New light on the Warwick Shaffron: understanding horse and shaffron size through the collections of the Royal Armouries

The Warwick Shaffron, held in the collections of the Royal Armouries (VI.446) is the earliest extant example of a European medieval shaffron on public display anywhere in the world. In addition to its early date, the shaffron is also considered exceptional because of its seemingly formidable size, coupled with apparent physical evidence of battle damage. This study sheds new light on the Warwick Shaffron through the application of an original measurement methodology developed through the AHRC-funded ‘Warhorse Project’ and applied to shaffrons in the collections of the Royal Armouries. Drawing upon the Warhorse Project’s wider findings on medieval horse stature, this paper critically discusses what shaffrons can tell us about the size of the mounts that wore them. A headline finding is that the dimensions of the Warwick Shaffron are entirely compatible with what we know about the size range of medieval horses at the time of its manufacture, albeit towards the upper end of the range.


Introduction
As well as offering protection to the mount, whether in tournament or battle, horse armour changed the outward appearance of the animal and enhanced the image of its rider.As such, horse armour represented the 'visual language of power', making the mount more menacing and accentuating aspects of its physique as well as displaying the wealth, status, sophistication and martial prowess of its rider. 1 Shaffrons (head defence) are the most commonly surviving elements of horse armour.This paper presents the results of a fresh examination of the shaffron renowned as the earliest surviving medieval example in Europe-the Warwick Shaffron (Figure 1), held in the collections of the Royal Armouries (VI.446), and taking its name from Warwick Castle (Warwickshire), where it formed part of the armoury of the Earls of Warwick. 2 It is often considered exceptional because of its seemingly formidable size, coupled with apparent physical evidence of battle damage.This paper casts some fresh light on the Warwick Shaffron by presenting the results of a new programme of research that has recorded detailed measurements of a large sample of medieval and early modern shaffrons, including the piece itself and the majority of other examples in the collection of the Royal Armouries, including those in storage and on display in the Tower of London, with the aim of testing what these pieces can tell us about the stature of the horses that wore them.
Horse armour remains seriously understudied relative to human armour.Building on essential groundwork in the eighteenth and nineteenth centuries, Viscount Dillon published a specific study of the subject in The Archaeological Journal, in 1902, which identified exemplars in major collections and collated key documentary references. 3The essential chronology of horse armour development has since been refined within the context of more general works on medieval arms and armour. 4Research specifically focused on horse armour has primarily been conducted from an art historical perspective, with the emphasis on charting and explaining stylistic development and celebrating the status of pieces as fine objets d'art rather than investigating what these pieces tell us about medieval horses and the ways they were used in war. 5 The finest and most fully developed studies of horse armour are found in museum catalogues.A recent example was published by the Metropolitan Museum of Art, New York, for a rotating exhibition on the subject, 6 while a monograph accompanying an exhibition of the armour of Emperor Maximillian (1459-1519) at the same museum collates information on his spectacular horse armour. 7Archaeologically focused approaches to medieval horse armour are rare by comparison, although the evidence from the Roman period has attracted attention, 8 while an experimental archaeological study of mail horse armour sheds new light on its protective value. 9he work reported on in this paper was conducted under the auspices of a collaborative research project 'Warhorse: Archaeology of a Military Revolution?', funded by the Arts and Humanities Research Council, based at the Universities of Exeter and East Anglia, and working in partnership with the Royal Armouries. 10Running between 2019-2023, the project conducted the first ever systematic study of warhorses, and horses generally, in medieval England, understood more broadly within its wider British and European context, in the period AD 800-1600.The point of departure for the project was the fact that established understandings of warhorses had been based almost entirely on historical scholarship-hinging on the use of documentary sources-while the archaeological evidence had been largely overlooked, despite its richness, diversity and, through new state-of-the-art methodologies, its capacity to challenge received wisdom and to create new knowledge and understandings.It set out to break new ground by collecting and analysing the fullest range of evidence available for medieval horses and integrating these otherwise disparate strands of evidence into a new narrative.Alongside historical documents, the evidence base investigated by the project comprised: the physical remains of horses ('zooarchaeological' materials: bones and teeth, including small samples taken from them for various scientific analyses); equine material culture (apparel and armour); horse breeding landscapes (studs and stables); and visual depictions of horses in medieval sculpture and art.The overarching aims of the work were not only to create a new and more rounded understanding of the medieval horse, but also to unpick its complex, fascinating and ever-evolving interrelationship with medieval society through the centuries. 11his paper begins with a summary of previous scholarship on the Warwick Shaffron, which prefaces a description of the piece and a discussion of known parallels in sculpture and manuscript art.Given that the methodology used to measure the shaffron is new, being developed specifically for the Warhorse research project, this is outlined in full in the hope that it can be applied more widely.The following section presents the results of the shaffron measuring exercise by highlighting how the metrics of the Warwick Shaffron compare to those of others within the collections of the Royal Armouries.A discussion then draws out how these findings can feed into a wider discourse and debate regarding the sizes of medieval horses, prior to a brief summary that underlines the central conclusion that the Warwick Shaffron is well within the size range for medieval shaffrons generally, but was clearly intended for a horse of far above average stature for the period.

The Warwick shaffron
The Warwick Shaffron, on display at the Royal Armouries Museum in Leeds, is an exemplary object embodying the power and technology of European warfare in the late-fourteenth and early-fifteenth century.As the earliest known example of a European plate shaffron, usually dated to about 1400, the Warwick Shaffron is significant in what it reveals about the form and construction of horse head defences in Europe at the advent of the development of plate armour. 12The Warwick Shaffron belongs to a group of horse head defences which appear in iconography around the end of the fourteenth century, and are identifiable by a long, inflexible plate extending backwards from the poll, and a 'bulbous' muzzle encompassing the nose and mouth. 13The closest example to the Warwick Shaffron seems to be that shown on a chess piece, probably English, and currently dated to around 1360, at the Metropolitan Museum of Art (see Figure 2).There is a key difference between the two: the eye guards are open on the chess piece example. 14The form of ear defences on the Warwick Shaffron is unknown as they are now lost, and it is unclear from the chess piece precisely what form of ear guard it originally represented.Parallels to the Warwick Shaffron in medieval manuscript and other art, defined by their apparently larger, more hood-like form, and with pierced eye, nose and sometimes also pierced ear guards, tend to survive, largely, in iconography of the first decades of the fifteenth century, mainly in manuscript illustrations but also sculpture, most notably Henry V's tomb in Westminster Abbey (Figures 2-3). 15However, the chess piece and other examples in Figure 3 demonstrate that there are earlier outliers. 16For the purposes of this study therefore, the authors accept the dating by Eaves and Richardson that the Warwick shaffron was probably made sometime towards the very end of the fourteenth century and probably no later than 1420. 17Post 1410, shaffrons appear to lose the rigid form of the rear plate at the top of the head, and the rounded, pierced termination at the nose, which becomes narrower and more 'guttershaped'. 18 Horse armour was in use in Europe from around the mid to late eleventh century. 19In this early period horses were protected by trappers of quilted fabric or mail (or a combination of both) and, later, by small plates sewn into textiles. 20The shaffron, either of leather or plate, first appeared in the mid-thirteenth century, followed by larger plate defences for the body. 21By the mid-fifteenth century horse armour comprised a shaffron for the head, crinet for the neck, peytral for the chest, flanchards to protect the sides, and a crupper worn over the horse's rump.Other armoured horse furniture included steel rein-protectors, and saddle steels.Horses were equipped in some or all these pieces depending on the occasion, and the wealth and preferences of the rider.Full surviving horse armour, also known as bards or barding, is extremely rare, the earliest dating to 1450. 22The difficulty in investigating horse armour in iconography prior to this period arises from the use of textile caparisons.These were long fabric covers which sometimes covered the horse from nose to hoof, in use from the late twelfth century. 23They were often brightly coloured and featured heraldry and other identifying symbolism, and were worn either alone or over mail, leather or plate bards.Where caparisons are depicted in surviving iconography it is difficult to ascertain, therefore, the form of horse armour, if any, worn beneath.
For the mounted knight, protecting his horse, as much as himself, was essential, and shaffrons provided some defence on the battle and tournament field against edged weapons and projectiles such as arrows, pikes, lances, other staff weapons, and sword blows.
The Warwick Shaffron possibly originally belonged to Richard Beauchamp, Earl of Warwick (1382-1439), whose seat was at the historic Warwick Castle. 24Richard Beauchamp gained active military experience, first in Wales, and, in the first decades of the 1400s, in France during the period now known as the Hundred Years' War. 25 He was also celebrated and recognised for his skills at tournaments, such as the Christmas festivities hosted in Calais in January 1415, which are depicted in the Beauchamp Pageant. 26t is likely the Warwick shaffron remained at Warwick Castle until the second half of the twentieth century. 27It can clearly be identified in an illustration of armour from Warwick Castle, by J. Hamilton, published in 1786. 28The earliest known mention of the shaffron at Warwick Castle is possibly in a description by Richard Dugdale in 1656. 29Following Grose and Dugdale's early descriptions the shaffron then featured in several publications by some of the nineteenth-and early twentiethcentury's leading scholars of arms and armour, including Sir Guy Laking, Sir James Mann, and Claude Blair, among others, and was also included in two notable London exhibitions. 30Until the late-twentieth century the Warwick Shaffron was largely discussed briefly, as a remarkably early example of a plate shaffron, used to illustrate and evidence the development of horse armour.
In 1987 a full precis of the Warwick Shaffron's provenance, historiography, and, most importantly, a detailed description and discussion of the object's form and use, was provided by Ian Eaves and Thom Richardson's comprehensive article in the Journal of the Arms and Armour Society. 31It was in this study that the shaffron was published in detail, interrogating and examining all aspects of the piece's individual and independent history, including how its form and damage might be interpreted, as well as where it sits within the wider history of shaffron development.
The shaffron has since appeared in publications which further highlight the object's importance as an historical treasure, and, as in in Paggiarino's dramatic photography works, also as a visual and sculptural masterpiece. 32For the purposes of this study it is important to note that whilst the Warwick Shaffron is described in previous material as 'voluminous', and 'of exceptional size', it has not yet been discussed for the purposes of understanding the stature of the horse or horses that wore it, or how this 'imposing' object is representative of, or an exception to, average horse size in the fourteenth and fifteenth centuries. 33he Warwick Shaffron (Figure 4) is formed of a large, embossed ferrous metal plate covering the face, with a prominent medial ridge.The eye guards and muzzle are domed and pierced with a series of large circular holes to enable the horse to see and breathe.At the top of the face plate are large holes for the horse's ears, which probably originally held upright ear guards.Two further plates, joined at the top in a continuation of the medial ridge, extend over the poll and down the sides of the shaffron.Three pairs of holes at the top edge of these additional pieces are for the attachment of a crinet, whilst a single row of small holes, called 'points', line the side edges of the shaffron.These were probably for the attachment of a textile liner which would be laced to the metal and therefore removable. 34he shaffron is undecorated, except for the scalloped edge of the face plate, where it overlaps the other pieces.Despite its plain appearance in comparison to the highly decorative shaffrons of later decades, the shaffron is of high quality.The carefully shaped face plate and sharp medial ridge speak of the armourer's skill.The eye defences are embossed, not as circular or semi-circular cups, as seen on other examples from the end of the fourteenth and turn of the fifteenth century, but instead like a large teardrop embossed from the face plate, and rounded at the bottom. 35Most of the metal is thin, and yet the shaffron is heavy and would have been robust in the period of its use.There is nothing to identify either maker or owner.The underside is rough from the hammer and reveals rows of surviving rivets hammered flat and largely invisible on the outer side.
There is damage and material loss to the top edge of the shaffron and to both eye cages, as well as a long dent above the right eye, and material loss the bottom edge of the nose.Three other remarkable features appear to indicate damage in battle or tournament (see Figure 4): 1. Near the top of the shaffron, 10 mm to the left of the medial ridge, as worn by the horse, is an impact point shaped like an isosceles triangle (1.5 mm Â 2 mm Â 2 mm), which at its centre point penetrates through the entire plate. 362. Mid-way down the shaffron, cutting across the medial ridge and angled slightly down from left to right, is a cut, 11 mm long Â 1 mm wide, which slices through the entire plate.The angle of the cut indicates that the blow which created it was struck downwards.3. Two-thirds of the way down the shaffron, above the nose guard, is a dent, with maximum dimension of 13 mm wide Â 7 mm long, which has damaged the plate but does not penetrate its full thickness.
In the 1786 illustration shown in Grose's A Treatise on Ancient Armour, the shaffron is shown already displaying many of the damage features noted above.Hamilton depicted the damage and loss of material from the eye cages, the depression above the right eye, and at least one depression across the medial ridge.Based on these early illustrations it seems the shaffron has remained in a fairly stable and largely unchanged state since the late eighteenth century. 37

Methodology
No systematic study has been made of the relationship between horse armour and horse size prior to the work of the Warhorse Project.While museum catalogue entries typically contain details of gross maximum dimensions and weights for pieces of armour, 38 finer measurements are mainly unavailable without first-hand inspection and recording.An original methodology was therefore developed to enable detailed recording of armour pieces in significant collections, including those of the Royal Armouries. 39iven that the purpose of the research was to explore the potential for armour to provide an index of horse size and conformation, the methodology focused on recording landmark positions on each piece that have a clear relationship with skeletal morphology.The method focused solely on shaffrons, which have a clear and obvious relationship with a discrete element of horse physiology (the head, including its length, breadth, volume and the distance between eyes and ears), and survive in sufficient numbers in major collections to make the exercise viable and ensure that the metrics produced from it are sufficiently voluminous and robust to repay analysis.A number of factors mean that shaffrons cannot be taken as an exact proxy of horse head size and shape, however.Most important of all, most shaffrons were originally padded or lined, to prevent excessive rubbing, meaning that they were manufactured slightly over-size.Fragmentary evidence of padding sometimes survives, especially around inside edges, where rivets can be found holding in place a strip of leather with the remains of padding fabric trapped beneath.While fibres have not been analysed, the material visibly resembles hessian, usually with coarse underlying layers but a finer fabric on the surface facing the horse.Another important consideration is that while some shaffrons were bespoke creations for specific (and sometimes known) patrons and their mounts, others would have been for more general usage and sized accordingly.
It should be noted that other elements of horse armour do potentially hold information on horse morphology but issues with their recording and interpretation mean that the Warhorse Project's work focused purely on shaffrons.Cruppers, covering the horse's hindquarters, are exceptionally rare survivals, and the fact that most are of two hinged plates means that the precise relationship with horse physiology is not straightforward, as they can be expanded and contracted.Peytrals (breastplates) are similarly unusual and, hanging quite loosely around the horse's chest, are a relatively imprecise indicator of size, while most crinets (covering the horse's neck) consist of conjoined elements that can be both contracted and extended, especially when mounted on model horses, meaning that their relationship with neck length is difficult to reconstruct.These caveats aside, measuring these elements as potential indicators of horse size might be a suitable area for future research.
The system for measuring shaffrons was designed to be non-intrusive, relatively rapid and repeatable and used a pro forma recording sheet for each separate piece (Figure 5). Figure 6 shows the recording programme in action.The standardised nature of the recording system means that the methodology can be applied in the future to items in other collections and the results compared and contrasted, although as discussed below not every single measurement can necessarily be taken on every single piece, with the morphology of half shaffrons and some exceptionally decorated pieces presenting particular challenges.On each pro forma was entered key background information on each individual piece, including date, location, and details of decoration (including the presence or absence of an escutcheon plate, plume-holder, rondel and central spike), alongside notes on any re-working or damage.The pro forma also recorded up to 22 separate linear measurements in millimetres.The locations of these measurements were designed to capture distances between landmark points that could be identified across a large number of shaffrons, regardless of artistic embellishment or design considerations, with a particular focus on measurements that were clearly related to the size and shape of the underlying horse's head.Measurements were taken using a combination of precision digital callipers and hand tapes.The locations of each measurement are illustrated in Figure 5, and detailed below: 1. Linear measurement between inner edges of ear-guards/ear-holes at top (poll) of shaffron (mm) 2. Linear measurement between centre-points of the front of ear-guards/ ear-holes (mm) 3. Linear measurement of max width of shaffron below ear-guards/earholes (mm) 4. Linear measurement of max width of shaffron at the back of eyeguards/eye-holes (mm) 5. Linear measurement of distance between centre-points of eye-guards/ eye-holes (mm) 6. Linear measurement of distance between inners edge of eye-guards/ eye-holes, not including any decorative element (mm) 7. Linear measurement of distance between outer edges of eye-guards/ eye-holes, including any decorative element (mm) 8. Linear measurement of max width of shaffron in front of eye-guards/ eye-holes (mm) 9. Linear measurement of narrowest part of nose (mm) 10.Curved measurement following the contours of the shaffron at the point of max width below ear-guards/ear-holes (mm) 11.Curved measurement following the contours of the shaffron at the point of max width behind eye-guards/eye-holes (mm)  The most useful of these measurements in terms of replicability between pieces and the potential for cross-comparison with measurements on modern, living horses (see below) are the following: 2 (between ear holes); 4 and 8 (maximum width, both above and below eye holes); 5 (between centre points of eye holes); 14 þ 15 (length of horses head from poll/back of cranium to nose immediately prior to nostril flaring) and 21 (centre of eye hole to front centre of ear hole).The other measurements are of lesser significance in terms of relating shaffron size to horse morphometry but were taken for their future research potential in comparing metrics between different pieces.It should be noted that design differences between pieces meant that not all measurements could be taken for all examples, however.Half shaffrons were recorded using the same system but with several measurements related to the lower part of the piece absent.In some cases there were two-part shaffrons, where the nose piece could be detached to create a half shaffron.In such cases, the specimen was measured as a full shaffron with the nose piece correctly fitted.The following sections plot the results of this programme of shaffron recording for examples within the collections of the Royal Armouries, including the Warwick Shaffron, and discuss the implications of these for our understanding of horse size.

Results
All 22 measurements taken from the Warwick Shaffron (VI.446) are presented in Table 1.
In total, 49 additional shaffrons in the care of the Royal Armouries at Leeds or the Tower of London were included in the metrical analyses presented in this paper.A small number more were examined and measured but excluded from this study because they were later replicas or their design made it impossible to obtain appropriate data.The measurements graphically compared in this paper have been selected as those that most reliably represent key points of fit to the horse that wore the shaffron.The addition of measurements 14 and 15 best represents the length of the horse's head from the poll (back of head) to just before the nostrils.However, in some shaffron designs, even if considered full shaffrons, the metal can terminate a little short of the nostril region, whilst in others flaring indicates that point has been reached.Thus the length measurement must be viewed as a minimum with some design variation.Measurement 4 best represents the width of the head plus any padding and at this point of measurement there is very limited error caused by design variation.Measurement 21 relates to the distance between specific anatomical landmarks (eye to ear) where openings in the shaffron need to be correctly aligned with the animal's facial features.This is perhaps the most reliable measurement for relating the metrics of the shaffron to the horse within.These selected measurements are presented in Table 2, for all the historic specimens.
To provide a frame of reference that allowed us to understand the relationship between shaffron size and horse stature more fully, we were able to locate four modern horse shaffrons that we knew had been specifically manufactured in an authentic early sixteenth century form to properly fit particular modern animals of different sizes.All these animals had been extensively ridden in re-enactment events, undertaking activities such as jousting, in full armour with an armoured rider. 40hese animals ranged from 14 to 16 hands high (hh) at the withers and their shaffron measurements are listed in Table 3. 41 All the modern shaffrons were modelled on early sixteenth-century shaffron designs.
The following three graphs compare (Figures 7-9) the Warwick Shaffron with the other specimens.The historic specimens have been divided by date with the Warwick Shaffron sitting on its own and categories for late fifteenth, early sixteenth and late sixteenth centuries (a few fall into the early seventeenth century).The dataset is dominated by early sixteenth-century examples.The four modern specimens are to the right of the graph.These plots are box, whisker and jitter plots.The jittered dots represent individual data points and the vertical line is the full range.The box represents the inter-quartile range containing a line showing the mean average.
The first graph (Figure 7) presents measurement 21 (eye to ear).This measurement was a challenge to take from the outside of Warwick shaffron due the eye guard design, but it could be taken completely accurately from the inside when the shaffron was inverted.The first point to note is that the Warwick Shaffron is well within normal ranges for this key measurement relating to anatomical landmarks.It is slightly larger than the interquartile ranges for other historical periods but far from unusual in size.It sits in the middle of the modern shaffron range, relatively close in size to the modern horse called Warlord, which was 15.2 hh.The graph also appears to suggest that shaffrons in the late fifteenth century were generally small, but caution is needed with any such conclusion.First, the sample size of four specimens is very small and, secondly, the smallest two in this period are blind shaffrons (eye position is marked but there is no hole to allow sight) associated with the Rennen joust. 42If these are removed as exceptions of a different type, then there is nothing remarkable about the other two shaffrons in terms of size.The metrics of the Rennen form may be worthy of further research in other collections to obtain a representative dataset.Figure 8 presents a similar graph representing the length of the horse's head (measurements 14 þ 15).Almost identical observations can be made with respect to head length from poll to nostrils.As noted above, the length measurement can be more affected by shaffron style and this may particularly affect the late sixteenth century examples when the design had a great tendency to terminate early, before flaring for the nostrils.
Width (measurement 4) is presented in Figure 9.In this case the Warwick Shaffron is an outlier, and it appears to be much wider that all specimens from later periods.This is fascinating, since the Warwick Shaffron is totally comparable in terms of key anatomical landmarks on a horse's head (poll, nostrils, eyes and ears), yet this design is very spacious in terms of width, being about 10 cm wider than expected from later designs.One key factor that could affect different measurements to different extents relates to the use of greater padding.This possibility is considered further in the discussion section below, alongside general issues of horse stature in the medieval period.

Stature of medieval warhorses and Warwick Shaffron
In popular culture and amongst some scholars, armoured and armoured knight-bearing medieval warhorses are assumed to be massive steeds; R.H.C Davis suggested  that great horses could have reached between 17 and 18 hh (hands high), equating to 1.83 m or 6 ft to the withers, for example. 43Mounts of this stature would be akin to the drum horses of the present day household cavalry which can be larger than 18 hh, 44 or police horses that are commonly over 17 hh 45 The Warwick Shaffron has been considered to be of 'exceptional size' 46 which implies also a horse of unusual stature, however, these assumptions remained largely untested.We can, in fact, see from Figures 8 and 9, which reflect key anatomical landmarks for facial length and eye-ear positions respectively, that the Warwick Shaffron is not exceptional, when compared to later shaffrons.It is upper quartile but within the normal range of shaffron size.Furthermore, it is midrange in size compared to shaffrons custom made for modern horses of between 14 and 16 hh.It is only the width (Figure 7) that is exceptional, a feature which could relate to the use of additional padding (see below), since other anatomical landmarks are normal.It is this additional width that perhaps makes for the shaffron's voluminous appearance.Given that the Warwick Shaffron's key anatomical metrics are in line with shaffrons made for known 14-16 hh horses, the implication is that many commonly held impressions are off the mark, but how does that size range compare to osteological evidence of horse size in the middle ages?
A recent study of a very large sample of medieval horses bones 47 was able to model withers heights from long bone length measurements (see Figure 10).This showed that the average English medieval horse was only around 13 hh (132 cm).For reference, the generally accepted modern cut-off point between a pony and a horse is 14.2 hh. 48Thus the majority of medieval horses are technically ponies, and animals in the range from 14.2 (147 cm) to 16 hh (163 cm) are a small minority with no surviving evidence of specimens exceeding that.Of the 268 withers heights reconstructed only 2.2% are above the pony/horse threshold, with the highest such frequency reaching 4.5% in the High Medieval period.Whilst rather modest-sized animals by modern standards, horses in this range would still have been notably larger than the vast majority of medieval equine stock, hence warranting the label 'great horses'.The size of the Warwick Shaffron is very much in line with animals in this size range.The successful modern use of horses from 14 to 16 hh in armoured reconstruction and jousting events demonstrates the physical feasibility of employing horses of this stature, if the conformation is correct.Detailed measurements taken by the Warhorse project on the wooden horses in the Line of Kings in the Tower of London, show that those in the earliest and original group, dating to the mid-sixteenth century, 49 measure in the range of 14.0-14.3hh, with an average stature of 1.47 m (14.3 hh).

Padding and shaffron metrics
From our observations, internal linings very occasionally survive attached within shaffrons of the sixteenth century and comprise a few layers of cloth amounting to a few millimetres in thickness. 50This does not preclude the use of additional padding which was either not attached to the shaffron or no longer survives.A clue to the potential nature of unattached padding can be found on a rare mid-sixteenth century miniature model of a Stechzeug armoured horse and rider in the collections of the Bavarian National Museum. 51In this instance, an inner layer of linen sacking, with ear and eye holes, is draped from the poll to just above the nostrils.A second, outer layer has ear holes, but no eye holes and stops short on the nose.This presents some potential evidence for the existence of additional padding, in this case to be worn under a blind shaffron, though it still represents just two layers of cloth.
Whilst such linings have negligible implications with respect to the size of horse the shaffron might fit, the use of thicker padding could have.The largest effect of thick padding would be on the width measurement (4), since the pad thickness would need to be fully accounted for on both sides of the shaffron.The head length from poll to nostrils (14 þ 15) would be less affected, since raising the plate at the end of the nose would not impact on length.However, if padding was present over the poll, the point of hinging for the poll plate might extend back further by the thickness of the pad.There is a more complex geometric relationship between pad thickness and measurement from eye to ear (21), but that effect is limited.Given that many sixteenth-century shaffrons already imply relatively modest horse size, it seems unlikely that very thick pads were employed.
The Warwick Shaffron, however, is exceptionally wide.It is up to 10 cm wider than one would expect given its other metrics.Does this imply some form of padding up to 5 cm in thickness, which would require extra space on both sides of the shaffron?If this were true, what would the metrical effect be on the three measures?Figure 11 shows the size of horse's head that would fit within the Warwick shaffron, if 5 cm padding was employed, and compares that to all other shaffrons for all three measurements.With this adjustment for padding, the width left for the horse's head is very much in line with other shaffrons.If 5 cm of padding was present over the poll then the length measurement (14 þ 15) becomes comparatively midrange, instead of upper quartile.The impact of padding on the eye to ear measurement (21)  was approximately modelled by taking a calliper measurement between points 5 cm inwards from the eye and ear holes.As with length, this adjustment makes the Warwick shaffron appear mid-range in size.It appears that thicker padding can reasonably explain the shape and size of the Warwick Shaffron.The exact nature of such padding remains speculative but thick padding could potentially involve rolls of material at key anatomical points rather than a full covering.One could also consider the space required for tack when using this design.The use of the Warwick shaffron design, with various types of padding, tack and fixings, could usefully be the subject of some experimentation with replicas.

Conclusion
Discussions with owners of modern jousting horses confirms that shaffrons do need to be manufactured to an appropriate size to fit their mounts.As we observed ourselves, practically, transferring a shaffron intended for a larger horse to a smaller one and vice-versa can result in misalignment with key anatomical features such as ear and eye holes.While horse head size certainly does not vary in direct proportion to stature at the withers, there is clearly a positive correlation.By comparison with the modern examples, it appears that the vast majority of shaffrons held by the Royal Armouries would be consistent with animals of between 14-16 hands high.Some may indeed fall below that.This is also consistent with metrical evidence obtained from archaeological medieval horse bones that suggests that animals around 13 hh were average and those above 14 hh were unusually large, and there is no current evidence for medieval English horses that exceed 16 hh.
Based upon shaffron length and positions of eye and ear holes, the Warwick Shaffron is not exceptional in size when compared to later shaffrons.It is larger than average in these respects, but is far from being an outlier.One of Henry VIII's earlysixteenth century shaffrons, object number VI.1, is in fact clearly the largest example held in the collections of the Royal Armouries.It is the width measurement that gives the Warwick Shaffron its bulky appearance, rather than key anatomical landmarks.We propose that the most parsimonious explanation for this is that this earlier form of heavy, all-enclosing, shaffron was worn with more padding than later forms.Up to a likely maximum of 5 cm of padding may have been employed at key locations on the head to support the shaffron.If the measurements are modelled to take approximate account of such additional padding, then the horse the shaffron was made for would appear to be mid-range in size on all key dimensions.Direct evidence for padding practices at c. 1400 is absent.However, the form and function of Warwick-type shaffrons may reasonably have necessitated such additional protection.Another example of a similar form of embossing over the eyes, although not completely covering the eyes, could be interpreted from the chess piece held at the Metropolitan Museum of Art, accession number 68.85, see Figure 3. 36 Eaves and Richardson, p. 219, who describe this damage as a 'small, slightly square, puncture' suggest this impact could have been caused by a bodkin head, and if so, could be evidence that the shaffron was used on the battlefield rather than for tournament, although they do reference Sir James Mann's assertion that the hammerwork, being coarse, and the lightness of the metal may instead suggest tournament use.Whilst the presence of a potential bodkin head impact mark does assert that the shaffron may have been worn on the battlefield, as arrows were not known to be used in mounted tournament events, it is not sufficient evidence that the shaffron was used exclusively in this context.It may be that the shaffron was used in both tournament and battlefield contexts. 37Whilst some of the damage to the Warwick Shaffron may be explained by possible contemporary weapons and use, it is important to note that all damage to historical arms and armour should be treated with caution.It is impossible to know for certain if damage is the result of actual battlefield or tournament blows, or if the object was damaged in some other way, either during its working life or in the centuries afterwards. 38

Figure 2
illustrates this development.

FIGURE 2 .
FIGURE 2. Parallels to the Warwick Shaffron in manuscript art and sculpture.Image created by Kate Kanne.

FIGURE 3 .
FIGURE 3. Sculptural depictions of Henry V and his warhorse, equipped with a shaffron similar in style to the Warwick Shaffron, in the chantry chapel adjacent to his tomb in Westminster Abbey.Photographs by Oliver Creighton.

FIGURE 5 .
FIGURE 5. Pro forma sheet for measuring shaffrons, developed by the Warhorse Project.

FIGURE 6 .
FIGURE 6. Shaffron measuring in action in the collections of the Royal Armouries (object number VI.408).

FIGURE 8 .
FIGURE 8. Box, whisker and jitter plots of measurements 14 þ 15 (length from poll to nostrils in mm) divided by period.

FIGURE 9 .
FIGURE 9. Box, whisker and jitter plots of measurement 4 (width behind eyes in mm) divided by period.

FIGURE 11 .
FIGURE 11.A 3D scatterplot of all three measurements (14 þ 15, 4, 21 in mm) comparing the Warwick Shaffron (red cross), as adjusted for padding, with all other shaffrons.With such adjustment the Warwick Shaffron is within the interquartile range for all measurements.

TABLE 1 .
Measurements from the Warwick Shaffron.

TABLE 2 .
A list of shaffrons included in this study along with their storage location, accession number, approximate date, and key measurements (mm): 14 þ 15, 4, and 21.

TABLE 3 .
Details of modern horses and their shaffron measurements.
FIGURE 7. Box, whisker and jitter plots of measurement 21 (eye to ear in mm) divided by period.
armoury of the Earls of Warwick, and, according to Eaves and Richardson, remained at Warwick Castle from the time of its use until about 1970.It was included in an exhibition Chaucer's London, London Museum, in 1972.It was then sold on 26th February 1987, via Howard Ricketts Ltd., on behalf of an anonymous vendor, and purchased by the Royal Armouries.The purchase was made possible by funds raised by the Royal Armouries, the National Heritage Memorial Fund, and a private benefactor.The shaffron was included in the Royal Academy's exhibition Age of Chivalry: Art in Plantagenet England 1200-1400 (Nov 1987-Mar 1988), and was then on permanent display at the Tower of London.The shaffron then moved up to the Royal Armouries Museum, Leeds, in 1996, and returned briefly back to the Tower for the Agincourt 600 exhibition (October 2015-Jan 2016).It is currently on display in the War Gallery at the Royal Armouries Museum in Leeds. 28Francis Grose, A Treatise on Ancient Armour and Weapons (London: S. Hooper, 1786), p. xxvii.pl.XLII. 29Sir William Dugdale, The Antiquaries of Eaves and Richardson, 217-19.It has previously been suggested that the seemingly enormous size of the Warwick Shaffron was to accommodate a thick, padded lining; see Impey, p. 68.
Pyhrr et al.The Metropolitan Museum of Art, New York; The Wallace Collection, London; The Styrian Armoury, Graz; The Imperial Armoury, Vienna; and Wien Museum, Vienna. 40We were kindly allowed access to the modern examples and details of the horses by Jason NEW LIGHT ON THE WARWICK SHAFFRON