Part 28

Hugh of Lincoln was the universally-acknowledged leader of the English Church in all matters of religion and morals; he had exercised in Henry II.’s later years such an influence over the king as no one, except perhaps Thomas Becket, had ever possessed; the whole Church and nation reverenced him as it had never reverenced any man since the death of S. Anselm. When he took up the position of Thomas and Anselm as a champion of constitutional liberty, the victory was sure. Strangely enough, his action seems to have taken the primate completely by surprise. For a moment Hubert stood speechless; then he turned to Bishop Herbert of Salisbury, and with quivering lips asked what he was minded to do for the king’s assistance. As a son of Richard of Ilchester and a kinsman of the great ministerial house founded by Roger of Salisbury,[1788] Herbert represented the traditions of an old and venerated political school, as Hugh represented those of the best school of ecclesiastics. The statesman’s reply was an echo of the saint’s: “It seems to me that, without grievous wrong to my church, I can neither do nor say aught but what I have heard from my lord of Lincoln.” The justiciar, hurling a torrent of reproaches at Hugh, broke up the assembly, and wrote to the king that his plan had been foiled through Hugh’s opposition.[1789] Richard in a fury ordered the property of the two recalcitrant bishops to be confiscated; in the case of Salisbury this was done, but no Englishman dared lay a finger on anything belonging to the saint of Lincoln, “for they feared his curse like death itself.” In vain did the king reiterate his command, till at last his own officers begged Hugh to put an end to the scandal by making his peace, for their sakes if not for his own; Hugh therefore went to seek Richard in Normandy, and literally forced him into a reconciliation on S. Augustine’s day. Herbert, on the other hand, had to purchase his restoration at a heavy price;[1790] but the king and his justiciar were none the less completely beaten. The death of Rees Ap-Griffith and a dispute between his sons for the succession in South Wales gave Hubert an opportunity of renewing his fading laurels by a brilliant expedition to the Welsh marches, where he succeeded in restoring tranquillity and securing the border-fortresses for the king.[1791] He had however scarcely had time to recover from his political defeat before he was overwhelmed by the bursting of an ecclesiastical storm which had long been hanging over his head. Pope Celestine died on January 8, 1198. On the morrow the cardinals elected as his successor a young deacon named Lothar, who took the name of Innocent III., and began at once to sweep away the abuses of the Roman court and to vindicate the rights of his see against the Roman aristocracy with a promptness and vigour which were an earnest of his whole future career.[1792] The monks of Canterbury lost no time in sending to the new Pope their list of grievances against their primate; and at the head of the list they set a charge which, in the eyes of such a pontiff as Innocent, could admit of no defence. Hubert, said they, had violated the duties and the dignity of his order by becoming the king’s justiciar, acting as a judge in cases of life and death, and so entangling himself in worldly business that he was incapable of paying due attention to the government of the Church. Innocent immediately wrote to the king, charging him, if he valued his soul’s health, not to suffer either the archbishop of Canterbury or any other priest to continue in any secular office; and at the same time he solemnly forbade the acceptance of any such office by any bishop or priest throughout the whole Church. Discredited as Hubert now was in the eyes of all parties, he had no choice but to resign, and this time Richard had no choice but to accept his resignation.[1793]

[1788] On Herbert’s antecedents and connexions see Stubbs, _Rog. Howden_, vol. iv. pref. p. xci, note 4.

[1789] _Mag. Vita S. Hug._ (Dimock), p. 250. Cf. the brief account in Rog. Howden (Stubbs), vol. iv. p. 40.

[1790] _Mag. Vita S. Hug._ (Dimock), p. 251.

[1791] On Rees’s death his two sons quarrelled over the succession, and Hubert had to go to the “fines Gwalliæ” and make peace between them. Rog. Howden (Stubbs), vol. iv. p. 21. At Christmas he was at Hereford, where he took the castle into his own hands, turning out its custodians and putting in new ones, “ad opus regis”; he did the same at Bridgenorth and Ludlow. _Ib._ p. 35. See also Gerv. Cant. (Stubbs), vol. i. p. 543, Gerald’s letter to Hubert after his victory, and Hubert’s reply: Gir. Cambr. _De Rebus a se gestis_, l. iii. cc. 5, 6 (Brewer, vol. i. pp. 96–102).

[1792] Rog. Howden as above, pp. 41–44.

[1793] _Ib._ pp. 47, 48.

The last few months of his justiciarship were however occupied with the projection, if not the execution, of a measure of great constitutional importance. Early in the spring he had, in his master’s name, laid upon England a carucage to the amount of five shillings upon every carucate or ploughland. The great increase in the rate of taxation, as compared with that of 1194, was not unjustifiable; for since that year the socage-tenants, on whom the impost fell, had paid no direct taxes at all, while two scutages had been exacted from the tenants-in-chivalry. But a far more important change was made in the assessment of the new impost. Until now, the carucate, like the hide, had been a term of elastic significance. It represented, as the literal meaning of the word implied, the extent of land which could be cultivated by a single plough; and this of course varied in different parts of the country according to the nature of the soil, and the number and strength of the plough-team. In general, however, a hundred acres seem to have been reckoned as the average extent both of the carucate and of the hide. In order to avoid the endless complications and disputes which under the old system had made the assessment of the land-tax a matter of almost more trouble than profit, Hubert Walter adopted this average as a fixed standard, and ordered that henceforth, for purposes of taxation, the word “carucate” should represent a hundred acres. It followed as a necessary consequence that the whole arable land of England must be re-measured. The old customary reckoning of hides, based upon the Domesday survey, would no longer answer its purpose: the venerable rate-book which had been in use for more than a hundred years, partially superseded since 1168 by the Black Book of the Exchequer, was now to be superseded entirely. Hubert therefore issued in the king’s name a commission for what was virtually a new Domesday survey. Into every shire he sent a clerk and a knight, who, together with the sheriff and certain lawful men chosen out of the shire, were, after swearing that they would do the king’s business faithfully, to summon before them the stewards of the barons of the county, the lord or bailiff of every township and the reeve and four lawful men of the same, whether free or villein, and two lawful knights of the hundred; these persons were to declare upon oath what ploughlands there were in every township--how many in demesne, how many in villenage, how many in alms, and who was responsible for these last. The carucates thus ascertained were noted in a roll of which four copies were kept, one by each of the two royal commissioners, one by the sheriff, and the other divided among the stewards of the local barons. The collection of the money was intrusted to two lawful knights and the bailiff of every hundred; these were responsible for it to the sheriff; and the sheriff had to see that it agreed with his roll, and to pay it into the Exchequer. Stern penalties were denounced against witnesses, whether free or villein, who should be detected in trying to deceive the commissioners. No land was to be exempted from the tax, except the free estates belonging to the parish churches, and lands held of the king by serjeanty or special service; even these last, however, were to be included in the survey, and their holders were required to come and prove their excuses at its conclusion, in London at the octave of Pentecost.[1794]

[1794] Rog. Howden (Stubbs), vol. iv. pp. 46, 47.

This was Hubert’s last great administrative act, and it had a far more important significance than he himself probably knew. In form, the application of the process of jury-inquest to the assessment of an impost on the land was only a return to the precedent of Domesday itself. In reality, however, it was something much more important than this. The jury-inquest had been introduced by the Conqueror in 1086 under exceptional circumstances, and for an exceptional purpose which could be attained by no other means. So far as its original use was concerned, the precedent had remained a wholly isolated one for more than a hundred years. But during those years the principle which lay at the root of the jury-inquest had made its way into every branch of legal, fiscal and judicial administration. It had been applied to the purposes of private litigation by the Great Assize, to the determination of individual liability to military duty by the Assize of Arms, to the assessment of taxation on personal property by the ordinance of the Saladin tithe; it had penetrated the whole system of criminal procedure through the Assizes of Clarendon and Northampton; and it had gained a yet fuller recognition in the judicial ordinances of 1194. Viewed in this light, its application to the assessment of taxation on real property was another highly important step in the extension of its sphere of work. But this was not all. The chief value of the jury-system lay in its employment of the machinery of local representation and election, whereby it was a means of training the people to the exercise of constitutional self-government. The commission of 1198 shews that, although doubtless neither rulers nor people were conscious of the fact, this training had now advanced within measurable distance of its completion. The machinery of the new survey was not identical with that used in 1086. The taxpayers were represented, not only by the witnesses on whose recognition the assessment was based, but by the “lawful men chosen out of the shire” who took their place side by side with the king’s officers as commissioners for the assessment, and by the bailiff and two knights of the hundred who were charged with the collection of the money. The representative principle had now reached its furthest developement in the financial administration of the shire. Its next advance must inevitably result in giving to the taxpayers a share in the determination, first of the amount of the impost, and then of the purposes to which it should be applied, by admitting them, however partially and indirectly, to a voice in the great council of the nation.[1795]

[1795] On this “Great Carucage” see Stubbs, _Constit. Hist._, vol. i. pp. 510, 511, and pref. to Rog. Howden, vol. iv. pp. xci–xcv.

We must not credit Hubert Walter with views so lofty or so far-reaching as these. The chief aim of his policy doubtless was to get for his master as much money as he could, although he would only do it by what he regarded as just and constitutional methods. Unluckily the commissioners’ report is lost, and there is not even any proof that it was ever presented; for before Whitsuntide the new Pope’s views had become known, and on July 11 a royal writ announced Hubert’s retirement from the justiciarship and the appointment of Geoffrey Fitz-Peter in his stead.[1796] Like Hubert, Geoffrey Fitz-Peter came of a family which had long been engaged in administrative work. His elder brother Simon had in Henry’s early years filled the various offices of sheriff, justice-in-eyre, and king’s marshal.[1797] Geoffrey himself had been sheriff of Northampton throughout the last five years of Henry’s reign, and had during the same period acted occasionally as an ordinary justice of assize, and more frequently as a judge of the forest-court.[1798] In 1189 Richard appointed him one of the assistant-justiciars, and in this capacity he supported Walter of Rouen in the affair of William of Longchamp’s deposition.[1799] In the early days of William’s rule, however, Geoffrey had made use of the latter’s influence to secure for himself the whole English inheritance of the earl of Essex, William de Mandeville, upon which his wife had a distant claim.[1800] Such a man was likely to be controlled by fewer scruples, as well as hampered by fewer external restraints, than those which had beset the justiciar-archbishop; and in truth, before the year was out, both clergy and people had cause to regret the change of ministers. Some of the religious orders refused to pay their share of the carucage; their refusal was met by a royal edict declaring the whole body of clergy, secular as well as monastic, incapable of claiming redress for any wrongs inflicted on them by the laity, while for any injury done by a clerk or a monk to a layman satisfaction was exacted to the uttermost farthing. The archbishop of Canterbury could hardly have published what was virtually a decree of outlawry against his own order; the new justiciar published it seemingly without hesitation, and the recalcitrant monks were compelled to submit.[1801] This act was followed by a renewal of the decree requiring all charters granted under the king’s old seal to be brought up for confirmation under the new one[1802]--a step which seems to imply that Richard’s former command to this effect had not been very strictly enforced by Hubert. Meanwhile three justices-errant, acting on a set of instructions modelled upon those of 1194, were holding pleas of the Crown in the northern shires;[1803] “so that,” says King Henry’s old chaplain Roger of Howden, “with these and other vexations, just or unjust, all England from sea to sea was reduced to penury. And these things were not yet ended when another kind of torment was added to confound the men of the kingdom, through the justices of the forest,” who were sent out all over England to hold a great forest-assize, which was virtually a renewal of that issued by Henry in 1184.[1804]

[1796] Rymer, _Fœdera_, vol. i. p. 71.

[1797] He was sheriff of Northamptonshire, Bedfordshire and Buckinghamshire from 1156 till 1160, and of Northamptonshire again from Michaelmas 1163 till Easter 1170. See the list of sheriffs in index to Eyton’s _Itin. Hen. II._, pp. 337, 339. He appears as marshal in 1165 (Madox, _Form. Angl._, p. xix), and as justice-errant in Bedfordshire, A.D. 1163, in the story of Philip de Broi (above, p. 21).

[1798] Eyton, _Itin. Hen. II._, list of sheriffs, p. 339; _ib._ pp. 265, 273, 281, 291, 298. Pipe Roll I. Ric. I. (Hunter) _passim_.

[1799] Rog. Howden (Stubbs), vol. iii. pp. 16, 28, 96, 153.

[1800] Stubbs, _Rog. Howden_, vol. iii., pref. p. xlviii, note 6.

[1801] Rog. Howden (Stubbs), vol. iv. p. 66.

[1802] _Ibid._ Mat. Paris, _Chron. Maj._ (Luard), vol. ii. p. 451. Ann. Waverl. a. 1198 (Luard, _Ann. Monast._, vol. ii. p. 251).

[1803] Instructions in Rog. Howden (as above), pp. 61, 62. The judges were Hugh Bardulf, Roger Arundel and Geoffrey Hacket; they held pleas in Lincolnshire, Nottinghamshire, Derbyshire, Yorkshire, Northumberland, Westmoreland, Cumberland and Lancashire.

[1804] Rog. Howden (as above), pp. 62–66.

Stern and cruel, however, as was the administration of the last eight months of Richard’s reign, it was still part of a salutary discipline. The milder chastenings which Richard’s English subjects had endured from Hubert Walter, the scorpion-lashes with which he chastised them by the hands of Geoffrey Fitz-Peter, were both alike stages in the training which Richard’s predecessor had begun, and whose value they were to learn when left face to face with the personal tyranny of his successor. For nearer at hand than they could dream was the day when English people and Angevin king were to stand face to face indeed, more closely than they had ever stood before. The nine generations of increasing prosperity promised to Fulk the Good were all numbered and fulfilled, and with their fulfilment had come the turn of the tide. The power of the Angevins had reached its destined limit, and had begun to recede again. From the sacred eastern land all trace of it was already swept away; in the west it was, slowly indeed as yet, but none the less surely falling back. Five years were still to pass before the tide should be fairly out; then it was to leave the Good Count’s heir stranded, not on the black rock of Angers, but on the white cliffs of England.

Richard had spent the first half of his reign in fighting for a lost cause in Palestine; he spent the other half in fighting for a losing cause in Gaul. The final result of the long series of conquests and annexations whereby the Angevin counts, from Fulk the Red to Henry Fitz-Empress, had been enlarging their borders for more than two hundred years, had been to bring them into direct geographical contact and political antagonism with an enemy more formidable than any whom they had yet encountered. In their earliest days the king of the French had been their patron; a little later, he had become their tool. Now, he was their sole remaining rival; and ere long he was to be their conqueror. Since the opening of the century, a great change had taken place in the political position of the French Crown; a change which was in a considerable measure due to the yet greater change in the position of the Angevin house. When Louis VI. came to the throne in 1109, he found the so-called “kingdom of France” distributed somewhat as follows. The western half, from the river Somme to the Pyrenees, was divided between four great fiefs--Normandy, Britanny, Anjou and Aquitaine. Four others--Champagne, Burgundy, Auvergne and Toulouse--covered its eastern portion from the river Meuse to the Mediterranean Sea; another, Flanders, occupied its northernmost angle, between the sources of the Meuse, the mouth of the Scheld, and the English Channel. The two lines of great fiefs were separated by an irregular group of smaller territories, amid which lay, distributed in two very unequal portions, the royal domain. Its northern and larger half, severed from Flanders by the little counties of Amiens and Vermandois, was flanked on the east by Champagne and on the north-west by Normandy, while its south-western border was ringed in by the counties of Chartres, Blois and Sancerre, which parted it from Anjou, and which were all linked together with Champagne under the same ruling house. Southward, in the upper valleys of the Loire and the Cher, a much smaller fragment of royal domain, comprising the viscounty of Bourges and the territory afterwards known as the Bourbonnais, lay crowded in between Auvergne, the Aquitanian district of Berry, and the Burgundian counties of Mâcon and Nevers and that of Sancerre, which parted it from the larger royal possessions north of the Loire. The whole domains of the Crown thus covered scarcely more ground than the united counties of Anjou, Touraine and Maine, scarcely so much as the duchy of Normandy. Within these limits, however, Louis VI. had in his twenty-nine years’ reign contrived to establish his absolute authority on so firm a basis that from thenceforth the independence of the Crown was secured. To destroy that of the great feudataries, and to bring them one by one into a subjection as absolute as that of the royal domain itself, was the work which he bequeathed to his successors.

[Illustration: Map VII.

FRANCE AND THE ANGEVIN DOMINIONS.

To illustrate the wars of Richard and John with Philip Augustus.

Key: _Royal Domain of Philip. A. D. 1194._

Wagner & Debes’ Geogˡ. Estabᵗ. Leipsic. London, Macmillan & Co. ]

We may set aside the temporary annexation of Aquitaine through the marriage of Louis VII. and Eleanor as forming no part of this process of absorption. In the plans of Louis VI. it was doubtless meant to be a very important part; but as a matter of fact, its historical importance proved to be of a wholly different kind. The marriage of Louis and Eleanor contributed to the final acquisition of Guienne and Gascony by the French Crown not a whit more than the marriage of Geoffrey Martel and Agnes had contributed to their acquisition by the house of Anjou. The Parisian king, like his Angevin follower of old, had work to do on his own side of the Loire before he might safely attempt the conquest of the south. By the middle of the century, the map of Gaul had undergone a marked transformation. Its eastern and central portions indeed remained unchanged; but the western half was utterly metamorphosed. Its four great divisions had been virtually swept away, and the whole land had become Angevin. In face of this altered state of things, the remaining powers of northern Gaul were of necessity driven into union, as a counterpoise to this enormous growth of Anjou; and the only possible centre of union, alike in a political and a geographical point of view, was the king of the French. He alone could claim to match in rank and dignity the crowned masters of the west; and under his leadership alone was it possible to face them all along the line from the mouth of the Somme to the source of the Cher with a front as unbroken as their own. The old Angevin march had ceased to be a marchland at all; its original character was now transferred to the counties of Chartres and Blois; while to north and south of these, from Nonancourt to Aumale and along the whole course of the Cher above Vierzon, the royal domain itself was the sole bulwark of north-eastern Gaul against the advancing power of Anjou. To secure Chartres and Blois was the first necessity for the king: but their counts needed his protection even more than he needed their fidelity, for the whole width of his domains parted them from Champagne, where the bulk of their strength lay. Accordingly Louis VII., by the matrimonial alliances which he formed first for his daughters and lastly for himself with the house of Blois and Champagne, easily succeeded in binding them to a community of personal interests with the royal house of France, whereby their subservience to the French Crown was for the future secured. The chain was too strong to be broken by the boyish wilfulness of Philip Augustus; and from the moment of his reconciliation with his mother and uncles in 1180, the whole military and political strength of Blois, Chartres and Champagne may be reckoned at his command as unreservedly as that of his own immediate domains.

Since that time, the royal power had made an important advance to the northward. At the opening of Philip’s reign the dominions of the count of Flanders stretched from the Channel to the borders of Champagne, covered the whole northern frontier of the royal domain, and touched that of Normandy at its junction with Ponthieu. Twelve years later, more than half this territory had passed, either by cession or by conquest, into the hands of the king. Vermandois was given up to him in 1186; and in 1191 the death of the Flemish count Philip made him master of all Flanders south of the river Lys, which had been promised to him as the dowry of his first queen, Elizabeth of Hainaut, niece of the dead count and daughter of his successor.[1805] This was in several respects a most valuable acquisition. Not only did it bring to the Crown a considerable accession of territory, including the whole upper valley of the Somme, the famous fortress of Péronne, and the flourishing towns of Amiens and Arras; but the power of Flanders, which a few years before had threatened to overshadow every other power in northern Gaul, was completely broken; and the effect upon the political position of Normandy was more important still. While Vermandois and Amiens were in Flemish hands, a league between the Flemish count and the ruler of Normandy would at any moment not only place the whole north-western border of France at their mercy, but would enable them to call in the forces of the imperial Crown to a junction which the French king could have no power to hinder, and which must almost certainly lead to his ruin. Now, on the other hand, such a junction was rendered well-nigh impossible; the whole territory between Normandy, Ponthieu and the German border was in the king’s own hands, and all that was left of Flanders lay in almost complete isolation between the Lys and the sea. In fine, as the dukes of Burgundy had for several generations been obedient followers of their royal kinsmen, now that Blois, Champagne and Vermandois were all secured, the power and influence of the French Crown north of the Loire was fully a match in territorial extent for that of the house of Anjou. South of the Loire the balance was less equal. The extensive possessions of the house of S. Gilles may indeed be left out of both scales; their homage for Toulouse was now secured to the dukes of Aquitaine, but it was a mere formality which left them practically still independent of both their rival overlords. It was indeed at the expense of Toulouse that the Angevin rulers of Poitou had made their last conquest, that of the Quercy. But since then the French king, too, had been gaining territory in Aquitaine; and his gains were made at the expense of the Poitevin duke. Richard had found it needful to buy Philip’s assent to his peaceful entrance upon his ancestral heritage after his father’s death by a renunciation of all claims upon Auvergne and a cession of two important lordships in Berry, Graçay and Issoudun.[1806] The sacrifice was trifling in itself, but it was significant. It marked Richard’s own consciousness that a turning-point had come in the career of his house. Hitherto they had gone steadily forward; now it was time to draw back. The aggressive attitude which had been habitual to the counts of Anjou for nearly three hundred years must be dropped at last. Henceforth they were to stand on the defensive in their turn against the advance of the French Crown.

[1805] See above, p. 234, note 7{1115}.

[1806] Rigord (Duchesne, _Hist. Franc. Scriptt._, vol. v.) p. 29. Will. Armor. _Gesta Phil. Aug._ (_ibid._), p. 75.

It was not the strength of that advance itself which made it so formidable to Richard; it was the knowledge that, side by side with the process of consolidation in France, there had been and still was going on in the Angevin dominions a process of disintegration which his father had been unable to check, and against which he himself was well-nigh helpless. The French monarchy was built up around one definite centre, a centre round which all the subordinate parts of the structure grouped themselves unquestioningly as a matter of course. Paris and its king, even when his practical authority was at the lowest ebb, had always been in theory the accepted rallying-point of the whole kingdom, the acknowledged head of the body politic, none of whose members had ever dreamed of establishing any other in its place. But the empire of Richard Cœur-de-Lion had no centre; or rather, it had three or four rival ones. In Angevin eyes its centre was Angers; in Norman eyes it was Rouen; to the men of the south, it was Poitiers. Even Henry Fitz-Empress had felt at times the difficulty of fulfilling two such opposite parts as those of duke of Normandy and count of Anjou without rousing the jealous resentment of either country against himself as the representative of the other; while as for Britanny and Aquitaine, he had only been able to keep an uncertain hold over them by sheer force, until Britanny was appeased by the marriage of Constance, and Aquitaine subdued by the vigour of Richard. But for Richard in his father’s place the difficulty was far greater. Chafe as they might against the yoke which bound them together--dispute as they might over their respective shares in their common ruler and their respective claims upon him--neither Angevin nor Norman could fail to recognize his own natural sovereign and national representative in the son of Geoffrey and Matilda. But the chances of this recognition being extended to the next generation expired with the young king. If the two Henrys were strangers in Britanny and in Aquitaine, yet on the banks of the Seine, the Loire and the Mayenne they were felt to be at home. But Richard was at home nowhere, though he was master everywhere, from the Solway to the Pyrenees. His Aquitanian subjects for the most part, if they counted him as a fellow-countryman, counted him none the less as an enemy; his subjects north of Loire counted him as a southern stranger. Normans and Angevins still saw in him, as they had been taught to see in him for the first twenty-six years of his life, the representative not of Hrolf and William or of Fulk the Red and Geoffrey Martel, but simply of his mother’s Poitevin ancestors. The Bretons saw in him the son of their conqueror, asserting his supremacy over them and their young native prince only by the right of the stronger. As Suger had laid it down as an axiom, more than half a century ago, that “Englishmen ought not to rule over Frenchmen nor French over English,” so now we begin to discern growing up in Richard’s continental dominions a feeling that Normans should not rule over Angevins, nor Angevins over Normans, nor either over Bretons and Poitevins, nor Poitevins over any of the rest; and that if one and all must needs submit to the loss of their ancient independence, it would be more natural and less humiliating to lay it down at the feet of the prince who had always been acknowledged in theory as the superior of all alike, the king of the French.

This feeling, however, had scarcely come into existence, much less risen to the surface of politics, when Philip Augustus came home from the Crusade at Christmas 1191. It is scarcely probable that any plan of actual conquest had as yet taken shape in Philip’s mind. But the very audacity of the demand which he made upon the credulity of the Norman constables when in the following spring he asked them to believe that Richard had ceded to him not only the whole Vexin, but also the counties of Aumale and Eu--a cession for which there was not a shadow of reason either in past history or in present circumstances, and which if carried into effect would have cut off the Norman communications with Ponthieu and Flanders, and given him at once a foothold upon the Channel and an invaluable coign of vantage for an attempt upon Rouen--seems to indicate that he was already forming some more definite design against the Angevins’ power than the simple system of lying in wait to steal from them any territorial or political advantage that could be stolen with impunity, with which he, like his father, had hitherto been content. The terms of his treaty with John in the following year point still more strongly in the same direction. As the price of John’s investiture with the rest of his brother’s dominions, Philip reserved to himself the whole Norman territory on the right bank of the Seine, except the city of Rouen; on the left bank, nearly half the viscounty of Evreux, including the castles of Vaudreuil, Verneuil and Ivry; and from the older Angevin patrimony, all that was most worth having in Touraine--Tours itself, Azay, Montbazon, Montrichard, Amboise and Loches--besides the transfer of the Angevin fiefs in the Vendômois from the count of Anjou to the count of Blois.[1807] Owing to the disorganized state of Richard’s dominions caused by his captivity, Philip’s endeavours to carry this bargain into effect by conquering Normandy in John’s interest and his own met for a while with considerable success. His first attempt at invasion was indeed repulsed by the Norman barons under the leadership of Earl Robert of Leicester;[1808] but a few weeks later treason opened to him the gates of Gisors and Neaufle; the rest of the Vexin was easily won,[1809] and secured thus against attack in his rear, he marched northward to the capture of Aumale and Eu.[1810] Thence he turned back to besiege Rouen, but soon retreated again into his own territories,[1811] taking Pacy and Ivry on his way.[1812] In July, finding that, according to his own phrase, the Angevin demon was after all to be let loose upon him once more, he thought it advisable to accept Richard’s overtures of peace; and Richard on his part--being still in prison--deemed it wise for the moment to sanction the French king’s recent conquests in Normandy and the liberation of Ademar of Angoulême, and also to let Philip have temporary possession of Loches, Châtillon-sur-Indre, Driencourt and Arques, as pledges for the payment of twenty thousand marks, due within two years of his own release.[1813]

[1807] Treaty in Rymer, _Fœdera_, vol. i. p. 57.

[1808] Rog. Howden (Stubbs), vol. iii. p. 205.

[1809] Rog. Howden (Stubbs), vol. iii. p. 206. Will. Newb., l. iv. c. 34 (Howlett, vol. i. pp. 389, 390). Rigord (Duchesne, _Hist. Franc. Scriptt._, vol. v.), p. 36. Will. Armor. _Gesta Phil. Aug._ (_ibid._), p. 77.

[1810] Will. Newb. as above (p. 390).

[1811] _Ibid._ Rog. Howden, as above. Cf. Chron. Rothom., a. 1193 (Labbe, _Nova Biblioth._, vol. i. p. 369).

[1812] Will. Newb. as above.

[1813] Rog. Howden as above, pp. 217–220. These were apparently the twenty thousand marks promised in 1189 and not yet paid.

Whether he intended to keep or to break these engagements is practically no matter; for, if he meant to break them, Philip took care to anticipate him. Seven months after the treaty was signed he again crossed the Norman border, took Evreux,[1814] which he handed over to John’s custody,[1815] and marched up by way of Neubourg and Vaudreuil, both of which he captured, to besiege Rouen. Thence, however, he again retired--scared, it may be, by tidings of Richard’s approach--and hurrying back to the southern border laid siege to Verneuil on May 10.[1816] Two days later Richard landed at Barfleur,[1817] and by the end of another fortnight he was encamped at L’Aigle,[1818] within a few miles of Verneuil. His presence there, coupled with the defection of John who had contrived to join him on the road,[1819] and the surprise and slaughter of the French garrison of Evreux by a body of Norman troops,[1820] alarmed Philip so much that on Whitsun Eve, May 28, he again fled into his own dominions.[1821] Richard was busy strengthening the walls of Verneuil when tidings came to him that “the Angevins and Cenomannians” were besieging Montmirail,[1822] a castle on the borders of Perche and Maine, famous as the scene of a stormy conference between Henry II. and S. Thomas. Who the besiegers actually were, or what was the ground of their hostility either to William of Montmirail[1823] or to his overlord King Richard, must remain undecided. It is plain, however, that in Richard’s ears the tidings sounded as a warning of disaffection in his patrimonial dominions. He hurried to the relief of Montmirail, but found it levelled with the ground.[1824] He wasted no time in pursuit of its destroyers, but pushed on direct to Tours, took up his quarters in Châteauneuf,[1825] and shewed his suspicions concerning the origin of the new mischief by driving the canons of S. Martin out of the abbey where they dwelt under the special protection of the French king.[1826] The burghers, on the other hand, made proof of their loyalty by a free-will offering of two thousand marks.[1827] Determined now to redeem his pledges to Philip not with gold but with steel, Richard marched on to Beaulieu,[1828] to join a body of Navarrese and Brabantines, sent by his brother-in-law Sancho of Navarre, in blockading the castle of Loches;[1829] a few days after his arrival, on June 13, it was surrendered by its French garrison.[1830] He was however standing between two fires. Bertrand de Born was again stirring up the south, singing and fighting ostensibly in Richard’s interest against his disaffected neighbours in the Limousin, but in reality kindling into a fresh blaze all the reckless passions and endless feuds which had been smouldering too long for the warrior-poet’s pleasure.[1831] Philip meanwhile was again threatening Rouen;[1832] the Norman archbishop and seneschal attempted to negotiate with him in Richard’s name, but without result;[1833] and at the end of the month he marched southward to meet Richard himself. On July 4 the two kings were within a few miles of each other--Richard at Vendôme, Philip at Fréteval.[1834] What followed is told so diversely by the English and French historians of the time that it seems impossible to reconcile the rival accounts or to decide between them. All that we know for certain is that Philip suddenly struck his tents and withdrew into the territories of the count of Blois; that Richard set off in pursuit, missed Philip himself, but fell at unawares upon the troops who were convoying his baggage towards Blois, routed them, and captured all the French king’s most precious possessions, including his royal seal and the treasury-rolls of the whole kingdom, besides a number of valuable horses, an immense quantity of money and plate, and--what would be scarcely less useful to Richard for political purposes--the charters of agreement between Philip and all the Norman, Angevin and Poitevin rebels who had plotted treason with him and John against their lord.[1835]

[1814] Will. Newb. as above, c. 40 (p. 403). Rigord (as above), p. 37. Will. Armor. _Gesta Phil. Aug._ as above; _Philipp._, l. iv. (_ibid._) p. 143.

[1815] Will. Armor. _Philipp._ as above.

[1816] Rigord as above. Will. Armor. _Gesta Phil. Aug._ as above. Cf. _Philipp._ as above; Rog. Howden (as above), pp. 251, 252; R. Diceto (Stubbs), vol. ii. pp. 114, 115; and Will. Newb., l. v. c. 2 (Howlett, vol. ii. p. 418). The date of the siege of Verneuil comes from Rog. Howden.

[1817] Rog. Howden (Stubbs), vol. iii. p. 251. R. Diceto (Stubbs), vol. ii. p. 114.

[1818] Will. Newb., l. v. c. 2 (Howlett, vol. ii. p. 418).

[1819] Rog. Howden (as above), p. 252. R. Diceto, as above, says they met “apud Bruis.”

[1820] This is all that Rigord says about the disaster (Duchesne, _Hist. Franc. Scriptt._, vol. v. p. 37). In the hands of the poet William of Armorica it becomes a horrible romance, wherein John, as commandant of Evreux, invites the unsuspecting Frenchmen to a banquet, and then brings in his “armed Englishmen” to massacre them (_Philipp._, l. iv., _ib._ p. 143; _Gesta Phil. Aug._, _ib._ p. 77). John has so many undoubted crimes to answer for that it probably seemed a mere trifle to add one more to the list, but for that very reason one cannot admit it on the sole testimony of the poet-historiographer. The English writers say nothing of the whole matter.

[1821] Rog. Howden and Will. Newb. as above. R. Diceto (as above), p. 115. Cf. Rigord and Will. Armor. as above.

[1822] “Andegavenses et Cenomannenses” says Rog. Howden as above. R. Diceto (as above), p. 116, has “Andegavenses” only; the Chron. S. Albin. a. 1192 (Marchegay, _Eglises_, p. 49), has “Andegavenses et alii.”

[1823] William “Gohet” as R. Diceto calls him; _i.e._ (see Bishop Stubbs’s note, _ibid._), “William of Perche Gouet, Goeth, or le petit Perche.”

[1824] Rog. Howden as above. R. Diceto as above, p. 117. Cf. Chron. S. Albin. a. 1192 (as above).

[1825] R. Diceto as above.

[1826] Rigord (Duchesne), _Hist. Franc. Scriptt._, vol. v. p. 38.

[1827] “Dono spontaneo,” Rog. Howden (Stubbs), vol. iii. p. 252; “nullâ coactione præmissâ,” R. Diceto (Stubbs), vol. ii. p. 117. The “burgenses” in question, as appears from R. Diceto, were those of Châteauneuf, not the _cives_ of Tours proper.

[1828] R. Diceto as above.

[1829] Rog. Howden (as above), pp. 252, 253.

[1830] _Ib._ p. 253 (with the date). R. Diceto as above. Cf. Chron. S. Albin. a. 1192 (Marchegay, _Eglises_, p. 49).

[1831] Clédat, _Bert. de Born_, pp. 83, 84.

[1832] Rog. Howden as above, p. 253. R. Diceto, p. 116.

[1833] Rog. Howden as above, pp. 253–255.

[1834] R. Diceto as above.

[1835] Cf. Rog. Howden (Stubbs), vol. iii. pp. 255, 256; R. Diceto (Stubbs), vol. ii. pp. 117, 118; Will. Newb., l. v. c. 2 (Howlett, vol. ii. p. 419); Rigord (Duchesne, _Hist. Franc. Scriptt._, vol. v.), p. 38; Will. Armor. _Gesta Phil. Aug._ (_ibid._), p. 77; _Philipp._, l. iv. (_ibid._), p. 144; and Chron. S. Albin. a. 1192 (Marchegay, _Eglises_, p. 49). Rog. Howden alone mentions the charters, and Will. Armor. the treasury-rolls and seal.

The repairing of this disaster gave Philip sufficient occupation for the rest of the year, and Richard was free to march upon the Aquitanian rebels. Sancho of Navarre was already wasting the lands of the ringleaders, Geoffrey of Rancogne and Ademar of Angoulême;[1836] and by July 22 Richard was able to report to his justiciar in England that he was master of all the castles of the Angoumois and all the lands of Geoffrey.[1837] From Angoulême he marched northward again, took measures for the security of Anjou and Maine,[1838] and then returned to Normandy, where he found that his representatives, headed by the chancellor, had just concluded a truce with the French king to last till All Saints’ day[1839]--a proceeding which served him as the pretext for that withdrawal of the seal from William and repudiation of all engagements made under it, which has been mentioned already.[1840] No further movement was however made by either party until the spring. Then the wearisome story of fruitless negotiations alternating with indecisive warfare begins again, and goes on unceasingly for the next four years. Save for an occasional attempt to make a diversion in Berry, the actual fighting between the two kings was confined to the Norman border.[1841] Normandy was the chief object of Philip’s attack, partly no doubt because, owing to its geographical position, he could invade it with more ease and less risk than any other part of Richard’s dominions, but also because it was the key to all the rest. A French conquest of Normandy would sever Richard’s communications not only with Flanders and Germany, but also with England; and the strength of the Angevins in Gaul now rested chiefly upon the support of their island-realm. Neither assailant nor defender, however, was able to gain any decisive advantage in the field. The armed struggle between them was in fact of less importance than the diplomatic rivalry which they carried on side by side with it; and in this, strangely enough, Richard, who had hitherto shewn so little of the far-sighted statecraft and political tact of his race, proved more than a match for his wily antagonist.

[1836] R. Diceto as above, p. 117. Will. Newb. as above.

[1837] Letter of Richard to Hubert Walter (date, Angoulême, July 22) in Rog. Howden as above, pp. 256, 257. Cf. R. Diceto as above, pp. 118, 119. Will. Newb. as above (p. 420).

[1838] “Rediit in Andegaviam, et redemit omnes baillivos suos, id est, ad redemptionem coegit. Similiter fecit in Cenomanniâ.” Rog. Howden as above, p. 267. At Le Mans “convocavit magnates omnes suæ jurisdictioni subpositos,” and apparently tried to shame them into more active loyalty--or more liberal gifts--by eulogy of their English brethren: “ubi fidem Anglorum in adversitate suâ semper sibi gratiosam, integram et probabilem plurimum commendavit.” R. Diceto as above, p. 119.

[1839] Rog. Howden as above, pp. 257–260. Cf. R. Diceto as above, p. 120, and Will. Newb., l. v. c. 3 (as above). This last gives a wrong date; that of the document in Rog. Howden is July 23.

[1840] Rog. Howden (Stubbs), vol. iii. p. 267. See above, p. 343.

[1841] It may be followed in Rog. Howden (Stubbs), vol. iii. pp. 301–305, vol. iv. pp. 3–7, 14, 16, 19–21, 24, 54–61, 68, 78–81; Rigord (Duchesne, _Hist. Franc. Scriptt._, vol. v.), pp. 38–40, 42; Will. Armor. _Gesta Phil. Aug._ (_ibid._), pp. 78, 79; _Philipp._, l. v. (_ib._), pp. 146–154.

That the foes in Richard’s own household should league themselves against him with Philip, as he had done in earlier days against his own father, was, so far as Richard himself is concerned, no more than retributive justice. Philip’s alliance with John had proved a failure; but it was not long before he saw a chance of securing a more useful tool in the person of little Arthur of Britanny. English historians tell us that when Richard and Philip made their treaty at Messina in March 1191 Richard obtained a formal acknowledgement of his rights, as duke of Normandy, to the overlordship of Britanny and the liege homage of its duke.[1842] The text of the treaty of Messina, however, contains not a word on this subject; the agreement, if made at all, must have been drawn up in a separate form; and it seems to have remained a dead letter, like another agreement made at the same place a few months earlier--the treaty with Tancred whereby Richard had engaged to recognize Arthur of Britanny as his successor in default of direct heirs. Although after five years of marriage Queen Berengaria was still childless, no such recognition had yet been made. Richard on his return to Europe probably perceived that Arthur’s succession would be impossible in England, and in Gaul would be fatal to the independence of the Angevin house. Accordingly, he was once more doing all in his power to win the attachment of John; and John, having at length discovered that his own interests could be better served by supporting his brother than by intriguing against him, proved an active and useful ally in the war against Philip.[1843] On the other hand, Richard seems never to have received Arthur’s homage for Britanny; and those who had the control of political affairs in that country were determined that he never should. The dispute between Henry and Philip for the wardship of the two children of Geoffrey and Constance had apparently ended in a compromise. Eleanor, the elder child, was now under the care of her uncle Richard;[1844] but Constance seems to have succeeded in keeping her infant boy out of the reach of both his would-be guardians, and, moreover, in governing her duchy without any reference to either of them, for nearly seven years after the death of her father-in-law King Henry. She had been given in marriage by him, when scarcely twelve months a widow, to Earl Ralf of Chester,[1845] son and successor of Earl Hugh who had been one of the leaders in the revolt of 1173. As the earls of Chester were hereditary viscounts of the Avranchin--the border-district of Normandy and Britanny--this marriage would have furnished an excellent means of securing the Norman hold upon the Breton duchy, if only Ralf himself could have secured a hold upon his wife. In this however he completely failed. Safe in her hereditary dominions, with her boy at her side, and strong in the support of her people rejoicing in their newly-regained independence, Constance apparently set Ralf, Richard and Philip all alike at defiance, till in 1196 Richard summoned her to a conference with himself in Normandy, and she set out to obey the summons. Scarcely had she touched the soil of the Avranchin at Pontorson when she was caught by her husband and imprisoned in his castle of S. James-de-Beuvron.[1846] It is hard not to suspect that Richard and Ralf had plotted the capture between them; for Richard, instead of insisting upon her release, at once renewed his claim to the wardship of Arthur, and prepared to enforce it at the sword’s point. The Bretons first hurried their young duke away to the innermost fastnesses of their wild and desolate country under the care of the bishop of Vannes,[1847] and then, after a vain attempt to liberate his mother, intrusted him to the protection of the king of France,[1848] who of course received him with open arms, and sent him to be educated with his own son.[1849]

[1842] _Gesta Ric._ (Stubbs), p. 161. Rog. Howden (Stubbs), vol. iii. pp. 99, 100.

[1843] See _e.g._ Rog. Howden (Stubbs), vol. iv. pp. 5, 16, 60; Rigord (Duchesne, _Hist. Franc. Scriptt._, vol. v.), p. 38; Will. Armor. _Gesta Phil. Aug._ (_ibid._), p. 77.

[1844] Rog. Howden (Stubbs), vol. iii. pp. 275, 278.

[1845] _Gesta Hen._ (Stubbs), vol. ii. p. 29.

[1846] Rog. Howden (Stubbs), vol. iv. p. 7.

[1847] Will. Armor. _Philipp._, l. v. (Duchesne, _Hist. Franc. Scriptt._, vol. v.), p. 149. Will. Newb., l. v. c. 18 (Howlett, vol. ii. pp. 463, 464).

[1848] Rog. Howden as above.

[1849] Will. Armor. as above.

Philip had now got the old Angevin patrimony between two fires; but the Bretons were so little accustomed to act in concert even among themselves, far less with any other power, that he found it impossible to make any real use of them as allies either for military or political purposes. The independent warfare which they carried on with Richard across the south-western border of Normandy[1850] had little effect upon that which Richard and Philip were carrying on along its eastern border; and upon the Angevin lands which lay directly between Britanny and France the Breton revolt had no effect at all. To the end of Richard’s life, we hear of no further troubles in Maine or Anjou. Nay more, we hear of no further troubles in Aquitaine. If Philip had in some sense turned Richard’s flank in the west, Richard had turned Philip’s flank far more effectually in the south. The unwonted tranquillity there may indeed have been partly due to the fact that one of the chief sources of disturbance was removed in 1196 by the withdrawal of Bertrand de Born into a monastery;[1851] but it was also in great measure owing to Richard’s quickness in seizing an opportunity which presented itself, in that same eventful year, of forming a lasting alliance with the house of Toulouse. His old enemy Count Raymond V. was dead;[1852] he now offered the hand of his own favourite sister, the still young and handsome Queen Jane of Sicily, to the new Count Raymond VI.;[1853] and thenceforth the eastern frontier of his Aquitanian duchy was as secure under the protection of his sister’s husband as its southern frontier under that of his wife’s brother, the king of Navarre.

[1850] Will. Newb. as above, c. 30 (p. 491). Rog. Howden as above.

[1851] Clédat, _Bert. de Born_, p. 92.

[1852] In 1194, according to Rigord (Duchesne, _Hist. Franc. Scriptt._, vol. v.), p. 38.

[1853] Rog. Howden (Stubbs), vol. iv. p. 13. Will. Newb., l. v. c. 30 (Howlett, vol. ii. p. 491). R. Coggeshall (Stevenson), p. 70.

Nor were Richard’s alliances confined within the boundaries of Gaul. His year of captivity in Germany had not been all wasted time. When he parted from his imperial jailor in the spring of 1194, they were, at any rate in outward semblance, close political allies; and at the same time Richard had succeeded in gaining over his bitterest foe, Leopold of Austria, by an offer of his niece Eleanor of Britanny as wife to Leopold’s son.[1854] The marriage-contract was however not yet executed when the Austrian duke met with a fatal accident and died in agony, owning with his last breath that his miserable end was a just retribution for his conduct towards the English king.[1855] The impression made by this event deepened the feeling of respect and awe which the captive lion had already contrived to inspire in the princes of the Empire. Meanwhile Henry VI. had made himself master of Sicily;[1856] and now the old dream by which the German Emperors never quite ceased to be haunted, the dream of re-asserting their imperial supremacy over Gaul, was beginning to shape itself anew in his brain. In the summer of 1195 he sent to Richard a golden crown and a message charging him, on his plighted faith to the Emperor and on the very lives of his hostages, to invade the French kingdom at once, and promising him the support and co-operation of the imperial forces. Richard, suspecting a trap, despatched William of Longchamp to inquire into the exact nature, extent and security of Henry’s promised assistance; Philip vainly tried to intercept the envoy as he passed through the royal domains;[1857] and the negotiations proved so far effectual that Henry remitted seventeen thousand marks out of the ransom, as a contribution to Richard’s expenses in his struggle with Philip.[1858] When, on Michaelmas Eve 1197, Henry VI. died,[1859] the use of that homage on Richard’s part which his English subjects had resented so bitterly was made apparent to them at last. While the English king was holding his Christmas court at Rouen there came to him an embassy from the princes of Germany, summoning him, as chief among the lay members of the Empire[1860] by virtue of his investiture with the kingdom of Arles, to take part with them in the election of a new Emperor at Cöln on February 22.[1861] Richard himself could not venture to leave Gaul; but the issue proved that his presence at Cöln was not needed to secure his interests there. He wished that the imperial crown should be given to his nephew Duke Henry of Saxony, eldest son and successor of Henry the Lion. This scheme, however, when laid before the other electors by the envoys whom he sent to represent him at Cöln, was rejected on account of the duke’s absence in Holy Land.[1862] The representatives of the English king then proposed Henry’s brother Otto, for whom Richard had long been vainly endeavouring to find satisfactory provision on either side of the sea,[1863] and who seems really to have been his favourite nephew. The result was that, on the appointed day, Otto was elected Emperor of the Romans,[1864] and on July 12 he was crowned king of the Germans at Aachen by the archbishop of Cöln.[1865]

[1854] Rog. Howden (Stubbs), vol. iii. p. 273. See above, p. 325.

[1855] Rog. Howden as above, pp. 276, 277. R. Diceto (Stubbs), vol. ii. p. 124. Will. Newb. as above, c. 8 (pp. 431–434). R. Coggeshall (Stevenson), pp. 65, 66.

[1856] In the autumn of 1194. Rog. Howden (Stubbs), vol. iii. pp. 268–270. Cf. R. Diceto (Stubbs), vol. ii. pp. 123, 124.

[1857] Rog. Howden (as above), pp. 300, 301.

[1858] _Ib._ pp. 303, 304.

[1859] _Ib._ vol. iv. p. 31.

[1860] “Sicut præcipuum membrum imperii.” _Ib._ p. 37.

[1861] _Ibid._

[1862] _Ib._ pp. 37, 38.

[1863] He appointed him earl of York in 1190, but as the grant was made after the king left England, some of the Yorkshire folk doubted its genuineness, and Otto never succeeded in obtaining possession. Rog. Howden (Stubbs), vol. iii. p. 86. The elaborate scheme for his endowment in the north, projected in 1195, has already been mentioned (above, p. 341). This having also failed, Richard in 1196 gave him the investiture of Poitou. Rog. Howden (Stubbs), vol. iv. p. 7; cf. _ib._ vol. iii. p. 86, and R. Coggeshall (Stevenson), p. 70.

[1864] Rog. Howden (Stubbs), vol. iv. pp. 37–39. R. Diceto (Stubbs), vol. ii. p. 163.

[1865] R. Diceto as above.

For a moment, at the mere prospect of beholding a grandson of Henry Fitz-Empress seated upon the imperial throne of the west, there had flashed across the mind of at least one friend of the Angevin house a fancy that the world-wide dominion which seemed to be passing away from the heirs of Fulk the Good was to be renewed for yet one more generation.[1866] There was indeed an opposition party in Germany, who set up a rival Emperor in the person of Philip of Suabia, a brother of Henry VI.;[1867] and he at once made common cause with his French namesake.[1868] This Suabian alliance, however, and the support of the count of Ponthieu--purchased two years before with the hand of the unhappy Adela, whom Richard had at last restored to her brother[1869]--could not much avail Philip Augustus against such a league as was now gathering around the English king. The vast sums which Hubert Walter had been sending, year after year, to his royal master over sea were bringing a goodly interest at last. Flanders, Britanny, Champagne, had all been secretly detached from the French alliance and bought over to the service of Richard;[1870] the Flemish count had already drawn Philip into a war in which he narrowly escaped being made prisoner;[1871] and in the summer of 1198, when the imperial election was over, not only Baldwin of Flanders, Reginald of Boulogne, Baldwin of Guines, Henry of Louvain, Everard of Brienne, Geoffrey of Perche and Raymond of Toulouse, but even the young count Louis of Blois and the boy-duke Arthur of Britanny himself, one and all leagued themselves in an offensive and defensive alliance with Richard against the French king.[1872] The immediate consequence was that Philip begged Hubert Walter, who being just released from his justiciarship had rejoined his sovereign in Normandy, to make peace for him with Richard; and he even went so far as to offer the surrender of all the Norman castles which he had won, except Gisors. Richard however would listen to no terms in which his allies were not included.[1873] At last, in November, a truce was made, to last till the usual term, S. Hilary’s day.[1874] When it expired the two kings held a colloquy on the Seine between Vernon and Les Andelys, Richard in a boat on the river, Philip on horseback on the shore;[1875] this meeting was followed by another, where, by the mediation of a cardinal-legate, Peter of Capua, who had lately arrived in Gaul, they were persuaded to prolong their truce for five years.[1876]

[Illustration: Plan VII.

LES ANDELYS AND CHATEAU-GAILLARD.

(From Deville, “Histoire du Château-Gaillard”)

Wagner & Debes’ Geogˡ. Estabᵗ. Leipsic. London, Macmillan & Co. ]

[1866] R. Diceto tells the story of the prophecy made to Fulk the Good in two places; in the _Abbreviationes Historiarum_ (Stubbs, vol. i. p. 149) and in the _Opuscula_ (vol. ii. pp. 267, 268). In the latter place he adds: “Quod quondam probavit regnum Jerosolimitanum; quod adhuc ostendit regnum Anglorum; quod suo tempore declarabit Romanum imperium.” This, as Bishop Stubbs notes, “looks like an anticipation of the election of Otto IV. to the empire.... As Bishop Longchamp died in 1197, before which date we must suppose MS. R to have been written” [the MS. from which the _Opuscula_ are printed, and which begins with a dedication to William of Longchamp], “it can scarcely be a prophecy after the event.” As William of Longchamp died January 31, 1197 (R. Diceto, vol. ii. p. 150; February 1 according to Gerv. Cant., Stubbs, vol. i. p. 543), it seems indeed to shew that the possibility of one or other of Richard’s nephews becoming Emperor at the next vacancy was already in contemplation more than eight months before the death of Henry VI. Or was Ralf dreaming rather of a transfer of the imperial crown to Richard himself? for it is to be observed that Otto can be included within the “nine generations” only by excluding from them Fulk the Good himself; but this mode of computing would fail if applied to the eastern branch of the Angevin house, where it would give only eight generations, so that we can hardly suppose it to have been adopted by Ralf. According to R. Coggeshall (Stevenson), p. 88, and Gerv. Cant. as above, p. 545, a party among the electors actually did choose Richard, and--much more strangely--another party chose Philip of France.

[1867] Rog. Howden as above, p. 39.

[1868] Treaty in Rymer, _Fœdera_, vol. i. p. 70; date, June 29 [1197].

[1869] Rog. Howden (Stubbs), vol. iii. p. 303. Rigord (Duchesne, _Hist. Franc. Scriptt._, vol. v.), p. 38. Will. Armor. _Gesta Phil. Aug._ (_ibid._) p. 77.

[1870] Cf. Rog. Howden (Stubbs), vol. iv. p. 19, R. Coggeshall (Stevenson), p. 77, and Will. Newb., l. v. c. 32 (Howlett, vol. ii. p. 495). Richard’s treaty with Flanders is in R. Diceto (Stubbs), vol. ii. pp. 152, 153, and Rymer, as above, pp. 67, 68; it has no date, but as R. Diceto (as above, p. 158) tells us that it was drawn up by Hubert Walter, and also that Hubert was in Gaul from September 14 (or 28, according to Gerv. Cant., Stubbs, vol. i. p. 574) to November 8 [1197], it must fall in that interval.

[1871] Rog. Howden as above, pp. 20, 21. Will. Newb. as above. R. Coggeshall (Stevenson), pp. 77, 78.

[1872] Rog. Howden as above, p. 54.

[1873] _Ib._ p. 61.

[1874] _Ib._ p. 68.

[1875] Rog. Howden (Stubbs), vol. iv. pp. 79, 80.

[1876] _Ib._ p. 80. Rigord (Duchesne, _Hist. Franc. Scriptt._, vol. v.), p. 42.

Yet all the while, there lurked in Richard’s heart a misgiving that, in the last resort, his diplomacy would prove to have been in vain; that, strive as he might to turn away the tide of war from his own borders by stirring up north and east and south to overwhelm the Crown of France, still, after all, the day must come when the Angevins would have to stake their political existence solely upon their own military resources, and to stand at bay, unaided, unsupported, alone, behind whatever bulwark they might be able to devise by their own military genius. It was the genius and the foresight of Richard himself which insured that when the crisis came, the bulwark was ready, even though it were doomed to prove unavailing in the end. The last and mightiest of the many mighty fortresses reared by Angevin hands since the first great builder of the race had begun his castle-building in the Loire valley was the Château-Gaillard, the “saucy castle” of Richard the Lion-heart. He “fixed its site where the Seine bends suddenly at Gaillon in a great semicircle to the north, and where the valley of Les Andelys breaks the line of the chalk cliffs along its banks. Blue masses of woodland crown the distant hills; within the river curve lies a dull reach of flat meadow, round which the Seine, broken with green islets and dappled with the grey and blue of the sky, flashes like a silver bow on its way to Rouen.”[1877] Some three-quarters of a league from the right bank of the river, in a valley opening upon it from the eastward and watered by the little stream of Gambon, stood the town of Andely. Between the town and the river stretched a lake, or rather perhaps a marsh,[1878] through which the Gambon and another lesser rivulet descending from the hills to the north of Andely found their way by two separate issues into the Seine, nearly opposite two islets, of which the larger and more northerly was known as the Isle of Andely.[1879] The space enclosed between the three rivers and the marsh seems to have been a tract of waste land, occupied only by a toll-house for the collection of dues from the vessels passing up and down the Seine[1880]--dues which formed one of the most important items in the revenue of the archbishop of Rouen, to whom Andely and its neighbourhood belonged.[1881] Over against this spot, on the southern bank of the Gambon, in the angle formed by its junction with the Seine, a mass of limestone crag rose abruptly to the height of three hundred feet. Its western side, almost perpendicular, looked down upon the great river, the northern, scarcely less steep, over the Gambon and the lake beyond; to the north-east and south-west its rocky slopes died down into deep ravines, and only a narrow neck of land at its south-eastern extremity connected it with the lofty plateau covered with a dense woodland known as the Forest of Andely, which stretches along the eastern side of the Seine valley between Andely and Gaillon. One glance at the site was enough to rivet a soldier’s gaze. If, instead of the metropolitan church of Normandy, a lay baron had owned the soil of Andely, we may be sure that long ago that lofty brow would have received its fitting crown; if the power of Fulk the Builder had reached to the banks of the Seine, we may doubt whether the anathemas of the Norman primate would not have availed as little to wrest such a spot from his grasp as those of the archbishop of Tours had availed to wrest from him the site of Montrichard. But a greater castle-builder than Fulk Nerra himself was the architect of Château-Gaillard.

[1877] I copy Mr. Green’s picture, _Hist. of the English People_, vol. i. p. 187.

[1878] Now dried up. See Deville, _Hist. du Château-Gaillard_, pp. 27, 28.

[1879] “Est locus Andelii qui nunc habet insula nomen.” Will. Armor. _Philipp._, l. vii. v. 29 (Deville, _Château-Gaillard_, p. 126; Duchesne, _Hist. Franc. Scriptt._, vol. v. p. 169).

[1880] See a charter of Archbishop Malger (11th century) and one of Pope Eugene III., a. 1148, quoted in Deville as above, p. 26, note 2.

[1881] The archbishops seem to have looked upon Andely as their most profitable territorial possession; Rotrou called it his “unicum vivendi subsidium” (Rotr. Ep. xxiv., _Rer. Gall. Scriptt._, vol. xvi. p. 632); Walter called it “patrimonium ecclesiæ solum et unicum” (R. Diceto, Stubbs, vol. ii. p. 148).

Richard’s historical connexion with the “rock of Andely” has its ill-omened beginning in a ghastly story of the fate of three French prisoners whom he flung from its summit into the ravine below, in vengeance for the slaughter of some Welsh auxiliaries who had been surprised and cut to pieces by the French king’s troops in the neighbouring valley.[1882] By the opening of 1196, however, he had devised for it a more honourable use. In a treaty with Philip, drawn up in January of that year, the fief of Andely was made the subject of special provisions whereby it was reserved as a sort of neutral zone between the territories of the two kings, and a significant clause was added: “Andely shall not be fortified.”[1883] As by the same treaty the older bulwarks of Normandy--Nonancourt, Ivry, Pacy, Vernon, Gaillon, Neufmarché, Gisors--were resigned into Philip’s hands, this clause, if strictly fulfilled, would have left the Seine without a barrier and Rouen at the mercy of the French king. The agreement in short, like all those which bore the signatures of Philip and Richard, was made only to be broken; both parties broke it without delay; and while Philip was forming his league with the Bretons for the ruin of Anjou, Richard was tracing out in the valley of the Gambon and on the rock of Andely the plan of a line of fortifications which were to interpose an insurmountable barrier between his Norman capital and the French invader. His first act was to seize the Isle of Andely.[1884] Here he built a lofty octagonal tower, encircled by a ditch and rampart, and threw a bridge over the river from each side of the island, linking it thus to either shore.[1885] On the right, beyond the eastern bridge, he traced out the walls of a new town, which took the name of the New or the Lesser Andely,[1886] a secure stronghold whose artificial defences of ramparts and towers were surrounded by the further protection of the lake on its eastern side, the Seine on the west, and the two lesser rivers to north and south, a bridge spanning each of these two little streams forming the sole means of access from the mainland.[1887] The southern bridge, that over the Gambon, linked this New Andely with the foot of the rock which was to be crowned with the mightiest work of all. Richard began by digging out to a yet greater depth the ravines which parted this rock from the surrounding heights, so as to make it wholly inaccessible save by the one connecting isthmus at its south-eastern extremity. On its summit, which formed a plateau some six hundred feet in length and two hundred in breadth at the widest part, he reared a triple fortress. The outer ward consisted of a triangular enclosure; its apex, facing the isthmus already mentioned, was crowned by a large round tower,[1888] with walls ten feet in thickness; the extremities of its base were strengthened by similar towers, and two smaller ones broke the line of the connecting curtain-wall. This was surrounded by a ditch dug in the rock to a depth of more than forty feet, and having a perpendicular counterscarp. Fronting the base of this outer fortress across the ditch on its north-western side was a rampart surmounted by a wall ninety feet long and eight feet thick, also flanked by two round towers; from these a similar wall ran all round the edges of the plateau, where the steep sides of the rock itself took the place of rampart and ditch. The wall on the south-west side--the river-front--was broken by another tower, cylindrical without, octagonal within; and its northern extremity was protected by two mighty rectangular bastions. Close against one of these stood a round tower, which served as the base of a third enclosure, the heart and citadel of the whole fortress. Two-thirds of its elliptical outline, on the east and south, were formed by a succession of semicircular bastions, or segments of towers, seventeen in number, each parted from its neighbour by scarcely more than two feet of curtain-wall--an arrangement apparently imitated from the fortress of Cherbourg, which was accounted the greatest marvel of military architecture in Normandy, until its fame was eclipsed by that of Richard’s work.[1889] This portion of the enclosure was built upon a rampart formed by the excavation of a ditch about fifteen to twenty feet in width; the counterscarp, like that of the outer ditches, was perpendicular; and a series of casemates cut in the rock ran along on this side for a distance of about eighty feet. On the western side of the citadel stood the keep, a mighty circular tower, with walls of the thickness of twelve feet, terminating at an angle of twenty feet in depth where it projected into the enclosure; it had two or perhaps three stages,[1890] and was lighted by two great arched windows, whence the eye could range at will over the wooded hills and dales of the Vexin, or the winding course of the river broadening onward to Rouen. Behind the keep was placed the principal dwelling-house, and under this a staircase cut out of the rock gave access to an underground passage leading to some outworks and a tower near the foot of the hill, whence a wall was carried down to the river-bank, just beyond the northern extremity of a long narrow island known as the “isle of the Three Kings”--doubtless from some one of the many meetings held in this district by Louis VII. or Philip Augustus and the two Henrys.[1891] The river itself was barred by a double stockade, crossing its bed from shore to shore.[1892]

[Illustration: Plan VIII.

CHATEAU-GAILLARD

(From Deville, “Histoire du Château-Gaillard”).

Wagner & Debes’ Geogˡ. Estabᵗ. Leipsic. London, Macmillan & Co. ]

[1882] Will. Armor. _Philipp._, l. v. (Duchesne, _Hist. Franc. Scriptt._, vol. v.), p. 151.

[1883] Treaty in Rymer, _Fœdera_, vol. i. p. 66. For date see Rigord (Duchesne as above), p. 39.

[1884] Letter of Walter of Rouen (a. 1196), R. Diceto (Stubbs), vol. ii. pp. 148, 149. Cf. Rog. Howden (Stubbs), vol. iv. p. 14, and Will. Newb., l. v. c. 34 (Howlett, vol. ii. p. 499).

[1885] Will. Armor. _Philipp._, l. vii. vv. 29–43 (Deville, _Château-Gaillard_, p. 126; Duchesne as above, p. 169).

[1886] A poet of the thirteenth century, William Guiart, calls it “le Nouvel-Andeli.” It is known now as “le Petit-Andely.” Deville as above, p. 26.

[1887] Will. Armor. _Gesta Phil. Aug._ (Duchesne, _Hist. Franc. Scriptt._, vol. v.), p. 81. Deville, _Château-Gaillard_, p. 27.

[1888] Now known as “tour de la Monnaie.” Deville as above, p. 30, note 1.

[1889] See Deville, _Château-Gaillard_, p. 34, and the passage there quoted from _Hist. Gaufr. Ducis_ (Marchegay, _Comtes d’Anjou_, p. 300).

[1890] See Deville as above, p. 38, note 2.

[1891] _Ib._ p. 36. The island is now joined to the mainland; _ib._ note 1.

[1892] For description see Will. Armor. _Gesta Phil. Aug._ (Duchesne, _Hist. Franc. Scriptt._, vol. v.), p. 81; _Philipp._, l. vii. vv. 48–85 (_ib._ pp. 169, 170; Deville as above, pp. 126, 127), and Deville as above, pp. 25–40.

All this work was accomplished within a single year.[1893] Richard, who had watched over its progress with unremitting care, broke into an ecstasy of delight at its completion; he called his barons to see “how fair a child was his, this child but a twelvemonth old”;[1894] he called it his “saucy castle,” “Château-Gaillard,”[1895] and the name which he thus gave it in jest soon replaced in popular speech its more formal title of “the Castle on the Rock of Andely.”[1896] The hardness of the rock out of which the fortifications were hewn was not the sole obstacle against which the royal builder had had to contend. Richard had no more thought than Fulk Nerra would have had of asking the primate’s leave before beginning to build upon his land; the work therefore was no sooner begun than Archbishop Walter lifted up his protest against it; obtaining no redress, he laid Normandy under interdict and carried his complaint in person to the Pope.[1897] Richard at once sent envoys to appeal against the interdict and make arrangements for the settlement of the dispute.[1898] Meanwhile, however, he pushed on the building without delay. Like Fulk of old, the seeming wrath of Heaven moved him as little as that of its earthly representatives; a rain of blood which fell upon the workmen and the king himself, though it scared all beside, failed to shake his determination; “if an angel had come down out of the sky to bid him stay his hand, he would have got no answer but a curse.”[1899] He had now, however, made his peace with the Church; in the spring of 1197 he offered to the archbishop an exchange of land on terms highly advantageous to the metropolitan see; and on this condition the Pope raised the interdict in May of the same year.[1900] The exchange was carried through on October 16,[1901] and ratified by John in a separate charter, a step which seems to indicate that John was now recognized as his brother’s heir.[1902]

[1893] That is, the castle on the rock, built 1197–1198. See the story of the rain of blood in May 1198 (R. Diceto, Stubbs, vol. ii. p. 162), which fixes its completion after that date. The tower on the island and the Nouvel-Andely were the work of the previous year, 1196–1197.

[1894] “Ecce quam pulcra filia unius anni!” J. Bromton, Twysden, _X. Scriptt._, col. 1276.

[1895] “Totamque munitionem illam vocavit Gaillardum, quod sonat in Gallico petulantiam.” Will. Armor. _Gesta Phil. Aug._ (Duchesne, _Hist. Franc. Scriptt._, vol. v.), p. 81.

[1896] “Castrum” or “castellum de Rupe Andeleii” or “Andeliaci,” it is called in the charters of Richard and John. The first document in which it appears as “Château-Gaillard” is a charter of S. Louis, “actum in Castro nostro Gaillard,” A.D. 1261; Deville, _Château-Gaillard_, p. 40. Will. Armor. however uses the name, and other writers soon begin to copy him.

[1897] Rog. Howden (Stubbs), vol. iv. p. 14. Cf. Will. Newb., l. v. c. 28 (Howlett, vol. ii. pp. 487, 488), R. Coggeshall (Stevenson), p. 70, and Gerv. Cant. (Stubbs), vol. i. p. 544.

[1898] The envoys were William of Longchamp, William bishop of Lisieux and Philip elect of Durham; Rog. Howden (as above), pp. 16, 17. They must have started early in 1197, for William of Longchamp died on the journey, at Poitiers, on January 31 or February 1; see above, p. 373, note 4{1866}.

[1899] Will. Newb., l. v. c. 34 (as above, p. 500). This is William’s last sentence. R. Diceto (Stubbs), vol. ii. p. 162, also tells of the portent, and gives its date, May 8, 1198.

[1900] Rog. Howden (Stubbs), vol. iv. pp. 17–19. Will. Newb., l. v. c. 34 (Howlett, vol. ii. pp. 499, 500).

[1901] Richard’s charter, of which Deville gives a fac-simile in his _Château-Gaillard_, p. 18, and a printed copy in his “pièces justificatives,” _ib._ pp. 113–118, is also in R. Diceto (Stubbs), vol. ii. pp. 154–156. According to this last writer (_ib._ pp. 158, 159), and Gerv. Cant. (Stubbs, vol. i. p. 544), the settlement was due to the mediation of Archbishop Hubert.

[1902] See Deville, as above, pp. 21, 22. John’s charter is in the “pièces justificatives,” _ib._ pp. 119–123.

It was probably about the same time that the treaty with Flanders, the corner-stone of the league which Richard was forming against the king of France, was signed within the walls of the new fortress.[1903] Yet, as has been already seen, the coalition was not fully organized till late in the following summer; and even then the complicated weapon hung fire. Want of money seems to have been Richard’s chief difficulty, now as ever--a difficulty which after Hubert Walter’s defeat in the council at Oxford and his resignation in the following July must have seemed well-nigh insurmountable. At last, however, in the spring of 1199, a ray of hope came from a quarter where it was wholly unexpected. Richard was leading his mercenaries through Poitou to check the viscount of Limoges and the count of Angoulême in a renewal of their treasonable designs[1904] when he was met by rumours of a marvellous discovery at Châlus in the Limousin. A peasant working on the land of Achard, the lord of Châlus, was said to have turned up with his plough a treasure[1905] which popular imagination pictured as nothing less than “an emperor with his wife, sons and daughters, all of pure gold, and seated round a golden table.”[1906] In vain did Achard seek to keep his secret and his prize to himself. Treasure-trove was a right of the overlord, and it seems to have been at once claimed by the viscount Ademar of Limoges, as Achard’s immediate superior. His claim, however, had to give way to that of his own overlord, King Richard; but when he sent to the king the share which he had himself wrung from Achard, Richard indignantly rejected it, vowing that he would have all. This Achard and Ademar both refused, and the king laid siege to Châlus.[1907]

[1903] R. Diceto (as above), p. 153.