Carolingian Arms and Armor in the Ninth Century

By Simon Coupland

From Viator: Medieval and Renaissance Studies v.21 (1990)

This study seeks to ascertain the nature of the armament carried by the Carolingian army in the ninth century by examining the written, iconographic, and archeological sources. The value of such an approach was demonstrated by Gessler's study of Carolingian weaponry published in 1908,1 but this work is now largely outdated, and more recent discussions of the subject have seldom contained a balanced evaluation of all three types of evidence. Thus Ferdinand Lot simply stated, “Il y a peu de chose a dire de l'armement," and subsequently included the briefest of discussions.2 Ganshof regarded the Psalterium aureum3 as the only reliable pictorial source and obtained almost all the rest of his information from capitularies.4 Verbruggen basically repeated Ganshof's findings, citing the tenth-century Leiden Maccabees 5 as additional iconographic evidence.6 Only Last has given detailed consideration to all three types of material, but he produced what was intended to be no more than a summary of current knowledge. The article's brevity was also a handicap, so that illustrations were reproduced without suitable discussion of parallels, possible models, or conformity with other types of evidence.7 Finally, although a recent book by Nicolle set out to provide an accurate portrayal of Carolingian arms and armor, it displayed little critical evaluation of the sources.8 This resulted in such bizarre features as an illustration of a ninth-century Breton cavalryman equipped with third-century Roman horse armor and a "fourth– to seventh-century Romano-Byzantine helmet"! 9

From this brief survey of the secondary literature, it is apparent that a detailed examination and comparison of the different types of primary material is long overdue. Among the written sources, ninth-century narrative texts rarely described arms or armor, and it was only in the capitularies of Charlemagne that Carolingian military equipment was discussed in detail. As a result, these few references have been extensively cited and studied.10

The armament carried by the cavalry was specified in two texts. The Capitulare missorum of 792-793 11 referred to benefice and office holders who were able to possess horses and armor, as well as shield and lance, longsword (spata) and sax (semispatum).12 In the letter sent by Charlemagne to Abbot Fulrad in 806, however,13 each horseman was commanded to have a bow and several quivers of arrows in addition to the shield and lance, sword and sax.14

The infantry was not required to be so heavily armed: The Capitulary of Aachen, issued in 802-803,15 expected the counts to supply each of their foot-soldiers with a shield and lance as well as a bow with a spare string and twelve arrows.16 The equipment for the royal war carts also included these weapons, doubtless for the driver's protection.17 The Aachen capitulary contained a further provision that no soldier should carry a cudgel (baculum), but rather a bow.18 This presumably related to those free peasants who had no rich lord to equip them.

            It is unfortunate that the capitularies prescribing armament all date from the reign of Charlemagne, since there were undoubtedly changes in military equipment over the course of the ninth century. For example, it is demonstrated below that short swords, or saxes, disappeared during this period, but other developments almost certainly occurred of which we are at present unaware.

The only other written source to provide a detailed description of ninth-century Frankish armament is Notker's biography of Charlemagne. This depicted the emperor and his entire army as iron men, each clad in iron helm, cuirass, thigh plates, greaves, and gloves, carrying iron shields, swords, and spears on iron-colored horses.19 Notker's aims, however, were to entertain his listeners with impressive stories and to provide an exemplum of a glorious Carolingian monarch, rather than to write an accurate historical account of early ninth-century conditions. Indeed, Lowe has written of Notker's "bereits ins Sagen- and Legendenhafte erhobene Bild" of Charlemagne.20  Not only is the equipment described much grander than that mentioned in other contemporary texts, but it would also have been far beyond the means of the vast majority of Frankish troops, as well as impractically heavy and cumbersome in battle. It is therefore almost certain that the Carolingian army was never equipped with such complete sets of heavy armor, not even in Notker's own lifetime.

With regard to the iconographic evidence, many Carolingian manuscripts and ivories depict soldiers and their armament in considerable detail. Unfortunately, art historians have often reached conflicting conclusions as to whether such figures were modeled on contemporary Frankish troops or on late antique or Byzantine miniatures. For example, it is generally agreed that the cavalrymen in the Psalterium aureum are portrayed wearing ninth-century Frankish armor.21  Yet their helmets are demonstrably of the same type as those worn by the soldiers depicted in the Vivian Bible,22 who are commonly judged to have been based on late antique pictorial models.23

The present study seeks to avoid this type of confusion by examining each item of Carolingian armament in turn, taking into account the full range of Carolingian pictorial sources, the possible late antique or Byzantine models, and the relevant written and archeological evidence. Throughout the study it is emphasized that even though certain aspects of an illustration may have been influenced by external tradition, individual details might still reflect contemporary local conditions.24 For instance, the miniature illustrating Psalm 3 in the Stuttgart Psalter25  displays an unmistakable similarity to the corresponding picture in the ninth-century Byzantine Chludov Psalter.26 The composition is almost identical, both groups of men are armed with spears, and some of the Carolingian troops are depicted wearing the same plain, rounded helmets as the Byzantine soldiers. Mutherich has rightly cited these parallels as evidence of Byzantine influence on the Frankish Psalter.27 Yet significant differences can also be observed, such as the wings on the Frankish spears and the banded helmets worn by the other Carolingian horsemen. Archeological evidence demonstrates that winged lances did not come into use on the Continent before the eighth century, while the banded helmets appear to be crude representations of a standard Carolingian type which is portrayed in a number of manuscripts.28 The depiction of these items therefore indicates that the Frankish artist incorporated at least some features from contemporary society in his work.

It is evident from this example that archeology can be of great assistance in confirming or complementing written and iconographic evidence. Unfortunately, there are all too few archeological remains from the Carolingian period. Grave goods, which provide the bulk of such material in early medieval northern Europe, ceased to be deposited west of the Rhine at the beginning of the eighth century, east of the Rhine in the mid-700s, and in the north of the empire another fifty years later.29 After these dates Carolingian objects are known only from stray finds, which are generally difficult to date due to the absence of any context, or as exports to regions beyond the imperial frontiers. As a result, the development of Frankish military equipment can usually be followed as far as the early eighth century, in some cases to the early ninth century, but rarely beyond that date.

HELMETS

Ninth-century written sources occasionally mention the wearing of a helmus or galea, although the shape of these helmets is never indicated, and the material from which they were made is seldom specified. It has already been demonstrated that Notker's description of "ferreus Karolus, ferrea galea cristatus"30 is not necessarily reliable, but the Breton leader Murman is also reported to have worn an iron helmet.31 The helms which Eberhard of Friuli bequeathed in 867 32 and those on which the Lex Ribuaria set a price of six solidi33 were surely also made of metal rather than leather, in view of their high value. A capitulary reference which implies that only army commanders were expected to own helmets likewise suggests that these were more than simple leather caps.34

Unfortunately, archeology has not furnished us with any helmets datable to the ninth century, nor indeed are any eighth-century specimens known. Hejdova argued that a hemispherical helmet found at Chamoson dated from the ninth century because she deemed its decoration to be Carolingian plant ornament.35 However, in a discussion of the identically shaped Niederrealta helmet Schneider demonstrated that both helms should be dated to the early twelfth century.36 Hejdova's further suggestion that the nasal and brow-piece of the so-called Wenceslas helmet, a conical helm of the tenth or eleventh century, were taken from a lost Carolingian helmet of the late eighth or early ninth century37 is likewise without foundation. The ornament on the nasal, which Hejdova believed to be early Carolingian, is in fact consistent with the dating of the rest of the helmet.38

It seems unlikely that such one-piece conical helms were worn on the Continent during the ninth century. To my knowledge, only one such helmet is portrayed in any Frankish pictorial source of the period, namely on one folio of the Bern Psychomachia, which dates from the late ninth or early tenth century.39 Stylistic analysis suggests that the artist who drew the miniature in question did not illuminate any of the other folios in the manuscript,40 and this explains why the same illustration also includes the only depiction of scale armor in the manuscript. Scale armor has demonstrable links with Byzantine pictorial tradition, and it is therefore significant that conical helmets similar to the one in the Bern Psychomachia can also be seen in contemporary Byzantine miniatures.41 These parallels, coupled with the absence of Frankish analogues, even in the same manuscript, suggest that the artist probably derived his inspiration from a Byzantine miniature rather than a Frankish cavalry helmet.

One conical helmet type which probably was worn by the Carolingians is the Spangenhelm. The name derives from the six or more metal Spangen, or strips, which joined the headband to a metal plate at the apex, forming a framework which was then filled with plates of metal or horn. The Spangenhelm is attested in the West from the third to the seventh centuries,42 and although there are no later Merovingian or Carolingian finds, a Spangenhelm portrayed in the Corbie Psalter of about 800 43 is strikingly similar to those found in sixth-century burials.44 Although it is possible that the Corbie artist copied an earlier illustration incorporating a Spangenhelm, the absence of any such models, together with the fact that the lance in the miniature seems to have been copied from life,45 makes it more likely that the Spangenhelm was still worn in the ninth century.

The standard Carolingian helmet appears to be most clearly portrayed in the Psalterium aureum.46  The helmet can be described as follows: the cap tapers toward a projecting neckguard, with an obvious rim encircling the entire helm. This rim appears to rise to a point at the forehead, where a button marks the intersection with a band descending from the apex. This band may form part of a crest running across the whole of the cap, which some sources depict bearing a plume as well.47

Most of these features can be observed on the helmets in the Stuttgart Psalter, although the neckguard is less pronounced, and the band displaced from the center of the forehead.48 The same helmet type is also portrayed in the Bible of San Paolo fuori le Mura,49 and in several of the decorated Bibles from the Tours scriptorium.50 Similar helmets can also be seen in the Utrecht Psalter51 and the Bern Psychomachia,52 and on ivories such as a worn tablet now in the Louvre53 and a diptych in Milan.54

Two factors suggest that the Psalterium aureum helmet represents a type which was genuinely worn by the Carolingians rather than one which originated in external pictorial tradition. First, the helmets depicted in surviving late Roman and Byzantine miniatures are not of this type.55 Although Dufrenne has claimed that late antique models can be found in the synagogue at Dura 56 and on the Theodosian column,57 in both cases the helmets are closer to those in the Vatican Virgil than to the Psalterium aureum type.58 Second, the most plausible explanation for the consistent portrayal of the same helmet type in such a variety of Frankish sources influenced by different iconographic traditions is that the artist copied from life. As is demonstrated below, the depiction of round shields with onionshaped bosses provides a parallel case where archeology has proved the contemporaneity of the type. As for the derivation of the Psalterium aureum helmet, several late Roman cavalry helms have been discovered which could represent the sort of model from which the Carolingian form ultimately developed.59

One of the helmets bequeathed by Eberhard of Friuli was described as "helmum cum hasbergha."60 Although this term evidently signified a mailcoat (hauberk) in later medieval texts,61 the fact that the will of Count Eccard of Macon referred to a "brunia cum alsbergo" implies that in the ninth century the two items were distinct.62 The association of the halsberga with a helmet in Eberhard's will, but above all the evident etymological derivation, from hair: "the neck" and berga: "a protector," suggest that the item in question was some sort of neck protector, possibly attached to the helmet. Oman proposed that the halsberga was a "thick leather covering hiding the ears and neck,"63 but the Spangenhelm found at Krefeld-Gellep had a mail curtain forming a neckguard,64 as did the late eighth-century Anglo-Saxon helmet recently discovered in York.65 As the Stuttgart Psalter also portrays what appears to be a helmet with a mail curtain, this seems the most likely interpretation of the "helmum cum hasbergha."66

SHIELDS

The cheapest piece of defensive equipment available to any Carolingian soldier was the shield. Whereas the price of a helmet was set at six solidi in the Lex Ribuaria, and that of a byrnie at twelve solidi, a shield and lance together cost only two solidi.67 Although the law text is a reliable indication of the relative value of these items, the price of shields and lances may well have risen during the ninth century, as a Lotharingian charter from 850 recorded a donation of "scutum cum lancea, valentem V solidos."68 Most Frankish troops could undoubtedly afford to possess shields, and contemporary written sources show that both foot soldiers and horsemen carried them. In Charlemagne's letter to Abbot Fulrad, for instance, it was laid down that "unusquisque cabalarius habeat scutum,"69 while at the siege of Bergamo the infantry held their shields above their heads in a testudo to ward off the hail of rocks from the city walls.70

The supply of shields was regarded as a matter of importance by both the king and the magnates. Charlemagne commanded royal stewards to ensure that shieldmakers were resident in every district,71 and charters of Louis the Pious and Louis the German reveal that some monasteries were required to send shields, horses, and lances to the treasury every year.72 At Saint Gall and Corbie there were scutarii in the monastic workshops73 who were presumably responsible for equipping the milites and perhaps also for providing the royal gift.

Since shields would not have lasted long in combat, shield-sellers accompanied the army on campaign, as we know from the report that "scuta vendentes" obstructed Charles the Bald's troops as they fled after the battle of Andernach.74

Unfortunately, written sources give no indication as to the size, shape, or construction of Frankish shields, but miniatures and ivories consistently portray a single type of shield, carried by both infantry and cavalry.75 Round and concave, the shield appears to have been about 0.8 m. in diameter, protecting the body from the neck to the thighs.76 However, Carolingian artists are known to have been influenced by late Roman and Byzantine artistic traditions, and both Byzantine and late antique manuscripts also depict round, concave shields of a similar size.77 It has therefore been claimed by some art historians that the shields portrayed in Carolingian sources were copies of such pictorial models and not contemporary Frankish armament.78 However, two factors suggest that this is not the case.

Firstly, most of the shields depicted in Frankish sources have an onion-shaped sugarloaf boss,79 while many are also decorated with a series of radial arcs.80 These features are not found in either late Roman or Byzantine illuminations. The arcs apparently formed part of Germanic tradition, since they can also be seen on the Hornhausen stone, from about 700,81 and on picture stones on Gotland.82 As for the onion-shaped shield boss, archeological evidence indicates that this form only developed in the eighth century, and thus that the Frankish artists who drew such bosses were reproducing contemporary conditions.83

Secondly, the few eighth-century shield remains which have been discovered testify that Frankish shields of the period were indeed round or oval, although the known specimens were somewhat less than 80 cm. in diameter, ranging from 52 to 70 cm.84

All of these eighth-century shields were wooden, and at least one was covered on both sides with leather. The low price of shields in the Lex Ribuaria also suggests that Carolingian shields were wooden. Notker's description of the shields carried by Charlemagne's army, "in clipeo nihil apparuit nisi ferrum," was almost certainly exaggeration or invention, as the weight of such a shield would have rendered it too unwieldy to be of value.85  Nevertheless, it seems like that the radial arcs depicted in many manuscripts represent metal strips, riveted to the wooden shields for extra strength and protection.86 Most of the shields in ninth-century manuscript illuminations and ivories also have a distinct rim, similarly fastened by rivets, and several illustrations portray lattice-work mounts attached to the shield just inside the rim.87 Although no ornaments of this design are known from the Continent, similar mounts have been found in the Swedish boat-graves of Valsgarde, where the shields were also reinforced with iron bands riveted to the surface.88

The shield was held by a grip running across the center of the rear of the boss and attached to the shield board on either side. This grip can be seen in several illustrations,89 and the remains of both wooden and iron examples have been discovered.90 In addition to this handle, the shield was also fitted with a carrying strap, enabling it to be slung across the shoulder when required.91 Abbo described King Odo hanging his shield round his neck before charging into battle ("id scutumque simul recipit colloque pependit"),92 and Ermold recalled serving with his shield over his shoulder ("Huc egomet scutum humeris ensemque revinctum / Gessi").93 The Psalterium aureum contains a clear depiction of afoot soldier with a shield slung across his back: the strap over his shoulder appears to be secured to the end of the shield grip.94

In battle, the primary function of the shield was of course to ward off hostile blows, and Ermold even claimed that shields could deflect a javelin in flight.95 It has also been suggested that the combination of a light wooden shield and a pointed sugarloaf boss could enable the shield to be used as an offensive weapon, to be thrust forward against an attacker.96

BODY ARMOR

The Carolingians obviously perceived body armor as a highly important piece of military equipment, and Charlemagne's capitularies in particular made repeated references to it. Royal vassals with twelve or more mansi were required to possess a brunia, and any man who owned such armor but failed to bring it on campaign would be deprived of both his armor and his benefice.97 Army commanders were encouraged to own helmets or body armor,98 but the high price of a brunia must have put it beyond the means of all but the wealthy.99 For example, Count Eccard of Macon included only one brunia in his will, although Eberhard of Friuli bequeathed no fewer than four.100 Ganshof's assertion that the Capitulary of Aachen "made it obligatory for counts to have brumae and helmets in reserve in order to equip horsemen destined to be armored knights"101 is not supported by the text, which neither mentioned the use of horses nor specified that the counts should supply anything more than light arms to their followers. It is evident from the lists of equipment cited earlier that the common infantry and cavalry were not expected to possess body armor, but churches and monasteries apparently owned stocks of armor from which to equip their milites.102

The strategic importance of armor is also indicated by the continued attempts made by the Carolingian rulers to prevent its export.103 Nevertheless, written sources imply that both the Saracens and the Bretons wore armor,104 and that the Vikings acquired bruniae from unscrupulous Franks.105

The most common Latin terms employed to denote body armor were brunia and lorica, although thorax is occasionally found.106 As the terms brunia and lorica were neither employed interchangeably nor contrasted, it is unclear whether they were synonymous or denoted different types of armor. However, the fact that no distinction was drawn between them in law codes and capitularies suggests that both words had the same meaning. For instance, the ban on sales of the brunia abroad presumably applied to all armor, not merely to one particular type. Notker's use of the word lorica also implied that it could be used as a general term for armor, since he wrote, "Coxarum exteriora, quae propter faciliorem ascensum in aliis solent lorica nudari, in eo ferrets ambiebantur bratteolis" (the outside of his hips, which on others are usually bare of armor for easier mounting, were covered by iron plates).107

Although Raban Maur's derivation of the lorica as "eo quod loris careat; solis enim circulis ferrets contexta est"108 was taken directly from Isidore of Seville,109 a subsequent reference to the lorica's hooks and iron rings ("hamis et circulis ferrets") was evidently the author's own composition.110 However, no ninth-century pictorial source appears to portray mail, and no other contemporary written source confirms that lorica or brunia could denote a mailcoat.111 Instead, contemporary miniatures and ivories depict two forms of body armor, although the two types are not normally portrayed in a single source.112 The Romanic cuirass, with pteryges forming a skirt and covering the upper arms, is found in works such as the Utrecht Psalter,113 the Milan diptych,114 and illuminated Bibles from Tours.115 Scale armor, apparently composed of small overlapping metal plates attached to a jerkin or skirt,116 can be seen in for instance the Stuttgart Psalter,117 the Psalterium aureum,118 or on an ivory now in Florence.119

Historians and art historians who have discussed Carolingian military dress have consistently stated that scale armor was genuinely worn by Frankish troops.120 However, the passage from Notker quoted earlier casts doubt on this theory, since although much of the account was exaggerated, the statement that Frankish troops did not normally wear armor on their hips seems to have been included to lend credibility to a somewhat unbelievable story, and is thus presumably reliable. Yet the illustrations of scale armor invariably show the skirt covering the hips. Furthermore, soldiers wearing a very similar type of scale armor to that portrayed in Carolingian illustrations can be seen both in contemporary Byzantine miniatures121 and in late antique manuscripts.122 In particular, an unusual costume in the Stuttgart Psalter, consisting of a mixture of scale armor and plate, is virtually identical to the armor portrayed in a contemporary Byzantine manuscript.123 Although it is possible that Carolingian armor was similar to that worn in contemporary Byzantium, these parallels, Notker's statement, and the lack of positive evidence for scale armor in Frankish written sources all suggest that Carolingian artists were probably copying from external pictorial traditions rather than life.

By contrast, most commentators have agreed that the cuirass and pteryger depicted in Carolingian illustrations were based on classical pictorial models.124 The closest parallels are to be found in late antique artifacts such as the Probus diptych 125 or the Scipio shield,126 or on one of the David plates from early seventh– century Byzantium.127  However, the fact that some Carolingian artists consistently added royal armigeri dressed in cuirass and pteryger to classically inspired composi–tions 128 suggests that they may have represented a contemporary Frankish feature which was deemed sufficiently familiar and important to require inclusion. If this was the case, it seems likely that the figures would have been depicted in Carolin–gian costume, as Kohler commented with regard to identical uniforms in the Vivian Bible.129 The present study shows that the artists certainly equipped the soldiers with contemporary helmets, shields, swords, and spears. The obvious resemblance to Roman armor might be explained by the strong element of continu–ity in early medieval western Europe, and perhaps the imperial pretensions of the Carolingian dynasty.

To summarize, the lack of positive evidence in the form of archeological finds or unambiguous written descriptions does not allow categorical conclusions about Carolingian body armor to be drawn.   Raban Maur's discussion of the lorica is the only piece of evidence to suggest that mailcoats were worn, but the author's work was largely based on that of Isidore. Some pictorial sources portray soldiers wearing Romanic cuirass and pteryger, while others depict scale armor, but a remark of Notker's implies that the latter was not usual Frankish military dress. Since some type of body armor was evidently worn, the most likely form would therefore appear to have been the cuirass.

Several other items of armor are also mentioned in ninth-century texts. Leg armor was listed both in Eberhard's will 130 and in the Lex Ribuaria 131 under the name of bagnbergas (bagn: "leg"; bergas: "guards"), while Notker employed the classical term ocreae.132  Notker's claim that greaves were made of iron is supported both by their relatively high price in the law code, 6 solidi, presumably for a pair, and by their inclusion among Eberhard's bequests. No capitulary included them in the lists of equipment which troops were expected to own, and they were almost certainly luxury items which few possessed. Two different forms of leg armor seem to be portrayed in pictorial sources, tight scale leggings in the Stuttgart Psalter,133 and greaves possibly depicted on an ivory tablet in Florence.134 There is no indication which, if either, of these were known as bagnbergas.

The hands and arms could be protected by armored gauntlets or armguards. Gauntlets, manicae or wanti, were mentioned by both Eccard and Eberhard,135 and the fact that the latter referred to "bruniam I, helmum I et manicam unam ad ipsum opus" indicates that this was a piece of armor rather than a simple glove. The sale of armguards, bauga, was prohibited along with the sale of bruniae in a capitulary of 803,136 and it is possible that the "brancale" listed in Eccard's will was an armguard, although it may simply have been an armring of precious metal.137  I know of no ninth-century portrayal of an armored gauntlet, although an ivory in Leipzig may depict an armguard.138

SWORDS

During the reign of Charlemagne, the Lex Ribuaria recorded the price of a sword and scabbard as seven solidi, and that of a sword alone as three solidi.139 By comparison, a text from late in Charles the Bald's reign put the price of a sword at five solidi.140 Swords were therefore relatively expensive items, and this explains why in Charlemagne's capitularies only the cavalry, who could afford to own and maintain a horse, were required to possess swords.141 A passage in Regino's Chronicle suggests that by the late ninth century at least, the sword was considered to be the mounted warrior's principal weapon.142

Two types of sword were referred to in the capitularies: the spata, or longsword, and semispatum, or sax. Both types are well known from archeological finds.

The sax was a single-edged weapon made of iron, some 65 to 80 cm. in length.143 In the seventh century it was a standard feature of weapon graves,144 but by the late eighth century the sax was already beginning to disappear.145 It seems to have become all but obsolete by the early ninth century,146 and none of the major ninth-century narrative sources refers to the use of short swords, nor are they apparently depicted in any contemporary illustration. In this respect the decree that horsemen should carry both sword and sax seems to have reflected a long established Frankish custom which was in the process of dying out. A possible explanation for this development will be proposed below.

Despite the disappearance of grave goods within the empire in the eighth century, a relatively large number of Carolingian longswords have been found, some in Continental rivers, others as exports to eastern and northern cultures which still practiced the custom of burying the dead with their weapons. As a result, it has been possible to study the development and chronology of the Carolingian sword in some detail.147 The spata was a double-edged weapon, usually between 90 and 100 cm. in length, of which the blade represented some 75-80 cm. Eighth-century blades were often damascened, but this practice began to die out in the ninth century, almost certainly as a consequence of improved forging techniques which produced blades of higher quality steel.148

The reign of Charlemagne also witnessed a significant change in the shape of the longsword blade. On earlier swords, the edges had run parallel for most of the length of the blade, then converged sharply a little way above the point. After about 800, however, the edges of the blade tapered gradually from hilt to tip, with the result that the centre of gravity shifted toward the sword grip, making the weapon significantly more maneuverable and facilitating swordplay.149 Such a development may account for the disappearance of the sax, which had hitherto presumably represented the weapon used for swift, agile fencing, while the heavy, unwieldy longsword was relied upon for its sheer weight and destructive power. Now that weight and maneuverability were combined in the spata, the sax would have become redundant.

Early medieval swords are classified and dated according to the shape of their pommels and guards, and the most common pommel forms in the late eighth and ninth centuries were either a simple triangle or a series of three or five rounded lobes.150 As these were native Frankish forms which did not appear before the eighth century, it is significant that miniatures in the Stuttgart Psalter,151 Utrecht Psalter,152 Lothar Gospels,153 and Bern Psychomachia,154 as well as a painting on the wall of the church at Malles,155 all depict swords with pommels of these types. This fact lends strong support to the contention that Carolingian artists reproduced contemporary conditions in their work.

Another development which can apparently be dated to the reign of Charlemagne is the addition of inlaid inscriptions to the blades, of which one of the earliest, and the best known, is the smith's name Ulfberht.156 The oldest known Ulfberht blade dates from about 800, though such blades continued to be produced over the following three centuries.157 The name Ulfberht suggests that the original smith came from the lower Frankish region,158 but the paucity of finds from within Francia itself makes it impossible to locate the area where Ulfberht blades, or indeed any other Carolingian swords, were manufactured. Nor unfortunately do any ninth-century texts record the activities of swordsmiths, apart from a reference to the presence of "emundatores vel politores gladiorum" in one of the workshops at Saint Gall.159 In addition, two men busy sharpening swords are portrayed on one folio of the Utrecht Psalter, one using a grindstone, the other a file.160

The distribution of Ulfberht blades, from Iceland to Russia, reveals the importance of the Frankish arms trade.161 Carolingian rulers attempted to insure that Frankish weapons were not exported to potential enemies,162 and in 864 Charles the Bald threatened anyone caught supplying arms to the Vikings with the death penalty.163 It is important to note, however, that the king took this step because the practice was already widespread, and the archeological evidence likewise suggests that export embargoes were unsuccessful. One early tenth-century Arab writer, Ibn Fadlan, even remarked that the Vikings in Russia carried swords of Frankish type.164

To the south of the empire, the Saracens likewise recognized the quality of Carolingian swords, as is indicated by their demand for one hundred fifty such weapons as part of the ransom for Archbishop Rotland of Arles in 869.165 However, the Franks seem to have prized Saracen swords equally highly, and one of Charlemagne's vassals is said to have captured from the Moors "spata india cum techa de argento parata."166  “Spata india" literally signifies an Indian sword, or a sword from India, but it was presumably the term used by the Carolingians to distinguish a curved Saracen weapon from the straight Frankish spata. Count Eccard of Macon also left a "spata indica" in his will, and the fact that the same individual who received this sword was also bequeathed "tabulas saraciniscas," perhaps of ivory or silver, reinforces the belief that a spata india was a Saracen sword.167

            To summarize, as a result of technological changes during the reign of Charlemagne, the ninth-century Frankish sword was a considerably stronger and more maneuverable weapon than its antecedents. The swords' signed blades and high cost both reflected the superior quality which made them greatly sought after by other peoples, including the Scandinavians. These improvements were presumably a by-product of the increased military activity of the period, and have only been identified because of the relatively large number of finds of ninth-century swords. If other artifacts were available in similar quantities, further discoveries about developments in ninth-century armament could probably be made.

The swords bequeathed by Eberhard of Friuli in his will must have been particularly valuable, since their hilts were decorated with gold, silver, and ivory.168 In one instance the point ("cuspis") was also said to be made of gold, and since this is unlikely to have referred to the blade, the most plausible interpretation is that it denoted the scabbard chape. An illustration in the Bible of San Paolo apparently depicts a gold chape,169 and several such fittings in gold or silver have been found on sax scabbards in eighth-century Frankish graves.170

Given that Carolingian scabbards were made of wood and leather,171 this sort of ornament offers one possible explanation for the very high price of scabbards in the Lex Ribuaria. At four solidi they appear to have been more costly than the swords themselves.172 However, the law codes contained normative values for standard objects, and most Carolingian scabbards were almost certainly unornamented, except perhaps for surface decoration of the wood and leather covering. This ornament can be seen in several manuscripts, such as the Stuttgart Psalter,173 Utrecht Psalter,174 and Vivian Bible,175 and on ivories such as the casket from the Tournai school now in Berlin.176 An alternative explanation must therefore be sought.

For the sword to be worn, it had to be attached to a sword-belt, and a number of miniatures show the mounts which performed this function.177 Although rarely found on the Continent,178 many such mounts have been discovered in Scandinavian silver hoards, and others in Croatian graves.179 The numerous finds in Scandinavia again suggest that the Vikings managed to obtain Frankish swords in relatively large numbers, either through trade or through plunder.

A complete set of fixtures seems to have consisted of two or three oval or halfoval mounts, one large strap-end, a buckle, and a trefoil mount.180 Menghin attempted to establish how these fittings were attached to the scabbard and belt, and thus how the sword was worn, from illustrations of Carolingian spatharii.181 However, the illustrations in question depict the sword and belt being carried, not worn, and do not offer a secure basis for reconstructing the arrangement of the mounts.182 Nevertheless, this arrangement can be reconstructed with the additional help of pictures in the Stuttgart Psalter which portray the sword-belt being worn.183 The sword is shown suspended from the belt by two converging straps, and the function of the trefoil mount, which is not visible, was almost certainly to attach these two straps to the sword belt. In one miniature the straps are attached to the scabbard by loops, but in the other case the characteristic half-oval mounts are depicted in the same position on the scabbard as in the spatharii illustrations. In addition, a miniature in the Saint Emmeram Gospels depicts the buckle and long strap-end which fastened the belt around the wearer's waist, thus accounting for the full set of fixtures as it is known from archeological finds.184 As many of the known fittings were made of gilt silver, a scabbard complete with sword-belt and mounts would have been of considerable value. Perhaps the expression "spata cum scogilo" in the Ripuarian law code therefore signified not only the sword and sheath, but also the sword-belt and fittings.

The introduction of the trefoil mount apparently dates from the early ninth century,185 and the fact that the Vivian Bible, Lothar Gospels, and San Paolo Bible all depict such mounts demonstrates once again that Carolingian illustrators reproduced at least some aspects of contemporary conditions in their work.

LANCES

The lance, lancea, was the cheapest weapon listed in the Lex Ribuaria, costing only two solidi together with a shield,186 though by 850 this price had risen to five solidi in places.187 The capitularies cited earlier recorded that lances were carried by the infantry and cavalry alike, and that a shield and lance were standard equipment on the royal war carts. Pictorial sources show the lance as the predominant weapon carried by both foot soldiers and horsemen, far more common than either the bow or the sword.188 The fact that the lances depicted in many of these illustrations are winged, a feature which only developed in the course of the eighth century,189 implies that the miniatures reflect contemporary conditions. This impression is reinforced by archeological evidence, since lances are also among the most common weapons in eighth-century graves, found far more often than arrows, and slightly more often than swords.190 It may therefore seem surprising that the Capitulary of Aachen decreed that soldiers should carry a bow, not a lance, instead of a cudgel.191 However, this measure may have been taken precisely because of a shortage of archers, and there is no evidence that the bow ever replaced the lance as the principal Frankish weapon.

Large numbers of lances were needed to equip the Carolingian army, and the desire to build up royal stocks is reflected by the specification of a lance and shield as the annual gift required from certain monasteries.192 These could in turn obtain the weapons from craftsmen living on their estates: for instance, the smith Ermenulf paid the abbey of Saint-Germain-des-Pres an annual rent of six lances for his half mansus.193

Contemporary texts suggest that the Franks primarily used their lances as thrusting weapons, rather than throwing them. Thus Regino recorded that the Franks were unused to hurling spears like the Bretons,194 while Ermold contrasted the heavy cuspis Franciscus with the light spears, hastae or missilia, thrown by the Bretons.195 Furthermore, the few references to Franks throwing spears almost all occur in descriptions of sieges, where close combat was excluded.196

The archeological evidence is apparently consistent with this view, in that the ango, or javelin, is rarely found after 800,197 while the lances found in eighth-century graves are large, heavy weapons which seem unsuited to being thrown.198 Nevertheless, that the Carolingians still used the javelin is evident from the will of Count Eccard of Macon, who bequeathed not only two sets of "scuto cum lancea," but also "dardo I," almost certainly a light throwing spear.199

In an influential article, Lynn White argued that the discovery of the stirrup greatly enhanced the value of the lance for the Frankish cavalry, who could now use it in "mounted shock combat."200 This entailed riding at the enemy with the lance held at rest beneath the arm, so that the weapon would strike home with the full force of the charging horse. White's theory was refuted by Ross, however, who demonstrated that the use of the couched lance is not attested by iconographic evidence before the eleventh century.201 By contrast, Carolingian illustrations only depicted horsemen throwing spears, or thrusting them downward or upward.202

Finally, certain archeologists claim to have discovered material evidence of the Carolingian custom of flying a banner from the lance shaft. Paulsen cited the remains of straps wrapped around the base of spear sockets,203 while Stein referred to the presence of what appeared to be lance shoes in graves which did not contain spearheads.204 However, the straps prove only that the lances were bound, not that banners were attached to them,205 and elsewhere Paulsen himself suggested that the binding on a lance shaft portrayed in the Corbie Psalter would have been added to prevent the wood splintering.206 As for Stein's `lance shoes', Steuer observed that they have variously been interpreted as parts of a bow or whip, or as arrowheads,207 and the absence of lance points in the graves also runs counter to the prevalent view that banners were flown from lances, not from plain staffs. There are therefore no definite archeological remains of Carolingian banners.

BOWS

Attention has already been drawn to the pronouncement in the Capitulary of Aachen that nobody in the army should carry a cudgel, but rather a bow.208 Another clause in the same capitulary listed the bow as a standard infantry weapon, and Charlemagne's letter to Fulrad implies that it was also regular equipment for the cavalry.209 Although it is unlikely that the bow ever replaced the lance as the basic Frankish weapon, written and pictorial sources alike indicate that both infantry and cavalry regularly used the bow in battle.210

Ganshof proposed that the discrepancy between the late eight-century Capitulare misrorum, which did not list the bow among cavalry weapons,211 and the letter to Fulrad of 806, which did, indicates that archery was only widely adopted by the Franks as a result of experience gained fighting the Avars and Slavs in the late eighth century.212 However, arrows were common in Merovingian graves of the sixth and seventh centuries,213 and it seems unlikely that the omission from the capitulary is significant. The passage referring to armament appears to have been a sort of means test ("qui . . . caballos, arma et scuto et lancea spata et senespasio habere possunt"), so that bow and arrows may have been omitted simply because they were among the cheapest of Frankish weapons: wood was readily available and manufacture presumably inexpensive. The fact that bows and arrows were not listed in the Lex Ribuaria may similarly reflect their low value rather than their rarity.

No remains of Carolingian bows have been discovered, but the one bow which has survived from the Merovingian period is D-shaped, made of wood, and some 2 m. long.214 In contrast to the Byzantines and Lombards, the Franks and Alamanni are not believed to have used the composite reflex bow at this time.215 However, some ninth-century pictorial sources seem to portray reflex bows,216 and Ermold was apparently describing a composite bow when he referred to an archer "cornea plectra tenens."217 Although it is impossible to be categorical, given the absence of archeological remains, this evidence does suggest that the Franks were familiar with the reflex bow in the ninth century.

D-shaped wooden bows were undoubtedly still common, however,218 and two contemporary texts refer to yew as being particularly suitable for the manufacture of both bows and arrows.219 A very small number of arrowheads have been found in excavations, all made of iron, some of them rhomboid, others barbed,220 and a miniature in the Stuttgart Psalter depicts both types of point, as well as feathered flights.221

Along with the bow, foot soldiers were also expected to carry a spare string and twelve arrows.222 Perhaps this represented the contents of one quiver, examples of which are depicted in several Carolingian illustrations. The Utrecht Psalter in particular shows that the quiver was equipped with a domed hood to protect the arrowheads, attached by a strap,223 and that another strap enabled the archer to sling the quiver across his back.224 Presumably Carolingian quivers were made of wood and leather, but no remains have survived.

SUMMARY

The present study has emphasized the distinction between the armament carried by the mass of poor Frankish infantry and that owned by the wealthier horseman. Foot soldiers were armed with a round wooden shield and a spear, which was normally used as a thrusting weapon, and perhaps carried a bow and arrows as well. Cavalrymen were similarly equipped with all these weapons, but could also afford to own a sword, which represented their standard weapon, and in some instances helmets and body armor, too.

Carolingian helmets, shields, and spears were all recognizably different from earlier forms, but it was only in the field of sword production that significant technological progress was made at this time. During the reign of Charlemagne, Frankish smiths discovered how to improve not only the strength of the blade but also its maneuverability, with the consequence that the short sword, or sax, became redundant. A less welcome consequence for the Franks was that the Carolingian spata also became highly prized by hostile neighbors such as the Saracens and Danes.

The most important conclusion of the study, however, is that Carolingian ivories and manuscript illuminations are a more reliable guide to contemporary armament than has hitherto been believed. It has been demonstrated that even though certain features may have been influenced by late Roman or Byzantine pictorial traditions, ninth-century Frankish illustrations depicted current forms of helmets, shields, swords, sword-mounts and spears. It is therefore probable that Carolingian artists also reproduced other contemporary items in their work, and that where similarities with late Roman pictorial models are apparent, this could merely reflect the continuity between the antique and medieval worlds.

End Notes

I am extremely grateful to Rosamond McKitterick, Betsy Sears and Julia Smith for their comments on an earlier draft of this article.

l. E. A. Gessler, Die Trutzwaffen der Karolingerzeit vom 8. bis zum 11. Jahrhundert (Basel 1908).

2. F. Lot, L'art militaire et les armies au moyen age en Europe et dans le proche Orient, 2 vols. (Paris 1946) 1.103-104.

3. St. Gall, Stiftsbibliothek MS 22.

4. F. L. Ganshof, "A propos de la cavalerie dans les armies de Charlemagne," Comptes-rendus de l'Academie des inscriptions et belles-lettres (1952) 531-536, esp. 532-535; idem, "Charlemagne's Army," in Frankish Institutions under Charlemagne (New York 1970) 57-68, esp. 65-66.

5. Leiden, Bibliotheek der Rijksuniversiteit MS Perizoni 17.

6. J. F. Verbruggen, "L'armee et la strategie de Charlemagne," in W. Braunfels, ed., Karl der Grosse, Lebenswerk and Nachleben, 4 vols. (Diisseldorf 1965) 1: Personlichkeit and Geschichte, 420-436, esp. 423-425.

7. M. Last, "Die Bewaffnung der Karolingerzeit," Nachrichten aus Niedersachsens Urgeschichte 41 (1972) 77-90.

8. D. Nicolle, The Age of Charlemagne (London 1984) 9-13; see also 34-39, the notes to the plates.

9. Ibid. 35, pl. B2.

10. For instance, Ganshof, "Charlemagne's Army" (n. 4 above) 65-67; Last (n. 7 above) 78-80; Verbruggen (n. 6 above) 423-426.

11. On the dating, see F. L. Ganshof, Was waren die Kapitularien? (Damstadt 1961) 164.

12. Capitulare missorum c. 4, MGH Capitularia regum Francorum 1.67 (hereafter MGH Cap. 1).

13. Ganshof (n. 11 above) 167.

14. Karoli ad Fulradum abbatum epistola, MGH Cap. 1.168.

15. Ganshof (n. 11 above) 167.

16. Capitulare Aquisgranense c. 9, MGH Cap. 1. 171.

17. Capitulare de villis c. 64, Ibid. 89.

18. Capitulare Aquisgranense c. 17, Ibid. 171.

19. Notkeri Gesta Karoli Magni 2.17: H. F. Haefele, ed., Notker der Stammler: Taten Kaiser Karls des Grossen, MGH Scriptores rerum Germanicarum n.s. 12 (Berlin 1959; rev. 1962) 83-84.

20. H.Lowe, "Das Karlsbuch Notkers von St. Gallen and sein zeitgeschichtlicher Hintergrund," Schweizerische Zeitschrift fur Geschichte 20 (1970) 269-302 at 279.

21. The references are assembled in n. 120 below.

22. Paris, Bibliotheque nationale MS lat. 1; the helmet type is discussed at length below.

23. E L. Ganshof, "De uitrusting van de lijfwachters der Karolingische koningen en keizers," Gentse bijdragen tot de kunstgeschiedenis 12 (1949-1950) 122-126; J. Adhemar, influences antiques dans Part du moyen age franfais, Studies of the Warburg Institute 7 (London 1939) 153 n. 4; Lot (n. 2 above) 1.104 n. 3; F. Miitherich and J. E. Gaehde, Carolingian Painting (London 1977) 79.

24. See in this context the comments of S. Dufrenne and C. Villain-Gandossi, "Bateaux figures dans des oeuvres carolingiennes," Archaeonautica 4 (1984) 243-260, esp. 243.

25. Stuttgart, Wurttemburgische Landesbibliothek, Biblia fol. 23: Der Stuttgarter Bilderpsalter, 2 vols. (Stuttgart 1965-1968) 1: Faksimile, fol. 3v.

26. Moscow, Historical Museum MS 129, fol. 3: Stuttgarter Bilderpsalter 2 fig. 32.

27. E Miitherich, "Die Stellung der Bilder in der fruhmittelalterlichen Psalterillustration," in Stuttgarter Bilderpralter 2.151-222 at 164, 167.

28. Winged lances and the standard Carolingian helmet type are both considered at greater length below.

29. J. Werner, "Sporn von Bacharach and Seeheimer Schmuckstuck," in K.-H. Otto and J. Herrmann, eds., Siedlung, Burg and Stadt, Studien zu ihren Anfdngen: Festschrift P. Grimm, Deutsche Akademie der Wissenschaften zu Berlin, Schriften der Sektion fur Vor- and Fruhgeschichte 25 (Berlin 1969) 497‑506 at 499.

30. Notker 2.17 (n. 19 above) 83.

31. Ermoldi Nigelli Carmen in honorem Hludowici, line 1710: E. Faral, ed., Ermold le Noir: Poeme .rur Louis le Pieux et epitres au roi Pepin, Les classiques de 1'histoire de France au moyen age 14 (Paris 1964 (1932) 130.

32. P. Schramm and F. Mutherich, Denkmale der deutschen Konige and Kaiser (Munich 1962) 93-94.

33. Lex Ribuaria 40.11: MGH Leges nationum Germanicarum 3.2.94.

34. Capitulare Aquisgranense c. 9: MGH Cap. 1. 171.

35. D. Hejdova, "Des sogenannte St-Wenzels-Helm" 3, Waffen‑ und Kostumkunde (1968) 15-30 at 15-16.

36. H. Schneider, "Des Helm von Niederrealta: Ein neuer mittelalterlicher Helmfund in der Schweiz," Waffen- and Kostumkunde (1967) 77‑90.

37. D. Hejdova, "Der sogenannte St-Wenzels-Helm" 1, Waffen‑ and Kostumkunde (1966) 95-110 at 103-106; idem (n. 35 above) 23.

38. D. M. Wilson, The Bayeux Tapestry (London 1985) 233.

39. Bern, Burgerbibliothek MS 264, fol. 31: R. Stettiner, Die illustrierten Prudentius-Handschriften, 2 vols. (Berlin 1895-1905) 2 pl. 129; for the dating, see E. J. Beer, "Uberlegungen zu Stil and Herkunft des Berner Prudentius-Codex 264," in O. P. Clavadetscher, H. Maurer, and S. Sonderegger, eds., Florilegium Sangallense: Festschrift furjohannes Duft zum 65. Geburtstag (St. Gall 1980) 15-70 at 20-22.

40. Beer 49.

41. E.g., London, British Library MS Add. 40731 (Bristol Psalter), fol. 231v: S. Dufrenne, L'illustration des psautiers grecs du moyen age, Bibliotheque des cahiers archeologiques 1 (Paris 1966), pl. 59; Paris, Bibliotheque Nationale MS grec 139, fol. 4v: A. Cutler, The Aristocratic Psalters in Byzantium, Bibliotheque des cahiers archeologiques 13 (Paris 1984), fig. 248.

42. J. Werner, "Zur Herkunft der fruhmittelalterlichen Spangenhelme," Praehistorische Zeitschrift 34-35 (1949-1950) 178-193.

43. Amiens, Bibliotheque municipale MS 18, fol. 123v: A. Boinet, La miniature carolingienne (Paris 1913), pl. CXLVIII.

44. See for instance V H. Elbern, ed., Das erste jahrtausend: Kultur and Kunst im werdenden Abendland an Rhein and Ruhr, 3 vols. (Diisseldorf 1962), Tafelband, pl. 149; R. Pirling, "Ein frankisches Fiirstengrab aus Krefeld‑Gellep," Germania 42 (1964) 188-216 and pl. 45.

45. P. Paulsen, "Fliigellanzen: Zum archaologischen Horizont der Wiener Sancta lancea," Fruhmittelalterliche Studien 3 (1969) 289-312 at 299.

46. Notably on pp. 140-141: C. Eggenberger, Psalterium aureum sancti Galli: Mittelalterliche Psalterillustration im Kloster St. Gallen (Sigmaringen 1987), pis. 13-14.

47. Even so, neither Notker's reference to "ferreus Karolus, ferrea galea cristatus" (n. 30 above) 83, nor the "helmo condericto" in the Lex Ribuaria (n. 33 above) signified a crested helmet. In the former case, the adjective "cristatus" agrees with "Karolus," not "galea," and evidently meant "crowned" or "topped." With regard to the latter instance, condirigere is attested elsewhere with the meaning "to use, to make use of": C. D. du Cange, Glossarium mediae et infimae Latinitatis, rev. L. Favre, 10 vols. (Niort 1883-1887) 2.487. This fact, coupled with the parallel descriptions of "equum videntem et sanum . . . brunia bona . . . bagnbergas bonas," suggests that "conderictus" should likewise be interpreted as "useful" or "serviceable."

48. See for instance Stuttgarter Bilderpsalter (n. 25 above) 1 fols. 71 v, 158 v.

49. Rome, Church of San Paolo fuori le Mura, Bible, e.g., fols. 83v, 243v: J. E. Gaehde, "The Pictorial Sources of the Illustrations to the Books of Kings, Proverbs, Judith and Maccabees in the Carolingian Bible of San Paolo Fuori Le Mura in Rome," Fruhmittelalterliche Studien 9 (1975) 359389, pis. 75, 100.

50. E.g., the Vivian Bible, fols. 215v and 423, or Paris, Bibliotheque nationale MS lat. 266 (Lothar Gospels), fol. lv: J. Hubert, J. Porcher, and W. F. Volbach, L'empire carolingien (Paris 1968), figs. 128, 129, 133. Although these helmets sometimes appear rectangular, the fact that a direct line runs from the forehead to the nearside angle of the neckguard whether the helmet is viewed from left or right indicates that the helm is the same shape as that in the Psalterium aureum.

51. Utrecht, Bibliotheek der Rijksuniversiteit MS eccl. 484; crested helmets can be seen on, e.g., fol. 13v, others without a crest on fol. 30v: E. T. de Wald, The Illustrations of the Utrecht Psalter (Princeton 1932), pis. XXI, XLIX.

52. For instance fol. 35v: Stettiner (n. 39 above) 2 pl. 137; see also O. Homburger, Die illustrierten Handschriften der Burgerbtbliothek Bern: Die vorkarolingischen and karolingischen Handschriften (Bern 1962) 143.

53. A. Goldschmidt, Die Elfenbeinskulpturen aus der Zeit der karolingischen and sdchsischen Kaiser, 4 vols. (Berlin 1914-1926) 1.67 and pl. LVIII 

54. Hubert et al. (n. 50 above), fig. 202.

55. Compare for example Vatican, MS Vat. lat. 3225 (Vatican Virgil), picturae 48‑49: J. de Wit, Die Miniaturen des Vergilius Vaticanus (Amsterdam 1959), pis. 35, 38; Vatican, MS Vat. lat. 3867 (Roman Virgil), fol. 74v: Picturae ornamenta complura scripturae specimina codicis Vaticani 3867, Codices e Vaticanis selecti 2 (Rome 1902), pl. X; W. F. Volbach, Elfenbeinarbeiten der Spdtantike and des friihen Mittelalters, Romisch-Germanisches Zentralmuseum zu Mainz, Katalog 7 (Mainz 1952), no. 151, pl. 51 and p. 74; or the Byzantine illustrations cited in n. 41 above.

56. C. H. Kraeling, The Synagogue, Excavations at Dura-Europos 8.1 (New Haven 1956) 196-197 and pl. LXIL

57. G. Beccati, La colonna coclide istoriata: Problemi storici iconografui stilistici (Rome 1960), pls. 7779.

58. S. Dufrenne, Les illustrations du Psautier d'Utrecht: Sources et apport carolingien (Paris 1978) 160 and n. 592.

59. R. Bruce‑Mitford, The Sutton Hoo Ship-Burial, 3 vols. (London 1975‑1983) 2: Arms, Armour and Regalia, fig. 168, particularly (b), Intercisa 2.

60. Schramm and Mutherich (n. 32 above) 94.

61. See du Cange (n. 47 above) 4.160 for examples.

62. M. Prou and A. Vidier (eds.), Recued des charter de Pabbaye de St-Benoit-.rur-Loire, 2 vols. (Paris 1900‑1937) 1.66.

63. C. Oman, A History of the Art of War in the Middle Ages, ed. 2, 2 vols. (London 1924) 1.126.

64. Pirling (n. 44 above) 200.

65. D. Tweddle, The Coppergate Helmet (York 1984) 19.

66. Stuttgarter Bilderpsalter (n. 25 above) 1, fol. 5v.

67. Lex Ribuaria 40.11 (n. 33 above) 3.2.94. The combination of shield and spear as a Waffenpaar is discussed at length by D. Hopper-Drage, Schild and Speer: Waffen and ihre Bezeichnungen im friihen Mittelalter, Germanistische Arbeiten zu Sprache and Kulturgeschichte 3 (Frankfurt 1983), esp. 182200.

68. L. A. J. W. Sloet, ed., Oorkondenboek der grafschappen Gelre en Zutfen, tot op den slag van Woeringen, 5 juni 1288 (The Hague 1872-1876) 43.

69. Karoli ad Fulradum abbatum epistola: MGH Cap. 1.168.

70. Annales Fuldenses (Regensburg continuation) 894: F. Kurze, ed., MGH Scriptores return Germanicarum in usum scholarum (Hanover 1891).

71. Capitulare de villis c. 45: MGH Cap. 1.87.

72. M. Quantin, ed., Cartulaire general de I'Yonne (Auxerre 1854) 42 (St. Remigius, Vareilles); H. Doniol, ed., Cartulaire de Brioude (Clermont 1863) 349 (Brioude); Ludowici Germanici diplomata 70: MGH Diplomata regum Germaniae ex stirpe Karolinorum 1.100 (St. Gall).

73.W. Horn and E. Born, The Plan of St. Gall, 3 vols. (Berkeley 1979) 2.190; J. Semmler, ed., "Consuetudines Corbeienses," in Corpus consuetudinum monasticarum (Siegburg 1963- ) 1.355-422 at 367.

74Annales Bertiniani 876: F. Grat, J. Vielliard, and S. C1emencet, eds, Annales de Saint-Bertin (Paris 1964) 209.

75. For cavalry and infantry together, see, e.g., San Paolo Bible, fol. 59v: C. R. Dodwell, Painting in Europe, 800-1200 (Harmondsworth 1971), pl. 43; or Psalterium aureum, p. 141: Eggenberger (n. 46 above), pl. 14

76. In addition to the references cited in the previous note, see also, e.g., Utrecht Psalter, fol. 13v: de Wald (n. 51 above), pl. XXI; Stuttgarter Bilderpsalter (n. 25 above) 1, fol. 150v; or Goldschmidt (n. 53 above) 1, pl. V.

77. See for instance Vatican Virgil, pictura 49: de Wit (n. 55 above), pl. 27; Roman Virgil, fol. 188v: Vat. 3867 (n. 55 above), pl. XVII; Mount Athos, Pantocrator 61, fol. 30 v: Dufrenne (n. 41 above), pl. 4; or Paris, BN MS grec 139, fol. 4v: Cutler (n. 41 above), fig. 248.

78. E.g., Dufrenne (n. 58 above) 159‑160; Mutherich and Gaehde (n. 23 above) 55; Adhemar (n. 23 above) 204.

79. Besides the illustrations in the San Paolo Bible, Psalterium aureum, and Utrecht and Stuttgart Psalters cited above, see also, e.g., Lothar Gospels, fol. Iv: Hubert et al. (n. 50 above), fig. 133; Bern P.rychomachia, fol. 39: Stettiner (n. 39 above) 2, pl. 143; or Goldschmidt (n. 53 above) 1, pl. V

80. References to the San Paolo Bible, Psalterium aureum, Stuttgart Psalter, Lothar Gospels, and Bern P.rychomachia as above; also Goldschmidt, 1, pis. V, LXXII.

81. G. Henderson, Early Medieval (Harmondsworth 1972; repr. rev. 1977) 159.

82. S. Lindqvist, Gotlands Bildsteine, 2 vols. (Stockholm 1941-1942) 1.75.

83. E Stein, Adelsgrdher des achten jahrhunderts in Deutschland, Germanische Denkmdler der V61kerwanderungszeit set. A 9, 2 vols. (Berlin 1967), Textband 75-76 (Galgenberg type); Dufrenne is wrong on this point (n. 58 above) 160.

84. Stein, Textband 19, 77.

85. Notker 2.17 (n. 19 above) 84.

86. J. R. Rahn, Das Pralterium aureum von Sanct Gallen: Ein Beitrag zur Geschichte der karolingischen Miniaturmalerei (St. Gall 1878) 43.

87. Lothar Gospels, fol. lv, San Paolo Bible, fol. 50v: Mutherich and Gaehde (n. 23 above), pls. 25, 43; Goldschmidt (n. 53 above) 1, pl. VI.

88. G. Arwidsson, Valsgarde 8, Die Graberfunde von Valsgarde 2 (Uppsala 1954), pls. 10, 11.

89. For instance, San Paolo Bible, fol. 50v, as in n. 87 above; Stuttgarter Bilderpsalter (n. 25 above) 1, fol. 138; Goldschmidt (n. 53 above) 1, pl. LXXIL

90. Stein (n. 83 above), Textband 19, 77.

91. Such a strap can be seen on fol. 50v of the San Paolo Bible (n. 87 above).

92. Abbonis Bella Farisiacae urbis 2, line 500; H. Waquet, ed., Abbon: Le siege de Paris par les Normands, Les classiques de 1'histoire de France au moyen age 20 (Paris 1942; reissued 1964) 104.

93. Ermold, lines 2016-2017 (n. 31 above) 154.

94. Psalterium aureum, p. 141 (n. 46 above), pl. 14.

95. Ermold line 1701 (n. 31 above) 128.

96. K. Tackenberg, "Uber die Schutzwaffen der Karolingerzeit and ihre Wiedergabe in Handschriften and Elfenbeinschnitzereien," Fruhmittelalterliche Studien 3 (1969) 277-288 at 281.

97. Capitulare missorum in Theodonir villa datum secundum, generale c. 6: MGH Cap. 1.123.

98. Capitulare Aquisgranense c. 9: Ibid. 171.

99. The relative prices are listed in the Lex Ribuaria (see n. 67 above).

100. Prou and Vidier (n. 62 above) 1.66; Schramm and Mutherich (n. 32 above) 93‑94.

101.Ganshof, "Charlemagne's Army" (n. 4 above) 66, citing the capitulary article referred to in n. 98 above.

102. Capitulare Bononiense c. 10: MGH Cap. 1. 167.

103. Capitulare missorum in Theodonis villa datum secundum, generale c. 7: ibid. 123; Capitulare Bononiense c. 10: ibid. 167; Edictum Pistense c. 25: MGH Cap. 2.321.

104. Ermold lines 574, 1710‑1711 (n. 31 above) 46, 130.

105. Edictum Pistense c. 25: MGH Cap. 2.321.

106. Lorica: see, e.g., Reginonis abbatis Prumiensis Chronicon 867: F. Kurze, ed., MGH Scriptores rerum Germanicarum in usum scholarum 92; Annales Fuldenses 876: (n. 70 above) 89; Capitulare Aquisgranense c. 9: MGH Cap. 1.171; brunia: see, e.g., Lex Ribuaria 40.11 (n. 33 above) 3.2.94; Capitulare Bononiense c. 10: MGH cap. 1.167; wills of Eberhard of Friuli and Eccard of Macon: as n. 100 above; thorax: Notker 2.17 (n. 19 above) 83. The fact that brunia, a Germanic loan word, occurs exclusively in legal texts, whereas lorica, a classical term, is found in narrative sources as well, perhaps reflects the difference between the consciously classical language of the clerical narrative sources and contemporary legal terminology, which was based more upon the spoken word.

107. Notker 2.17, p. 84.

108. That which lacks straps, for it is solely interwoven from iron rings "Hrabanus Maurus, De universo 20.13: PL 111.543; cited by P. Contamine, La guerre au moyen age (Paris 1980) 324.

109. Isidore, Etymologiae 18.13: W. M. Lindsay, ed., Isidori Hispalensis episcopi etymologiarum sive originum libri XX, 2 vols. (Oxford 1911), vol. 2; the work has no page numbers.

110. See n. 108 above.

111. Though Gregory of Tours wrote of a lance being "repulsa a circulis luricae"‑Historiae Francorum 7.38: MGH Scriptores return Merovingicarum 1.1.319.

112. The only instance of which I am aware is the Bern Psychomachia, e.g., fols. 31, 41: Stettiner 2 (n. 39 above), pis. 129, 147.

113. For example, fols. 13v, 64v, 76v: de Wald (n. 51 above) pis. XXI, CI, CXVIII.

114. Hubert et al. (n. 50 above), fig. 202.

115. For instance, Vivian Bible, fol. 423: Ibid. pl. 129; compare San Paolo Bible, fol. 170v: Boinet (n. 43 above), pl. CXXIV.

116. This seems to be what Isidore 18.13 (n. 109 above) described as a squaraa, "lorica ferrea ex lamminis ferreis aut aeneis concatenata in modum squamae piscis."

117. Stuttgarter Bilderpsalter 1 (n. 25 above), e.g., fols. 5v, 12v, 150v.

118. Notably on pp. 140-141: Eggenberger (n. 46 above), pis. 13-14.

119. Goldschmidt 1 (n. 53 above), pl. V.

120. See for instance Ganshof, "Charlemagne's Army" (n. 4 above) 158 n. 47; Last (n. 7 above) 8184; Miitherich and Gaehde (n. 23 above) 124; W. Braunfels, Die Welt der Karolinger and ihre Kunst (Munich 1968) 21; E. H. Kantorowicz, "Gods in Uniform," Selected Studies (New York 1965) 7-24 at 19; A. Merton, Die Buchmalerei in St Gallen vom neunten his zum elften jahrhundert, ed. 2 (Leipzig 1923) 58.

121. F. Mutherich, "Die verschiedenen Bedeutungsschichten in der fruhmittelalterlichen Psalterillustration," Fruhmittelalterliche Studien 6 (1972) 232-244 at 243; Pantocrator 61, fols 30v, 89v: Dufrenne (n. 41 above), pis. 4, 12.

122. For instance Vatican Virgil, picturae 48, 49: de Wit (n. 55 above), pis. 35, 38.

123. Stuttgarter Bilderpsalter 1 (n. 25 above), fols. 70v, 150v; compare Pantocrator 61, fol. 89: Dufrenne (n. 41 above), pl. 12.

124. Ganshof (n. 23 above) 122-126; Lot 1 (n. 2 above) 104 n. 3; Miitherich and Gaehde (n. 23 above) 21, 79; Dufrenne (n. 58 above) 161; Adhemar (n. 23 above) 153 n. 4.

125. Volbach (n. 55 above) no. 1, pl. 1.

126. A. Venturi, Storia dell'arte italiana, 11 vols. (Milan 1901-1939) 1, fig. 436.

127. K. Weitzmann, ed., Age of Spirituality: Late Antique and Early Christian Art, Third to Seventh Century (New York 1979) 481.

128. J. E. Gaehde, "Carolingian Interpretations of an Early Christian Picture Cycle to the Octateuch in the Bible of San Paolo Fuori Le Mura in Rome," Fruhmittelalterliche Studien 8 (1974) 351-384 at 354, and idem (n. 49 above) 364 n. 24.

129. W Kohler, Die karolingirchen Miniaturen 1: Die Schule von Tours, 3 vols. (Berlin 1930-1933),2.229; see also 226.

130. Schramm and Mutherich (n. 32 above) 93-94

131. Lex Ribuaria 40.11 (n. 33 above) 3.2.95.   

132. Notker 2.17 (n. 19 above) 84.

133. Stuttgarter Bilderpsalter 1 (n. 25 above) fol. 12v.

134. Goldschmidt 1 (n. 53 above), pl. V

135. Prou and Vidier (n. 62 above) 1.66; Schramm and Mutherich (n. 32 above) 93-94.

136. Capitulare misrorum c. 7: MGH Cap. 1.115. For some reason Ganshof interpreted bauga as metal leg guards: "Charlemagne's Army" (n. 4 above) 65-66 n. 49; but see du Cange (n. 47 above) 1.609.

137. The term is clearly similar to brachiale, brachile, both of which are attested with the meaning "armring, armguard": du Cange 1.729-730.

138. Goldschmidt 1 (n. 53 above), pl. VI.

139. Lex Ribuaria 40.11 (n. 33 above) 3.2.94.

140. Gesta ranctorum Rotonensium 1.6: new edition by C. Brett, Ph.D. diss. (University of Cambridge 1986); on the dating, see F. Lot, Melanges d'histoire bretonne (Vle-Xle siecle) (Paris 1907) 10.

141. Capitulare mitsorum c. 4: MGH Cap. 1.67; Karoli ad Fulradum abbatum epistola: ibid. 168.

142. Reginonis Chronicon 860 [recte 851] (n. 106 above) 79.

143. Stein (n. 83 above), Textband 13.

144. J.Werner, "Bewaffnung and Waffenbeigabe in der Merowingerzeit," in Ordinamenti militari in occidente nell'alto medioevo, Settimane di studio del centro italiano di studi sull'alto medioevo 15, 2 vols. (Spoleto 1968) 1.95