Chapter II: The Beginnings of Anjou (2)

The receiver of Fulk’s famous letter had gone before him to the grave; Louis From-over-sea, the grandson of Eadward the Elder, the last Karolingian worthy of his race, had died in 954. His death brought the house of France a step nearer to the throne; but it was still only one step. Lothar, the son of Louis, was crowned in his father’s stead; two years later the king-maker followed the king; and thenceforth his son, the new duke of the French, Hugh Capet, steadily prepared to exchange his ducal cap for a crown which nevertheless he was too prudent to seize before the time. In the face of countless difficulties, Louis in his eighteen years’ reign had contrived to restore the monarchy of Laon to a very real kingship. His greatest support in this task had been his wife’s brother, the Emperor Otto the Great. The two brothers-in-law, who had come to their thrones in the same year, were fast friends in life and death; and Otto remained the faithful guardian of his widowed sister and her son. So long as he lived, Hugh’s best policy was peace; and while Hugh remained quiet, there was little scope for military or political action on the part of his adherent Geoffrey of Anjou. In 973, however, the great Emperor died; and soon after he was gone the alliance between the Eastern and Western Franks began to shew signs of breaking. Lothar and Otto II. were brothers-in-law as well as cousins, but they were not friends as their fathers had been. In an evil hour Lothar was seized with a wild longing to regain the land which bore his name,--that fragment of the old “Middle Kingdom,” known as the duchy of Lotharingia or Lorraine, which after long fluctuating between its attachment to the imperial crown and its loyalty to the Karolingian house had finally cast in its lot with the Empire, with the full assent of Louis From-over-sea. Lothar brooded over its loss till in 978, when Otto and his queen were holding their court at Aachen, his jealousy could no longer endure the sight of his rival so near the border, and he summoned the nobles of his realm to an expedition into Lorraine.[268] Nothing could better fall in with the plans of Hugh Capet than a breach between Lothar and Otto; the call to arms was readily answered by the duke and his followers, and the grey tunic of the Angevin count was conspicuous at the muster.[269] The suddenness of Lothar’s march compelled Otto to make a hasty retreat from Aachen; but all that the West-Franks gained was a mass of plunder, and the vain glory of turning the great bronze eagle on the palace of Charles the Great towards the east instead of the west.[270] While they were plundering Aachen Otto was preparing a counter-invasion.[271] Bursting upon the western realm, he drove the king to cross the Seine and seek help of the duke, and before Hugh could gather troops enough to stop him he had made his way to the gates of Paris. For a while the French and the Germans lay encamped on opposite banks of the river, the duke waiting till his troops came up, and beguiling the time with skirmishes and trials of individual valour.[272] But as soon as Otto perceived that his adversaries were becoming dangerous he struck his tents and marched rapidly homewards, satisfied with having inflicted on his rash cousin a far greater alarm and more serious damage than he had himself suffered from Lothar’s wild raid.[273]

[268] Richer, l. iii. c. 68.

[269] Chron. Vindoc. a. 954 (Marchegay, _Eglises_, p. 163).

[270] Richer, l. iii. c. 71.

[271] The exact date of Lothar’s attack on Lotharingia seems to
be nowhere stated. That of Otto’s invasion of Gaul, however,
which clearly followed it immediately, is variously given as
977 (Chronn. S. Albin. and Vindoc., Marchegay, _Eglises_, pp.
21, 163) and 978 (Chronn. S. Flor. Salm. and S. Maxent., _ib._
pp. 186, 381). The later date is adopted by Mr. Freeman, _Norm.
Conq._, vol. i. p. 264.

[272] Among these the Angevin writers (_Gesta Cons._,
Marchegay, _Comtes_, pp. 79, 80) introduce Geoffrey Greygown’s
fight with a gigantic Dane, Æthelwulf. It seems to be only
another version, adorned with reminiscences of David and
Goliath, of Richer’s account (l. iii. c. 76) of a fight between
a German champion and a man named Ivo; and the whole story of
this war in the _Gesta_ is full of hopeless confusions and
anachronisms.

[273] Richer, l. iii. cc. 72–77.

From that time forth, at least, Geoffrey Greygown’s life was a busy and a stirring one. It seems to have been in the year of the Lotharingian raid that he married his second wife, Adela, countess in her own right of Chalon-sur-Saône, and now the widow of Count Lambert of Autun.[274] By his first marriage, with another Adela, he seems to have had only a daughter, Hermengard, who had been married as early as 970[275] to Conan the Crooked, count of Rennes. There can be little doubt that this marriage was a stroke of policy on Geoffrey’s part, intended to pave the way for Angevin intervention in the affairs of Britanny. The claims of Fulk the Good to the overlordship of Nantes had of course expired with him; whatever rights the widow of Duke Alan might carry to her second husband, they could not pass to her stepson. Still Geoffrey could hardly fail to cherish designs upon, at least, the debateable ground which lay between the Mayenne and the original county of Nantes. Meanwhile the house of Rennes had managed to establish, by the right of the stronger, its claim to the dukedom of Britanny. Hoel, a son of Alan Barbetorte, remained count of Nantes for nearly twenty years after Fulk’s death; his career was ended at last by the hand of an assassin;[276] and as his only child was an infant, his brother Guerech, already bishop of Nantes, was called upon to succeed him, as the only surviving descendant of Alan who was capable of defending the state. Guerech was far better fitted for a secular than for an ecclesiastical ruler; as bishop, his chief care was to restore or rebuild his cathedral, and for this object he was so eager in collecting contributions that he made a journey to the court of Lothar to ask help of the king in person. His way home lay directly through Anjou. Geoffrey felt that his opportunity had come; and he set the first example of a mode of action which thenceforth became a settled practice of the Angevin counts. He laid traps in all directions to catch the unwary traveller, took him captive, and only let him go after extorting homage not merely for the debateable land, but also for Nantes itself; in a word, for all that part of Britanny which had been held or claimed by Fulk as Drogo’s guardian.[277]

[274] See note C at end of chapter.

[275] Morice, _Hist. Bret._, vol. i. p. 63. See note C at end
of chapter.

[276] Chron. Brioc., Morice, _preuves_, vol. i., p. 31. Chron.
Namnet., _Rer. Gall. Scriptt._, vol. viii. p. 278. “C. 980,”
notes the editor in the margin.

[277] Chron. Brioc., Morice, _Hist. Bret., preuves_, vol. i.
col. 32.

Geoffrey had gained his hold over Nantes; but in so doing he had brought upon himself the wrath of his son-in-law. Conan, as duke of Britanny, claimed for himself the overlordship of Nantes, and regarded Guerech’s enforced homage to Geoffrey as an infringement of his own rights. His elder sons set out to attack their step-mother’s father, made a raid upon Anjou, and were only turned back from the very gates of Angers by a vigorous sally of Geoffrey himself.[278] Conan next turned his vengeance upon the unlucky count-bishop of Nantes. The Angevin and his unwilling vassal made common cause against their common enemy, who marched against their united forces, bringing with him a contingent of the old ravagers of Nantes--the Normans.[279] The rivals met not far from Nantes, on the _lande_ of Conquereux, one of those soft, boggy heaths so common in Britanny; and the issue of the fight was recorded in an Angevin proverb--“Like the battle of Conquereux, where the crooked overcame the straight.”[280] Conan was, however, severely wounded, and does not appear to have followed up his victory; and the Nantes question was left to be fought out ten years later, on the very same ground, by Geoffrey’s youthful successor.

[278] See note D at end of chapter.

[279] Chron. Brioc., as above.

[280] See note D at end of chapter.

The death of Lothar, early in March 986, brought Hugh Capet within one step of the throne. The king’s last years had been spent in endeavouring to secure the succession to his son by obtaining for him the homage of the princes of Aquitaine and the support of the duke of the French--two objects not very easy to combine, for the great duchies north and south of the Loire were divided by an irreconcileable antipathy. In 956 William “Tête-d’Etoupe,” or the “Shockhead,” strong in his triple power as count of Poitou, count of Auvergne and duke of Aquitaine--strong, too, in his alliance with Normandy, for he had married a sister of his namesake of the Long Sword--had bidden defiance not unsuccessfully to Lothar and Hugh the Great both at once.[281] In 961 Lothar granted the county of Poitiers to Hugh;[282] but all he could give was an empty title; when William Shockhead died in 963,[283] his son William Fierabras stepped into his place as count of Poitou, duke of Aquitaine, and leader of the opposition to Hugh Capet.

[281] Richer, l. iii. cc. 3–5.

[282] _Ib._ c. 13.

[283] Chron. S. Maxent. ad ann. (Marchegay, _Eglises_, p. 381).

It was now evident that the line of Charles the Great was about to expire in a worthless boy. While the young King Louis, as the chroniclers say, “did nothing,”[284] the duke of the French and his followers were almost openly preparing for the last step of all. The count of Anjou, following as ever closely in the wake of his overlord, now ventured on a bold aggression. Half by force, half by fraud, he had already carried his power beyond the Mayenne; he now crossed the Loire and attacked his southern neighbour the count of Poitou. Marching boldly down the road which led from Angers to Poitiers, he took Loudun, and was met at Les Roches by William Fierabras, whom he defeated in a pitched battle and pursued as far as a place which in the next generation was marked by the castle of Mirebeau. Of the subsequent details of the war we know nothing; it ended however in a compromise; Geoffrey kept the lands which he had won, but he kept them as the “man” of Duke William.[285] They seem to have consisted of a series of small fiefs scattered along the valleys of the little rivers Layon, Argenton, Thouet and Dive, which furrow the surface of northern Poitou.[286] The most important was Loudun, a little town some eighteen miles north-west of Poitiers. Even to-day its gloomy, crooked, rough-paved streets, its curious old houses, its quaintly-attired people, have a strangely old-world look; lines within lines of broken wall wind round the hill on whose slope the town is built, and in their midst stands a great square keep, the work of Geoffrey’s successors. He had won a footing in Poitou; they learned to use it for ends of which, perhaps, he could as yet scarcely dream. Loudun looked southward to Poitiers, but it looked northward and eastward too, up the valley of the Thouet which led straight up to Saumur, the border-fortress of Touraine and Anjou, and across the valley of the Vienne which led from the Angevin frontier into the heart of southern Touraine. Precious as it might be in itself, Loudun was soon to be far more precious as a point of vantage not so much against the lord of Poitiers as against the lord of Chinon, Saumur and Tours.

[284] “Ludovicus qui nihil fecit” is the original form of the
nickname usually rendered by “le Fainéant.”

[285] See note D at end of chapter.

[286] Fulk Nerra’s Poitevin castles, Maulévrier, Thouars, etc.,
must have been built on the ground won by Geoffrey.

The little marchland had thus openly begun her career of aggression on the west and on the south. It seems that a further promise of extension to the northward was now held by Hugh Capet before the eyes of his faithful Angevin friend. Geoffrey’s northern neighbour was as little disposed as the southern to welcome the coming king. The overlordship of Maine was claimed by the duke of the Normans on the strength of a grant made to Hrolf in 924 by King Rudolf; it was claimed by the duke of the French on the strength of another grant made earlier in the same year by Charles the Simple to Hugh the Great,[287] as well as in virtue of the original definition of their duchy “between Seine and Loire”; but the Cenomannian counts owned no allegiance save to the heirs of Charles the Great, and firmly refused all obedience to the house of France. Hugh Capet, now king in all but name, laid upon the lord of the Angevin march the task of reducing them to submission. He granted Maine to Geoffrey Greygown[288]--a merely nominal gift at the moment, for Hugh (or David) of Maine was in full and independent possession of his county; and generation after generation had to pass away before the remote consequences of that grant were fully worked out to their wonderful end. Geoffrey himself had no time to take any steps towards enforcing his claim. Events came thick and fast in the early summer of 987. King Louis V. was seized at Senlis with one of those sudden and violent sicknesses so common in that age, and died on May 22. The last Karolingian king was laid in his grave at Compiègne; the nobles of the realm came together in a hurried meeting; on the proposal of the archbishop of Reims they swore to the duke of the French a solemn oath that they would take no steps towards choosing a ruler till a second assembly should be held, for which a day was fixed.[289] Hugh knew now that he had only a few days more to wait. He spent the interval in besieging a certain Odo, called “Rufinus”--in all likelihood a rebellious vassal--who was holding out against him at Marson in Champagne; and with him went his constant adherent Geoffrey of Anjou. At the end of the month the appointed assembly was held at Senlis. Passing over the claims of Charles of Lorraine, the only surviving descendant of the great Emperor, the nobles with one consent offered the crown to the duke of the French. From his camp before Marson Hugh went to receive, at Noyon on the 1st of June,[290] the crown for which he had been waiting all his life. Geoffrey, whom he had left to finish the siege, fell sick and died before the place, seven weeks after his patron’s coronation;[291] and his body was carried back from distant Champagne to be laid by his father’s side in the church of S. Martin at Tours.[292]

[287] Chron. Frodoard, a. 924 (_Rer. Gall. Scriptt._, vol.
viii. p. 181).

[288] See note E at end of chapter.

[289] Richer, l. iv. cc. 5 and 8.

[290] Richer, l. iv. c. 12. On this Kalckstein (_Geschichte
des französischen Königthums unter den ersten Capetingern_,
vol. i. p. 380, note 2), remarks: “Aus Rich. iv. 12 wäre zu
schliessen, dass Hugo in Noyon gekrönt wurde ... aber eine
gleichzeitige Urkunde von Fleury entscheidet für Reims. Richer
gibt wohl in Folge eines Gedächtnissfehlers den 1 Juli (wie für
Juni zu verbessern seine wird) als Krönungstag. Hist. Francica
um 1108 verfasst, Aimoin Mirac. S. Bened. ii. 2 (Bouq., x. 210
u. 341).” The _Hist. Franc. Fragm._ here referred to places
the crowning at Reims on July 3. Aimoin, however, places it at
Noyon and gives no date. The question therefore lies really
between Richer and the Fleury record referred to, but not
quoted, by Kalckstein; for the two twelfth century writers
are of no authority at all in comparison with contemporaries.
We must suppose that the Fleury charter gives the same date
as the _Hist. Franc. Fragm._ But is it not possible that Hugh
was really crowned first at Noyon on 1st June, and afterwards
recrowned with fuller state at Reims a month later?

[291] Chronn. S. Albin., S. Serg., and Vindoc., a. 987; Rain.
Andeg. a. 985; S. Maxent. a. 986 (Marchegay, _Eglises_, pp. 21,
134, 164, 9, 382). Fulk Rechin (Marchegay, _Comtes_), p. 376.

[292] Fulk Rechin, as above, and _Gesta Cons._ (_ib._), p. 89,
say he was buried in S. Martin’s. R. Diceto (Stubbs, vol. i. p.
165) buries him in S. Aubin’s at Angers.

The century of preparation and transition was over; the great change was accomplished, not to be undone again for eight hundred years. The first period of strictly French history and the first period of Angevin history close together. The rulers of the marchland had begun to shew that they were not to be confined within the limits which nature itself might seem to have fixed for them; they had stretched a hand beyond their two river-boundaries, and they had begun to cast their eyes northward and dream of a claim which was to have yet more momentous results. In the last years of Geoffrey Greygown we trace a foreshadowing of the wonderful career which his successor is to begin. From the shadow we pass to its realization; with the new king and the new count we enter upon a new era.

NOTE A.

ON THE SOURCES AND AUTHENTICITY OF EARLY ANGEVIN HISTORY.

Our only detailed account of the early Angevins, down to Geoffrey Greygown, is contained in two books: the _Gesta Consulum Andegavensium_, by John, monk of Marmoutier, and the _Historia Comitum Andegavensium_, which goes under the name of Thomas Pactius, prior of Loches. Both these works were written in the latter part of the twelfth century; and they may be practically regarded as one, for the latter is in reality only an abridgement of the former, with a few slight variations. The _Gesta Consulum_ is avowedly a piece of patchwork. The author in his “Proœmium” tells us that it is founded on the work of a certain Abbot Odo which had been recast by Thomas Pactius, prior of Loches, and to which he himself, John of Marmoutier, had made further additions from sundry other sources which he enumerates (Marchegay, _Comtes d’Anjou_, p. 353. This “Proœmium” is there printed at the head of the _Historia Abbreviata_ instead of the _Gesta Consulum_, to which, however, it really belongs; see M. Mabille’s introduction, _ib._ p. xxxi.). The _Historia Comitum Andegavensium_ (_ib._ p. 320) bears the name of Thomas of Loches, and thus professes to be the earlier version on which John worked. But it is now known that the work of Thomas, which still exists in MS., is totally distinct from that published under his name (see M. Mabille’s introduction to _Comtes d’Anjou_, pp. xviii., xix.), and, moreover, that the printed _Historia Comitum_ is really a copy of a series of extracts from Ralf de Diceto’s _Abbreviationes Chronicorum_--extracts which Ralf himself had taken from the _Gesta Consulum_ (see Bishop Stubbs’ preface to R. Diceto, vol. ii. pp. xxiii.–xxix). There is, however, one other source of information about the early Angevins which, if its author was really what he professed to be, is of somewhat earlier date and far higher value, although of very small extent. This is the fragment of the _Angevin History_ which goes under the name of Count Fulk Rechin. Its authorship has been questioned, but it has never been disproved; and one thing at least is certain--the writer, whoever he may have been, had some notion of historical and chronological possibilities, whereas John of Marmoutier had none. Fulk Rechin (as we must for the present call him, without stopping to decide whether he has a right to the name) gives a negative testimony against all John’s stories about the earlier members of the Angevin house. He pointedly states that he knows nothing about the first three counts (Marchegay, _Comtes_, p. 376), and he makes no mention of anybody before Ingelger. Now, supposing he really was Count Fulk IV. of Anjou, it is fairly safe to assume that if anything had been known about his own forefathers he would have been more likely to know it than a monk who wrote nearly a hundred years later. On the other hand, if he was a twelfth-century forger, such a daring avowal of ignorance, put into the mouth of such a personage, shews the writer’s disregard of the tales told by the monk, and can only have been intended to give them the lie direct.

The two first members of the Angevin house, then--Tortulf of Rennes and his son Tertullus--rest solely on the evidence of these two late writers. Their accounts are not recommended by intrinsic probability. We are roused to suspicion by the very first sentence of the _Gesta Consulum_:--“Fuit vir quidam de Armoricâ Galliâ, nomine Torquatius. Iste a Britonibus, proprietatem vetusti ac Romani nominis ignorantibus, corrupto vocabulo Tortulfus dictus fuit” (Marchegay, _Comtes_, p. 35). When one finds that his son is called Tertullus, it is impossible not to suspect that “Torquatius” and “Tertullus” are only two different attempts to Latinize a genuine Teutonic “Tortulf.” For the lives of these personages John of Marmoutier gives no distinct dates; but he tells us that Torquatius was made Forester of Nid-de-Merle by Charles the Bald, “eo anno quo ab Andegavis et a toto suo regno Normannos expulit” (Marchegay, _Comtes_, p. 35). Now this is rather vague, but it looks as if the date intended were 873. We are next told that Tertullus went to seek his fortune in France “circa id temporis quo Karolus Calvus ... ex triarcho monarchus factus, non longo regnavit spatio” (_ib._ pp. 36, 37), whatever that may mean. The next chronological landmark is that of the “reversion” of S. Martin, which John copies from the Cluny treatise _De Reversione B. Martini_, and copies wrong. Then comes Fulk the Red, on whom he says the whole county of Anjou was conferred by Duke Hugh of Burgundy, guardian of Charles the Simple, the county having until then been divided in two parts; and he also says that Fulk was related to Hugh through his grandmother (_ib._ pp. 64, 65).

There are several unmanageable points in this story. 1. The pedigree cannot be right. It is clear that John took Hugh the Great (“Hugh of Burgundy,” as he calls him) to be a son of the earlier Hugh of Burgundy (one copy of the _Gesta_, that printed by D’Achéry in his _Spicilegium_, vol. iii. p. 243, actually adds “filius alterius Hugonis”), and this latter to have been the father of Petronilla, wife of Tertullus.

The chronology of the life of Fulk the Red, long a matter of mingled tradition and guess-work, has now been fairly established by the investigations of M. E. Mabille. This gentleman has examined the subject in his introduction to MM. Marchegay and Salmon’s edition of the _Chroniques des Comtes d’Anjou_, and in an article entitled “Les Invasions normandes dans la Loire,” in the _Bibliothèque de l’Ecole des Chartes_, series vi. vol. v. pp. 149–194; to each of these works is appended by way of _pièces justificatives_ a series of charters of the highest importance for establishing the facts of the early history of Anjou and Touraine. The first appearance of Fulk is as witness to a charter given at Tours by Odo, as abbot of S. Martin’s, in April 886. (Mabille, introd. _Comtes_, p. lxix. note). Now if Fulk the Red was old enough to be signing charters in 886, his parents must have been married long before the days of Louis the Stammerer--in 870 at the very latest, and more likely several years earlier still. His grandparents therefore (_i.e._ Tertullus and Petronilla) must have been married before 850. It is possible that Hugh the Abbot who died in 887 may have had a daughter married as early as this; but it does not seem very likely.

2. The story of Ingelger’s investiture with Orleans and the Gâtinais is suspicious. His championship of the slandered countess of Gâtinais (Marchegay, _Comtes_, pp. 40–45) is one of those ubiquitous tales which are past confuting. Still the statement that he somehow acquired lands in the Gâtinais is in itself not impossible. But the coupling together of Gâtinais and Orléans is very suspicious. Not one of the historical descendants of Ingelger had, as far as is known, anything to do with either place for nearly two hundred years. There is documentary proof (see the signatures to a charter printed in Mabille’s introd. _Comtes_, p. lxiv, note 1; the reference there given to Salmon is wrong) that in 942, the year after the death of Fulk the Red, the viscount of Orléans was one Geoffrey; and he belonged to a totally different family--but a family which, it seems, did in time acquire the county of Gâtinais, and in the end became merged in the house of Anjou, when the son of Geoffrey of Gâtinais and Hermengard of Anjou succeeded his uncle Geoffrey Martel in 1061. It is impossible not to suspect that the late Angevin writers took up this story at the wrong end and moved it back two hundred years.

3. Comes the great question of Ingelger’s investiture with half the county of Anjou.

In not one of the known documents of the period does Ingelger’s name appear. The only persons who do appear as rulers of the Angevin march are Hugh the Abbot and his successor Odo, till we get to Fulk the Viscount. Fulk’s first appearance in this capacity is in September 898, when “Fulco vicecomes” signs a charter of Ardradus, brother of Atto, viscount of Tours (Mabille, Introd. _Comtes_, p. xciii). He witnesses, by the same title, several charters of Robert the Abbot-Count during the next two years. In July 905 we have “signum Fulconis Turonorum et Andecavorum vicecomitis” (_ib._ p. xcv); in October 909 “signum domni Fulconis Andecavorum comitis” (_ib._ p. xcviii); and in October 912 he again signs among the counts (_ib._ p. lxi, note 4). But in May 914, and again as late as August 924, he resumes the title of viscount (_ib._ pp. c and lxii, note 2). Five years later, in the seventh year of King Rudolf, we find a charter granted by Fulk himself, “count of the Angevins and abbot of S. Aubin and S. Licinius” (_ib._ p. ci); and thenceforth this is his established title.

These dates at once dispose of R. Diceto’s statement (Stubbs, vol. i. p. 143) that Fulk succeeded his father Ingelger as second count in 912. They leave us in doubt as to the real date of his appointment as count; but whether we adopt the earlier date, in or before 909, or the later one, between 924 and 929, as that of his definite investiture, we cannot accept the _Gesta’s_ story that it was granted by Hugh the Great on behalf of Charles the Simple. For in 909 the duke of the French was not Hugh, but his father Robert; and in 924–929 the king was not Charles, but Rudolf of Burgundy.

But the chronology is not the only difficulty in the tale of Count Ingelger. The _Gesta_-writers admit that “another count” (_i.e._ the former count, Duke Hugh) went on ruling beyond the Mayenne. This at once raises a question, very important yet very simple--Did the Angevin March, the March of Robert the Brave and his successors, extend on both sides of the Mayenne? For the assumption that it did is the ground of the whole argument for the “bipartite” county.

The old territory of the Andes certainly spread on both sides of the river. So also, it seems, did the march of Count Lambert. The commission of a lord marcher is of necessity indefinite; it implies holding the border-land and extending it into the enemy’s country if possible. It appears to me that when Lambert turned traitor he carried out this principle from the other side; when Nantes became Breton, the whole land up to the Mayenne became Breton too. This view is distinctly supported by a charter in which Herispoë, in August 852, styles himself ruler of Britanny and up to the river Mayenne (Lobineau, _Hist. Bretagne_, vol. ii. col. 55); and it gives the most rational explanation of the Breton wars of Fulk the Good, Geoffrey Greygown and Fulk Nerra, which ended in Anjou’s recovery of the debateable ground. If it is correct, there is an end at once of the “bipartite county” and of Count Ingelger; “the other count” cannot have ruled west of the Mayenne, therefore he must have ruled east of it, and there is no room for any one else.

The one writer whose testimony seems to lend some countenance to that of the _Gesta_ need not trouble us much. Fulk Rechin (Marchegay, _Comtes_, p. 374) does call Ingelger the first count; but his own confession that he knew nothing about his first five ancestors beyond their names gives us a right to think, in the absence of confirmatory evidence, that he may have been mistaken in using the title. He says nothing about the county having ever been bipartite, and his statement that his forefathers received their honours from Charles the Bald, not from the house of Paris (_ib._ p. 376), may be due to the same misconception, strengthened by a desire, which in Fulk Rechin would be extremely natural, to disclaim all connexion with the “genus impii Philippi,” or even by an indistinct idea of the investiture of Fulk I. For, if this is regarded as having taken place between 905 and 909, it must fall in the reign of Charles the Simple, and might be technically ascribed to him, though there can be no doubt that it was really owing to the duke of the French. Every step of Fulk’s life, as we can trace it in the charters, shows him following closely in the wake of Odo, Robert and Hugh; and the dependance of Anjou on the duchy of France is distinctly acknowledged by his grandson.

The latter part of the account of Ingelger in the _Gesta_ (Marchegay, _Comtes_, pp. 47–62) is copied bodily from the _Tractatus de reversione B. Martini a Burgundiâ_, which professes to have been written by S. Odo of Cluny at the request of his foster-brother, Count Fulk the Good. The wild anachronisms of this treatise have been thoroughly exposed by its latest editor, M. A. Salmon (_Supplément au Recueil des Chroniques de Touraine_, pp. xi–xxviii), and M. Mabille (“Les Invasions normandes dans la Loire et les pérégrinations du corps de S. Martin,” in _Bibl. de l’Ecole des Chartes_, ser. vi. vol. v. pp. 149–194). It is certain, from the statement of S. Odo’s own biographer John, that the saint was born in 879 and entered religion in 898; at which time it is evident that Fulk the Good, the Red Count’s youngest son, must have been quite a child, if even he was in existence at all. The letters in which he and the abbot address each other as foster-brothers are therefore forgeries; and the treatise which these letters introduce is no better. The only part of it which directly concerns our present subject is the end, recounting how the body of the Apostle of the Gauls, after a thirty years’ exile at Auxerre, whither it had been carried to keep it safe from the sacrilegious hands of Hrolf and his northmen when they were ravaging Touraine, was brought back in triumph to its home at Tours on December 13, 887, by Ingelger, count of Gâtinais and Anjou, and grandson of Hugh, duke of Burgundy. Now there is no doubt at all that the relics of S. Martin were carried into Burgundy and afterwards brought back again, and that the feast of the Reversion of S. Martin on December 13 was regularly celebrated at Tours in commemoration of the event; but the whole history of the adventures of the relics as given in this treatise is manifestly wrong in its details; _e.g._ the statements about Hrolf are ludicrous--the “reversion” is said to have taken place after his conversion. M. Salmon has gone carefully through the whole story: M. Mabille has sifted it still more thoroughly. These two writers have shewn that the body of S. Martin really went through a great many more “peregrinations” than those recounted in the Cluny treatise, that the real date of the reversion is 885, and in short that the treatise is wrong in every one of its dates and every one of the names of the bishops whom it mentions as concerned in the reversion, save those of Archbishop Adaland of Tours and his brother Raino, who, however, was bishop of Angers, not of Orléans as the treatise says. The passages in the Tours chronicle where Ingelger is described as count of Anjou are all derived from this source, and therefore prove nothing, except the writer’s ignorance about counts and bishops alike.

The mention of Archbishop Adaland brings us to another subject--Ingelger’s marriage. Ralf de Diceto (Stubbs, vol. i. p. 139) says that he married Ælendis, niece of Archbishop Adaland and of Raino, bishop of Angers, and that these two prelates gave to the young couple their own hereditary estates at Amboise, in Touraine and in the Orléanais. The _Gesta Consulum_ (Marchegay, _Comtes_, p. 45) say the same, but afterwards make Raino bishop of Orléans. This story seems to be a bit of truth which has found its way into a mass of fiction; at any rate it is neither impossible nor improbable. The author of the _De Reversione_ is quite right in saying that Archbishop Adaland died shortly after the return of the relics; his statement, and those of the Tours Chronicle, that Adaland was consecrated in 870 and died in 887, are borne out by the same charters which enable us to track the career of Fulk the Red. As to Raino--there was a Raino ordained bishop of Angers in 881 (Chron. Vindoc. ad ann. in Marchegay, _Eglises d’Anjou_, p. 160). The version which makes Orléans his see is derived from the false Cluny treatise.

Fulk the Red was witnessing charters in 886 and died in 941 or 942. He must have been born somewhere between 865 and 870; as the traditional writers say he died “senex et plenus dierum, in bonâ senectute,” it may have been nearer the earlier date. There is thus no chronological reason why these two prelates should not have been his mother’s uncles; and as the house of Anjou certainly acquired Amboise somehow, it may just as well have been in this way as in any other.

NOTE B.

THE PALACE OF THE COUNTS AT ANGERS.

Not only ordinary English tourists, but English historical scholars have been led astray in the topography of early Angers by an obstinate local tradition which long persisted in asserting that the counts and the bishops of Angers had at some time or other made an exchange of dwellings; that the old ruined hall within the castle enclosure was a piece of Roman work, and had served, before this exchange, as the synodal hall of the bishops. The date adopted for this exchange, when I visited Angers in 1877 (I have no knowledge of the place since that time) was “the ninth century”; some years before it was the twelfth or thirteenth century, and the synodal hall of the present bishop’s palace, with its undercroft, was shown and accepted as the home of all the Angevin counts down to Geoffrey Plantagenet at least. The whole history of the two palaces--that of the counts and that of the bishops--has, however, been cleared up by two local archæologists, M. de Beauregard (“Le Palais épiscopal et l’Eglise cathédrale d’Angers,” in _Revue de l’Anjou et de Maine-et-Loire_, 1855, vol. i. pp. 246–256), and M. d’Espinay, president of the Archæological Commission of Maine-et-Loire (“Le Palais des Comtes d’Anjou,” _Revue historique de l’Anjou_, 1872, vol. viii. pp. 153–170; “L’Evêché d’Angers,” _ib._ pp. 185–201). The foundation and result of their arguments may be briefly summed up. The first bit of evidence on the subject is a charter (printed by M. de Beauregard, _Revue de l’Anjou et de Maine-et-Loire_, as above, vol. i. pp. 248, 249; also in _Gallia Christiana_, vol. xiv. instr. cols. 145, 146) of Charles the Bald, dated July 2, 851, and ratifying an exchange of lands between “Dodo venerabilis Andegavorum Episcopus et Odo illustris comes.” The exchange is thus described:--“Dedit itaque præfatus Dodo episcopus antedicto Odoni comiti, ex rebus matris ecclesiæ S. Mauricii, æquis mensuris funibusque determinatam paginam terræ juxta murum civitatis Andegavensis, in quâ opportunitas jam dicti comitis mansuræ sedis suorumque successorum esse cognoscitur. Et, e contra, in compensatione hujus rei, dedit idem Odo comes ex comitatu suo terram S. Mauricio æquis mensuris similiter funibus determinatam prænominato Dodoni episcopo successoribusque suis habendam in quâ predecessorum suorum comitum sedes fuisse memoratur.” As M. de Beauregard points out, the traditionary version--whether placing the exchange in the ninth century or in the twelfth--is based on a misunderstanding of this charter. The charter says not a word of the bishop giving up his own actual abode to the count; it says he gave a plot of ground near the city wall, and suitable for the count to build himself a house upon. Moreover the words “sedes fuisse memoratur” seem to imply that what the count gave was not his own present dwelling either, but only that which had been occupied by his predecessors. There can be little doubt that the Merovingian counts dwelt on the site of the Roman citadel of Juliomagus; and this was unquestionably where the bishop’s palace now stands. That it already stood there in the closing years of the eleventh century is proved by a charter, quoted by M. d’Espinay (_Revue historique de l’Anjou_, vol. viii. p. 200, note 2) from the cartulary of S. Aubin’s Abbey, giving an account of a meeting held “in domibus episcopalibus _juxta S. Mauricium Andegavorum matrem ecclesiam_,” in A.D. 1098.

So much for the position of the bishop’s dwelling from 851 downwards. Of the position of the count’s palace--the abode of Odo and his successors, built on the piece of land near the city wall--the first indication is in an account of a great fire at Angers in 1132: “Flante Aquilone, accensus est in mediâ civitate ignis, videlicet apud S. Anianum; et tanto incendio grassatus est ut ecclesiam S. Laudi et omnes officinas, deinde comitis aulam et omnes cameras miserabiliter combureret et in cinerem redigeret. Sicque per Aquariam descendens,” etc. (Chron. S. Serg. a. 1132, Marchegay, _Eglises_, p. 144). The church of S. Laud was the old chapel of S. Geneviève,--“capella B. Genovefæ virginis, infra muros civitatis Andegavæ, ante forum videlicet comitalis aulæ posita,” as it is described in a charter of Geoffrey Martel (_Revue Hist. de l’Anjou_, 1872, vol. viii. p. 161)--the exact position of a ruined chapel which was still visible, some twenty years ago, within the castle enclosure, not far from the hall which still remains. A fire beginning in the middle of the city and carried by a north-east wind down to S. Laud and the Evière would not touch the present bishop’s palace, but could not fail to pass over the site of the castle. The last witness is Ralf de Diceto (Stubbs, vol. i. pp. 291, 292), who distinctly places the palace of the counts in his own day--the day of Count Henry Fitz-Empress--in the south-west corner of the city, with the river at its feet and the vine-clad hills at its back; and his description of the “thalami noviter constructi” just fits in with the account of the fire, the destruction thereby wrought having doubtless been followed by a rebuilding on a more regal scale. It seems impossible to doubt the conclusion of these Angevin archæologists, that the dwelling of the bishops and the palace of the counts have occupied their present sites ever since the ninth century. In that case the present synodal hall, an undoubted work of the early twelfth century, must have been originally built for none other than its present use; and to a student of the history of the Angevin counts and kings the most precious relic in all Angers is the ruined hall looking out upon the Mayenne from over the castle ramparts. M. d’Espinay denies its Roman origin; he considers it to be a work of the tenth century or beginning of the eleventh--the one fragment, in fact, of the dwelling-place of Geoffrey Greygown and Fulk the Black which has survived, not only the fire of 1132, but also the later destruction in which the apartments built by Henry have perished.

NOTE C.

THE MARRIAGES OF GEOFFREY GREYGOWN.

The marriages of Geoffrey Greygown form a subject at once of some importance and of considerable difficulty. It seems plain that Geoffrey was twice married, that both his wives bore the same name, Adela or Adelaide, and that the second was in her own right countess of Chalon-sur-Saône, and widow of Lambert, count of Autun. There is no doubt about this second marriage, for we have documentary evidence that a certain Count Maurice (about whom the Angevin writers make great blunders, and of whom we shall hear more later on) was brother at once to Hugh of Chalon, son of Lambert and Adela, and to Fulk, son of Geoffrey Greygown, and must therefore have been a son of Geoffrey and Adela. A charter, dated between 992 and 998 (see Mabille, Introd. _Comtes_, pp. lxx–lxxi), wherein Hugh, count of Chalon, describes himself as “son of Adelaide and Lambert who was count of Chalon in right of his wife,” is approved by “Adelaide his mother and Maurice his brother.” Now as R. Glaber (l. iii. c. 2; _Rer. Gall. Scriptt._, vol. x. p. 27) declares that Hugh had no brother, Maurice must have been his half-brother, _i.e._ son of his mother and her second husband; and that that second husband was Geoffrey Greygown appears by several charters in which Maurice is named as brother of Fulk Nerra.

It is by no means clear who this Adela or Adelaide of Chalon was. Perry (_Hist. de Chalon-sur-Saône_, p. 86) and Arbois de Jubainville (_Comtes de Champagne_, vol. i. p. 140) say she was daughter of Robert of Vermandois, count of Troyes, and Vera, daughter of Gilbert of Burgundy and heiress of Chalon, which at her death passed to Adela as her only child. But the only authority for this Vera, Odorannus the monk of S. Peter of Sens, says she was married in 956, and Lambert called himself count of Chalon in 960 (Perry, _Hist. Chalon, preuves_, p. 35. See also Arbois de Jubainville as above), so that if he married Vera’s daughter he must have married a child only three years old. And to add to the confusion, Robert of Troyes’s wife in 959 signs a charter by the name of “Adelais” (Duchesne, _Maison de Vergy, preuves_, p. 36). What concerns us most, however, is not Adela’s parentage, but the date of her marriage with Geoffrey Greygown; or, which comes to much the same thing, the date of her first husband’s death. The cartulary of Paray-le-Monial (Lambert’s foundation) gives the date of his death as February 22, 988. If that were correct, Geoffrey, who died in July 987, could not have married Adela at all, unless she was divorced and remarried during Lambert’s life. This idea is excluded by a charter of her grandson Theobald, which distinctly says that Geoffrey married her after Lambert’s death (Perry, _Hist. Chalon, preuves_, p. 39); therefore the _Art de vérifier les Dates_ (vol. xi. p. 129) proposes to omit an x and read 978. Adela and Geoffrey, then, cannot have married earlier than the end of 978. Geoffrey, however, must have been married long before this, if his daughter Hermengard was married in 970 to Conan of Britanny (Morice, _Hist. Bret._, vol. i. p. 63. His authority seems to be a passage in the Chron. S. Michael. a. 970, printed in Labbe’s _Bibl. Nova MSS. Librorum_, vol. i. p. 350, where, however, the bride is absurdly made a daughter of Fulk Nerra instead of Geoffrey Greygown). And in Duchesne’s _Maison de Vergy, preuves_, p. 39, is the will, dated March 6, 974, of a Countess Adela, wife of a Count Geoffrey, whereby she bequeathes some lands to S. Aubin’s Abbey at Angers; and as the Chron. S. Albin. a. 974 (Marchegay, _Eglises_, p. 20) also mentions these donations, there can be little doubt that she was the wife of Geoffrey of Anjou. M. Mabille (Introd. _Comtes_, p. lxx) asserts that this Adela, Geoffrey Greygown’s first wife, was Adela of Vermandois, sister of Robert of Troyes, and appeals to the will above referred to in proof of his assertion; the will, however, says nothing of the sort. He also makes the second Adela sister-in-law instead of daughter to Robert (_ib._ p. lxxi). It seems indeed hopeless to decide on the parentage of either of these ladies; that of their children is, however, the only question really important for us. Hermengard, married in 970 to the duke of Britanny, was clearly a child of Geoffrey’s first wife; Maurice was as clearly a child of the second; but whose child was Fulk the Black? Not only is it a matter of some interest to know who was the mother of the greatest of the Angevins, but it is a question on whose solution may depend the solution of another difficulty:--the supposed, but as yet unascertained, kindred between Fulk’s son Geoffrey Martel and his wife Agnes of Burgundy. If Fulk was the son of Geoffrey Greygown and Adela of Chalon, the whole pedigree is clear, and stands thus:

1 2
Lambert = Adela = Geoffrey
| |
Adalbert = Gerberga Fulk
of Lombardy | |
| |
Otto William |
| |
Agnes = Geoffrey.

The two last would thus be cousins in the third degree of kindred according to the canon law. The only apparent difficulty of this theory is that it makes Fulk so very young. The first child of Adela of Chalon and Geoffrey cannot have been born earlier than 979, even if Adela remarried before her first year of widowhood was out; and we find Fulk Nerra heading his troops in 992, if not before. But the thing is not impossible. Such precocity would not be much greater than that of Richard the Fearless, or of Fulk’s own rival Odo of Blois; and such a wonderful man as Fulk the Black may well have been a wonderful boy.

NOTE D.

THE BRETON AND POITEVIN WARS OF GEOFFREY GREYGOWN.

The acts of Geoffrey Greygown in the _Gesta Consulum_ are a mass of fable. The fight with the Dane Æthelwulf and that with the Saxon Æthelred are mythical on the face of them, and the writer’s habitual defiance of chronology is carried to its highest point in this chapter. From him we turn to the story of Fulk Rechin. “Ille igitur Gosfridus Grisa Gonella, pater avi mei Fulconis, cujus probitates enumerare non possumus, excussit Laudunum de manu Pictavensis comitis, et in prœlio superavit eum super Rupes, et persecutus est eum usque ad Mirebellum. Et fugavit Britones, qui venerant Andegavim cum prædatorio exercitu, quorum duces erant filii Isoani (Conani). Et postea fuit cum duce Hugone in obsidione apud Marsonum, ubi arripuit eum infirmitas quâ exspiravit; et corpus illius allatum est Turonum et sepultum in ecclesiâ B. Martini” (Fulk Rechin, Marchegay, _Comtes_, p. 376).

Whoever was the author of this account, he clearly knew or cared nothing about the stories of the monkish writers, but had a perfectly distinct source of information unknown to them. For their legends he substitutes two things: a war with the count of Poitou, and a war with the duke of Britanny. On each of these wars we get some information from one other authority; the question is how to make this other authority tally with Fulk.

1. As to the Breton war, which seems to be the earlier in date.

No one but Fulk mentions the raid of Conan’s sons upon Angers; and M. Mabille (Introd. _Comtes_, p. xlviii) objects to it on the ground that Conan’s sons were not contemporaries of Geoffrey.

Conan of Rennes was killed in 992 in a battle with Geoffrey’s son. He had been married in 970 to Geoffrey’s daughter Hermengard (see above, pp. 121, 135). Now a daughter of Geoffrey in 970 must have been almost a child, but it by no means follows that her husband was equally young. On the contrary, he seems to have been sufficiently grown up to take a part in politics twenty years before (Morice, _Hist. Bret._ vol. i. p. 62). It is certain that he had several sons; it is certain that two at least of them were not Hermengard’s; it is likely that none of them were, except his successor Geoffrey. Supposing Conan was somewhat over fifty when killed (and he may have been older still) that would make him about thirty when he married Hermengard; he might have had sons ten years before that, and those sons might very easily head an attack upon their stepmother’s father in 980 or thereabouts. Surely M. Mabille here makes a needless stumbling-block of the chronology.

If no other writer confirms Fulk’s story, neither does any contradict it. But in the _Gesta Consulum_ (Marchegay, _Comtes_, pp. 91–93) an exactly similar tale is told, only in much more detail and with this one difference, that Fulk Nerra is substituted for Geoffrey Greygown, and the raid is made to take place just before that other battle of Conquereux, in 992, in which Conan perished. The only question now is, which date is the likeliest, Fulk’s or John’s? in other words, which of these two writers is the better to be trusted? Surely there can be no doubt about the choice, and we must conclude that, for once, the monk who credits Greygown with so many exploits that he never performed has denied him the honour of one to which he is really entitled.

Fulk Rechin’s account of Geoffrey’s Breton war ends here. The Breton chroniclers ignore this part of the affair altogether; they seem to take up the thread of the story where the Angevin drops it. It is they who tell us of the homage of Guerech, and of the battle of Conquereux; and their accounts of the latter are somewhat puzzling. The Chron. Britann. in Lobineau (_Hist. Bret._, vol. ii. col. 32) says: “982. Primum bellum Britannorum et Andegavorum in Concruz.” The Chron. S. Michael. (Labbe, _Bibl. Nova_, vol. i. p. 350; _Rer. Gall. Scriptt._, vol. ix. p. 98) says: “981. Conanus Curvus contra Andegavenses in Concurrum optime pugnavit.” But in the other two Breton chronicles the Angevins do not appear. The Chron. Namnetense (_Rer. Gall. Scriptt._, vol. viii. p. 278) describes the battle as one between Conan and Guerech; the Chron. Briocense (Morice, _Hist. Bret., preuves_, vol. i. col. 32) does the same, and moreover adds that Conan was severely wounded in the right arm and fled defeated. This last is the only distinct record of the issue of the battle; nevertheless there are some little indications which, taken together, give some ground for thinking its record is wrong. 1st. There is the negative evidence of the silence of the Angevin writers about the whole affair; they ignore the first battle of Conquereux as completely as the Bretons ignore the unsuccessful raid of Conan’s sons. This looks as if each party chronicled its own successes, and carefully avoided mentioning those of its adversaries. 2d. In the _Hist. S. Flor. Salm._ (Marchegay, _Eglises_, p. 260) is a proverb “Bellum Conquerentium quo tortum superavit rectum”--an obvious pun on Conan’s nickname, “Tortus” or “Curvus.” It is there quoted as having arisen from the battle of Conquereux in 992--the only one which it suits the Angevin writers to admit. But this is nonsense, for the writer has himself just told us that in that battle Conan was defeated and slain. Therefore “the crooked overcame the straight,” _i.e._ Conan won the victory, in an earlier battle of Conquereux.