explicitly that there was no truce at the end of 1202;1 and there was certainly no break at the end of 1203. Hence it is impossible to connect the operations of 1203 exclusively with Arthur's death or the condemnation of John. So far as this argument goes, it shows that the condemnation. might have been passed in 1203 or 1204 or 1205, or any other year. At the same time Philip, who had been urging on war all the more fiercely because of his suspicions, became convinced that Arthur was dead. In reply to every suggestion of peace he said, 'Either produce Arthur, or, if you have killed him, surrender all your continental possessions.' At last he felt sure. He had heard, says Ralph of Coggeshall, that he was drowned. This was in the spring of 1204, and the condemnation, if it was passed, would most naturally follow then. Philip did not know the exact details, nor do I think that he knew them until some years had gone by.

Our chief authority for this summary has been the abbot of Coggeshall. All historians, except Miss Norgate, are convinced of the value of this writer. His narrative is at bottom annalistic, embroidered by tales of visitors and neighbours. There is no attempt at continuous history, but, mixed with jejune summaries, we find two kinds of story, both of which show the sort of authority upon which they are based. One of them is the religious marvel, the other the striking political incident. We do not need the

3

abbot's explicit statement to know that a special source a visitor, a monk who has been on business, a neighbouring baron-has produced these stories. The vivid narrative of Richard's capture was related by the royal chaplain, Anselm. Another eye-witness, Hugh de Nevill, brought back a story of the crusade. In spite of Miss Norgate's 1. Rigord, ed. Delaborde, i, 153. Superveniente veso hyeme uterque sine pace et treuga, marchiis munitis, a bello cessavit. This is the more significant, since John attempted to bring about a truce.

2. Cf. Petit-Dutaillis, p. 111.

3. Coggeshall, p. 54.

4. Ibid, p. 45.

Above, p. 241.

criticism the account of the first condemnation of John in 1202 has been amply verified by French scholars; nor is there any reason to disbelieve the circumstantial relation of the events at Falaise, though they are not mentioned by any other writer. Now it seems to me to be a valid argument that, if the widespread tale of Arthur's supposed death at Falaise has only come down in one chronicle, his mysterious fate would be still more likely to pass unchronicled, or would only be revealed accidentally through the gossip of the few people who knew what had happened. It is only when a chance discovery, like that of the biography of the Marshal, brings some unknown authority to light that we can realise faintly what a vast story lies untold. By accident or good fortune a chronicler here and there heard one thing out of a hundred, or a rhyming biographer put down the reminiscences of his hero. Except in rare and definite cases the argument e silentio is invalid for the medieval historian. Further, when there is reason for secrecy, the chances of truth are of course less. Arthur subito evanuit, said Roger of Wendover. We must not think of Arthur at this time as a popular hero, except in Brittany. When John's crime was made a political question by Philip and Louis in 1216, the pope did not trouble himself to deny it. He made little of it. The chronicles, he said, tell us of the murder of innocent persons by many princes, the kings of France as well as others, but we do not read that the murderers were ever condemned to death. Arthur was no innocent victim; he was captured at Mirebeau, a traitor to his lord, to whom he had sworn homage (cui homagium et liganciam fecerat), and he could rightly be condemned without a formal trial to die the most shameful of deaths.1

In the spring, then, of 1204 Philip was becoming convinced that Arthur was dead. If the Breton tradition be

1. Matt. Paris, Chron. Mai., ii 659 (from Wendover). For remarks on the effect of Arthur's disappearance see above, p. 247. The papal view of treason was evidently more comprehensive than the feudal view (above, p. 418.)

correct and we have seen reason to believe that it is based on truth—he had long been urged to condemn John for the murder. If and when he condemned him is, so far as this body of evidence goes, uncertain. If he did, the natural date would be early in 1204, before the last campaign and the fall of Rouen. Those writers who state or imply that the condemnation took place in 1203 are either late, like the chronicle of Lanercost, or are joining several events together in the usual medieval way. Nothing is more common in the historical writing of all ages than to anticipate events for the sake of clearness or through the natural association of ideas, and in the medieval chronicles, with their short annalistic entries, events are often transferred to a wrong date for the same reason. There is an excellent illustration of this in an important reference to Arthur in a chronicle of Rouen. The chief of three small chronicles of Rouen, which were first thrown into one in 1546, was the chronicle of St. Catherine. Part of this was, according to M. Chéruel, written in the first half of the thirteenth century. Its local character lends it value. Under the year 1201-an entirely wrong date-after referring to the death of Arthur, the chronicler says of John super quo a baronibus apud regem Franciae, cuius vassallus erat, quum comparere nollet, post multas citationes per iudicium parium exhaeredatus est.1

The authorities with which I have dealt hitherto may be regarded as contemporary, or as going back to a contemporary source. The Coggeshall chronicle was written up from time to time. The portion comprising the years 1202-1205 was composed before the death of Abbot Ralph in 1207, and forms a separate part. Rigord of St. Denis

1. Normanniae nova Chronica e tribus chronicis MSS. Sancti Laudi, Sanctae Catharinae, et Maioris Ecclesiae Rothomagensium collecta, nunc primum edidit e ms. codice Bibliothecae publicae Rothomagensis A. Chéruel (Mém de la Société des Antiq. de Normandie (1850), xviii, 156, separately paged, published under the final editorship of MM. Charma and Delisle).

died about the same time. Neither of them knew of Arthur's real fate. The former gives valuable details showing that Philip's suspicions had become certainties by Easter 1204; the latter says nothing at all. The Breton tradition is largely borne out by Coggeshall and shows when suspicion was first aroused. The charters are of course contemporary. On the strength of this evidence I think we might assume that Philip had sufficient cause for calling his court together to condemn John, but we could not be certain whether he did so or not. And there we should have to leave the matter.

II.

Twelve years later the English barons urged Louis of France to come over and help them. King Philip had twice before been baulked in an attempt to invade England, and he was not prepared to let this third chance slip. Both in France and at Rome the French case was justified -in France before the legate Gualo, in Rome before the pope himself. One argument upon which great stress was laid was thus expounded by Louis' proctor a fortnight after Easter at Laon, before king and legate and all the assembled barons and clergy: My lord king, it is well known (res notissima) to all that John, styled king of England, was condemned to death in your court by the judgment of his peers for his treachery to his nephew Arthur, whom he slew with his own hands, and that afterwards, because of his many crimes, he was repudiated by his barons in England,' &c. It is round this text that a famous literary controversy has been fought. M. Bémont, arguing from the silence of most authorities, from the late date of others, and from the charters of Philip Augustus, declared that Philip and Louis told a bold lie in 1216, and that it was on the strength of this assertion, and not upon other evidence, that later chroniclers believed in the con

1. The documents of 1216 are preserved by the St. Albans chronicle of Roger of Wendover, and are best seen in Matthew Paris, Chron. Mai., ii, 647.

demnation of John. Unless the proof be very positive this view is hard to maintain. It seems such a stupid lie, so easily refuted. Unless we put aside as fabrications all the documents preserved by Roger of Wendover which deal with the negotiations, it is clear that the pope and everybody else believed the story. Innocent's view was that the condemnation was not justified. The argument that these documents, somehow preserved at St. Albans, are the source of the other evidence upon the subject can only be considered when we have examined this evidence. The evidence is twofold-a marginal commentary in Matthew Paris (who follows Wendover for these years) and a rather long bit of narrative in the annals of Margam, a Cistercian abbey in Glamorganshire. Let us consider the latter first.

Like the Coggeshall chronicle, the chronicle of Margam is a brief record amplified by narrative passages. It exists in a manuscript of Trinity College, Cambridge (0. 2. 4. no. 1108). The chronicle ends abruptly and imperfectly in 1232; the manuscript belongs to about 1240. It does not seem to be the original,' and there is little evidence as to the dates of the original composition, but the part with which we are concerned was put together after 1210.2 This is noteworthy, since it reminds us that the narrative of what happened in 1203 could be connected with later events. The monks of Margam had heard, circumstantially, how John had killed Arthur in a drunken fury, on a certain day, in a certain place, at a certain time (in turre tandem Rothomagensi, feria quinta ante Pascha, post prandium, ebrius et daemonio plenus, propria manu interfecit). He had tied a stone to the body and thrown it into the Seine. It was discovered by a fisherman, recognised, and, for fear

1. There is a similar MS. with the same diagram of parhelia, ending at the same date, in the library of Trinity College, Dublin. For the Cambridge MS., see M. R. James, The Western Manuscripts in the Library of Trinity College, Cambridge, iii, 83, 84.

2. Under the year 1199 reference is made to the exile and death of William of Briouze in 1211; Ann. Monastici, ed. Luard, i, 24.