the abbot wrote before 1207. Yet, on the other hand, as William the Breton wrote his poem in three years, this part could not have been composed much earlier than 1214, in any case after the flight of William of Briouze to France. He was in almost constant attendance upon Philip, and likely to hear what was going on. He would be interested in the famous fugitive who had experienced such a turn of fortune and fled like a beggar from the English coast. Is it not possible that at last the full story of the murder was known at the French court, and that in the Philippid we get the tale—naturally favourable to William of Briouze—which is found elsewhere only in the chronicle of a Welsh abbey? This would partly account for the terror and atrocities of John during these years, for the alliance between Philip and the English barons, and for the projected invasion. It would be tempting to suggest that it was then that Philip summoned John to appear for his crime; but this is impossible.

This analysis has, I think, enabled us to form a juster idea of the value of the Margam chronicle, and to trace to some extent the origin of the most detailed account which has come down to us of Arthur's death. I have maintained that the Margam narrative is to be regarded as a whole, and therefore, unless very serious evidence were brought against it, we are forced to the belief that Philip's court probably did condemn John a second time. Louis' proctor in 1216 said he was condemned to death; the Margam chronicle and later tradition are content to say that he was sentenced to lose all his continental possessions. It is quite possible that, after the revelations and awful crimes of 1210, when John was excommunicated, and Philip had been urged by the pope to deprive him entirely, Philip's court had proceeded to a sentence of death. The language used in 1216 suggests that the repudiation of allegiance by the English barons followed the French judgment after no very long interval. Still, this is only possible. What seems impossible is that Louis told a lie in 1216 and that the annals of Margam, the tradition in Brittany, and the independent testimony of Matthew Paris are at fault. With the argument that the condemnation must have taken place in 1203 I have dealt already; it depends on the partly erroneous belief of our authorities that it caused the loss of Normandy. Yet everybody would agree that the death of Arthur gave strength to the French king, and if so a formal sentence of confiscation, as soon as he was sure of Arthur's death, would strengthen him much more. The other arguments against the condemnation are negative—the late and unsatisfactory nature of the authorities and the silence of the chief records and chronicles. But we have seen that the annals of Margam are not so very unsatisfactory after all. There is

There is very late testimony to the condemnation, which has been rejected by M. Petit-Dutaillis with some contumely. This is the marginal note inserted by Matthew Paris in the documents preserved by Roger of Wendover. Matthew breaks in to tell the true story. What really happened, he says, was this : King John sent the bishop of Ely and Hubert de Burgh to Philip to say that he was ready to stand a trial, but Philip insisted on his presence without a safe-conduct. The embassy replied that, even if the duke of Normandy could attend, the king of England could hardly do so without a guarantee of safety. And so the magnates Francie proceeded to condemn him unjustly in his absence. It is probable that this late story is not quite true. Eustace of Ely was certainly one of the embassy of 1204, and may have been sent on a special errand as well. That Hubert de Burgh went is not so likely. But the story is not to be dismissed summarily because Matthew Paris

1. Revue Historique, lxxi (1899), p. 35. 2. Chron. Mai., ii, 658. For Eustace, bishop of Ely, see Coggeshall (p. 144), whose narrative is not at all a bad parallel to Matthew Paris. Hubert de Burgh was at this time custodian of Chinon, but it is quite possible that he was engaged in another capacity in the early months of 1204. Note how studiously vague the Marshal's biographer is about the proposals of peace (iii, 176).

sometimes makes a blunder; for it clearly represents an independent tradition-independent, that is, of the document of 1216-and therefore corroborates, so far as it is worth anything, the Margam annals. 1

III.

I should say a word about the last important argument used by M. Bémont and his followers, the argument from silence. It may be admitted that this is invalid so far as the chroniclers were concerned; if the murder passed unrecorded, the condemnation obviously would also; but what about the French registers and the papal registers; and why did not William the Breton, who says so much about Arthur, enlarge upon the condemnation? But the French registers were not kept systematically like the English records, and there is no mention of any condemnation upon them or in Philip's charters. Philip wrote about the first trial to the pope, but our only authority is the pope's answer; no official record would tell us anything. The French court of 'peers ' was like the English curia regisin its broadest sense—in this, that its proceedings could pass unnoticed by the ordinary man if they were not recorded. John's trial after Richard's return passed almost unnoticed in England. Everything was very informal, and the trial of John is really of importance to the French historian and jurist because it seems to suggest

1. Since this article appeared in the English Historical Review, it has occurred to me that Matthew Paris may have been confusing his dates, and be really describing the negotiations which preceded the admitted condemnation in 1202 (above, p. 219). The bishop of Ely, as John's letters show (Rot. Pat., 10b) was one of John's spokesmen with Philip, and at this very time Hubert de Burgh was evidently in high favour with the king, and entrusted with important tasks (Rot. Pat., 6b, 7b, 9, 9b, 11). If this view be correct, it is of course impossible to use the passage in Matthew Paris in support of a second condemnation. But I prefer the view that the chronicler is correctly referring to the

а

a

the beginnings of something a little more formal. I have purposely avoided all the juridical arguments of M. Guilhiermoz; if the historical evidence is lacking, the judicial can hardly be adduced; but although I think the historical evidence is sufficient to allow us to believe in the condemnation, I would also urge that these semi-legal, semi-political, proceedings would easily escape the attention of contemporaries. They hardly form a theme for the chaplain's epic. He was content to say that Philip hastened to take vengeance, that Iohanni retribui possit pro morte nepotis, and this is not altogether unjuridical.2 Since John did not appear, the trial would be short, and all the more easily disregarded.

Great stress, again, has been laid on the silence of the papal letters of 1203. If the trial took place later this is not surprising. 3 And after all, it is not hard to see why Innocent should refrain from mentioning the subject. The point is that he does not mention the disappearance of Arthur, of which he must have heard. It is certain that Arthur disappeared, yet there is no allusion to him; surely then it is rather illogical to say that John was not tried for the death of Arthur, because the pope does not refer to the trial. At this time Innocent was anxious to bring about peace between Philip and John in the interests of the king of the Romans, Otto. He was also in the midst of his efforts to rescue the unfortunate wife of Philip, Ingeborg, from her imprisonment. So far as he took sides he was certainly supporting John rather than Philip. One

3

1. How relatively unimportant the undeniable (first) trial was is seen from any consecutive account of the French court, e.g., Viollet, Hist. des Institutions Politiques, iii, 301-2.

2. v. 16 (Delaborde, ii, 177). 3. Innocent, in his well known letter to the Norman bishops in 1205, refers to the sentence of Philip's court as Philip's plea in justification of his attack on Normandy (above, p. 405), but it is probable that the pope was referring to the condemnation of 1202.

4. Scheffer-Boichorst in Forschungen zur deutschen Geschichte, viii (1868), 511-6.

a

English chronicler, who was in the way of knowing, states definitely that it was part of the papal legate's duty to find out exactly what had happened to Arthur. The documents of 1216 show that the pope had got some information, and professed to think that John's action was justified. After his quarrel with John he doubtless may have made much of the death of Arthur; but here a significant fact appears to show us how vain is this argument from silence. On 31 October 1213 he wrote to Nicholas, bishop of Tusculum, his legate in France, ordering him to collect and destroy by fire every letter which he had written against John to the English bishops, whether before or after the interdict of March 1208, and especially one letter which had been distributed through France, England, Scotland, Ireland, and in the bishoprics of Liége and Utrecht.2 Surely we can no longer wonder that Innocent's letters tell us nothing of the fate of Arthur. It is a curious and noteworthy fact that the chancery rolls for the very years when John was busiest in his furious attacks on the clergy and barons have also been destroyed.

IV. In the previous inquiry I have taken up and examined the arguments used by M. Bémont to controvert the statements that King John was condemned for the death of Arthur by the court of Philip of France. The whole evidence with regard to the murder of Arthur has in this way been brought before the reader, and we have seen that the chronicle of Margam, which is most explicit in affirming the fact of the trial of the murderer, is also best informed on the details of the murder. Setting on one side the problem of the trial, I will, in conclusion, bring together the scattered evidence which tends to confirm the story of the crime as told by the annalist of Margam and William the Breton.

1. Gervase of Canterbury, ii, 95.

2. See Epist., xvi, 133, in Migne, Patrol. Lat., ccxvi, 926. Potthast, no. 4837.