The Development of Southern French and Catalan Society, 718-1050
Archibald R. Lewis
The Military System of the Midi and Catalonia
 In the last two chapters we have seen how, during the first three quarters of the tenth century, principalities based on family rule tended to disintegrate in regions south of Poitou and Burgundy and how, when this happened, the public court system of the Carolingians tended to disappear almost everywhere and to be replaced by an informal boni homines system of justice. But what of the military aspects of government? What happened to the military system and in what form did it emerge by the year 975?
Obviously we must examine this military system in terms of the demands placed upon it by a need for defense against outside attack and against internal disorders. Let us first consider defense against outside attack. Earlier, in the ninth century, as we have noted, there were three regions which were particularly subject to invasion. They were Western Gascony and Aquitaine where Viking pirates concentrated their activities, the Spanish March close to the hostile Moors of Spain, and Provence and Alpine regions east of the Rhone where Moslem freebooters and other brigands seem to have operated with relative freedom. By the first years of the tenth century the problem presented by the Vikings ceased to be a serious one in Gascony and Western Aquitaine. As far as we can tell, after 900 these raiders left Gascony alone(1) and by 906 Bordeaux had recovered sufficiently to be able to send trading ships to Christian Asturias.(2) As for Aquitaine, about 930 King Raoul and Count Ebles  Manzur of Poitou defeated the last Viking bands who were operating in this part of France.(3) Hereafter, except for occasional coastal raids which continued down into the eleventh century, Aquitaine ceased to be disturbed by Viking activity.(4)
The story of Moslem Spanish aggression is a somewhat similar one. Early in the tenth century, still reacting to Count Guifred's activity in advancing his frontiers to the south, Moslem lords of the Ebro Valley continued to carry on warfare with the Christian Spanish March. Llop ibn Mohammed, who had slain Count Guifred in battle about 897 or 898,(5) continued hostilities after the latter's death. In 904, according to our sources, he attacked Pallars in force, capturing some seven hundred prisoners, among them Isarn, grandson and heir of Count Ramón of Pallars-Ribagorça.(6) Four years later his contemporary, Al-Tamil, lord of Huesca, invaded Ribagorça and occupied Rhoda.(7) He followed this up, the next year, with a raid on Catalonia in which the frontier castles of Alguaire, Gualter, and Oliola were captured and in which some three hundred prisoners and much booty were taken.(8) In 912, after assaults on Navarre and Aragón, he turned again to Catalonia and launched another campaign against this region, in the course of which Count Sunyer of Barcelona was routed near Tarragona.(9) Two years later he was less fortunate and met his death during a similar razzia into Catalan territory.(10)
After some two decades of constant war, the death of Al-Tamil in 914 brought relative peace to the Spanish March. This was due, however, less to Catalan strength than to a new political situation which had developed in Moslem Spain, one caused by the rise to power of Abd-ar-Rahman III. Abd-ar-Rahman III about 918 began an effort, eventually successful, to unite all Moslem Spain under his rule, which meant, among other things, ending the de facto independence of Moslem marcher lords in the Ebro  Valley. Facing such a Cordovan threat, which they viewed with as much concern as the Christian kings of Léon, Castile, and Navarre did, the rulers of Saragossa, Huesca, and Lérida felt it wiser to keep peace along the frontiers of Pallars, Ribagorça, and Catalonia.(11) And when, about 940, these Moslem marcher lords had been absorbed into Abd-ar-Rahman's centralized caliphate, by hastening to accept Cordova's suzerainty, the counts of Barcelona were able to continue this peaceful co-existence until almost the end of the century.(12) After 914, then, peace rather than war generally prevailed along the frontiers of the Spanish March.
Our third region, Provence and Alpine areas east of the Rhone, presents a somewhat different picture. In this part of the Midi the great Hungarian raid of 938 did some damage, particularly in the Lyonnais.(13) The principal problem, however, and a continuing one, was caused by the activities of those Moslem freebooters who had established themselves at Fraxinetum and were raiding the interior. Though it is difficult to estimate the extent of the damage they caused, we do know that their raids forced the canons of Saint-Victor of Marseille to abandon this city about 923 and to transfer this establishment to land protected by the walls of the castle of Fos.(14) We also learn from later records of certain damage which the raiders did in the region of Aix.(15)
Finally as the rule of Louis the Blind and his viceroy, Hugh of Arles, proved to be increasingly ineffective, they moved north into Alpine regions, where, aided by local brigands and perhaps even some local nobles, they became a menace to traffic crossing the clusae into Italy. In 972 they were bold enough to capture the celebrated St. Mayeuil, abbot of Cluny, on his way back to his abbey after a visit to Rome.(16) It was this action which probably at last forced action against them. Though King Conrad the Peaceful appears to have done nothing, his principal vassals were spurred into action. Bishop Isarn of Grenoble began to build castles and  to restore order to Dauphiny.(17) Even more important, William, count of Arles, rallied the local nobility of Provence and in 972 led them to a victory over the Moslems of Fraxinetum. He not only expelled the raiders from this pirate base, but also, by encouraging the building of castles in remote regions of Provence, began to bring order at last to Alpine regions which had not known it since the time of Louis the Pious.(18) After more than a century in a disorganized state, Alpine regions east of the Rhone at last got something like effective government.
Let us turn from this examination of external enemies to the problem of internal disorder. Our sources contain less information than we would like to have concerning this aspect of the Midi's development. For many regions we have no information at all. For others, however, charters of the tenth century, like those of the ninth, by telling us of land acquired by "conquests," that is by means other than inheritance, give us some indication of local conditions. For the Limousin and Quercy, for instance, such charters date from the period 923 to 945, with mention of land which was called "ex conquestu" or "per conquestem" in documents which date from the years 923, 926, 927, 930, 932, 937, and 945.(19) Ademar of Chabannes gives us some additional information. He tells us how Bernard, count of Périgord, killed Viscount Lambert of Marillac and his brother Bernard soon after 920, and how his uncle Count William Taillefer of Angoulême restored this honorem a little later to their brother, Viscount Odalric.(20) He also informs us of how, late in this period, about 973-974, Boson the Old, count of La Marche, tore down the castle of Brosse, which Duke William of Poitou had erected in the former's domains.(21)
When we examine charters from Auvergne and Rouergue we get much more evidence of instability than we find in the Limousin and Western Aquitaine. Some thirty charters from Conques dating from the period 901-974 mention land whose owners acquired it "ex conquestu." Five of these date from the period before 925, twelve from the period 925-950, and fifteen from the years 951-974.(22) Evidence from Auvergne  seems to be even more complete. Some seventy-six charters from this region mention "ex conquestu" land -- most in the central and southern portions of this province. Thirty-one date from the period down to 925, twenty-eight from the years 926 to 950, and eighteen from the years 951 to 975.(23) In contrast to this our sources from the nearby regions of the Albigeois and Velay give us little information about such instability of land ownership.
The basic political situation which helped create this instability of landholding in the Limousin and Quercy for some two decades and which caused what seems to have been a state of endemic disorder in Rouergue and Auvergne is known to us. It was the conflict which took place over these regions between the House of Toulouse and the Houses of Bernard of Auvergne and of Ebles Manzur of Poitou. This long conflict helped to create factions and what might be termed chronic civil war in these parts of the Midi. By 950 such disorders appear to have ceased in the Limousin, but they continued unchecked in Auvergne and Rouergue because no political authority strong enough to stop them existed.(24) Nor was any to appear here until the twelfth century.
When we consider the Valley of the Rhone proper our charters also give us some evidence of similar disorders. We find certain references to "ex conquestu" land in charters dating from 908 and 923(25) which concern Dauphiny; in ones from 937, 947-948, and 956-957(26) which deal with the Vienne region; and in ones from 968 and 970(27) from the Lyonnais. Such land ownership changes in the Vienne region may well have  been associated with a change in dynasties after the death of Louis the Blind.(28) Those found in Dauphiny in part seem related to the rise of a local family, that of Sylvius, to a position of authority.(29) Those in the Lyonnais were probably the result of Artald's assumption of the power and authority of count about the year 970.(30) In contrast to Auvergne and Rouergue, though, our sources seem to show that the Valley of the Rhone suffered only occasional disorders in specific areas for short periods of time -- except in Alpine areas and Provence, of course.
Such then is an analysis of what we can discover concerning internal and external disorders in Southern France and the Spanish March during the period 900 to 975, in the light of which we must consider military aspects of government. But though such an analysis helps to explain why some form of military organization was necessary in most of the Midi and Catalonia, it tells us nothing more. To get some answers to the question of what form this military organization took we must turn to what we can discover about castles and their use during the period. Before we do so, however, we must remember that our sources are incomplete. All we can hope for, as in previous periods, is some indications of military organization. We must also, generally, take it for granted that most of the fortresses and fortifications which were to be found in these regions in late Carolingian times continued to be used during this period, unless we have specific evidence to the contrary.
With this in mind let us first consider Auvergne and Rouergue where, as has been pointed out, disorders were almost constant during these decades. Curiously enough, though, our sources reveal fewer fortresses in this part of Central France than we might expect. We do nevertheless find some. One is a castellum near Sauxillanges mentioned in a charter which dates from about 928.(31) Still another is the important fortress of Castelluc, referred to several times in our sources.(32) Others are a place called Casterole, found in a document of 945(33) and 958, and that rocca  which, we are told, was held in 961 by two Auvergnat nobles, Genesius and Aimery.(34) Late in our period we also find references to a castrum called Yconensium near Monte Asenario(35) and another called Brézac or Bresantium located near the abbey of Aurilac.(36) For Rouergue, charters of the region speak of a castle of Salmegro in 937(37) and another called a turre or castrum not far away in 957.(38)
We have more evidence of castles and other fortifications in use during this period in Quercy, the Limousin, and other regions of Western Aquitaine. Ademar of Chabannes tells us that the walls of Angoulême which we know were in existence in 918(39) were rebuilt by Count Aldouin sometime before 916,(40) and the Cartulary of Uzerche mentions the work of refortifying Limoges by Bishop Turpio in the same period.(41) In 922 and 930(42) our sources mention the castrum of Fouilloux as well as one called Bellum located near Cahors.(43) We learn of three more located near the abbey of Tulle during this period, Echelles or Scalas,(44) one at Stanquillières(45) and one called Moncieux,(46) as well as that of Uxelladuno which King Raoul gave to this monastery in 935.(47) In 941, 944, 947, and 975 we learn of a rocca owned by a certain Guitard and his family.(48) And we hear of a place called Castres given to the abbey of Beaulieu in 943.(49) From the pen of Ademar of Chabannes we learn how Bishop Ebles of Limoges, about 944, built the castle of Saint-Hilaire near this city and surrounded the abbey of Saint-Martial with fortifications.(50) West of the  Limousin in Saintonge we learn of the castle of Ostend in 956,(51) south of the newly erected castrum of Melle.(52)
When we look further south toward an area in Rouergue, Quercy, and the Toulousain which was definitely controlled directly by the House of Toulouse, we find references to a large number of such fortifications in the will of Raymond I of Rouergue, and a little later in the period in those of Countess Garsinde and Hugh, bishop of Toulouse. Raymond's will mentions a rocheta near Albi, five castles near Moissac, and some twelve others scattered through his domains which he is leaving to his son, Raymond II, and his other heirs.(53) Garsinde's will of 972 refers to a castle near Saint-Pons of Thomières and one near Castres, as well as one in Lodève owned by Bishop Frotaire.(54) Bishop Hugh of Toulouse in his will mentions the castle of Durimane belonging to Bishop Frotaire and that of Sexago in Razes held by the count and countess of Carcassonne.(55)
Charters from Septimania give us still more information concerning such fortresses and fortifications. We have mention of the walls of Nimes in 925(56) and of castles at Anduze and Roquador(57) nearby at about this same time, as well as one called Salaves mentioned in a charter of 959.(58) Our records show that the castle of Substantion was still in use in 923 and 926.(59) So was a castrum called Lunet near Béziers, which is mentioned in a charter of 909,(60) one near Agde referred to in a document of 929,(61) and a turre near the church of Saint-Martin to which we find other references in a charter dating from late in the period.(62) Further to the west near Narbonne and Carcassonne we learn of a castle located in 902 and  904 near the abbey of Lagrasse,(63) a nearby one called Leorte found in 950,(64) one at Montolieu in 925,(65) and a castrum called Aizonne referred to in a charter dating from 918.(66) Somewhat later about 959 we find a source which mentions a turre sold to the archbishop of Narbonne by Viscount Matfred as well as the castrum of Boxia which he left to his son Raymond in 966.(67)
When we turn to the Valley of the Rhone again we find evidence of a number of these fortifications. Our sources reveal that Lyon and Vienne during this period were still fortified,(68) as was LePuy whose burg belonged to the bishop of the city.(69) We learn of a castle at Mont Burton near Vienne in 965,(70) of one at Crussal in 943(71) and of another one called Mezengo in Velay at about this same period.(72) In 950 there seems to have been at least one in Vivarais.(73) A little earlier, in 947, the bishop of Valence gave to a certain Gibernus the right to build a castle in this region,(74) and a little later, between 950 and 976, Bishop Isarn of Grenoble, according to a posterior source, began constructing such castles in Dauphiny.(75) Further in the interior we find castra mentioned in 928 near Gap, and nearby in 950 at Barret and in 967 at Lurs.(76) In addition to pre-existing fortifications we hear of a castle at Fos in 923,(77) of a turre near Marseille in 950,(78) and in 970 of four castra in various parts of this province, one near Marseille, two near Aix, and one near Fréjus.(79)
 When we turn from the Midi to the Spanish March we find even more evidence of such fortresses, as we would expect upon approaching the Moslem frontier. Leaving aside for a moment the border marches of Catalonia and Pallars and Ribagorça let us consider information from counties which were fairly far removed from the frontier. When we do so our sources mention a turre which was located in Roussillon not far from Elne(80) and other fortifications nearby in 947,(81) two castra called Castellano and Turres Betses referred to in documents from Confluent in 952 and 958,(82) a Castrum Falconis in Vallespir in 936,(83) and two called Bertin and Castellaris in Perelada in 948.(84)
When we approach the frontier such castles multiply. Our sources give us information of some twenty-two located near the Moslem border in the counties of Barcelona, Ausona, and Manressa during these years. In the order of their appearance by date in our documents they are Kastellionem, Cervallo, Cleriana, Cervilione, Subirats, Besaura, Rosidors, Marro, Miralès, Queralt, Palumbi, Vidde, Manressa, Arampufia, Olerdula, Castelito, Gellito, Masquefa, Turre Hostalde, Mulier, Castelfels, and Montmell.(85) For the interior regions of Berga-Cerdanya-Besalu we have a record of some thirteen of them: Gratos, Llaiers, Casserès, Lucano, Usalito, Ravinas, Nuret, Olone, Montgrony, Nebule, Spugnole, Saint-Martin, and Agremundo.(86) For Pallars and Ribagorça, in addition to the numerous turres which dotted the mountainsides,(87) we find mention of  castles at Leovalles, Castellous, and Lemignano,(88) to name only a few. Obviously the Spanish March was a land of castles.
We might sum up our evidence of the existence of castles, then, as follows. Except for Gascony, for which we have no evidence of the existence of such fortress, our documents show that a great many more castles existed by the year 975 than were to be found in 900. How many more it is difficult to say. The preceding pages, though, give proof of the existence of some one hundred castles and other fortifications north of the Pyrenees and between forty and fifty in the Spanish March. Though in our sources only a few castles are definitely stated to be ones which were newly built during these decades, it seems fair to say that this was an important period for the construction of such fortresses, particularly in Aquitaine, Septimania, Provence, and the Spanish March.
This brings us to an even more important question. Who owned and controlled these castles that were the basis of military power south of Poitou and Burgundy during this period? Often we cannot be sure. In other cases our information is more precise. Many of the castles north of the Pyrenees seem to have been the outright allodial possessions of the principal counts, viscounts, and principes of these regions. Some, no doubt, became so in the course of these years. This seems to have been the case for those seventeen castles and one rocheta scattered through Languedoc which are mentioned in 961 in the will of Raymond I of Rouergue(89) or the three fortresses mentioned by Countess Garsinde in 972.(90) Ademar, count of Scalas, seems to have owned at least some of the castles found in Quercy and the Limousin about 930,(91) as did Count Acfred in Auvergne at about the same time.(92) During this period our sources show us that Viscount Bernard of Albi acquired at least two castles, one at Solenegra in 937 and another castrum or turre in 957.(93) By 928, Frédélon, progenitor of the later marquises of Anduze, seems to have been in full control of his castle.(94) In Narbonne the viscontal family disposed of at least three castles which were their allodial possessions, in 950, 959, and 966.(95) In the Valley of the Rhone, for which our information  concerning castles is less precise, at least one such fortress, that of Treslia, was owned by King Conrad of Burgundy-Provence about 950.(96)
In Catalonia where the counts tended to control most castles, we still find other individuals who owned such fortresses. Examples of this are the castle of Ravinas, which Viscount Leopardo sold in 928,(97) or that of Spugnole given away by Countess Adelaise in 950,(98) or Masquefa which Hennego, a border marcher lord, acquired as an allod in 963.(99) Sometimes such an individual owned only a small fortress, or turre, like those mentioned in 956 and 959(100) which were given to the abbey of San-Cugat. In Catalonia, as in the rest of the Midi then, a number of powerful nobles controlled castles which were their allodial property.
If members of the secular nobility owned a number of these castles, so did the Church, which, as a matter of fact, seems to have steadily increased the number of such fortresses in its possession. In the Limousin and Quercy, for example, we know of some five castles belonging to the abbey of Tulle, as well as one owned by the abbey of Beaulieu,(101) and several which were built at Limoges about 944 by Bishop Ebles.(102) Further to the south the will of Raymond I of Rouergue, dating from 961, makes it clear that this important noble contemplated that a number of his castles would ultimately go to the churches of Albi and Cahors or to abbeys like Figeac, Moissac, Marcillac, and Brioude.(103) Eleven years later Countess Garsinde's will left similar fortresses to Saint-Vincent of Castres and Saint-Pons of Thomières,(104) as well as consenting to a gift of one to Saint-Michael of Gaillac.(105) About 975 a similar castle became the allodial possession of the abbey of Gellone.(106)
In the Carcassonne-Razès region in 900 we find that the castrum of Por is the allodial possession of the abbey of Saint-Jacques de Joucan,(107) while Lagrasse owned two castles, one near the monastery itself,(108) and another  it acquired about 950.(109) Montolieu, as has been noted, remained fortified throughout this period.(110) The archbishop of Narbonne in 959 acquired a turre as part of his possessions.(111) To the south we learn of another turre belonging to the abbey of Cuxa,(112) of a castrum owned by Saint-Joan de les Abadesses,(113) and several turres which belonged to the abbey of San Cugat near Barcelona.(114)
A similar situation seems to have existed in the Valley of the Rhone. From 924 on in Velay the burg of LePuy belonged to the bishops of this city,(115) while, as we have noted, Bishop Isarn of Grenoble was responsible for the construction of castles throughout his domains sometime between 950 and 974.(116) Even in Provence by 950 the abbey of Saint-Victor of Marseille had acquired a turre or fortress, that of Treslia.(117) In this period as in Carolingian times, the Church, like other landowners owned a large number of the fortresses which are to be found in the Midi and Catalonia.
In addition to the ownership of such fortresses, however, there is another question which we need to consider -- that which concerns the occupation of them. Many of those who owned such fortresses as allods were in no position to actually occupy them personally, particularly if they were nobles or churchmen who owned more than one. What then did they do with them? Various solutions of this problem are to be found during this period. One of the more interesting appears to have been the use of a system of joint ownership, a sort of divisio which continued into later periods. Under this system the ownership of a castle, or its allod, to use a contemporary phrase, was divided. One half belonged to the occupant: one half to an owner who did not live in it. An example of such castle ownership is to be found in a charter, dating from 957 which tells us of a castle acquired by Bernard, viscount of Albi. When Bernard took this over, one-half of this fortress was owned by a certain Senegunde and her four sons "ad proprium allodum," the other half by Bligardis and her two sons in the same manner.(118) The castle of Cenceno which Countess Garsinde gave to Saint-Pons of Thomières in 972 was similarly divided --  as an allod between this countess and her relative Adelaise of Narbonne and the latter's two sons.(119) In 961, in his will, Raymond of Rouergue mentions one-half of certain castles that he was giving to the abbey of Moissac and his part of the castle of Gordo, which suggests the same kind of joint ownership.(120) So does his mention of a rocca in Auvergne which he says Genesius and Aimery hold in common.(121) Joint ownership, particularly by members of the same family, then, was one way in which the problem of controlling castles was solved.
Still another was a system similar to that used by the Carolingians in the Midi and Catalonia, that of giving out castles as benefices or fiefs. The abbey of Tulle and some of the secular lords of the Limousin seem to have used this system for some of the castles they owned.(122) And in Valence it was the method contemplated by Bishop Aurilbert in about 943, when he allowed Gibernus to build a castle which he would then hold as a precarium from this pontiff.(123) The castle of Ulmo, which Bishop Frotaire held and which he gave to Gaillac in 972, seems to have been a similar -- benefice given to the bishop by the House of Toulouse.(124) So was the castle of Sexago, which, we learn the count and countess of Carcassonne held as  a fief from Bishop Hugh of Toulouse,(125) the castle of Parisio which a certain Malbert had "a feo," or the fortress held by Isarn and Arnold in the same way in 961 from Raymond of Rouergue.(126) It seems probable also that at least some of the castles which were the subject of an agreement between the viscounts of Urgell and Cerdanya in 954 were held as fiefs too, since Count Ermengol of Urgell is mentioned specifically in the agreement, in what seems to be his capacity as overlord.(127)
Sometimes such arrangements could be very complex, judging from our sources. The allodial owner of the castle of Sexago was the bishop of Toulouse, Hugh of the Toulouse-Rouergue family. The House of Carcassonne held this castle as a fief from the bishop.(128) Yet the actual occupant was neither of these, but a certain Gilbert, called a vicar, who in 958 presided over a court held in the castle, which met to decide a case involving a dispute over land in the nearby region.(129) Did Gilbert hold this castle in sub-infeudation as castellan from the count and countess of Carcassonne? We cannot say for sure, though it seems probable that he did. One thing though does seem dear. Castles given as benefices, like other land given out in this fashion, had a way of being lost to their owners and tended to become the allodial possession of those who held them. Thus we see attempts to put a time limit of one or two lifetimes on such grants.
Perhaps it was this fact-that castles so easily tended to slip away into the full possession of their occupants -- that explains still a third method of giving out grants, that is called the guarda or baillem system, which we begin to find in this period. Under this system, judging from certain specific examples found a little later, a castle was given out to an individual castellan, but the grant was revokable and limited to a certain period of time, probably less than a lifetime. Such a guarda or castle ward is mentioned in the will of Bishop Hugh of Toulouse which speaks of a castle "in the hands" of a certain Frotaire but "in guarda" of Bernard and Gausbert.(130) Again in a document from the Spanish March we learn that the two castles of Miralès and Queralt were held "in baillem" by a certain Berengar de Araquel.(131)
 Such then is the picture of the ownership, control, and occupation of castles in the Midi and Catalonia during this period. From it we can see that a variety of methods were used. Castles were owned allodially by important secular nobles, by counts, viscounts, and even minor vicars and other landowners. They belonged to various church establishments also. Sometimes these owners actually lived in the fortresses they owned as allods. Sometimes they kept a measure of control over them by joint ownership with others. Sometimes they granted them out as benefices. Sometimes they used a more flexible and less dangerous system of guarda, baillem, or castle ward. What is striking is how many types of castle ownership and control were in use during this period, which seems to be a reflection of the disorganized state of the military and governmental systems of the period. Yet it is worth noting too that we seldom hear during these years of a castle which was the subject of dispute or declared to be illegally built of occupied. The two exceptions are the destruction by Queen Matilda of Provence-Burgundy in 965 of a castle built near Vienne,(132) and the action of Count Boson of La Marche in destroying a castle which Duke William of Poitou built in his domains in 973-974.(133) Castles may have been held or owned in a variety of ways, but down to 975, at least, the direct problems represented by them had not yet become very serious ones for the security of the Midi.
When we probe beneath the surface, however, we begin to see by 975 that the increase in the number of such castles and the methods used in their ownership and control had already begun to bring certain changes to this society. Dimly we can see the outline of a new system of militarism which was arising in Southern France and the Spanish March. For as castles spread everywhere we begin to find along with them a special class of warriors or milites spreading too. These milites seem to differ from their predecessors in the period of Charlemagne or Charles the Bald in one important respect. The warriors of the previous period in the Midi and Catalonia were first important landowners and officials and then warriors. They seem to have been essentially a militia, or part-time warriors, to use a modern term. The milites whom we find appearing in this tenth century are first of all professional fighters and then landowners, judges, and officials. The emphasis had changed, hence their title of milites. Their homes were not the unfortified manor houses or the villas in the countryside, but the castles and fortifications which steadily increased in number  throughout this period.(134) Such fighting men were not a completely new phenomenon. They had been found earlier. But now they were found everywhere in the Midi and Catalonia in regions where they had been unknown before.
Perhaps our earliest documentary reference to this new class, interestingly enough, comes from Quercy and the Limousin early in the tenth century where a charter of about 925 speaks of a certain man as a miles.(135) A little later, about 935, in a charter giving to Tulle the castle of Uxelladuno, King Raoul speaks of the milites inhabiting such castles who were going outside of the walls and were oppressing "loca monarchia."(136) By 971 the number of such milites must have grown, since in that year the abbot of Beaulieu forbade his judices servum, who administered the domains belonging to the abbey, to make themselves into milites and to carry the lance and the spear. If they did, he threatened, they would be returned to the serfdom from which they had emerged.(137) This interesting document not only shows us the new milites in the Limousin, but also  gives us some idea of how they were recruited from the ranks of the minor officialdom and even serfs of the period. It also shows us the opposition to this class which already existed in this part of France.
Such milites were not confined to the Limousin. We find them elsewhere. In 931 a charter of King Raoul speaks of Dalmatius, viscount of LePuy, as a miles,(138)and in 975 another charter from nearby Auvergne mentions land belonging to "belli homini."(139) A document dating from 968, from the Lyonnais, mentions milites near the city of Lyon.(140) We find them in Roussillon according to charters which date from 925 and 953.(141) They are mentioned in the Toulousain in 972.(142) A charter of 954 from Urgell gives us still more information and is the earliest which we possess embodying an agreement between those who own a castle and those who occupy it-agreements which were to be common later on in the Midi and Catalonia. In this agreement between the viscounts of Urgell and Cerdanya we find mention of the milites of the castle of Saint-Martin and those of the castra of Miralès and Queralt. There is also a reference to the military service which was expected of these milites, the hostes et cavalcades.(143)Here we see dearly a reference to such professional warriors and their military duties as they existed along the frontiers of Catalonia.
Even where we do not have a direct reference to such milites we do have indirect evidence of their presence in the castles of the period and of their effect upon the surrounding countrysides. Such garrisons of professional warriors required special provision for support in dues levied on lands located near their fortresses, in contrast to Carolingian warriors who appear to have been scattered about the countryside on land from which they could support themselves. Hence we begin to find in our documents references to such special dues which were levied in addition to the ordinary cens. The most common appear to have been what were called the receptum or albergum, but north of the Pyrenees they are called by other names too, such as usaticos, questas, tallias, or firmancias, this last apparently representing service in labor which was exacted as well as service in kind. In 918, for instance, at a court held in the castle of Alzonne near Carcassonne there is a mention of such services, which, it is said, were owed by an allod belonging to the abbey of Montolieu.(144) In 933 in the same region we find Count Raymond Pons giving up exactions which his men, probably those from nearby castles, were levying on an allod of the same abbey -- dues in the form of levies in wine, bread, and pigs, which were probably such albergos or receptas.(145)
In the Lodève region near the abbey of Gellone, where castles were also common, we hear of a place in 930 which is called "Malos Albergos,"(146) surely a suggestive reference to such dues and the popular reaction to them. In the Albigeois whose many castles made it one of the more militarized parts of the Midi during this period, a charter issued by Raymond Pons in 936 mentions a variety of such exactions.(147) So does one of 942 in which Viscount Aton of Albi gives certain lands to Saint-Pons of Thomières -- which, incidentally lumps such dues under the general term potestatum.(148) In 972 Countess Garsinde of Toulouse mentions similar rights of "usaticos, tallias, et albergos" which she has over land in this same region.(149) As we have already noted, we know of similar exactions levied from castles in the Limousin in 935(150) and of receptas owed the abbey of Beaulieu in 971.(151)
As might be expected, however, Catalonia, a land covered with castles, provides us with a good deal of evidence of such dues, many of which, no doubt, developed directly from those which were owed to Carolingian officials by aprisio holders and which were now taken over by those who controlled the fortresses of this region. In Ausona, whose government Count Guifred organized around its castles, as early as 899 we find a  mention of such dues, which a charter calls "obsequia et servícia."(152) In a later document dating from the period 898-917 they were simply called servicia. In 913 a vicar of Count Miró's, who no doubt controlled one of the castles of the region, in attempting to levy them upon land belonging to Saint-Joan de les Abadesses, called them "hostem et alium regale servicium."(153) A charter from nearby Urgell dated 927 mentions a receptum,(154)and one of 942 speaks of "tasca et servicium" which were levied in the vicinity of the castle of Laiers.(155) By 957 similar dues are to be found in the frontier marches of Barcelona where castles bad multiplied, for in this year Hennego, an important marcher lord, mentions the rights and justice he possesses in the territory of the castle of Arampuña.(156) Our best description of such dues though comes from a charter which Bishop Vivas of Barcelona gave to the inhabitants of his castle of Montmell in 974. In this charter, which is called a pactum, he gave special privileges to those who lived under the jurisdiction of this castle. They included not only a right to freely sell or exchange their land, but also exemption from any dues which might be levied on their houses, their pigs, their horses, their cattle, and their sheep.(157) By 975, then, throughout Catalonia and nearby Pallars and Ribagorça, it had become the custom for the milites of a castle to levy dues on the inhabitants of the surrounding countryside -- dues, which, while originally Carolingian regalian ones, now pertained to a particular castle and those who owned and occupied it.
This leads us to our final question. Did this building of castles and the appearance of a class of professional milites who lived in them, and the dues and services levied in the counties of such fortresses, result in a spread of mandamenta throughout the Midi and Catalonia during this period? Did the castle and its surrounding territory begin to be a governing unit? We have already noted several cases for the late ninth century: one from Savoy and several from Ausona, in which this seems to have been the case. Now there can be little doubt that such mandamenta were beginning to appear in many more parts of Southern France and the Spanish March. In 928, for instance, we learn of land held in the "territory" of the castle of  Garza, located in the county of Gap,(158) and about 970 a charter from Provence mentions land "in termino de Castro Marigno," "in termino Castro Lambisco," and "in termino Castro Petro Castellano."(159) Even before Moslems had been expelled from Fraxinetum, Provence and neighboring regions were being divided into mandamenta.
We find a similar system in Aquitaine where a charter dated 930 from Quercy mentions "Scalas castrum cum omni caslania."(160) Another from Auvergne dating from 975 refers to land "in territorio Castelluc."(161)At about this same period we hear of land at Nîmes located "in vicario Ariensi sub Castro Exunatis."(162) Those rights of justice over neighboring regions possessed by the vicar who was castellan of the castrum of Sexago in Razes in 958 suggest the same kind of territory under the control of a castle.(163)
But again it is the Spanish March which makes dear how widespread this system had become by the year 975. Here along the entire frontier from Tarragona to Ribagorça, where castles were numerous, the castle had become the unit of government controlling its vicinity which was known as its territorium, its appendicium, or its termino. One could cite dozens of charters showing such an organization of land along these frontiers.(164) They make clear that here mandamenta already existed, in which vicars and others who served as castellans were exercising rights of government and exacting dues from the homines of their castra and the territory nearby.
We might sum up our conclusions concerning
the military system of the Midi and Catalonia, then, as follows: between
900 and 975, as principalities disintegrated and the legal system of the
Carolingians disappeared, almost everywhere a growing militarization took
place. This was marked by a building of castles in every part of the Midi
and Catalonia, the appearance  in many regions of a special
class of milites,
and the levying of special dues near these castles
to support the garrisons of milites.
Finally in various regions
this resulted in such castles and the territory becoming what one might
already call mandamenta in which the castellan who controlled such
a fortress began to exercise a certain independent jurisdiction or government
of his own. The process was by no means complete by the year 975, except
perhaps along the frontiers of the Spanish March and in Alpine regions
of Provence. But it had begun, despite opposition to such mandamenta
the exactions they encouraged, in such regions as the Limousin, the Carcassonne
area, and the Viennois. Nor had any uniform system yet been developed to
own and to control such fortresses. But already in 975 the outlines of
this new military system in the Midi and Catalonia are discernable, a military
system which was to increase in importance in the decades which were to
1. The brief chronicle found in the cartulary of Auch says that in the early tenth century, at the time of Count Bernard Lupo, Auch was a "civitas de ruina murorum" and that "cetera civitates Vasconie destructae erant" (Cart. d'Auch, no. 77, pp. 77-79). Then rebuilding began.
2. See the mention of the fleet of Bordeaux and of Amalvinus, dux of Bordeaux in 906 in Bibliothèque Nationale, Mélanges Colbert, XLVI, 97.
3. A charter from the cartulary of Tulle mentions damage done by the Vikings in this region (Cart. de Tulle, no. 289). See the charter of 935 which mentions the restoration of order in the Limousin too (ibid., no. 598).
4. See A. Lewis, The Northern Seas A.D. 300-1100, pp. 405-406, on late Viking raids along the coasts of Aquitaine.
5. P. Ponsch, "A propos de la date du décès de Wilfred le Velu," in Études Roussillonnaises, II, 132-134.
6. Abadal, Els Comtats de Pallars i Ribagorça, p. 123.
7. R. de Abadal i de Vinyals, Els Primers Comtes Catalans, pp. 313-314.
9. Ibid., p. 314.
10. See the accounts of Moslem historians in. J. Millás-Vallicrosa, Els textos d'historiadors musalmans referents a la Catalunya carolingia, in Cuaderns d'Estudi, XIV, 151-155.
11. Abadal, Els Primers Comtes Catalans, pp. 316-318.
12. The best account of the reign of Abd-ar-Rahman III is to be found in E. Lévi-Provençale, L'Espagne Musulmane au Xe siècle. See also Abadal, Els Primers Comtes Catalans, pp. 318-321.
13. Several charters mention Hungarian devastation of this region. For the Vienne region see a charter of 938 (Cart. de Vienne, no. 23). For the Lyonnais a charter of 949 (Cart. de Savigny, no. 38).
14. Cart. de Saint-Victor, no. 1.
15. See later charters in the Cart. de Saint-Victor and the Cart. de Lérins, for mention of the damage done by Moslems in Provence.
16. Raoul Glaber, Historiarum libri V, I, IV, 9, p. 10.
17. Cart. de Grenoble, no. 16.
18. Raoul Glaber, Historiarum libri V, I, IV, 9, p. 12. See also E. Duprat, in Bouches du Rhône, pp. 255-258.
19. Cart. de Beaulieu, nos. 38, 56, 132, 174, and Cart. de Tulle, nos. 229, 596.
20. Ademar de Chabannes, III, 23, p. 145.
21. Ibid., III, 29, p. 150.
22. Cart. de Conques, nos. 6, 25, 89, 91, 109, 110, 113, 114, 124, 132, 151, 157, 189, 200, 208, 227, 230, 235, 250, 288, 324, 340, 357, 391, 413, 424, 434, 436.
23. Both the Cart. de Brioude, and the Cart. de Sauxillanges contain numerous examples in charters dating from this period.
24. Contemporary charters reflect this political instability by ceasing in many cases to use the term county, which earlier was normal. See for instance the charter issued by King Raoul which calls Ademar a count "in partibus Cahors" in 935 (Cart. de Tulle, no. 598); or that dating from 936-942 which mentions the "optimates Limovicensi pago" (Cart. de Saint-Etienne de Limoges, no. 15); or that from Quercy dating from 969-974 which mentions the "urbem Cahors," though it refers to property in the countryside (Cart. de Beaulieu, no. 148). For Auvergne we have a charter of 901 which speaks of "in orbe Arvernio" (Cart. de Brioude, no. 32). Another of 930 mentions "in orbe Arvernico" (ibid., no. 4). Then two of 935 speak of the "patria Arvernica" (ibid., nos. 153, 186). The same expression is found in charters of 939, 940, 945, and 946 in ibid., nos. 40, 158, 246, and 3.
25. Cart. de Saint-Barnard de Romans, nos. 7 (908), 10 (923).
26. Chartes de Cluny, I, no. 476 (937); Cart, de Saint-André-le-bas, nos. 99 (947-948), 145 (956-957).
27. Cart. de Savigny, nos. 59 (968), 123 (970).
28. On this change see R. Poupardin, Le Royaume de Bourgogne, pp. 163-197, and R. Busquet, "Rois de Bourgogne et comtes de Provence," in Provence Historique, III (1950-1951), 147-150.
29. Charters dating from 922, 932, 969, 970, and 971 show the power which Sylvius exercised in Dauphiny (Cart. de Saint-Barnard de Romans, nos. 10, 13, 38, 40, and Cart. de Saint-Chaffre, no. 322.
30. See E. Fournial, "Recherches sur les comtes de Lyon au IXe et Xe siècles," in Le Moyen Age, LVIII (1952), 233-249.
31. Cart. de Sauxillanges, no. 13.
32. Cart. de Brioude, nos. 75, 81, 170, and Cart. de Sauxillanges, nos. 376, 182.
33. Grand Cart. de Brioude, nos. CCCXXXV, CCCCVII.
34. Hist. Gén. de Lang., V. no. 111.
35. Cart. de Sauxillanges, no. 145.
36. Cart. de Saint-Flour, no. 1.
37. Hist. Gén. de Lang., V. no. 71.
38. Ibid., no. 102.
39. Cart. d'Angoulême, no. 2.
40. Ademar de Chabannes, III, 23, p. 144.
41. "Ex historia monasteri Usercensis," in Cart. d'Uzerche, ed. Champéval, pp. 13-14.
42. Cart. de Tulle, no. 12 (922), 14 (930).
43. Ibid., no. 14.
44. Ibid., no. 290.
45. Cart. de Beaulieu, no. 56.
46. Cart. de Tulle, no. 14.
47. Ibid., no. 598.
48. Ibid., nos. 537, 472, 556.
49. Cart. de Beaulieu, no. 159.
50. Ademar de Chabannes, III, 25, p. 147.
51. Cart. de Saint-Jean d'Angély, no. 62.
52. Chartes et documentes pour servir a l'histoire de l'abbaye de Saint-Maixent [Poitou], ed. A. Richard, no. 67, in Archives Historiques de Poitou, XVI (1886), 85.
53. Hist. Gén. de Lang., V. no. 111.
54. Ibid., nos. 125, 126.
55. Cart. de Saint-Sernin, no. 280.
56. Cart. de Nîmes, no. 25.
57. Ibid., no. 2, and Cart. de Gellone, no. 117 (Anduze); Cart. de Nîmes, nos. 32, 33 (Roquador).
58. Cart. de Nîmes, no. 57.
59. Cart. de Gellone, no. 72.
60. Hist Gén. de Lang., V. no. 37.
61. Cart. de Gellone, no. 73.
62. Ibid., no. 281.
63. Cart. de Carcassonne, II, 218, and Hist. Gén. de Lang., V. no. 30.
64. Cart. de Carcassonne, II, 223-224.
65. Cros-Meyrévielle, Documents, no. 21.
66. Cart. de Carcassonne, I, 188-189.
67. Hist. Gén. de Lang., V. nos. 104, 115.
68. For Vienne see charter of 924 in Chartes de Cluny, I, no. 241. Somewhat later see mention of walls of the city in ibid., II, no. 900. For walls of Lyon see "Cart. d'Ainay," no. 192, in Cart. de Savigny, pp. 697-698.
69. Hist. Gén. de Lang., V. no. 49.
70. Cart. de Vienne, no. 28.
71. G. Letonnelier, "Essai sur les origine des châtelains et des mandaments en Dauphiné" in Annales de l'Université de Grenoble, I (1889), 10.
72. Cart. de Saint-Chaffre, no. 106.
73. Hist. Gén. de Lang., II, no. 208.
74. G. Columbi, Episcopi Valentiniani, VII, 61.
75. Cart. de Grenoble, no. 16.
76. Cart. de Conques, no. 431 (Gap); Gallia Christiana, Nov. I,, Inst. 73 (Barret), Inst. 443 (Lurs).
77. Cart. de Saint-Victor, no. 1.
78. Ibid., no. 1041.
79. Ibid., no. 23.
80. Hist. Gén. de Lang., V, no. 83.
81. Cart. roussillonnais, no. 7.
82. Catalunya Carolingia, II, 92-93 (952), 96-98 (958).
83. Hist. Gén. de Lang., V, no. 65.
84. Catalunya Carolingia, II, 229-231.
85. El Archivo Condal de Barcelona, nos. 9 (Kastellione), 17 (Marro), 47 (Cervallo), 67 (Besaura), 93 and 100 (Rosidors), 130 (Manressa), 132 (Palumbi, Queralt); Cart. de San Cugat, nos. 2 (Cervallo), 5 (Cervilione), 7 (Cervallo), 9 (Subirats), 45 (Vidde), 51 (Cleriana), 56 (Olerdula), 65 (Castelito, Gellito, Masquefer), 67 (Subirats), 87 (Mulier), 88 (Castelito), 95 (Gellito), 96 (Aram. pufla), 97 (Arampuña, Castelfels), 98, 99, and 105 (Castelfels), 107 (Aram. puña); Liber Feudorum, nos. 445 and 447 (Cleriana); Hut. Gén. de Lang., II, no. 209 (Miralès, Queralt); Archivo Catedral de Barcelona, nos. 23 (Turre Hostalde), 24 (Montmell).
86. El Archivo Condal de Barcelona, nos. 16 (Gratos), 72 (Llaiers), 73 (Casseres), 78 (Llaiers), 80 (Usalito), 81 (Agremundo), 91 (Ravinas), 113 (Nuret), 121 (Llaiers), 123 (Montgrony), 129 (Nuret), 130 (Nebule, Spugnole), 167 and 168 (Usalito), and appendix B., p. 444 (Lucano); Cart. roussillonnais, no. 7 (Olone) ; Hist. Gén. de Lang., II, no. 209 (Saint Martin).
87. See Abadal, Els Comtats de Pallars i Ribagorça, for the large number of turres in this region.
88. Hist. Gén. de Lang., V. nos. 82 (Leovalles); 95 (Castellous, Lemignano).
89. Ibid., no. 111.
90. Ibid., nos. 125, 126.
91. Cart. de Tulle, no. 14.
92. Cart. de Sauxillanges, no. 13.
93. Hist. Gén. de Lang., V, no. 102.
94. Cart. de Nîmes, nos. 32, 33.
95. Cart. de Carcassonne, II, 22 3-224 (950); Hist. Gén. de Lang., V. nos. 105 (959), 115 (966).
96. Cart. de Saint-Victor, no. 1041.
97. El Archivo Condal de Barcelona, no. 91.
98. Ibid., no. 130.
99. Cart. de San Cugat, no. 65.
100. Ibid., nos. 45 (956), 56 (959).
101. Cart. de Tulle, nos. 12, 14, 290, 598; Cart. de Beaulieu, no. 56.
102. Ademar de Chabannes, III, 25, p. 147.
103. Hist. Gén. de Lang., V. no. 111.
104. Ibid., no. 126.
105. Ibid., no. 123.
106. Cart. de Gellone, no. 281.
107. Cart. roussillonnais, no. 3.
108. Hist. Gén. de Lang., V. no. 38.
109. Cart. de Carcassonne, II, 223-224.
110. Ibid., p. 218, and Cros-Meyrévielle, Documents, no. 21.
111. Hist. Gén. de Lang., V, no. 105.
112. Catalunya Carolingia, II, 96-98.
113. El Archivo Condal de Barcelona, no. 9.
114. For example see a charter of 959 in Cart. de San Cugat, no. 56.
115. Hist. Gén. de Lang., V, no. 49.
116. Cart. de Grenoble, no. 16.
117. Cart. de Saint-Victor, no. 1041.
118. Hist. Gén. de Lang., V, no. 102.
119. Ibid., no. 125. See also a charter of 950 in which a certain Ermengardis gives her part of the castle of Leorte to the abbey of Lagrasse (Cart. de Carcassonne, II, 223-224).
120. Hist. Gén. de Lang., V. no. 111.
122. A charter of Tulle gives the history of such a castle during the tenth and eleventh centuries, in this case the castle of Moncieux, which was given out as a benefice. The castle originally was given to Tulle by Ademar, viscount of Scalas, who, along with his wife, kept possession of it as a lifetime benefice or usufruct. Then, about 935 the monks gave the castle as a benefice to a certain Donarel, son of Viscount Ademar. A little later Bernard, viscount of Turenne received the castle as a benefice with the surrounding area ("cum caslania"). He held it only as a lifetime usufruct and paid the monks a cens for it. Then Viscount Archimbaud of Comborn and his son were given the castle as a benefice and paid the abbey a cens. This carries the castle down to the end of this period (Cart. de Tulle, no. 290). It is worth noting that Tulle granted this castle as a benefice on a lifetime basis only during these years. A cens, proof of the abbey's allodial ownership of the castle, regularly was paid by its possessors. And thirdly the abbey shifted possession of the castle pretty much to suit itself, or as prudence, in terms of a strong defensor in the neighborhood, dictated.
123. G. Columbi, Episcopi Valentiniani, XVII, 21.
124. Hist. Gén. de Lang., V, no. 123. This charter shows that the castle of Ulmo and other honores were only given to Gaillac with the permission of Count Raymond II of Rouergue, his lord.
125. Cart. de Saint-Sernin, no. 280.
126. Hist. Gén. de Lang., V. no. 111.
127. Hist. Gén. de Lang., II, no. 209.
128. Cart. de Saint-Sernin, no. 280.
129. Hist. Gén. de Lang., V. no. 104.
130. Cart. de Saint-Sernin, no. 280.
131. Hist. Gén. de Lang., II, no. 209. See also the mention of a "guarda" in a charter given to the abbey of San Cugat in 938 by King Louis IV (Catalunya Carolinga, II, 183-193).
132. Cart, de Vienne, no. 28.
133. Ademar de Chabannes, III, 29, p. 150.
134. With the possible exception of three cases -- one a miles mentioned in a charter dating from 925 from the Limousin (see note 135), one a royal charter of 931 which calls Dalmatius, viscount of LePuy, a miles (see note 138), and one dating from 925 which calls Landry of Roussillon a miles (see note 141) -- the term miles or milites found in charters of this period are not substitutes for the words vassus or vassi. They do not denote social rank and cannot be translated by the term knight or knights. They clearly are occupational terms which denote fighting men and those garrisoning castles. See, for example, reference in a charter of 935 from the Limousin to the milites of the castle of Uxelladuno or, according to a charter of 972, the milites living in the castle of Ulmo (Hist. Gén. de Lang., V, no. 123), or those mentioned in the charter of 954 who are called milites of the castra of Miralès and Queralt located near the Moslem border of Catalonia (ibid., II, 209). Their duties seem to be essentially military rather than feudal, during this period, or at least these two duties have not yet coalesced to form the feudal fighting man. If one doubts this, the abbey of Beaulieu's refusal to allow its lay officials (judices) to become milites would be conclusive. What the abbey refused to allow them to do in 971 was to arm themselves as milites (Cart. de Beaulieu, no. 50). Down to 975 then in the Midi and Catalonia, those called milites were mounted, armed men living in fortresses with others of their kind. They were not yet a fighting and governing feudal class in the society of these regions.
135. This miles is a certain Analgorius who received a villa and a church "ad fevum" from Viscount Ademar of Scalas (Cart. de Tulle, no. 289). See however earlier references to such milites in this same region in Odo, Vita Sancti Geraldi, Auriligiensis, I, 8, 18, 31, 37-39, cols. 647, 654, 660-665.
136. Cart. de Tulle, no. 598.
137. Cart. de Beaulieu, no. 50.
138. Cart. de Carcassonne, I, 76.
139. Cart. de Conques, no. 228. These were probably milites, so called, who lived on land near a tract given to Conques during this period. See also charters of 913 and 916 which mention a certain cavalarius who lived near the abbey of Brioude (Cart. de Brioude, nos. 5, 151, 100).
140. "Cart. d'Ainay," no. 3, in Cart. de Savigny.
141. "Landry, miles" is the expression used in a charter of 925 (Cart. de Carcassonne, H, 221). See also a charter of 953 which mentions in a general way "boni homines, clericos, et milites" who were present at the dedication of a church in Vallespir (Cart. roussillonnais, no 8). The use of a term such as milites in this region seems to be a Carolingian survival, since such terms are also used in documents of the late ninth century originating in this part of the Midi.
142. Hist. Gén. de Lang., V, no. 123.
143. Ibid., II, no. 209.
144. These services, "circa, quarta et cavalcata," are said to be those which the spani (that is aprisio holders) are expected to render to constituted authorities.
145. Hist. Gén. de Lang., V, no. 57.
146. Cart. de Gellone, no. 36.
147. Hist. Gén. de Lang., Y, no. 67.
148. Ibid., no. 77.
149. Ibid., no. 125.
150. Cart. de Tulle, no. 598.
151. Cart. de Beaulieu, no. 50. Beaulieu, however, forbids laymen to be given receptas belonging to the abbey in this charter.
152. See Abacial, Els Primers Comtes Catalans, p. 88, on the duties which new inhabitants of Ausona were expected to render to constituted authorities. See also Catalunya Carolingia, II, 289-299.
153. El Archivo Condal de Barcelona, nos. 8, 38, and appendix A.
154. Ibid., no. 92.
155. Ibid., no. 121.
156. Cart. de San Cugat, no. 49.
157. Archivo Cathedral de Barcelona, no. 24.
158. Cart. de Conques, no. 431.
159. Cart. de Saint-Victor, no. 23.
160. Cart. de Tulle, nos. 14 and 290.
161. Cart. de Sauxillanges, no. 376.
162. Cart. de Gellone, no. 111.
163. Hist. Gén. de Lang., V, no. 104.
164. See Abadal, Els Primers Comtes Catalans, pp. 87-88, on this system in Ausona after it was repopulated by Count Guifred. Among others see references in 921 to the appendicil of the castle of Besaura (El Archivo Condal de Barcelona, no. 67); or land which in 941 was located in the appendicii of the castle of Castelloi (ibid., no. 129); or land said to be located in 955 "in termino" of the castrum of Cervello (Cart. de San Cugat, no. 40); or that which in 974 is located "in termino" of the castrum of Arampuña (ibid., no. 103).