits of the suit symbols.

All the circles together define what is needed to draw the new scimitars.

Ten of Polo Sticks:

In some examples the original design is less neat. The curves for the dragons in the Ten of Polo Sticks can't be approximated by two circles. Three circle parts allow at least a rough approximation. Here the design has been straightened a fair bit, so that each curve is made from two circle parts.

For the ornaments along the dragon's curved necks to look regular one would draw circles along the outer lines with the circle centers on the line and the circles just touching each others. Repeated forms are added where the circles touch.


Most of the colors on the cards are faded which makes it difficult to guess what the original colors have been like. There seemed to have been seven or eight colors present in the deck:

The color that has faded the most seems to be red. Green often is faded a lot es well. It is hard to tell whether brown/bronze also faded or whether what looks like faded brown/bronze actually is faded red. Green is not too hard to identify as it only appears on the flowers and mostly looks like dark brown where it is faded. Faded red though can hardly be distinguished from yellow/gold. It is not clear at all if there is any faded brown/bronze at all, ans if so it would look like yellow/gold as well.

For the reconstruction I assumed faded brown or red to be mainly red where it appears on the suit symbols and brown where it appears on the ornamental patterns close to the borders of the cards. This helps distinguishing suit symbols and other bits but might well be entirely wrong. Actually the only parts where the existence of brown is obvious are the outer borders of the pip cards and some stripes on the court cards.

The image above is an example where all the colors are preserved well. The brown band at the bottom is a little faded, but still differnt from the yellow/gold above it. Also yellow/gold and red are clearly distinguishable. Green looks green and blue looks blue. There definitely is red in the ornamental bit on the right half of the image.

On this image the yellow/gold pentagram-like center of the flower on the left half looks darkened, like most of those other bits on the card do that must have been yellow/gold in the first place. However it still looks different from the brown border and from the red bits. Green looks darkened too. The red parts of the flower look partly faded into yellow. This yellow looks more yellow/gold-ish than the actual yellow/gold bits do. Probably the bits in the upper right corner would have been red as well, although it is not clear why the faded red parts would be so clearly separated from the still reddish parts, unlike in the flowery bits to the left.

Here hardly any color but blue can be distinguished

On this card there are faded traces of red inside the ornaments on the left as well as inside the suit symbol on the left right of the image, and also on some flower bits. The red looks almost the same everywhere.

However on this bit of the same card the red in the flower is well preserved while inside the suit symbol on the right half of the image there is no red at all where we would strongly expect it to appear. It is possible that the suit symbols have been drawn first and the red has been added on top of the yellow layer, while with the flowers it might have been painted directly on the paper. This would explain why the red insight of the suit symbols often is missing, but of course it remains speculation.

There is no doubt though that the original cards must have contained a lot more red than it remains present on the cards now. Still there was a lot of guessing involved when choosing the colors for the diferent elements of the cards.

Card titles

The card titles on the original cards are described in the Inscriptions section. For the reproduction only the plain titles like "King of coins" were used. As the titles for the deputies translate to "deputy" and "second deputy" the obvious extension for the new fourth court cards would be "third deputy". This gives card titles as follows:

Flower Patterns

The flower patterns proved to be more work than the whole rest of the deck because these patterns don't repeat as much as the rest. Circles and rectangles as basic shapes were used to make a variety of derived shapes:

The derived shapes were used as elements of the flowers:

These elements were combined to build the actual flowers:

Those flowers were numbered and printed on transparent paper to cut them out and arrange them manually on draft prints of the cards. This turned out to be much quicker than using the vector graphics program.

Also the flowers are connected with curved stems. They had to be added, positioned precicely, cut into shape and merged with the flowers.

Progress so far

The design of the reconstructed deck is now finally complete. The next step is to look into printing options. The plan is to start a kickstarter campaign in september and to order the prints in november or december. With some luck the cards may be available around christmas, however I guess early 2019 will be more realistic. In the meantime I will do some minor corrections to some of the cards.

There is now also a facebook-Page about the Mamluk project, were I will post more information.


Here are some more examples of the completed cards:


The images on this site are copyright protected.

Contact: write to trzes at spiorad dot net.

Last change: 17-Jun-2018

Created: Sep 2015