Neglected British History

NEGLECTED BRITISH HISTORY
By W. M. FLINDERS PETRIE, F.R.S.

FELLOW OF THE ACADEMY

Read November 7, 1917

By any one reading the best modern authorities on history, it would hardly be expected that the fullest account that we have of early British history is entirely ignored. While we may see a few, and contemptuous, references to Nennius or Gildas, the name of the so-called Tysilio’s Chronicle is never given, nor is any use made of its record. Yet it is of the highest value, for, as we shall see farther on, the internal evidence shows that it is based on British documents extending back to the first century. The best MS. of it appears to be in the Book of Basingwerk (W F Skene, Four Ancient Books of Wales, 11, 24.); it was printed in Welsh in the Myvyrian Archaiology, of which a second edition appeared in 1870. It was translated into English by Peter Roberts, and published in 1811, and a second edition in 1862. This translation is now so rare that I cannot hear of any obtainable copy, and could only work on it by having one of the British Museum copies type-written. Sir John Rhys, who had edited the Welsh, had never heard of an English translation, but found a copy of the first edition in the Bodleian, when I inquired of him. There is no mention of this chronicle, or use of it, by the Encyclopaedia Britannica, S. R. Gardiner, C. I. Elton, J. R. Green, Rice Holmes in his study of Caesar, Haverfield in the Cambridge Mediaeval History, nor in the great collection of the Monumenta Historica Britannica. Even Dr. Hodgkin, when discussing both Geoffrey who copied from Tysilio, and also the Breton Brut from which Tysilio originates, ignores Tysilio; as also do the recent studies by Baldwin Brown, Munro Chadwick, T. W. Shore, J. W. Jeudwine, E McClure, and Henry Sharpe. When specially dealing with Arthurian writings both Thomas Wright and Ernest Rhys refer back to Geoffrey without a hint of his source in Tysilio. Strangest of all, a recent study by Professor Lewis Jones, on the earliest Arthurian records, spends pages on Nennius, never mentions Tysilio, and then suggests that Geoffrey had a large stock of popular traditions to draw upon, although in reality there is scarcely anything about Arthur in Geoffrey that he did not draw from Tysilio. The Professor deals with Geoffrey at length, stating that ‘his use of the Brutus legend constitutes the claim of his History to rank as the first of a long series of Bruts’; yet the whole Brut legend comes from Tysilio, and the still earlier Brut in Breton of AD 940. He states that the British history said to be in Armorica ‘has never yet been discovered'; yet it is known there at least as far back as 940. He adds: ‘No document either in Welsh or in Breton has’ yet been found even remotely resembling that which Walter the archdeacon is said to have brought over from Brittany’; yet the whole document is published, with Walter’s colophon complete. Such an ignoring of public documents seems impossible; yet this is issued authoritatively by the Cambridge Press in 1911. It is justifiable, then, to speak of the Neglect of British History. This general disappearance of a book of primary importance, of which two English editions were issued in the last century, shows how easily historical material may be lost to use, even while many writers are handling the subject.

The only excuse for this neglect of Tysilio is an occasional allegation that his work is an abridgement of Geoffrey. To judge of this I have prepared a copy in parallel columns of Tysilio, Gildas, Nennius, and Geoffrey. The close connexion of Tysilio and Geoffrey is obvious throughout; the test lies in the definite statements in each which are omitted by the other. The statements peculiar to Tysilio are the lengths of reigns of four British kings and a few details; for these Geoffrey had no use in his flowery style; but if Tysilio had copied from him, why should such obscure points be introduced or invented by an abbreviator? On the other hand, two important passages occur in Geoffrey—the long account of the Diocletian persecution and the description of Maxentius, neither of which are hinted at in Tysilio, but which would have been as suitable as any others for him. One of these Geoffrey has taken from Gildas, the other I have not traced, but it might be drawn from any Roman history. Thus the test of inclusion and omission confirms the first impression, and the express statement, that Geoffrey is a flowery expansion, rather than Tysilio being an abbreviation. In this view Stephens agrees, in his Literature of the Kymry, 1876.

In naming the original authorities in this paper, the names of Tysilio, Nennius, and Gildas are merely used as brief labels, which are enough to specify certain known works. The questions of real authorship, of original dates of compositions, and of successive MSS. are quite outside of my scope here, which is only to call attention to the historical value of writings which are at present ignored. In any case the name of Tysilio has merely been given to a. chronicle by guesswork, but it is useful as a label.

It is a misfortune that the Celtic mind prefers literature to history. Celtic writers of the present day may be greatly attracted by the later Arthurian legends, and their mythologic connexions, and write on them at great length; but they will not give any of this attention to the historical discussions of the real facts, on which the immense pile of romance has been raised. The fiction occupies twenty times the space of the historical material in the Encyclopaedia. It is this constitutional frame of mind in both Welsh and Irish which, from ancient to modern times, has prejudiced the solid information which rests in their hands. Had Geoffrey not so largely dressed up the chronicle of Tysilio as literature, it would have stood a much better chance of a hearing as history; and when once Geoffrey became discredited by his method, he impaired reliance on his source.

Comparison of Caesar and Tysilio.
If the history of Tysilio be regarded as a mediaeval compilation, it must have been drawn from some classic source. Taking for comparison the most detailed part, the account of Caesar’s invasions, we may set aside at once Paterculus, Appian, and Plutarch, as they scarcely mention Britain. Livy, book cv, might have been a possible source if not drawn from Caesar, and if we can suppose this lost book to have been known in the west of England in the twelfth century, while no other MS. of his history is known here. Cotta mentioned the invasion in his work on Roman polity, but there is no reason to suppose that he wrote a history of the second invasion, in which he took part. In Dion Cassius there is very little that could not have been drawn from Caesar, and was probably so derived, though written without a Caesarian bias. It is, therefore, Caesar’s account alone that can be used to compare with Tysilio, or could have served as material to a Welsh compiler. As during this period there is nothing in Geoffrey which is not based on Tysilio, it is sufficient to compare Tysilio with Caesar, in order to see if the British or Welsh account was based upon Caesar, or if it drew from other sources. It must be expected that accounts written by opposed races should differ, not only by making intentional omissions, and by the natural tendency to dwell on successes - modern bulletins show the same, - but also by ignorance about the personages of the enemy, and ignorance about their actions behind the fighting front, about their intentions, and their plans. It is, then, not only in correspondence as to main facts, but also in one-sided discrepancies, that we may look for evidence of the truth and originality of an account.

In Tysilio the letters of summons by Caesar, and reply by Caswallon or Cassivellaunus, are like the speeches in Thucydides and Livy - what the compiler thought likely. But there is an idea of the age put in: ‘the excessive avarice of the Romans cannot suffer the inhabitants of an island, remote as this,. . . to live in peace.’ Caesar, in his recital, suppressed the plunder motive, and only lightly names tribute at the last, though he never got any. The later Romans, when there was little in the world left to plunder, impressed others by their power and tradition; but the plunder motive was the mainspring in the earlier time, and is here put forward. It is certainly not a mediaeval view of Caesar.

The gathering ground of the Britons is stated by Tysilio to have been at Doral, in Geoffrey Dorobellum. This Doral appears to be the British form of Durolevum; and as in Low Latin minuscule l might easily he mistaken for b, and u for ll, Durolevum could pass into Dorobellum. Durolevum was midway between Rochester and Canterbury. It would be an excellent rendezvous in the uncertainty whether Caesar was striking at the Channel coast, the Medway, or the Thames. Such a rendezvous would be unknown to Caesar, and naturally not mentioned by him. Tysilio represents that the landing had already taken place during the British gathering - that is to say, the main forces and leaders were not present at the landing, but only local levies, which he ignores. Now in Caesar is a long and very spirited account of the landing, the great difficulties, the dismay of the legionaries, their great confusion, and the very successful opposition of the Britons riding into the waves. Is it conceivable that a strongly British writer could have ignored all this if he were compiling from Caesar? And would he, in an imaginative work, have represented all the British leaders as being absent at such a landing? Caesar himself agrees that he was by no means happy in the business. He could barely repel the Britons, and could not pursue them because his cavalry had been unable to land. This prevented his usual good fortune, as he complacently writes. He lays stress on his difficulties, the wreck of ships at the high tide, the hopes of the Britons to cut short negotiation and attack him again, and his remaining in the dark about the British movements, which he could only suspect might happen; he describes the attack of the Britons upon the foragers, and gives another spirited description at length of the mode of fighting on chariots, the extraordinary ability of driving, and the dismay of the Romans at being thus attacked. Is it in the least credible that Tysilio, if he ever saw this account, should not have triumphantly copied it? Then storms set in, Caesar demands hostages, not one of whom are given, and only two states sent over hostages afterwards to Gaul, probably as spies. Lastly, Caesar hurried away without any material result.

The natural conclusion of the Britons is that ‘Caesar himself fled with disgrace, and with much difficulty’, and that the Gauls were against him, hearing that he had been defeated. Exactly so Caesar states that the Gauls rose, and the troops from Britain had a hard fight for four hours against 6,000 men, not the suitable greeting for a conqueror’s return.

Tysilio then states that Caesar began to build the fort of Odina, at some distance from the sea of Moran, or the Morini. There is no place mentioned with the name of Odina; but Caesar states that - among other dispositions - he had sent troops to the Lexovii (Lisieux), and the river Olina there suggests the original of Odina. If so, this gives a presumption that the British account was in Greek letters, confounding Λ and Δ. Yet the name cannot have been borrowed from Caesar, as he does not mention it. Tysilio next describes the rejoicing over the British victory, as they reasonably deemed it. This was checked by the death of Nyniaw, the brother of Caswallon, after a fight at the embarkation, which he states was with Caesar himself. As Caesar does not name it, this was probably with some subordinate commander.

The second expedition is stated by Tysilio to have been two years later; really only one winter elapsed. Here comes in the story of the stakes of iron sunk below the water to protect the passage up the Thames, and the wreck of Caesar’s ships upon them. According to Caesar he never tried to pass up the Thames, and the stakes were at a ford across the Thames. This shows confusion of detail, but it entirely disproves copying. It may be that Caesar had seized some ships on the south shore of the Thames, and tried to use them for crossing, but was checked by the stakes. Such might be turned by tradition into a defence of the Thames by stakes against shipping. Caesar states that upon landing he pushed back the Britons, ‘but forbade his men to pursue them in their flight any great distance.’ The next morning, as he was setting out, comes the crushing news of the second wreck of his transport base. He recalls all his men, fortifies his base camp, working night and day for ten days, and drawing all the ships up into the camp for safety. He thus withdrew entirely from British view. Tysilio states that, being ‘compelled to fly’, he returned to Gaul; this was a very natural inference, but shows how little the Britons knew of the Roman movements. Again, copying from Caesar or any Roman source is impossible.

Here Caesar skilfully breaks the narrative, and describes the country and people, to draw off attention from his difficulties. He confesses to vigorous attacks on his camp, and the death of a tribune, Laberius, which is probably the source of the previous British claim to the death of Labienus. The shift of date may be due to tradition; it cannot agree with copying.

Caswallon regarding the retirement of Caesar as a victory, held a great feast in London. Daring this there was tilting, in accord with the custom shown by the tilting casques (See Curle, Roman Frontier Post, 170 - 80). One nephew of Caswallon was accidentally killed by a nephew of Avarwy, who was also a nephew of Caswallon. This led to treachery, and Avarwy went over to Caesar, and offered to betray Caswallon. There is no evidence that he is the Mandubratius named by Caesar, and the parentage and the date of joining with Caesar are against it. If Caesar’s narrative had been the original of Tysilio these differences in name and detail would scarcely occur.

Caesar then states that though he had his cavalry he was no better off than before. He found that the Britons drew them on, and then dismounting, attacked the horses on foot. His men ‘were little suited to this kind of enemy’, and the Britons arranged relays of fresh men to take up the fight. On all this skill in war Tysilio is entirely silent. It seems impossible that he had ever read of it, but it is natural that a British account would not dwell on methods which were usual. Caesar next describes his attack on the city, or forest fortress, of Caswallon, which he took. He states that Caswallon wrote to the kings of Kent to attack the base. Tysilio knew the other side, which was unknown to Caesar, that Caswallon himself went to Kent to make the attack.

Caesar states the British attack on the base, and its repulse, with the result of Caswallon sending ambassadors to treat with Caesar. Tysilio states that Caswallon was personally in the defeat near Canterbury, fled to a hill fort, and after two days’ siege sent to entreat the traitor Avarwy to make peace with Caesar. Peace being concluded, with promise of a tribute of 3,000 lb. of gold and silver, they all went to London, and Caesar wintered there. This is far less favourable to the Britons than Caesar’s account, that he demanded hostages, prescribed tribute (none of which was paid), and went back to Gaul as quickly as he could. If the record of such a retreat was before Tysilio, would he have said that Caesar stayed in London? Thus it appears that the British account is in its main lines substantially in accord with Caesar, but with frequent minor discrepancies and side-lights, all naturally due to opposite points of view. Such, however, entirely disprove copying, either from Caesar or any other Latin source. The passages of Caesar which are most favourable to the Britons - the hard-fought landing against skilled horsemen, the brilliant chariot fighting later, the skilful relays in fighting and sudden dismounting, rendering Caesar’s cavalry useless - all these passages, which would have been golden to a British compiler, are never even hinted. On the other hand, Tysilio knew nothing of the two great storms, nor of Caesar’s difficulties; he does not name Mandubratius, nor any of the tribes named by Caesar; he lets out that Caswallon was personally defeated in Kent, and had to surrender; and he states that Caesar stayed the winter in London. It seems on every account to be entirely impossible to suppose that Tysilio, or his sources, were compiled from Caesar’s narrative. If not, then, as no other Latin narrative is known or would be applicable, we are bound to refer this strongly British account to a British source.

The British source was not quite contemporary, the small errors, as to Laberius being killed in the first campaign, as to the use of the stakes, and Caesar staying in London, show that some time had passed before writing. But the narrative is too close to place it much beyond the actual eyewitnesses.

With some probability we may learn more about the original document. Various places are named in it naturally, as scenes of important events, such as York, London, Winchester, Silchester, and Cirencester. But one place is named most often, and yet without any necessity. Claudius is stated to have founded Gloucester. Gweyrydd (Aruiragus) was buried at Gloucester. Lles (Lucius) died and was buried at Gloucester. Coel, who fought Asclepiodotus, was Earl of Gloucester. Eidiol, Earl of Gloucester, killed many Saxons at the Ambresbury massacre of Britons; again, he captured Hengest; again with his brother the Bishop of Gloucester, he condemned Hengest; at last, he executed him. Then the Bishop of Gloucester was elected Archbishop of London. Here in eight passages Gloucester is named in details not necessary to the history. This points to the original document of Tysilio being the chronicle of the kingdom of Gloucester. That the Roman conquest had reached the Severn at Gloucester at the beginning of the Claudian war is shown by Dion Cassius; thus there is no improbability in a Romanized Briton, such as one of the hostages educated by Augustus, having started a chronicle by A.D. 45, or just a century after the attack by Julius. Those contemporary with such a writer would have heard the personal accounts of the Britons who fought Julius. Such a chronicle, kept up at the back of the British position, could continue unbroken till far on in the Saxon conquest, and would finally pass into Wales for safety. It is in agreement with this Western source that the great revolt of Boudicca is never mentioned in Tysilio, again showing his independence of Tacitus.

The account by Dion Cassius seems to distrust a large part of Caesar’s narrative. The advance across the Thames, and the capture of the town of Cassivellaunus, are not mentioned, and in this the British account agrees. Yet this latter is not derived from Dion, as it differs in the interval between the two invasions (two years for one), in the return to Gaul after the second storm, and in Caesar staying the winter in London. Dion, therefore, is not the source of Tysilio, though both agree in disregarding Caesar’s attack on Cassivellaunus.

Notes on Tysilio, Nennius, and Gildas.
The Caesarean war, which has been discussed above, is briefly summarized by Nennius, who mistook Durobellum for a General Dolobellus, showing that the corruption from Durolevum was earlier than his writing. He keeps the mistake about Caesar having made a third transit to Britain, owing to not knowing of his stay on the coast. This is proof that Nennius only followed Tysilio, and not Caesar, and he does not name Caesar in his authorities. A fragment from Roman history then comes in about Julius. The reason that Caesar’s invasion does not appear in Gildas is obvious - it had no place in his theological tract, and it is absurd to suppose that he was therefore ignorant of it.

The Claudian war is all attributed to Claudius personally, a mixture of principal and agent which is venial. The reduction of the ‘Orkneys’ is a natural confusion of name; it merely means the seals’ island, and seals’ island, Selsey, or any other sealing island might thus get the name from the ‘Scythian’ adventurers. The affair of Genuissa, daughter of Claudius, who was brought from Rome and married to Gweyrydd, son of Cynvelin (Cunobelin), was just a piece of state policy to keep hold of a distant ruler. The ‘daughter’ was no doubt some intriguing Italian girl, nominally adopted by Claudius to give her status and protection in her remote outpost. She earned her place later by reconciling her rebellious husband to Vespasian, and getting him to swear-in to Rome, and perhaps to find the promised tribute. Of the Roman campaigns in northern Britain there is scarcely a trace in Tysilio. Meurig apparently took part as ally with the Romans, for the repulse of the Pictish invasion is credited to him. There is no trace of the great revolt of Boudicca, either in Tysilio or Nennius, true to the western nature of the chronicle. Gildas, with his Roman bias, described it, as it suited his denunciations. All through these periods there is nothing in Geoffrey which is beyond a verbal amplification of Tysilio.

The Lucius question next arises. To judge of this we must look at the whole of the statements about the rise of the British Church. We must carefully keep to the authorities, as confusion has arisen by modern authors making arbitrary identification of the east British or London family of Caswallon with the west British or Silurian family of Caradog. The actual statements of the triads name two generations before Caradog (Caratacus) and three after him - Llyr, Bran, Caradog, Cyllin, Coel, Lleirwg. From triads 18 and 35 Bran was seven years a hostage in Rome for his son Caradog - implying that Caradog was sent back to rule in Britain. The seven years, therefore, would be from A.D. 51 to 58. From Rome he ’brought the faith of Christ to the Cambrians’. Looking at the Epistle to the Romans, written A.D. 58, the obvious strength of Christianity then, its hold in Caesar’s household, where Bran was a hostage, and its political position under Nero, there is nothing in the least improbable in a British hostage in Rome being among converts by A.D. 58. In triad 62 Lleurwg, the great-grandson of Caradog, ‘first gave land, and the privilege of the country (i.e. position of native free-men) to those who first dedicated themselves to the faith of Christ’, and he founded the first archbishopric, that of Llandav. This would be about A.D. 130 to 160. Three generations for such a spread of influence, from one of the royal family is certainly not too short a time.

Next comes the account in Tysilio and the Liber Pontificalis that Lles (Lucius) sent to Eleutherius, ‘soon after his entrance upon the pontificate’, or about A.D. 180, for missioners from Rome. If the west British rulers had already started official Christianity a generation or two earlier, there is nothing unlikely in this movement. That Christianity was firmly established in even remote parts of Britain at the close of the second century is shown by Tertullian stating that ‘the Britons in parts inaccessible to the Romans, Christ has truly subdued’ (Adv. Iud.; p 180, edit. 1664). Collateral with this is the great importance of the Gallic Church under Irenaeus A.D. 180. The later stage, of the British bishops in A.D. 814 attending the Council of Arles, brings the development into the full course of ecclesiastical history. In this growth thus recorded there is not a single stage that is historically inconsistent or improbable. Further agreeing with this is the genealogy of Vortigern in Nennius (49), where, amid purely British names, Paul occurs at about A.D. 175.

The Lucius mission is named under Eleutherius in the Liber Pontificalis. Bede has the same information. Platina (Bart. Sacchi) In 1479 gives the names of the missioners as Fugatius and Damianus, and states that there were 25 flamens, of whom 3 were arch-flamens, i.e. 22 + 3, in place of whom bishops and archbishops were appointed. Tysilio gives the names Dyvan and Fagan, and the numbers as 30 and 3 superiors, i.e. 30 + 3, stating also that the three archbishops were of London, York, and Caerleon. Geoffrey copies this, except that his numbers are 28 +3. Clearly Tysilio and Platina have a common source, or the latter copied from the former. Can it be supposed that Platina, about 1475, drew from Tysilio, or Geoffrey, these details to amplify the Liber Pontificalis? Is not this the case of British history surviving at Rome, as in the work of Ponticus Virunnius, who quotes writings of Gildas which are now lost? A good example of the ignorance of editors occurs when Bede here names Marcus Antoninus Verus and his brother Aurelius Commodus, on which the comment is that no such emperors ever reigned together. Yet Marcus Verus had the name Antoninus by adoption (commonly called Aurelius), and Lucius Commodus had the name of Aurelius (commonly called Verus). That Bede gives the legal names, and not the popular names, proves that he was quoting from an official document, and knew more than his editor.

The myth of Bassianus having a British mother is a confusion, as in the succeeding account about Caron (Carausius) defeating Bassianus. This has been looked on as a total anachronism, on the supposition that there was but one Bassianus, Caracalla. There was, however, a second Bassianus of great importance, brother-in-law of Constantine, who, when on the threshold of the Empire, was executed on suspicion of a plot in AD 314(Anon. Valesii, see Gibbon). Supposing him to have been fifty years of age then, he would have been twenty-five at the revo1t of Carausius, when he might quite possibly have been in command in Britain; and he would have been born in 264 and might be a son of either of the Bassi, consuls in 258, 259. With such a personage high in the Imperial court we cannot accuse Tysilio of certain error in writing of Carausius overcoming Bassianus. The difficult question of the sources of Hector Boece here arises. He describes at length the rise of Carausius, and definitely names the Roman governor Quintus Bassianus. John of Fordun also names the governor Bassianus as sent by the emperors from Rome. It seems unlikely that these details were concocted four centuries ago out of the meagre and confused account of Bassianus in Tysilio and Geoffrey. If this be independent, we must grant original value to the sources of the Scottish historian.

Allectus is called by Boece ‘the Roman legate’, and this agrees with the Senatorial commission to Allectus stated by Tysilio. The whole of this confused period of Carausius, Allectus, Asclepiodotus, and Constantius, needs rewriting with a critical appreciation of the sources of the British and Scottish chroniclers. For the first time Geoffrey oversteps Tysilio and borrows from Gildas. It is impossible in a paper to discuss all the contacts of the several accounts, but a few points of importance should be noted. Rarely Geoffrey gets facts additional to Tysilio, such as the account of Maxentius (v. 7); hence we must not reject the statement of 100,000 Britons and 30,000 soldiers being emigrated to Gaul. The strongly Cornish character of Breton was held by Dr. Hodgkin to justify the belief in a large migration (Royal Cornwall Polytechnic Soc., 1910). These numbers are not in Tysilio; but, strikingly, Tysilio gives the numbers of the later female migrations as 1,100 and 6,000, just a tenth of those in Geoffrey’s copy. Yet these latter can hardly be his invention, as the 11,000 instead of 1,100 appear in the Brut chronicle in Brittany of about 940 (See note in Hodgkin R.C.P.S. p.12). We must look to the historical conditions. The large migration of men is often referred to later, as a cause of the weakness in face of the Picts. It must therefore have been a large part of the fighting population. They required their women-folk to follow. How many could go in a voyage? The channel shipping had been raised to a high condition by Carausius, a couple of generations earlier. If we allow that it equalled that under Edward III it cannot be an overestimate. At the battle of Sluys about 300 vessels were engaged. Caesar raised 800 vessels in the channel for his transports (v. 8) and could carry 150 men in each (iv. 22, 37). Hence there is not the least difficulty as to 10,000 or 20,000 emigrants having been afloat at once in the channel. In triad 68 the British fleet is put at 360 ships of 120 sailors each, total 43000.

With regard to the mythical matter in Geoffrey, his own declarations seem to have been disregarded. In all the period that we have been noting there is nothing more than a florid expansion of Tysilio, except in a few fresh passages, mainly from Nennius and Gildas. But at the beginning of book vii he writes: 'I had not got thus far in my history, when the subject of public discourse happening to be concerning Merlin, I was obliged to publish his prophecies at the request of my acquaintance.’ He then gives book vii, which is not in Tysilio; and continues with viii to x, including all the Arthurian French legend, which is based on Tysilio. Not till book xi does he care to vouch for his history again: ‘Of the matter now to be treated of Geoffrey of Monmouth shall be silent; but will. . . briefly relate what he found in the British book above mentioned.’ Thus he very clearly withdraws from vouching as history the whole of books viii - x. This is Herodotean caution. In book xi onward to the close Geoffrey gives a mere expansion of Tysilio. It is therefore pretty clear that Tysilio is the essential basis of Geoffrey, expanded much as Livy might have expanded his sentences. In the middle some use is made of Nennius and of Gildas, showing direct verbal copying; a few authentic pieces come from some other sources. The bulk of the mythical matter, and also much that seems least certain in Tysilio, is distinctly borrowed by Geoffrey, as romance introduced by request, but not drawn from his ancient sources. Thus Geoffrey is fully justified when he begins by stating ‘Walter archdeacon of Oxford. . . offered me a very ancient book in the British tongue which. . . related the actions of’ all the British kings: and ending that he advises other writers ‘to be silent concerning the kings of the Britons since they have not that book written in the British tongue which Walter archdeacon of Oxford brought out of Brittany’. For at the end of the Welsh Tysilio is the colophon, ‘I Walter, archdeacon of Oxford, translated this book from the Welsh into Latin, and in my old age have again translated it from the Latin into Welsh.’ That such a work did exist long before Walter is guaranteed by the ‘Brut y Brenhined, written in Brittany in the Breton dialect in the time of Athelstan (925 - 941) by an insular Briton. . . . All the main points of the story, the bringing over Maximus from Rome. . . down to the fable of the 11,000 virgins, all these are to be found in the Brut y Brenhined’. Thus writes Dr. Hodgkin, quoting from the Biographie Bretonne (R.C.P.S. 1910, p. 12). There is no reason whatever, therefore, to doubt Walter’s statement that he brought the book out of Brittany, nor Geoffrey's statement that he used Walter’s manuscript. The material plainly lies before us.

General Character of the Saxon Immigration.
When we see the strong reasons for accepting Tysilio as contemporary history, we must largely modify the current views as to the Saxon immigration, it is represented by Tysilio as a long and gradual process, fluctuating in extent, always supported by a large party of the natives, and therefore always open freely to mixture with the native population. This is entirely in accord with the various statements of the triads. ‘The Saxons came to this island in peace and by the permission of the tribe of the Cambrians, and. . . in the protection of the country and of the tribe, and by treachery. . . confederated themselves in Lloegria (England) and Alban (Scotland) where they still reside’ (9). ‘Vortigern. . . first invited the Saxons to the Island as his defenders’ (21, 100). Medrawd. . . united with the Saxons. . . who violently usurped the sovereignty of the Isle of Britain, and murdered and cruelly used every person of the Cambrian race who would not join them’ (100, 21, 45). ‘The three arrogant ones. . . brought anarchy in the Isle of Britain; and those who were influenced by this anarchy, united with the Saxons, and finally became Saxons’ (74). ‘Aeddan, the traitor of the north, who with his men made submission to the power of the Saxons, so that they might be able to support themselves by confusion and pillage under the protection of the Saxons ‘(45). ‘The Coranians are settled about the Humber. . . . The Coranians and the Saxons united and by violence and conquest brought the Lloegrians (of England) into confederacy with them. . . . And there remained none of the Lloegrians that did not become Saxons, except those that are found in Cornwall, and in the commot of Carnoban in Deira and Bernicia’ (7). From all these references it is clear that the contemporary British view was that the population submitted in most parts to the Saxons and became mixed with them. The view given by the Anglo-Saxon Chronicle is only that of the scattered series of successful battles by the Saxons, where peaceful penetration broke down; it is the most favourable view for them, yet what does it amount to? The Saxon was never more than a day’s march from a creek for his boats till after the collapse following Arthur. It is thus only a record of short raids and coast-squatting for a century. There is therefore no contradiction with the British accounts, which show that the Saxons were repeatedly beaten and ejected when they tried to hold the interior of the country. This bears strongly on the mixture of race, which left many parts more Celtic than Saxon. Abstracting the later Danish migration, the most un-Celtic regions, Sussex and Hampshire, were mainly forest and without a fixed Celtic population. The Saxon dominated where there was least existing habitation. Another point strongly shown by the internal view given by Tysilio is the frequent and ready change of allegiance. In the tribal state a personal quarrel breaks the bond, which in the national state is too strong to be cut through. Hence we see Avarwy going over to Caesar, Vortigern and his party agreeing with the Saxons, Cadwallon fighting side by side with Penda, as freely as Alcibiades or Themistocles changed sides in older days. We cannot from our present national standpoint at all estimate what men were likely to do in a state of society so different to our own. All of this again enforces the probability of continual fusion of Saxon and Briton during the immigration.

Another point of view which has grown up from unfortunately reading only the Saxon Chronicle, is that Continental immigration began suddenly with the ‘three keels’. The evidence of tradition, and of tribal names, shows that there had been a continual flow of population into Britain before the Roman age. The Atrebates, the Belgae, the Parisii, the Brigantes, and others, are equally familiar names on both sides of the channel. Nor was this process stopped even by Rome: it was only regulated. Rome brought over masses of troops largely recruited from the Continent, even to the Huns on the Wall. Aurelius brought multitudes of the Marcomanni to settle in Britain. Similarly did Probus, with the colonies of Vandals and Burgundians. The Franks raided the south and occupied London under Allectus. Constantine was accompanied by the king of the Alamanni - and doubtless a good following - when he came over to Britain. Valentinian removed Fraomar and his tribe of Alamanni into Britain. Ammianus describes the Saxons and Franks ravaging Britain in 364 and 368, and a defeat of the Saxons in 374. This last was probably connected with the settlement mentioned in Nennius, of Saxons in 374 being received into Britain (discussed in H. 163). After all this continual flow of immigrants it seems impossible to refuse the direct evidence of Continental immigration, such as the frequent finds of coins of the second and third century with burials of Continental type. By clinging to the Saxon Chronicle, and its ignorance of all that went on before 449, the archaeological evidence has been rejected and a water-tight compartment of Britain has been formulated, which was never true of any century of its history. To enter on the evidence in detail is outside the present scope, but the general view of Tysilio accords fully with the recorded migrations and the archaeological evidence.

Summary.
The general conclusions to which we are led are:
 * 1) That there was a British record of Caesar’s attack written in entire ignorance of Caesar’s account, but closely according with it.
 * 2) That this British account was the basis of the chronicle of the kingdom of Gloucester, and passed on into the history known by the name of Tysilio.
 * 3) That the Brut legend was written about the time of Claudius.
 * 4) That there is nothing improbable in all the relations with Rome, at least down to the fifth century, as represented in Tysilio.
 * 5) That statements of marvels by Geoffrey are carefully withdrawn by him from historic materials and treated as fabulous.
 * 6) That there is no doubt as to the dependence of Geoffrey on Walter, and of Walter on an earlier manuscript, probably Breton, for the British history, as stated by those writers.
 * 7) That the Hengest invasion is dated by Celtic sources to A.D. 428, and the Saxon date is in error. Arthur reigned from 467-493, thus rendering possible the account of his French expedition.
 * 8) That the Continental immigration, and mixture with the native population, was continuous from long before the Roman age, onward to our own day.
 * 9) That the historical triads were compiled from before A.D. 450 down to the twelfth century, but received no accretions since then.
 * 10) That the laws of Moelmud show the pagan British civilization, at least as early as the Roman age.

The present requirement for British History, so much neglected, is a scholar in Old Welsh, Breton, Irish, and late Latin, accustomed to palaeography, who will deal as an historian, and not as a mythologist, with the following sources:- The Brut y Brenhined of A.D. 940, in Breton; all MSS. of Tysilio, of the Historia Britonum or Nennius, and of Gildas, tracing their descent and various dates of issue; the chroniclers, as Henry of Huntingdon, Hector Boece, John of Fordun, &c., to discriminate how far other sources of material - now perished - were used by them; the Irish Annals; the Mabinogion, the triads, the laws, and other literature which may embody historical detail. From these a consecutive narrative should be framed, from which suitable outlines might some day penetrate the general school books.

The following books will be useful in preliminary studies: -

ROBERTS, Peter, Chronicle of the Kings 1811. (Bodleian, Douse T., 301.) (Tysilio.) 2nd edit. 1862. (Brit. Mus. 9510 e 2.)

GUNN, W., Historia Britonum, 1819. Original and translation.

MOMMSEN, Historia Britonum.

GILES, J. A., Historia Britonum, or Nennius, in Six Chronicles, 1885. (Bohn.)

ZIMMER, Nennius Vindicatus, 1893. (An outline of the latter is In McClure (Edmund), British Place-Names, 128, 152.)

GILDAS and GEOFFREY, in Giles’s translations, Six Chronicles, 1885. (Bohn.)

HAIGH, Daniel H., Conquest of Britain by the Saxons, 1861. Useful in parts for references to a large amount of incidental material, original results on dating, and defence of Arthur’s French wars.

POSTE, Beale, Britannic Researches, 1883. Useful references to incidental material; see pp. 194-206 on Tysilio and Geoffrey.

POSTE, Beale, Britannia Antiqua, 1857. Much incidental material. The general historical constructions of Poste and Haigh should be considered, though much vitiated by untrustworthy methods.

HODGKIN, Thomas, Cornwall and Brittany: Royal Cornwall Polytechnic Society, 1910. Valuable for reference to the Armorican Brut y Brenhined of 940, and the Riotamus Expedition.

WEBB, Percy H., ''Coins of Carausius, Numis. Chron.'' 1908. Text of authorities for the period and full statement of coins.

SHARPE, Henry, Britain B.C. 1910. A geographical study of coast changes in relation to Caesar’s invasions which are discussed with all classical authorities.

PROBERT, William, Ancient Laws of Cambria 1823. Laws of Moelmud and of Howel, and the historical triads.

SKENE, W. F., The Four Ancient Books of Wales. 1868.

STEPHEN, Thomas, Literature of the Kymry. 1876.