The Wallace Collection


Collection Blog

The Puzzle Armour

Toby Capwell, our Curator of Arms and Armour, reports back from his trip to Germany and the successful attempt to solve the mystery of the ‘Puzzle Armour’.

A few years ago the Bavarian Army Museum in Ingolstadt, Germany, acquired the very intriguing fragments of a mid fourteenth-century body armour, or ‘coat-of-plates’. Since very few examples of this once common form of medieval armour survive, this find was extremely important. The pieces were found in the ruins of a German castle by an amateur archaeologist, who did not record anything about the context of the find. He also, unadvisedly, attempted to clean and reassemble them himself, having no apparent knowledge either of historical armour or modern conservation practices.



Individual fragments of the armour


When the pieces were offered for sale in Munich, the fragments were hastily assembled on a torso mount. The pieces remained in this assembled configuration until early this year, when the BAM invited a small group of arms and armour specialists from other museums to come to Ingolstadt to formulate a new and better informed reconstruction of these fascinating bits of medieval metal. I was pleased to join the team, and off we went to Bavaria.

Joined by, among others, my fellow curators of arms and armour Dirk Breiding (Philadelphia Museum of Art), Alfred Geibig (Kunstsammlungen Veste der Coburg) and Dr. Raphael Beuing (Bavarian National Museum, Munich), and hosted by the BAM’s curator Dr. Tobias Schönauer, we began by taking the previous assemblage apart completely. This way we can examine and interpret each of the forty or so fragments separately, preventing the examination from being influenced by anything other than the nature of the given piece itself.



Examining test results


The attempt to put all of the fragments back together in something resembling the correct locations was challenging to say the least. It felt very much like trying to put together a jigsaw puzzle with forty or fifty percent of the pieces missing, the available pieces having had their original external surfaces completely removed, and with no box cover image to work from.



Deliberating the puzzle


Nevertheless, quite a lot of progress could be made. The pieces still could tell us quite a lot. Their degree and type of curvature, the presence or absence of rivets (and if present, their number and orientation), and their basic shapes all could potentially reveal quite a lot about where on the armour they came from.



The armour is slowly assembled


Certain parts, such as what was obviously the main chest plate, could immediately be located with certainty. The nature of the original edges on some other of the plates helped us to understand where they belonged on the body- whether they came from an edge of the garment, or from somewhere in its midst, whether they were likely to belong on the front or the rear, and so forth. Comparing what we were beginning to understand about the original garment’s type, design and date, we were also able to start matching up the developing new assembly to mid fourteenth-century depictions of similar armours in German medieval art, in manuscript illustrations but most importantly on sculpted funerary effigies.



The main chest plate


One of the most significant features of this armour are its surviving ‘arming chains’. These were mounted on the main chest plate and originally used as ‘safety lines’ of a sort, to attach the wearer’s sword, dagger, shield and helm to his body, to prevent them from being lost in combat. While students of arms and armour are familiar with such chains from artistic depictions, these are the first surviving ones to be discovered. This important aspect makes this armour a crucial type-specimen and primary reference for our understanding of the evolution of medieval armour.

When completed, the new reconstruction had several important advantages. First, it made use of all the surviving pieces, which the previous attempt did not. Secondly it revealed that parts of the sides and back of the garments apparently had survived, while the previous assembly had concentrated solely on the front. Finally, enough of the armour was shown to survive that a much enhanced overall impression of the piece could be formed.


The arming chains


The reconstruction project was a wonderful example of how curators of different collections, with different interests and different kinds of experience can collaborate in order to better understand an enigmatic object which might seem utterly confounding and mysterious to any one person working on their own. It also demonstrates how a curator’s duties can extend out far beyond one’s own museum, and how collections of a similar type all over the world are dependent on each other for mutual inspiration and support.



A true collaborative effort!


I am very grateful to Dr. Schönauer and to Dr. Ansgar Reiß, Director of the Bayerisches Armeemuseum, for inviting me to contribute to the project and for the rare opportunity to handle this extraordinary object. During the recent visit we also discovered a number of fascinating connections between our two collections, which I hope we will be able to explore in future research projects.

Dr. Tobias Capwell
Curator of Arms and Armour


Great to know that this spectacular piece of armour got the appreciation it deserved. Will there be a documentation of sorts for it? I´m especially interested because I´m personally involved – the amateru archaeologist contacted me for help in dating the armour.

Thorsten Piepenbrink

Thorsten Piepenbrink on 4 April 2014

Excellent. Congratulations on this splendid achievement. If only I could have been there, secretly observing.

Dierk Hagedorn on 4 April 2014

Terrific work! I’m glad to see scholars revisiting old hack jobs and setting them right! However, I noticed the lack of any overlap in the plates. Perhaps it was intentional that they are not displayed that way, but virtually all of the evidence we have (from he other 25 extant coats of plates and other mid-14th C armour varieties made in a pieced-construction) indicates that the plates ought to be overlapped. Was this a consideration in the reconstruction?

Keith Kempenich on 5 April 2014

Fascinating discovery. A few questions: 1.Any speculation on the function of the spindles included at the top of each chain? 2. The absence of rivet holes was mentioned; were the plates sewn into the leather perhaps? 3. Are there any plans to create scale drawings of this armour? I am interested in making a full size replica to study how it moves when worn. Many thanks in advance for your time.

Respectfully, Scott

Scott Kershner on 8 April 2014

Thank you very much for your comments, Toby is actually on holiday at the moment but I will pass this on to him as soon as he gets back! Marie Stirling (Marketing and Development Assistant)

Marie on 9 April 2014

Scott, I have some likely answers to your questions. 1) The toggles on the chains hook into various pieces; sword, dagger, great helm. The spindle can be pushed through a loop or a hole and when drawn snug, it will catch and hold fast. This can be seen in many manuscripts from the period. 2)No rivet holes are evident because there are still rivets in them! You can see the rivet heads still in place on some of the plates. They would have been riveted to a shell. In my experience using wearing coats of plates, stitching them would be insufficient to hold it together for very long. 3) Also from experience, if the coat of plates is assembled with the plates overlapping, there is relatively little movement in them and the entire garment acts mostly as a solid whole. There is usually a bit of accordion-like action if you bend forward or side-to-side at the waist, but not a dramatic amount. For this reason, the armour probably did not cover the hips, which would restrict movement.

Keith Kempenich on 11 April 2014