The Laws of Howel the Good/Introduction

by Hywel ap Cadell

Rogo ut omnis lector, qui legerit hunc librum, det veniam mihi, qui ausus sum post tantos haec tanta scribere quasi garrula avis vel quasi quidam invalidus arbiter. Cedo illi qui plus noverit in ista peritia satis quam ego.—NENNIUS.

Y mae e'n wir yn orchwyl dyrus ddigon i chwilio allan Ddechreuad ein Cenedl ni yn gowir ac yn ddiwyrgam, a'i holrhain o'i haberoedd i lygad y ffynnon. Ond mi a amcanaf symud ymaith y niwl oddiar y ffordd, fel y bo ein taith at y gwirionedd yn eglur.—THEOPHILUS EVANS.

The notion that the Welsh came to the Isle of Britain with the grasshopper has been dispelled by modern research.—EGERTON PHILLIMORE.

INTRODUCTION

Not one of the law books bearing the name of Howel Dda, which have come down to us from the Welsh medieval age, is older than the last quarter of the twelfth century, that is, about 250 years after Howel's death. The earliest of all, the Peniarth MS. 28,^[1] is written in Latin with many Welsh terms, phrases, and short passages left untranslated. Next to this comes the Peniarth MS. 29 (MS. A), sometimes known as the Black Book of Chirk, and written in Welsh about 1200. Neither of these professes to be the original codex of the White House, nor does that claim appear to be made by any MS. of the laws now extant.
These Welsh medieval law books bear so strong a general resemblance to one another that it can hardly be doubted but that they are all based on some one ultimate original, which, in our present state of information, we may suppose to have been a 'Book of the White House'.^[2] Those written in Welsh, however, certainly fall into three distinct classes, each of which begins with its own peculiar type of preface. They may be distinguished as follows : —
(a) Those which refer exclusively to the King of Aberffraw in North Wales,^[3] and which give other indications that they pertain to the kingdom of Gwynedd in N.W. Wales, of which Aberffraw was the chief royal residence. Aneurin Owen dubbed them the 'Venedotian Code', that is, the code of Venedotia or Gwynedd, a name with which we need not on the whole quarrel. As it will be necessary, however, to diverge from Owen's other designations, this class will be distinguished here as the Book of Gwynedd. The chief exemplar is the Peniarth MS. 29 (MS. A) referred to above. References to a certain lorwerth ap Madog^[4] indicate his influence as a jurist on this class, but they are such as show that the Book of Gwynedd was regarded as existing before his time.
(b) Those which refer exclusively to the King of Dinevwr in South Wales,^[5] but are void of any other reference such as would lead one to associate them in any special degree with that Deheubarth of which Dinevwr was held to be the chief royal residence. From a passage in the preface it appears that their original was written not only outside Deheubarth but in Powys and by a Powysian.^[6] Is it possible that they represent what Aneurin Owen would have called the 'Powysian Code'? Unfortunately he styled them the 'Gwentian Code' as being the code adapted to Gwent or Southeast Wales, for which there appears to be no evidence of any kind.^[7] A peculiarity of the preface of this class of law book is that it refers its compilation to a certain Morgeneu and his son Cyvnerth. Elsewhere^[8] it is Cyvnerth ab Morgeneu who is referred to as a well-known 'jurist', for which reason this class will be distinguished here as the 'Book of Cyvnerth'. The text adopted by Aneurin Owen as the basis of his amalgam of this type of law book is the Peniarth MS. 37 (MS. U), 'not from any superiority but as being the simplest.'^[9]
(c) Those which refer both to the King of Dinevwr and to the King of Aberffraw,^[10] stating expressly that of all the kings in Wales gold is payable to these two only. The King of Dinevwr, however, is mentioned first in order, and there is a special section devoted to him. One of the Dinevwr kings is also mentioned by name, viz. Rhys ab Gruffydd,^[11] sovereign of Deheubarth, who died in 1197. There is also found a section dealing with the seven bishop-houses in Dyved,^[11] one of the patrias included in the Deheubarth. This class therefore appears to represent a 'Book of Deheubarth'. Unfortunately again Aneurin Owen named them the 'Dimetian Code', that is, the code of Dyved, although there appears to be no reason why they should be confined to this particular patria, and not made to cover the whole of that Deheubarth which was held to be subject to the King of Dinevwr. The preface of this class mentions a certain Blegywryd who is described as the cleric appointed in Howel's convention at the White House to reduce the revised laws to writing. Nothing more appears to be known of this Blegywryd beyond what is stated in this preface. Hence this particular group of law books may not inaptly be styled the Book of Blegywryd. Aneurin Owen adopted the Cotton MS., Titus D IX (MS. L), of the second quarter of the fourteenth century,^[12] as his representative text of this group. Dr. Gwenogfryn Evans regards the Peniarth MS. 36 A (MS. O), as the oldest now known, having been written shortly after 1282^[13] but according to Owen 'the variations in the manuscripts of this class are but few', for which reason he is of opinion that 'they perhaps exhibit the nearest affinity to the original compilation sanctioned by Howel'.^[14] It should be noticed that the earliest of all the existing law books bearing Howel's name, viz. the Latin Peniarth MS. 28, is of the form of the Book of Blegywryd, as also the important Latin Cotton MS., Vespasian E XI, written about 1250, 1 although the name of Blegywryd is absent from both.^[15]
   The following passage^[16] from the preface to the Book of Blegywryd is very suggestive as to these three types of law books, and appears also to throw some light on the puzzling matter of King Howel's dominions.

Guedy hynny yd erchis gwnneuthur tri llifuyr kyureith : vn vrth y lys peunydyaul pressuyl y gyt ac ef ; aral y lys Dineuur ; y trydyd y lys Aberffraw, megys y caffey teir rann Kymry, nyt amgen, Gwyned, Pwys, Deheubarth, audurdawt kyureith yn eu plith vrth eu reit yn wastat ac yn parawt.
After that he ordered three law books to be made: one for the daily court to be always with him; another for the court of Dinevwr; the third for the court of Aberffraw, so that the three divisions of Cymru, to wit, Gwynedd, Powys, and Deheubarth, should have the authority of law amongst them at their need, always and ready.

Here it is clearly implied that the king's daily court was not in Deheubarth but in Powys. The common opinion is that Howel Dda lived in Deheubarth and especially in Dyved,^[17] but in the light of the above passage he generally dwells in Powys. Moreover, the manner in which the White House is spoken of as being the king's hunting-lodge 'when he came to Dyved' seems to bear out the same idea. It is true that the preface to the Book of Cyvnerth appears to restrict Howel's dominions outside Powys, and it is curious that Powys appears to be the very division of Wales wherein that compilation had its origin. How to reconcile these apparent contradictions does not at present appear. The above passage would seem to suggest that there were three types of law books, those of Gwynedd, Powys, and Deheubarth respectively, although, as the passage now stands, it means no more than that three copies of one original were made for the three divisions of Cymru. It may be that in time they each underwent such modifications as adapted them more perfectly to the varying customs of each division. If, however, in the case of the Book of Cyvnerth, we are dealing with a 'Powysian Code', how shall we explain the mention of Dinevwr and the absence of all reference to any chief royal residence in Powys such as the Mathraval mentioned in later texts?^[18] It seems therefore advisable for the time being to abandon territorial designations for the two Dinevwr classes of law books, and to style them after the names of the 'jurists' preferred in their respective prefaces. The designations therefore tentatively proposed for the three kinds of Welsh law books in lieu of those invented by Aneurin Owen are as follows : —
Book of Gwynedd for Venedotian Code,
Book of Cyvnerth for Gwentian Code,
Book of Blegywryd for Dimetian Code.

I

Our present text, the Harleian MS. 4353 (MS. V) belongs to the second of these classes, viz. the Book of Cyvnerth, being, according to Dr. Gwenogfryn Evans,^[19] 'the oldest and most important' of this kind. Aneurin Owen had six codices of this class before him, which he denominated U, V, W, X, Y, and Z respectively. He noticed that some of them closely resembled the Book of Blegywryd, so much so indeed in the case of Y and Z that he soon ceased to use them in his edition. Two others, viz. V (our present text) and W, which are very closely allied, also resemble in some respects the Book of Blegywryd, especially V, which actually contains the passage on the bishop-houses of Dyved. The leading peculiarity of these two MSS., however, is this, that in their prefaces the name of Blegywryd appears in lieu of that of Cyvnerth and Morgeneu, and indeed appears to have been substituted for them. One was almost tempted on this account to distinguish these two codices, V and W, by some such name as the 'Composite Book of Cyvnerth and Blegywryd', a description which further investigation may yet substantiate. The two remaining codices, U and X, are much smaller in bulk than the two last, and might be supposed to approximate nearer to the original Book of Cyvnerth. If they were as closely allied as V and W, one might indeed think so, but they differ considerably in their arrangement, appear to be much condensed, and are both somewhat carelessly written. Their matter is practically all comprised within the present text, the few additions, which each contains, being given in the appendix.

V and W^[20]

V = Harleian MS. 4353. Vellum ; 7¾ x 5 inches ; folios 1-3, (4-5), 6-27, (28), 29-45, the three folios in round brackets being insertions on later material in a hand of about 1600, copied from X; written according to Dr. Gwenogfryn Evans about 1285 by the same scribe as wrote Peniarth MSS. 2 and 6 Part iv, and Mostyn MS. 117; 'it is curious that all the MSS. written in the same hand are imperfect'; coloured capitals, generally in red and chocolate alternately; 25 lines to the page (except 16b and 40a which have 24 and 26 lines respectively) ; the first and last pages are so stained and worn that it is difficult to be always certain of the reading ; one pagination in ink till folio 37, after which till last folio but one there are two paginations, one in ink (39 to 45), and one in pencil (38 to 44), which last is the official numbering of the British Museum, followed in this present work, the last folio being paginated in ink as 45 ; half bound in morocco. Most of the marginal index words are in the hand of Jaspar Griffith. ' Liber Humfredi Wanley A.D. 1714' (I b) ; 'Sum liber Jaſpar Gryffyth 1586[-1714=128] f (2 a); 'Yma y gellir craſſu a gweled dau beth. * I. Yn gyntaf pan yſcrifenned y llyfr hwn ſod yr offeirieid yn berchen gwragedd priawd, o ran bod breint yma wedi ei oſod i ferched offeirieid. 2. Yn ail mae yn gyffelyp yſcrifennu y llyfr hwn cyn gwahardd priodas ir offeirieid. Yr hon waharddedigaeth a gymmerth rym (?) yn Eglwys Loegr ynghylch y flwyddyn 1100 yn niwedd teyrnafiad Willm Rufus, edrych fol 44 a ' (3 b and 4 a bottom margins. The asterisk refers to 4 a, line 3) ; ' Rys ap howel ap Jeuan ap gwalter ddugan cof (?) ' (18 a right margin from top to bottom) ; ' Mae yma ddalen yn eifiau ' (31 b and 37 b) ; ' Timothy Middleton ' (32 a) ; ' Timothy Middleton his booke douth Owe '(43 a).
W = Cleopatra A XIV. Vellum ; 6½ x 4¾ inches ; folios 34-107 in pencil pagination ; first quarter of the fourteenth century ; coloured capitals in blue and red alternately, save that the larger ones are in blue with red foliations ; ẏ dotted throughout, and gu for gw ; 21 lines to the page, except 55 a from line 5, 55 b from line 13, 94 a from line 7, 94 b, 95, 107 a from line 9, and 107 b, which were left blank by the original scribe ; portions of the text are in the margins on folios 42 a, 42 b, 57 a and 83 b ; three paginations, two in ink and one in pencil, which last is the official one of the British Museum, followed in this present work ; two and a half lines on 101b( = V45a 10-12) are almost stained out by some prudish person ; bound in calf along with some Latin MSS. Besides occasional marginal index words, we have ' Liber Cardiff de Conſuetudinibus Walliae ', ' Leges Howeli Dha Wallice ' Robertus Cotton Bruceus ' (34 a) ; also much scribbling on folios 43, 44, 55, 94, 95 and 107, wherein occur proper names— ' Sciant prefentes et futuri quod Ego Johannes filli dedo conceſſi in hac prefenti Carta ' (43 b) ; ' Sciant prefentes et futuri quod Ego johannes (?) vabe ll (?) dedi conceſſi ' (44 a) ; ' Johannes vechan ' Jeuan ap phelippe hir dd ap fillippe hir ' (55 a) ; 'Johannes ap gwill (?) ', ' Willmus' (?) (55 b) ; '—vabe rimẏ', 'Jeuan ap dd ap—', 'Handrods dekerfilly in die martis,' ' Roberto ', 'Th et buon anne coſe nant per ta ' (94 b) ; '—ap blethyn joŕ ap r . . . . ap—' [k]arfill die—' ' Hoell ap—' (95 b). These names (says Dr. Evans) are in a fifteenth-century hand, but more or less intentionally deleted by rubbing.

Y and Z

Y = 'a manuscript presented by the Rev. Mr. Conybeare to the Literary Society of Neath, by whom the use of it for this work [viz. Ancient Laws and Institutes of Wales, MDCCCXLI] was kindly afforded to the Editor. It may be attributed to the middle of the fourteenth century.' So writes Aneurin Owen in the preface to his book on May 1, 1841. Mr. Egerton Phillimore in a note in Y Cymmrodor, vol. IX, states of this codex that it has been ' lost since before 1860 '.^[21] It appears to have contained the first part on the Laws of the Court as far as V 12 a 19, proceeding immediately to the Laws of the Gwlad, but agreeing so closely with the Book of Blegywryd, that Aneurin Owen ceases altogether to refer to it.^[22] One can therefore only surmise that it followed the Book of Cyvnerth as far as the point referred to, after which it followed the Book of Blegywryd.

Z^[23] = Peniarth MS. 259. Paper; 11x8 inches; folios a-e, 1-103 ; imperfect (folios b-e, 6-7, 13-20, 44, 47-8, 51, 56, 59-60, 99-101 being blank leaves inserted by binder) ; bicolumnar ; in two distinct hands ; first half of sixteenth century ; bound in leather with Peniarth MS. 259 A (MS. P). 'The text of folios 1-46 belongs to the class of which V or Harleian MS. 4353 is the prototype. This copy is a kind of selection arranged differently ; it is imperfect and corrupt. . . . The order of the text is very different.' A fresh hand begins at folio 49, being contemporary with the first. The latter inserts the following note on a passage written by the former— 'Hyn ydoedd wydi i scrivenu yny llyfr y copied hwn o hono. Y llyfr hwnn a gavos Einiawn ap adda pan vv yngharchar ymhwmfred gan y cunstabyl ai kavas gan brior y vynachlog a hanoedd o dehevbarth, ac nid oes athrondyst ar gyfraith namyn y sydd yn y llyfr hwn kysdal a hwn.' Aneurin Owen in his edition of the edition of the 'Gwentian Code' ceases using this codex at the very same point where he metes out like treatment to Y ; and he states of Z at the beginning of the Laws of the Gwlad that it ' is carelessly transcribed and has many chasms ', for which reason he leaves it. He inserts variant readings, however, from Z in vol. II of his work. Z is the codex which with S (the Brit. Mus. Additional MS. 22356 of the late fifteenth century) provides Owen with an interesting but extremely untrustworthy addition to the preface of his ' Dimetian Code '.^[24]

U and X

U = Peniarth MS. 37. Vellum ; 5⅜ x 4⅛ inches ; 156 pages (pp. 153-6 being in court hand) ; late thirteenth century, in the same hand apparently as Peniarth MS. 35 (MS. G) with very numerous sectional initials and titles in rubrics, and also rubricated letters ; 18 lines to the page ; partly gall-stained but complete ; in old binding newly covered with pigskin. The text of pp. 131-52 is no part of the Book of Cyvnerth, but is taken from the Book of Gwynedd, being found in A and its important transcript E. Dr. Evans, however, finds that it is in such close agreement with the corresponding part in G that both must be from the same archetype or the one is a copy of the other, both MSS. belonging to the same school of writing and being possibly the work of the same scribe. It will be found reproduced with translation in Y Cymmrodor, vol. XVII. The Book of Cyvnerth, properly so called, covers the first 120 pages, and was adopted by Aneurin Owen as the basis of his ' Gwentian Code '. On the whole his edition is trustworthy, following the order of his original and giving adequate notice when he fails to do so. He rarely or never expands contractions, and does not even reproduce them, but in the present case this involves no serious consequences as they are rarely of greater importance than ran for rann, or edlig for edling. The following are the only serious discrepancies :— p. 712, 1. 5. abu(vch ei)thyr (eu teithi) for abuuvch. p. 722, 1. 13. Add Tavlbort o afgvrn Moruil. dec arugemt y gwerth. p. 764, 1. 29. Section XII is erroneously bracketed, as it appears in U 49 a 6-10 as follows:— Or cledir pvll odyn ar tir dyn arall heb ganhyat pedeir. k. k'. ageift perchennavc y tir gan yneb ae cladho althri buhyn camlvrv yr bren.

X = British Museum Cotton MS. Cleopatra B V. Vellum; 7½x5½ inches; folios 165-222 (222 a 8-22 added by another hand) ; written about 1350 ; coloured capitals, generally red and blue alternately, five of which are illuminated, viz. 165 a, 184 b (cut out), 185 a (two on this page) and 200 a ; 20 lines to the page ; two paginations in ink and pencil, both the same, the latter being the official numbering of the Brit. Museum ; v^[25] = u or v, and not w ; bound with other works in Russian leather. ' Leges Howelli Boni principis Walli(ae) in Lingua Britanica ' (165 a) ; ' Cyfnerth mab Morgeneu yn gyntaf a ſcrifennodd ac a ddoſparthodd y llyfr yma ar y dull ar wedd hon. Jaſp. Gryff. 1600' (165 b) ; ' [ — ] y gwelir [ — ] or offeiriaid [ — ] briodol y [ — ] [pr]yd hynny ' (168b); the catchwords Morwyn yſtauell in a fish cut through by bookbinder (176b) ' habet hie liber quinquaginta & octo folia ' (222 a in Jasper Gryffyth's hand). As compared with V, ' the wording is often changed and abbreviated, many passages being omitted and a few others inserted.'

It will be noticed that none of the codices now extant of the Book of Cyvnerth and the Book of Blegywryd date from before the last quarter of the thirteenth century, probably after the death of Llewelyn ap Gruffydd in 1282. Those which are antecedent to this period are the following, which are enumerated in order of time : —

1. Peniarth MS. 28. Latin; 1175-1200.
2. Peniarth MS. 29 (= MS. A). Welsh; Book of Gwynedd ; about 1200.
3. Harleian MS. 1796. Latin; 1200-1250.^[26]
4. Brit. Mus. Additional MS. 14931 (= E). Welsh ; Book of Gwynedd ; about 1250.^[26]
5. Caligula A III (= MS. C). Welsh; Book of Gwynedd ; about 1250.^[26]
6. Vespasian E XI. Latin ; about 1250.^[26]

All the earliest and best MSS. extant therefore of the Laws of Howel Dda were written at a time when the Normans had long interfered with Welsh affairs and had taken permanent possession of the majority of the patrias of South Wales. It is very important to bear this in mind, inasmuch as the codices, which are confessedly in a state of flux, cannot fail to reflect the political situation in Wales as it was at the time of writing.

II

Throughout the twelfth and thirteenth centuries what may be called Welsh Wales, as distinct from Norman Wales,^[27] was divided into the three main divisions of Gwynedd, Powys, and Deheubarth. As the result of the important battle of Mynydd Carn in 1079,^[28] Gwynedd and Deheubarth henceforth remained under the rule of the House of Gruffydd ap Cynan and that of Rhys ap Tewdwr, respectively. Powys continued as before to be governed by the House of Bleddyn. These three families were all sprung from Rhodri the Great and were consequently of the true Cymric lineage of Cunedda Wledig. Norman Wales throughout the same period comprised the patrias of Morgannwg with Gwent, Brycheiniog, and Dyved, being roughly equivalent to the modern counties of Glamorgan with Monmouth, Brecon, and Pembroke respectively. The districts now known as Cardiganshire, Radnorshire, and Flintshire fluctuated, being sometimes held by the Welsh and sometimes by the Normans. Seeing, then, that our earliest codex dates from well within this period, and that its successors clearly show that the codification was subject to continual re-arrangement and other modifications, it must be allowed, as we have said above, that in reading them the political situation as it was in these two centuries is by no means to be disregarded. The law is the law of Howel, but it is the law of Howel as modified and amplified both by the varying customs of different parts of Wales and also by the changes which are taking place throughout three and a half critical centuries in the general life of the people.

What share King Howel had in the codification of Welsh law and custom in the tenth century is not easy to determine, especially as the earliest account of the convention which he is said to have assembled at the White House is over two centuries later than his time.^[29] Our earliest chronicle also, the so-called Annales Cambriae,^[30] completed only a few years after his death, is silent as to any activity he may have displayed in this direction, and contains no reference of any kind to the alleged convention. All the codices, however, agree in associating his name with the formulation of the laws of Cymru, frequently appealing to his authority and indicating the fact when they have occasion to depart from it or to add thereto. This unanimous testimony of the codices is corroborated by the nature of the few facts which are known of his career. By the death of his father and paternal uncles, the sons of Rhodri the Great, he rose steadily in power. He had married Elen, the daughter of the King of Dyved, by which he became king of that country.^[31] There is evidence which goes to show that he was by inheritance ruler of Powys, and as we find him laying claims to portions of Gwent in the far south-east,^[32] this, with other indications,^[33] makes it almost certain that Brycheiniog, which lay between him and Gwent, was also in his grasp. After the death of his cousin, King Idwal Voel of Gwynedd, in 943, he must have been easily supreme throughout the whole of Wales, although the realm of the king of Morgannwg appears not to have been brought under the sway of the family of Rhodri in the sense that the rest of Wales was subject to that house. Howel therefore between 943 and 950 was clearly in an excellent position to move with regard to the revision and codification of Welsh law and custom, if so minded ; and the evidence that he was so minded is ample. In the year 928 he had made a pilgrimage to Rome. He frequently attended the meetings of the Witenagemot of the Wessex kings, for his name appears as witness to several charters ranging from 931 to 949.^[34] He was thus clearly on intimate terms with the royal house of Wessex, and was thereby under the direct influence of the traditions of Alfred the Great, not to mention the general effect in the same direction which Asser must have produced on the life of Wales, particularly in Dyved.^[35] For Asser would spend six months with Alfred and six months in his own Britannia in his native Dyved.^[36] Through the same traditions there was operating also the influence of Charlemagne, to say nothing of this same influence as it may have operated through Howel's own grandfather, Rhodri the Great. Indeed, it can hardly be doubted that the fame and character of Charlemagne, Alfred the Great, and his own grandfather Rhodri acted powerfully on the mind of Howel, whose own life appears to be in emulation of theirs. We find that our earliest Welsh chronicle, accompanied by thirty-one invaluable pedigrees with other material, and attached to a copy of the historical compilation which goes under the name of Nennius, was completed (probably at St. David's)^[37] a few years after his death in 950 a fact which points to its having been accomplished under his patronage, if not at his direction. He stands unique among the kings and princes of old Wales as being the only one who is known to have struck coin.^[38] His reign was marked by unusual peace. And that he was in general an enlightened and a beneficent ruler we need no surer proof than the noteworthy fact that he is known in history as Howel the Good. It is only, however, as seen in the general history of Wales up to his time that the significance of his reign becomes apparent, how in particular it marks a noteworthy advance in the emergence of the entity we now know as Wales from the conditions which prevailed in the dim centuries of Roman Britain. It would require far more space than is at our present command to provide any adequate presentation of this subject, even if this were as yet possible. The main outlines, however, of the story are quite clear.

III

Roman Britain was treated as a single province till Severus (who died in A.D. 211) divided it into two, called Lower and Upper Britain, Britannia Inferior and Britannia Superior,^[39] so that henceforward the term Britannia came to be used not only for the island or even for Roman Britain, but also for portions of Roman Britain which was now known as Britanniae or the Britains. Dion Cassius^[40] gives us to understand that the legions at Caerlleon on the Usk and Chester on the Dee, were in Upper Britain, while that located at York was in Lower Britain. As the Romans, like other people, allowed the ready test of running water to decide what was upper and what was lower, it is natural to suppose that Upper Britain was mainly that part of Roman Britain which the legions had to approach by marching in the direction of the sources of the Thames and of the streams which meet to form the Humber. When, however, Upper and Lower Britain came to be distinguished as provinces, the question of what was expedient would also play its part in the new arrangements. And as the territory north of Chester would go more conveniently both for geographical and military reasons with that north of the Humber, the whole of this district falling under the surveillance of the official who resided at York, which we know to have been in Lower Britain, it is in no way improbable that Upper Britain as a province would be entirely excluded from what is now the north of England and would be confined to a territory south of Chester and including it. This then leaves us the country around the upper reaches of the Thames, and all to the west of it, including Wales plus the Devonian peninsula. Without for the moment attempting to define closely its eastern boundary we identify Upper Britain, Britannia Superior, with the territory west of a line drawn from Chester (which is included) to the Wiltshire Avon or thereabouts. The western portion of the Devonian peninsula, especially the country beyond the river Exe, was one of the least Romanized parts of Roman Britain, and Wales being a purely military district was similar in this respect, so that they would not inappropriately go together, being connected by the more Romanized region round about the estuary of the Severn.^[41] In 397 Diocletian divided Roman Britain into four provinces instead of two and called them Britannia Prima, Britannia Secunda, Flavia Caesariensis, and Maxima Caesariensis. As the names clearly imply, we have here nothing more than a renaming of the two old provinces into Britannia and Caesariensis, which are subdivided into Prima and Secunda, and Flavia and Maxima respectively. And as it is certain that Cirencester was in Britannia Prima,^[42] we conclude that by Diocletian's arrangement Upper Britain became exclusively known as Britannia, whilst Lower Britain was given the new name of Caesariensis. Moreover, as Cirencester was in Britannia Prima, we would also conclude that it was the Severn Sea which was the cause of the subdivision, and that therefore Wales was included in Britannia Secunda. Each of these Britannias was ruled by a governor called praeses or president, but the military command was in the hands of another official, who was called the Comes Brittaniae.
Whether the reasoning just elaborated will be substantiated or otherwise by fresh discoveries, this at least is certain, that it is unquestionably to the kings and ecclesiastics of the smaller Britannia which we have just delineated that St. Gildas, who died after the middle of the sixth century, addresses his well-known Epistola.^[43] Beginning with the words Reges habet Britannia, sed tyrannos (Britannia hath kings but they are tyrants), he proceeds to address five of the principal ones by name, commencing with him of Devon, and going in regular order until he reaches him of Anglesey, whom God hath 'made superior to almost all the kings of Britannia both in kingdom and in stature ', Maelgwn Gwynedd, insularis draco, dragon of the Isle of Mona.^[44] This famous king, who was the head of the house of Cunedda Wledig, is also said by Gildas to have had as instructor one who is described as 'the refined teacher of almost the whole of Britannia', a statement which with the other indications makes it quite clear that the Britannia, with which St. Gildas and his readers are familiar, is neither the island nor Roman Britain, but that western Britannia in Britain which I have given reason to show was the Britannia Superior of the Romans to which afterwards the term Britannia became more exclusively applied.

For it must not be supposed that the Roman provincial system in Britain crumbled away at the departure of the legions from the island. The divisions had been far too long established to perish in a night, especially those into Upper and Lower Britain, but it is probable in view of the troubles, which would afflict the land both from within and without, that the leading civil officials had to give way to the military governors, who alone persisted to protect the Roman tradition. These were the Dux Britanniae in the north, now probably in charge of the land from the Wall of Hadrian to the Humber and Mersey, constituting perhaps one of the provinces of Caesariensis or Lower Britain ; the Comes Littoris Saxonici in the south-east, from the Wash to the Wiltshire Avon or thereabouts, now likewise in probable charge of the whole of the other province of Caesariensis, and finding successors in the Saxon and Anglian Bretwaldas ;^[45] and lastly the Comes Britanniae in the west protecting the whole of Upper Britain, or, as it was now called, Britannia.

This Britannia, by the withdrawal of the legions from Chester and Caerlleon, became exposed to the incoming of Picts and Scots, which were the general names given by the Romans to the barbarians who dwelt beyond the Wall of Hadrian and in Ireland respectively. Given that a people dwelt beyond the Wall, it would be commonly classed with the Picts whether it was racially Pictish or otherwise. These two peoples entered Britannia from over the water,^[46] the Scots invading the west coast and effecting settlements in various districts ;^[47] and the Picts starting from due north and landing on the seaboard from Anglesey to the mouth of the river Dee. Owing to the limitation of the term Picts in later times to the people properly so called, the fact was lost sight of that the Picts, who entered Britannia at this period, were no other than those who are called in Welsh literature Gwyr y Gogledd, the Men of the North,^[48] including Cunedda and his Sons, who occupied the districts lying between the river Dee and the river Teify, having Scots to their north-west and south-west, and the original inhabitants (also interspersed with Scots) in occupation of the land south and east of the Dee and Teify.^[49] The 'Men of the North' were almost certainly for the most part Britons both by race and language, but all who were free amongst them called themselves at a later period, even if not already, by the name Cymry, that is, compatriots.
Cunedda is one of the very few to whom Welsh literature assigns the rare title of gwledig,^[50] a term which denotes the ruler of a territory, apparently as distinct from that of a community of persons, which is a very important distinction in view of ' tribal ' custom. The expression Cunedda Wledig in this case would point to Cunedda as a ruler of territory (gwlad) whilst Cunedda and his Sons would indicate his character as a ' tribal ' king. Almost all who are known to have borne the title of gwledig can be proved to have lived within a century or so about the end of Roman rule in Britain. The three best known, Maxen, Cunedda, and Emrys, are all credited with being in some way connected with the Roman officialdom or race, so that there can be little doubt that gwledig is a Welsh rendering for a Roman title, perhaps the Comes Britanniae. Maxen, who was very early confounded with the usurper Maximus, is associated with the three military centres of Caerlleon, Carmarthen, and Carnarvon. He marries Elen, daughter of Eudav,^[51] into whose family Cunedda marries at a later date. Emrys Wledig, otherwise known as Ambrosius Aurelianus, is associated with the patria of Glywysing^[52] in south-eastern Wales, and was a contemporary of Vortigern, on one of whose sons he as overlord of 'all the kings of the Britannic race' bestows the two patrias of Buallt and Gwrtheyrnion in the modern counties of Brecon and Radnor.^[53] Cunedda comes in point of time after Maxen and apparently before Emrys. His immediate ancestors all bore Roman names, and one of them was almost certainly a Roman official.^[54] His great achievement in Wales was the crushing of the Scotti,^[55] and it may be that it was on this account he became recognized as gwledig. The occupation of so much land, however, by his followers could hardly have been acceptable to the older inhabitants, especially to the Romani about the estuary of the Severn, whose supremacy would now be constantly challenged by these new comers. From this time also dates a close connexion between North Wales and that further and transmarine North whence Cunedda and his Sons had come, a connexion which can be traced for centuries afterwards.^[56] It is possible that Cunedda may have been a gwledig before he entered Wales, and that he might even have held the post of Dux Britanniae^[57] which implied the military leadership of the northern province, but the place and time in which he lived, his ' uncouth ' name, and the so-called ' tribal ' character of his settlement in Wales, all mark him as a 'barbarian' who may indeed have received honours from the imperial government, but only as the usual last and desperate remedy in the face of a ruin which was inevitable. The fact, however, remains that the House of Cunedda henceforth continued supreme in Wales for nine centuries, providing the Welsh with the greatest names in their history for the whole of that period. With its advent in Britannia about the end of the fourth century Welsh national history commences, and with the death of its last important representative, Llewelyn ap Gruffydd, in 1282, the first half of the same history closes.
The occupation of the northern and western portions of Britannia by Picts and Scots threw the old population of south-eastern Wales and the country between the Severn Sea and the Wiltshire Avon into a state of alarm. The Britons of the Devonian peninsula began to migrate in large numbers to Armorica on the mainland, where they founded Britanny. Already in 469 we find Apollinaris Sidonius speaking, as a matter of course, of the inhabitants of that region as Britons.^[58] In this way the south-eastern portion of Britannia beyond the Severn Sea was thinned of its population and thereby made ready for the West Saxon victory of Deorham in 577, which brought the old Roman province of Upper Britain definitely to a close and at the same time exposed the whole of the Devonian peninsula to that process of Saxonization which does not even yet appear to be complete. The Britons west of the Severn, on the other hand, are found in the third decade of the fifth century torn into two factions, the one under the celebrated Vortigern and the other under Ambrosius Aurelianus or Emrys Wledig.^[59] Vortigern is found in the country east of the river Usk and north of it along a line drawn from about the town of Monmouth to that of Llanidloes;^[60] and Ambrosius, as we have already seen, in Glywysing, roughly equivalent to modern Glamorganshire. Things reached a long-remembered crisis when Vortigern in the fourth year of his reign, being the year marked by the consulship of Felix and Taurus, that is, A.D. 428, invited the Saxons^[61] of the ' Saxon Shore ' to his assistance. The details of trie story have been rendered obscure by the misconceptions^[62] of later times, which transfigured Vortigern into a King of Britain who received continental supplies in the island of Thanet in order to withstand enemies who were threatening his country at the Wall of Hadrian ! Vortigern's invitation to the Saxons has consequently been magnified out of all reason, and completely torn from its true setting. It was certainly a blow aimed at the Romani of Britannia, which appears to have met with no small success seeing how the memory of Vortigern was afterwards execrated ; and it is clearly one of the remnant of the Roman faction who is originally responsible for the Excidium Britanniae of the pseudo-Gildas towards the close of the seventh century.

The office of gwledig, like that of the English bretwalda, does not appear to have passed from father to son. None of the descendants of Cunedda is known to have held it after Cunedda himself, not even the powerful Maelgwn. It certainly involved some sort of overlordship extending over all the kings of a given territory, and it is won by such military prowess as would ensure the protection of that territory, theoretically perhaps of Britannia.^[63] Cunedda protects Britannia from the Scots. Emrys likewise protects Britannia from the anti-Britannic policy of Vortigern and his allied Saxons. It represents the Roman tradition as opposed to the barbaric or 'tribal' interest of the native kings. And perhaps, above all, it in some way symbolizes the unity of Britannia, which in this case is what every gwledig would seek to preserve as the Roman legacy handed over to his special care. It would devolve on him to guard Britannia against all invasion and insult whether from the west, north, or east. Hence, when we read of Arthur being chosen to act for the kings of the Britons as their dux bellorum, we cannot be far wrong in suspecting that we have here the historic basis of that hero's renown. That he is never styled gwledig is true, but such equivalents as Arthur Miles, Dux Bellorum, Penteyrned (Chief of kings), and even Ameraudur (Imperator), are sufficient to assure us of the nature of his office.^[64] It is expressly stated that there were many of more noble descent than himself, which is corroborated by the absence of his pedigree in all lists prior to Geoffrey of Monmouth's romance.^[65] He was killed at Camlan ten years before the death of Maelgwn Gwynedd, and therefore shortly^[66] before St. Gildas wrote his Epistola. It is significant that in this work there is a total absence of any sign of fear or apprehension as to external enemies on the part of Britannia, whence we may safely gather that Arthur had not lived in vain.

The old provincial system of Roman Britain, however, was of necessity doomed to disappear. It ran on for a while by means of the power which had set it in motion, but, as that power was generated from without and not from within, its cessation was bound to bring the system to an end. With the removal of external pressure, internal forces began to bear on the situation and later to control it. Chief among these in the Britannia of the west was the reappearance, and, as it were, the renewed activity of native and primitive modes of life such as those which Julius Caesar had attempted to portray five centuries before. These, of course, could not but have undergone modification, but they were not obliterated. There is evidence to show that archaic social conditions, such as are associated with matriarchy and totemism, still lingered on, notwithstanding the Roman regime and the growth of Christianity.^[67] Throughout the fifth century we discern Wales dividing or already divided into a number of small kingdoms, which remain very much the same till Norman and post-Norman times. They war against one another, like the Saxons against the Jutes of Kent and Hampshire or against the Angles, the smaller and weaker kings seeking to preserve their independence, and the stronger kings anxious to make themselves paramount. Add to this the invasions from the west and north, the emigration of the Bretons, the isolation from the civilizing centres of the mainland and the consequent decay of commerce and culture—and we have ample explanation of the increasing difficulties of maintaining the old official unity of Britannia together with the final abandonment of the same.

Moreover, if the official unity of Britannia was impossible, much more so was any national unity of which it might have been capable, were it only for geographical reasons. Even officially it had apparently been found necessary to divide it into Prima and Secunda. A state west of a line drawn from the Dee to the Wiltshire Avon^[68] or thereabouts, divided as this territory is by the Severn Sea and exposed along the whole of its eastern boundary to hostilities from the English lowlands, was an absurdity. It tended to part asunder of itself. Sooner or later a strong attack from the east would capture the Severn shore from Gloucester to Bristol, which eventually took place in 577, the year of the Battle of Deorham by which Gloucester, Cirencester, and Bath fell into the hands of the West Saxons. Thus the unity of Roman Britannia became definitely a thing of the past. Henceforth Wales is free to evolve its own life. The unity of Cymric Britannia will now replace that of Roman Britannia, with this difference, that the latter was possibly never more than an official idea to be preserved, whereas the former becomes a national ideal to be attained.

It must not be supposed, however, that the memory of the Roman Britannia of the fifth century was lost, for it is this Britannia of the 'Roman' which becomes the Britannia of Romance. Its traditions, clustering around the figure of Arthur, become transfigured into a great national dream, a kind of golden age in the past, which grows more and more radiant in the minds of the Britons as they contrast it with the comparative insignificance of their actual position in the world. In Wales it had two very debilitating effects. In the first place, by putting the golden age in the past it made the Welsh regard themselves as decadent, a notion of course which their enemies never failed to encourage. So intensely indeed was this sense of racial decay felt that it forced into existence the counter-notion of a return of Arthur, a kind of messianic dream, which served to counterbalance the depressing and devitalizing effect of the other. In the second place, by substituting romance for history, it has surreptitiously concealed the steady and unbroken development of Cymric nationality from the day that Cunedda and his Sons established themselves in Wales at the commencement of the fifth century. Not only have authentic traditions been distorted to make them fit with the romance, not only has the memory of important historic events been for ever lost, but the very idea of the evolution of Wales from the primitive little kingdoms of the fifth century has been blurred in the national consciousness. It would be difficult to find a story more clear and simple in its main outlines than the growth of modern Wales from its earliest conscious beginnings in the fifth and sixth centuries, where we discern a number of small patriotic communities gradually cohering as they become more and more conscious of their common life. But when for all this there is substituted a golden age wherein Britannia is converted into the Isle of Britain and the Britons masters of the same from end to end ; where wicked Vortigern calls in the heathen from Germany, who drive the Britons pell-mell from the eastern districts of England into the midlands, and out of the midlands into Wales, there to relapse into barbarism ; where every step in the Cymric advance from age to age, marked by such names as Cadwallon, Gruffydd ap Llewelyn, and the post-Norman princes, is regarded as a convulsive effort of a dying people to regain some of the glory of the past—it can readily be understood how the history of Wales has suffered and how its national vigour has been enfeebled.

After the death of Arthur, who is commonly reputed to have perished in a civil war, we hear of no other military leader whom we may regard as the gwledig of Britannia in power as well as in title (that is, allowing that Arthur did really bear the title). Aurelius Caninus, one of the five kings addressed by Gildas, is also known as Cynan Wledig,^[69] so that it is possible that he was regarded as one of Arthur's successors. One gathers from the Epistola that he ruled east of Devon in the country 'between the Severn Sea and Poole Harbour',^[70] which was the part of Britannia where, with the south-east of Wales, the Roman interest was strongest. As late as the close of the seventh century it is still possible for a writer in that neighbourhood to be conscious of Roman imperial sentiment and to speak of Latin as nostra lingua. In view of the general decay of things Roman his life is embittered. The descendants of Ambrosius are still there but how ' greatly degenerated from their ancestral nobleness ' !^[71] In this neighbourhood therefore we should perhaps expect the office of gwledig to linger on until the catastrophe of the year 577. But already, with the death of Arthur, the centre of political interest in Wales has passed permanently in the person of Maelgwn Gwynedd to the House of Cunedda. Henceforth the political history of Wales may be treated quite apart from that of the Devonian peninsula, although the actual cleavage did not take place till the Battle of Deorham.

At the time when Gildas writes his Epistola, Maelgwn Gwynedd is certainly the leading king in Wales as was afterwards his son Rhun.^[72] In the seventh century also we find the House of Cunedda holding the same com- manding position in the person of Cadwallon^[73] (the fifth in descent from Maelgwn) who was killed by Oswald in 635. Between Rhun and Cadwallon, however, the supreme power may have passed for a while into the hands of the house of Cadell Ddyrnllug of Powys, for we find Cynan Garwyn, the head of that family, battling against Anglesey, Dyved, Glywysing, and Gwent.^[74] It is this house also which appears to have withstood Ethelfrith of Northumbria at the Battle of Chester in 617, in which Selyf ap Cynan Garwyn fell. This event was famous in ancient times because of the slaughter of about 1,200 monks of Bangor Iscoed, which was an incident of the fight.^[75] It has become famous in modern times because of ' the decisive character which it has been the fashion to ascribe to it of late '.^[76] For it is nowadays commonly and even dogmatically asserted that it divided the Britons of the North from those of Wales, whereas there is no evidence forthcoming that these were ever united by land. Late Glamorganshire legends ascribe the name of Teyrnllwg^[77] to a supposed Cymric patria lying appar- ently between the river Dee and the river Derwent in Cumberland, a name based on erroneous etymology as to Durnluc in Catel Durnluc, that is, Cadell Ddyrnllug, the king who founded the royal stem of Powys.^[78] But apart from this there is no real evidence for the presence of Cymry (or of any Britons) between the river Derwent and the river Dee further south than Cartmel below Windermere and the river Leven.^[79] That there was a close connexion between the Cymry of ' Cumberland ' and those of Wales is amply evident, but it was maritime and not terrene.

Cadwallon was succeeded by his son Cadwaladr, whose fame is due not to any known merits of his own, but to the imaginative genius of Geoffrey of Monmouth, who in his romantic History of the British Kings makes Cadwaladr the last of his list.^[80] The reign of this king becomes in consequence the appropriate finale of a long and glorious era of Welsh history. All this of course is purely fictitious, as Cadwaladr's death marks no known break of any kind in the perfectly clear development of Welsh nationality. Geoffrey's Cadwaladr in fact is a composite personage created out of Geoffrey's own confusion of Cadwaladr and his father, Cadwallon, and Ceadwalla of Wessex . As there were kings in Wales before Cadwaladr, so there were kings, and far greater kings, after him. He died in the second year of the great plague of 664-5,^[81] and was succeeded by his son Idwal. Of his immediate descendants little is known. They appear to sink into comparative insignificance by the side of Maelgwn, Rhun, and Cadwallon, and other than they may possibly have loomed larger in the life of Britannia and its Britons. But whenever the mists rise which conceal the affairs of these centuries from our view, we always discern the main stem of Cunedda Wledig towering amid the rest of the royal stems of Wales, and generally paramount. Moreover, we may be certain, in view of its prestige in the ninth century, that its history in the preceding centuries is that of a house which has been gradually gaining strength until it is now in a position to effect a change in Welsh political conditions which will mark the beginning of a new era in the slow and steady development of Cymric nationality.

We have seen that the first period in the history of post-Roman Wales must have come to an end in the year 577, although many years before this date the centre of political interest in Wales was shifting from the representatives of the Roman tradition in Britannia [that is, the gwledigs] to the House of Cunedda, which stood for the predominance of the Cymric kindreds. For in Maelgwn we seem to discern the progress of a policy which aims at bringing all the royal stems, from Anglesey to the river Wye, into subjection to the main stem of the family of Cunedda. This continues until in the first quarter of the ninth century there begins a new policy, which will bring almost the whole of Wales under the sole and immediate rule of this main stem of Cunedda. The many royal stocks are to give way to one royal stock, and in this manner is the unity of the Cymric Britannia to be achieved.

In 816 the main stem of Gwynedd ceased on the male side with the death of King Cynan Tindaethwy, the great grandson of Cadwaladr. His daughter, Etthil, had married Gwriad ap Elidyr, King of the Isle of Man, and now their son, Mervyn Vrych, comes from that island to claim the throne of Gwynedd.^[82] Mervyn is ominously surnamed in Welsh tradition Camwri, that is, Oppression.^[83] He is bent on asserting the old overlordship of Cunedda, Maelgwn, and Cadwallon over the whole of the Welsh kin from Anglesey to the river Wye. But in addition to this, he proceeds by diplomatic marriages to bring the land more directly under the sway of his house. By his marriage with Nest, sister of Cyngen, the last King of Powys of the line of Cadell Ddyrnllug, his son Rhodri becomes the immediate ruler of that kingdom in addition to his own. By the marriage of the same son, Rhodri, to Angharad, sister of Gwgon, the last King of Seisyllwg, a kingdom comprising the two patrias of Ceredigion and Ystrad Tywi, these lands also fall under the direct sway of his house.^[84] Thus when Rhodri comes into full possession of his dominions, his immediate rule extends from the Irish to the Severn Seas, including roughly the whole of that Welsh Wales which remained under native rulers throughout the Norman period, together with those portions which are described above as fluctuating between Welsh and Norman control. Dyved, Brycheiniog, Glywysing, and Gwent are the only patrias which remain outside the immediate rule of his house, and against these he adopts or rather continues the aggressive policy which aims at bringing them also under the same immediate control of his family. Rhodri was killed by the English in 877,^[85] but he remained in the memory of Wales as one who had achieved more real power over the Welsh kin than any who had gone before him, being known in history as Rhodri the Great. It is important to remember at this juncture that we are now in the century which saw Charlemagne reigning as Emperor of the West.

There was a legend current in later times that Rhodri the Great, erroneously regarded as king of all Wales, damaged the Welsh cause irretrievably by sharing his kingdom among his three sons, giving, according to one version, Gwynedd to Mervyn, Powys to Anarawd, and Deheubarth to Cadell.^[86] Nothing can be further from the truth than the impression left by this tale. For as we have seen, Rhodri's aim was to consolidate Wales by substituting the rule of his own family for that of many families. Princes of the blood of Rhodri alone were to govern the land directly from one end to the other. The legend of course echoes the ideas and possibilities of later times when men had come to see that, conducive as was the rule of one family instead of several families to keeping folk of the same kin together, yet the rule of one man was still more conducive to that desirable result. Consequently they wondered how it was that Rhodri could have divided his kingdom, forgetting that, unsatisfactory as the policy of Rhodri would have been in their day, yet in his own time it was a new thing in Wales, a fresh development, which had then become practicable, being an immense improvement on what had preceded it. The obstacles in the way of the unity of Wales were stupendous, such as no bare coercion could overcome. We have seen their like on a modern and larger scale in the story of Italian and German unity. In the Dark Age the difficulty was accentuated by the fact that, even given a unity achieved by a capable ruler, the mind of the age as reflected in the Leges Barbarorum, of which the Laws of Howel are the Welsh exemplar, compelled that unity to be divided after his death among his sons. Charlemagne himself had so to divide his empire ; the same necessity rested on Rhodri the Great. The policy therefore inaugurated by Mervyn Vrych, and continued by Rhodri and his successors, marks the beginning of a fresh epoch in our travail as a people to the full consciousness of our national entity.

The possessions of Rhodri then after his death in 877 were divided among his sons, of whom the best known, and those whose posterity played the largest part in later Wales, were Anarawd and Cadell. From Anarawd (died 915) the later kings of Gwynedd traced their descent, and from Cadell (died 909) both those of Powys and those of Deheubarth. It appears therefore that in the division of territories after Rhodri's death, the kingdom of Powys sooner or later fell into the hands of Cadell, together with Seisyllwg. The policy of bringing all Wales by politic marriages under the direct control of the family of Rhodri was now continued by one of the greatest princes whom the House of Cunedda had hitherto produced, namely, Howel the Good, the son of Cadell. Howel by his marriage with Elen, daughter of Llywarch, the last king of Dyved, who died in 903, became the immediate ruler of that kingdom ; and as the line of Dyved had claims on Brycheiniog through Cathen, son of Ceindrech, a lady who in her day appears to have been the sole representative of the ancient stem of Brychan, after whom Brycheiniog had its name,^[87] it is hardly probable that Howel in view of the policy of his family, would fail to assert those claims. In this manner the whole of Wales was gradually falling under the immediate sway of Rhodri's house.

Howel, however, inaugurated a still newer policy, which aimed at the unification of Wales ; and herein consists his prime importance in Welsh history. Not only did he continue and encourage the methods of Rhodri the Great, but added to them a method of his own. For as Rhodri would bring all Wales under the direct sway of one family, so Howel would bring the whole of the Welsh people under one law. A common rule implied a common law, and in order that men might know what this common law was, it had to be codified and thereby reduced to writing. This was the task to which Howel applied himself, and by having laid a sound foundation he occupies a foremost place not only amongst the rulers of the Welsh people, but also amongst all those who have distinguished themselves throughout the centuries by their devotion to the cause of Wales.

The following, which are the two earliest accounts ot the work which Howel took in hand, describe concisely both the way in which he proceeded and the nature and extent of his undertaking.

Preface to Peniarth MS. 28.

Incipit prologus in libro legum Howel Da.
Brittanie leges rex Howel qui cognominabatur bonus .i.
da . regni sui . s . Gwynedotorum Powyssorum atque Dextralium sapientium et in uno loco ante suum tribunal congregatorum uno consensu et diligenti quia ex omni natione medio circiter^[88] temperateque constituit. Acciuit de quolibet pago per suum regnum sex uiros auctoritate et scientia et omnes episcopos archiepiscopos abbates et sacerdotes totius Wallie pollentes ad locum qui dicitur Ty Gweyn ar Taf et ibi demorati sunt XL diebus et XL noctibus in pane et aqua et tune temperauerunt reditionem forefacti .i. cosp superflua diminuere que erant in pluribus reditionibus forefacti ita fecerunt pretium uniuscuiusque rei et iuditium congruum de qualibet re. Tunc surrexerunt omnes archiepiscopi episcopi abbates et sacerdotes induerunt uestes suas et insteterunt bacculis cum crucibus et candelis et ex communi consilio excommunicauerunt transgredientes leges istas et similiter obseruantes benedixerunt. Hec iudicia scripta sunt.

Here begins the preface to the book of the laws of Howel Dda.

King Howel, who was surnamed Good, that is, da, put together the laws of Britannia moderately and temperately with the unanimous consent and after the careful consideration of the wise men of his kingdom, namely, the men of Gwynedd, Powys, and Deheubarth, who had assembled together in one place before his tribunal. He summoned from every pagus throughout his kingdom six men who excelled in authority and knowledge, and all the bishops, archbishops, abbots, and priests of the whole of Wales to the place which is called Ty Gweyn ar Taf, and there they lived forty days and forty nights on bread and water ; and then they regulated the indemnity for wrong-doing, that is, cosp, and diminished the excesses which prevailed in many of the indemnities for wrong-doing by determining the worth of every particular thing and the decision suitable in every case. Then all the archbishops, bishops, abbots and priests rose up together, and assumed their robes, and leaned on their croziers with crosses and candles, and by common consent excommunicated those who should violate those laws, and likewise blessed those who should keep them.

Preface to Peniarth MS. 29.

Heuel da uab Kadell teuyhauc Kemry oil a uelles e Kemry en kam arueru or kefreythyeu, ac adeuenus atau uy guyr o pop kemud en y tehuyokaet e pduuar en lleycyon ar deu en scolecyon. Sef achaus e uennuyt er escleycyon rac gossod or lleycyn dym a vey en erbyn er escrftur Ian. Sef amser e doythant eno e Garauuys, Sef amser achaus e doyant e Garauuys eno urth delehu o paup bod en yaun en er amser glan hunnu, ac na guenelhey kam en amser gleyndyt. Ac o kyd kaghor a kyd synedycaeth e doython a doytant eno er hen kefreythyeu a esteryasant a rey onadunt a adassant y redec a rey a emendassant ac ereyll en kubyl a dyleassant ac ereyll o neuuyt a hosodassant. A guedy honny onadunt e kefreythyeu a uarnassant eu cadu, Heuel a rodes y audurdaut uthunt ac a orckemenus en kadarn eu kadu en craf. A Heuel ar doythyon a uuant y kyd ac ef a ossodassant eu hemendyth ar hon Kamry holl ar e nep eg Kemry a lecrey heb eu kadu e kefreythyeu. Ac a dodassant eu hemendyt ar er egnat a kamero dyofryt braut ac ar er argluyt ay rodhey ydau ar ny huypey teyr kolhouen kefreyth a guerth guellt a dof a pop pedh reyt y dynyaul aruer arnau.

Howel the Good, the son of Cadell, prince of all Cymru, perceived the Cymry abusing the laws, and summoned to him six men from every cymwd in his principality, four of them of the laity and two of the clergy. The reason that the ecclesiastics were summoned was lest the laics should insert anything contrary to Holy Writ. The time that they arrived there was Lent, and the reason that they came there in Lent was that it behoved all to be upright in that holy season and to avoid evil in a time of holiness. And with the mutual counsel and deliberation of the wise men who there assembled, they examined the old laws, some of which they allowed to continue, some they amended, and others they completely abolished, and others again they ordained afresh. And when they had promulgated the laws, which they had decided to establish, Howel gave his authority to them and strictly commanded that they should be scrupulously observed. And Howel and the wise men, who were with him, imposed their curse and that of all Cymru on any one in Cymru who perverted the laws and kept them not ; and they imposed their curse on the judge who should take a vow' to administer justice, and on the lord who should grant him authority without that judge knowing the Three Columns of Law, and the Worth of Wild and Tame, and everything necessary for the use of man.

IV

The leading work so far concerned with the laws of Howel is that edited by Aneurin Owen in 1841 for the Public Record Commissioners, entitled Ancient Laws and Institutes of Wales. It contains the three early Latin books, and also the three classes of Welsh books ; the additions made to the latter from the thirteenth to the sixteenth centuries are given with other legal matter under the heading of Anomalous Laws. The Welsh texts are provided with an English translation. The Books of Gwynedd, Blegywryd, and Cyvnerth, however, are produced in such a way that the various MSS. of each particular class are interblended, so that it is with the greatest difficulty that any particular one may be distinguished. Indeed, in the case of the majority of the MSS., it is impossible to do so. Moreover, by arranging the texts so that they fall into books, chapters, and sections, and by consequently attempting to bring them into harmony, the confusion becomes hopeless. The table of contents also and the indices are most jejune, misleading every beginner who takes up the book. There are besides other serious defects, so that, valuable as the work undoubtedly is, and great as is our indebtedness to this early and scholarly editor, it has become imperative that it should be done afresh. Until at least the oldest Latin law books and the best MSS. of the Books of Gwynedd and Blegywryd have been so reproduced with analytical summaries and indices that the reader may readily discover what they contain (a task here essayed with regard to the Book of Cyvnerth) the study of native Welsh law must suffer, and every treatise professing to deal with it as a whole must prove inopportune. It is not proposed, therefore, to deal with it here beyond what is attempted in the Glossary, mainly from the material afforded by the present text.

The Book of Cyvnerth, however, by itself is sufficient to provide the student with a door of entrance into the Welsh Dark Age. Remembering that it represents a late thirteenth-century form of Howel's codification of Welsh law and custom in the tenth century, he will enter safely into the midst of the social and political conditions of pre-Norman Wales. It befits him, however, to be wary, for he treads enchanted ground, and it will not be long before he meets Cadwaladr and Arthur and all the heroes of the Mabinogion and kindred tales. Many are they who have boldly entered here only to succumb to the charm of this realm of phantasy and illusion. But let him keep closely to the laws of Howel as interpreted by our Cyvnerth, and peruse the Pedigrees, the Annales Cambriae, the Historia Brittonum, the Vitae Sanctorum, the Excidium Britanniae of the pseudo-Gildas, and the Epistola of the true Gildas, in the light of the said laws, and below the Britannia of romance he will soon discern the no less interesting Britannia of history as it slowly emerges from the archaic conditions of the primitive inhabitants of Roman Wales into the life of the Middle Age. For be it remembered by the beginner that these laws are leges barbarorum, laws of the barbarians or natives of Wales as distinct from the civil law of imperial Rome and the canon law of the Church. The latter are from without, the former are from within. And it is largely because these laws of Howel have been so undeservedly neglected that the history of pre-Norman Wales is still so unsatisfactorily treated in our textbooks.

It should be noted that the term ' tribal system ' has been advisedly avoided in this work whilst dealing with the Welsh society of the Dark Age, seeing that there exists no satisfactory explanation of what precisely is meant by the word ' tribe '. Its Welsh equivalent llwyth, used, for instance, when speaking of the tribes of Israel, is nowhere found in the law books. We have cenedl, kindred ; teuln, household ; and gwlad, patria ; but nowhere llwyth, tribe, or any apparent equivalent of the same.

1. For a full account of this and other Peniarth MSS., see Report on Manuscripts in the Welsh Language, vol. I, Pts. II and III, by Dr. J. Gwenogfryn Evans, printed for the Historical Manuscripts Commission.
2. Lleuyr e Ty Guyn mentioned in the thirteenth-century Pen. MS. 30. See Report on MSS. in Welsh I. 361, note. viii.
3. Anc. Laws 1. 1-335 ; II. 1-36.
4. Ibid. I. 104, 218, 292.
5. Ibid. I. 620-797.
6. See Glossary under Deheubarth.
7. Anc. Laws 1.viii. Gwent was a patria between the lower courses of the river Usk and the river Wye, included in modern Monmouthshire.
8. Anc. Laws I. 218, 340.
9. Ibid. I. xxxi.
10. Ibid. I. 338-617.
11. Ibid. I. 574; 556-9.
12. This date is that of Dr. Gwenogfryn Evans as given in his report on the MSS. in Welsh in the British Museum, the relevant portions of which report (not yet published) he has kindly forwarded to me for perusal.
13. Report on MSS. in Welsh I. 369.
14. Anc. Laws I. xxx.
15. The third old Latin text, viz. Harleian MS. 1796, seems to be of the form of the Book of Gwynedd. See Glossary under taeogtrev.
16. Anc. Laws I. 340.
17. Seeing that he married Elen, daughter of the last king of Dyved, whereby he became immediate ruler of that kingdom.
18. Anc. Laws II. 50, 380, 584.
19. [Supra.]
20. I must again express my obligation to Dr. Gwenogfryn Evans for the invaluable help received from him in drawing up these descriptions of the particular MSS. in question. I have myself examined U, V, W, and X. [See note I on p. X.]
21. p. 299.
22. Anc. Laws I. 670, note 21, 686, note II.
23. Report on MSS. in Welsh, vol. I. 1074-5.
24. Anc. Laws I. 340-2.
25. [Here as elsewhere, "v" is used for the Middle-Welsh v (U+1EFC & U+1EFD). It was used variously to carry the values of Latin "v"/"u" and Anglo-Saxon wynn (i.e., "w").]
26. See note I on p. X. As to the form of the Latin books see p. X with note 4.
27. Outlines of the History of Wales by Prof. J. E. Lloyd, 164. See also the valuable article by the same writer in the Transactions of the Cymmrodorion Society for 1899-1900, entitled 'Wales and the Coming of the Normans '.
28. This as well as the majority of other dates in early Welsh history must be regarded as tentative only, until the whole subject of Welsh chronology has been thoroughly examined.
29. i.e. the preface to Peniarth MS. 28.
30. Reproduced by Mr. Egerton Phillimore, together with Pedigrees, &c., in Y Cymmrodor, vol. IX. 141-83. This reproduction is indispensable to every student of early Welsh history.
31. Y Cymm. IX. 171, Peds. I, II.
32. Y Cymm. IX. 325.
33. See Glossary under Deheubarth.
34. Transactions of the Cymm. Soc. 1905-6, pp. 11-13. It should be stated here however that there was a Howel, king of the West Welsh, flourishing at this time whose name appears in the Saxon Chronicle s. a. 926. See Plummer's Two Sax. Chrs. II. viii.
35. Where Howel could hardly fail to have lived, at least at the time when he became its king through marriage.
36. Stevenson's Asser, pp. 64, 65.
37. Brit. Mus. Harleian MS. 3859.
38. Transactions of Cymm. Soc. 1905-6, pp. 1-30.
39. Herodian III. 24.
40. Iv. 23. See Rhys's Celtic Britain, 3rd ed. 97, c. ; also The Welsh People, 103, &c.
41. Prof. Haverfield's Romanization of Roman Britain, 8 and note 2, 27.
42. An inscription found near Cirencester proves this. Eng. Hist. Review, July, 1896.
43. I would refer the reader at this point to my articles on the authorship of the Excidium Britanniae as distinct from the Epistola Gildae in the Celtic Review (Edinburgh) for April, July, and October, 1905 ; also in the St. David's College Magazine for December, 1904. Mr. E. W. B. Nicholson has replied in the Celtic Review for April, 1906, in an article which for the moment can well be left alongside of the original contributions. The contention is that the first twenty-six chapters of the work, now commonly attributed to Gildas, formerly constituted a distinct book known as Excidium Britanniae, which was written by a 'Roman' Briton towards the close of the seventh century somewhere in the neighbourhood of the mouth of the Severn. This work was considerably ' edited ' by some one who ignorantly or deliberately misunderstood it, probably both. In this form it passed into the hands of Bede, who used it as his chief and almost only authority for what he had to say of fifth-century Britain. Almost all that Bede professes to know of this period is taken from the Excidium, which he seems to ascribe to Gildas (H. E. I. 22), although he gives no evidence that he was familiar with the genuine work of that monk, viz. the Epistola Gildae, to which the Excidium was subsequently prefixed.
44. Epistola Gildae, cc. 34-36 (Chr. Min. III. pp. 41-7)
45. Bede's Ecc. Hist. II. 5 ; Saxon Chronicle under 827 ; Stevenson's Asser, 147, note I.
46. ' Duabus primum gentibus transmarinis vehementer saevis, Scotorum a circione, Pictorum ab aquilone calcabilis.' Excidium Britanniae c. 14 (Mommsen's Chr. Min. III. p. 33). Bede, who bases almost everything he has to say concerning the early centuries of post-Roman Britain on the Excidium, and indeed incorporates whole passages into his text, completely misunderstands the term transmarini as applied to the Picts, which he explains as being applicable to them in that they came from beyond the Firths of Forth and Clyde (H. E. I. 12). The only part of southern Britain which could be approached over the water from the north-west and the north is North Wales, which proves that the Britannia underlying that of the 'edited' Exadium, which came into Bede's hands, was the Britannia of the genuine Gildas, including Wales plus the Devonian peninsula.
47. Vita S. Carantod and Vita S. David in Rees's Cambro-British SS. pp. 97, 1 01, 124 ; the De Situ Brecheniauc and Cognado Brychan in Y Cymmrodor, vol. XIX ; the Hist. Britt. (Chr. Min. III. 156). See also Bury's Life of St. Patrick, 325.
48. Skene's Four Anc. Bks. I. 165-83.
49. It is very noteworthy and confirms the view expressed above that the Picts as a distinct race of northern invaders in Wales are nowhere mentioned, as are the Scots, in early Welsh literature outside the Excidium Britanniae and works influenced by it. Thus the only reference to them in the Book of Llanddv is in the Life of Teilo (pp. 99, 100), where the ' Historia Gildae ' (i. e. the Excidium) is expressly referred to as the authority. There is no reference what- ever to them in the Cambro-British Saints. It appears, however, that the identity of the invading Picts and the Cymry was not completely forgotten, for in the Peniarth MS. 118 the statement appears that ' the Picts were none other than the old Cymry ' (nid oedhynt y Picteit onyd yr hen Gymry.). Rep. on MSS. in Welsh I. 724.
    ' Hec sunt nomina filiorum Cuneda quorum numerus erat IX : Typiaun primogenitus qui mortuus in regione que uocatur Manau Guodotin et non uenit hue cum patre suo et cum fratribus suis pre[dictis] ; Meriaun filius eius diuisit possessiones inter fratres suos ; ii, Osmail ; iii, Rumaun ; iiii, Dunaut ; v, Ceretic ; vi, Abloyc ; vii, Enniaun Girt ; viii, Docmail ; ix, Etern.
    ' Hie est terminus eorum a flumine quod uocatur Dubr duiu usque ad aliud flumen Tebi et tenuerunt plurimas regiones in occidentali plaga Brittanniae.' These valuable sections are appended to the Pedigrees which follow the Annales Cambriae in Harleian MS. 3859 (Y Cymm. IX. 182-3)
50. In the indices to the Oxford Red Book of Hergest there are about ten names associated with this title, of the majority of which nothing whatever seems to be known. They are nearly all, however, made contemporaries of persons who are known to have lived before 577. Thus Tared Wledig is described as the father of Twrch Trwythy who appears in the tale of Kulhwch and Olwen as the wild boar pursued by Arthur and his men (Oxford Mab. 123, &c.).
51. See the tale entitled Breudwyt Maxen Wledic (Oxford Mab. 82-92), in which it is amply evident, if the author's identification of Maxen with Maximus is eliminated, that Maxen is a dweller in Britain. The Welsh word for Rome, viz. Rhufain, older Rumein, is from Romania and not from Roma. This fact will explain many a marvel in old Welsh literature of journeys to the Roman city. It is curious that Geoffrey does not bring Arthur to the city of Rome, although he brings him as far as the Alps (Hist. Reg. Brit. X. 13 ; Oxford Brut, 229), so that it is by no means improbable that what Geoffrey had before him was an account of Arthur's wars in Romania, that is, some part of Britain where the Roman interest was sufficiently strong to cause it to be distinguished as Romania. The word actually appears in the Excidium, ch. 7 (Chr. Minora III. 30).
52. 'Et ipse [i.e. Vortigern] legates ex consilio magorum per universam Brittanniam misit utrum infantem sine patre invenirent. Et lustrando omnes provincias regionesque plurimas venere ad campum Elleti qui est in regione quae vocatur Gleguissing. . . . Et rex ad adolescentem dixit, Quo nomine vocaris ? Ille respondit, Ambrosius vocor, id est, Embreis Guletic ipse videbatur. Et rex dixit, De qua progenie ortus es ? At ille Unus est pater meus de consulibus Romanicae gentis.' Hist. Britt. c. 41, 42 (Chr. Min. III. 182, 186).
53. 'Pascent qui regnavit in duabus regionibus Buelt et Guorthegirniaun post mortem patris sui [i. e. Vortigern] largiente Ambrosio illi qui fuit rex inter omnes reges Brittannicae gentis.' Hist. Britt. c. 48 (ibid. III.I92).
54. Cunedda, son of Eternus, son of Paternus, son of Tacitus. Y Cymm. IX. 170. Paternus is given the epithet Peisrudd, or him of the red tunic. Celtic Britain, 3rd ed. 118.
55. ' Filii autem Liethan obtinuerunt in regione Demetorum et in aliis regionibus id est Guir Cetgueli donee expulsi sunt a Cuneda et a filiis eius ab omnibus Brittannicis regionibus.' Hist. Britt. c. 14 (Chr. Min. III. 156).
    ' Mailcunus magnus rex apud Brittones regnabat id est in regione Guenedotae quia atavus illius id est Cunedag cum filiis suis, quorum numerus octo erat, venerat prius de parte sinistrali, id est, de regione quae vocatur Manau Guotodin . . . et Scottos cum ingentissima clade expulerunt ab istis regionibus et nusquam reversi sunt iterum ad habitandum.' Hist. Britt. c. 62 (ibid. III. 205-6).
56. See, for instance, the remarkable passage in the Book of Gwynedd, where Rhun, son of Maelgwn Gwynedd, is described as fighting in the North, apparently on the banks of the river Forth (Anc. Laws I. 104 ; Celtic Britain, 3rd ed. 126). Add to this the exploits of Cadwallon and the North Welsh usurper Cadavael between the Humber and the Forth. Hist. Britt. cc. 61, 64, 65 (Chr. Min. III. 204, 207-8) ; Celtic Britain, 3rd ed. 131-5.
57. Celtic Britain, 3rd ed. 118-20.
58. Hodgkin's Political Hist. of England to 1066, p. 106, and also note, where the reference is given as Ep. i. 7. See also Y Cymmrodor XI. 69.
59. ' Guorthigirnus regnavit in Brittannia et dum ipse regnabat urgebatur a metu Pictorum Scottorumque et a Romanico impetu nec non et a timore Ambrosii.' Hist. Britt. c. 31 (Chr. Min. III.)
60. Vortigern was the founder of the royal stem of the little kingdom of Gwrtheyrnion (in modern Radnorshire), which is called after his name (Gwrtheyrn). He therefore stands to Gwrtheyrnion as Brychan to Brycheiniog, Glywys to Glywysing, Ceredig to Ceredigion, and so forth. In other words, he is clearly one of the founders of the numerous little patrias or kingdoms into which we find post-Roman Wales divided. His father and grandfather bear the Roman names of Vitalis (Guitaul) and Vitalinus (Guitolin) respectively, being traditionally connected with the city of Gloucester. Hist. Britt. cc. 48, 49 (Chr. Min. I II. 192-3). Geoffrey of Monmouth describes him as consul Gewisseorum, represented in the Welsh version by iarll oed hwnnw ar Went ac Ergig ac Ettas (earl was he over Gwent and Erging and Ewyas). Hist. Regum Brit. VI. 6; Oxford Brut, 127. We find elsewhere a dux Wisseorum given in the Welsh as iarll Ergig ac Euas ; and Cadwaladr's mother, who is in the Latin described as sprung ex nobili genere Gewisseorum, is in the Welsh wreic wonhedic o Euas ac Ergig (a noble lady of Ewyas and Erging). Hist. Reg. Britt. V. 8, XII. 14 ; Oxford Brut, 109, 252.
    Erging, in English Archenfield, is the district now in Herefordshire west of the river Wye. In early times it must have included the whole of the territory from Monmouth to Moccas, east of the river Munnow and the river Dore. Ewyas lay to the west of Erging, having the river Dore as its eastern boundary as far, perhaps, as the river Grwyne Fawr. Gwent was the district south of Erging and Ewyas (which were known as ' the two true sleeves of Gwent uch Coed '), between the river Usk and the river Wye in modern Monmouthshire. Owen's Pembrokeshire I. 199, n. 5, 208, n. I ; III. 264, note E. As Glywysing, in which the boy Ambrosius Aurelianus was discovered, includes the territory west of the river Usk as far as the western confines of Gower, we may roughly locate Vortigern east and north of the river Usk, and Ambrosius west and south of it.
61. ' Guorthigirnus autem tenuit imperium in Brittannia Theodosio et Valentiniano consulibus et in quarto anno regni sui Saxones ad Brittanniam venerunt Felice et Tauro consulibus quadringentesimo primo anno [a passione] domini nostri lesu Christi.' Hist. Britt. c. 66 (Chr. Min. III. 209 cum apparatu critico). ' Vortigern, moreover, was ruling in Britannia when Theodosius and Valentinianus were consuls [i.e. 425], and the Saxons came to Britannia in the fourth year of his reign, when Felix and Taurus were consuls, and in the 401st year from the [Passion] of our Lord Jesus Christ [calculating according to Victorius of Aquitaine, that is, 28 + 400 = A. D. 428].' See the article entitled ' The Exordium of the "Annales Cambriae" ' by Mr. Alfred Anscombe in Eriu (January, 1908), where Mommsen's text of the Hist. Britt. c. 66, is subjected to severe criticism.
62. These misconceptions originated with the ' edited ' copy of the Excidium Britanniae placed in Bede's hands, where Britannia was ignorantly or maliciously identified with Roman Britain, or rather with the island of Britain ! It cannot be too much insisted upon that we learn from the Excidium Britanniae almost all that Bede knew or chose to know of fifth-century Britain, and that the former therefore, and not the latter, is the ' original authority ' with which the student has to deal in his researches into this period of history.
63. Cf. the description of Ambrosius as 'rex inter omnes reges Brittannicae gentis'. Hist. Britt. c. 48 (Chr. Min. III. 192). Also the passage in Maxen's Dream (Oxford Mab. 89), where Elen, on the morning after her marriage with the gwledig, being asked to mention the agweddi she desired, demanded ' ynys prydein yw that o vor rud hyt ym mor Iwerdon ar teir rac ynys y dala dan amherpdres ruuein a gwneuthur teir prif gaer idi hitheu yn y lie y dewissei yn ynys prydein ', which Lady Guest translates ' the Island of Britain [Britannia] for her father from the Channel to the Irish Sea, together with the three adjacent islands [that is, presumably, Wight, Anglesey, and Man], to hold under the empress of Rome; and to have three chief castles made for her in whatever places she might choose in the Island of Britain [Britannia].' The three castles or caers mentioned are Caermarthen, Caerlleon, and Caernarvon. Surely all this implies that Eudav, Elen's father, is to hold the whole of Britannia as gwledig under the emperor. Bede also, in the account which he gives (H. E. II. 5) of the overlords, who in the Chronicle are called Bretwaldas, describes them as the kings who ruled over all the southern provinces which are divided from the northern by the Humber, &c. (' qui tertius quidem in regibus gentis Anglorum cunctis australibus eorum prouinciis, quae Humbrae fluuio et contiguis ei terminis sequestrantur a borealibus, imperauit ').
64. See Sir John Rhys's Introduction to Malory's Le Morte D'Arthur in the Everyman's Library.
65. ' Et licet multi ipso nobiliores essent ipse tamen duodecies dux belli fuit ' (Chr. Min. III. 199, MSS. M and N).
66. That is, assuming that the two following anni to be reckoned from the same initial year. 'Annus XCIII. Gueith Camlann in qua Arthur et Medraut corruerunt. Annus CIII. Mortalitas magna in qua pausat Mailcun Rex Guenedotae.' Ann. Camb. (Y Cymm. IX. 154-5). The following will assist us to determine the period we are dealing with. It appears from the Vitae that St. David was born in the thirtieth year after St. Patrick went to Ireland as Bishop, which makes 433 + 29 = 462 ; and this date is confirmed by MS. B of the Ann. Camb., which places David's birth opposite Annus XIV. For if this be computed from the false Bedan date of the Saxon Advent, we get 449 + 13 = 462. We may therefore regard A.D. 462 for David's birth as tolerably well established. St. David was a descendant of Cunedda Wledig, but whether in the fourth remove like Maelgwn or in the third is uncertain. His father was Sant or Sanddef, who was the son either of Cedig ap Ceredig ap Cunedda, or of Ceredig ap Cunedda. The expression ' Dewi Sant ' for Saint David appears to be a late misreading of Dewi ap Sant, the position of Sant being also apparently unique in Welsh hagiography.
67. 1 Rhys and Jones's The Welsh People, 36-74 ; Y Cymmrodor XIX. 20-3.
68. Avon being the generic Welsh word for ' river ' there can be little doubt that the Wiltshire Avon was at one time a boundary line between Welsh and non-Welsh peoples, as would be the case also with regard to the Bristol and Tewkesbury Avons. The presence of Britons in the district roughly enclosed by these Avons is convincingly evident. The western boundary of the Saxon shore with its Saxon inhabitants is uncertain. If Portus Adurni is Porchester, we certainly bring it as far west as the Solent. In any case, it is significant that the earliest clashing of Britons and Saxons is traditionally stated to have taken place in this neighbourhood in the country immediately east of the river Avon (Sax. Chr. s. a, 495, 5oi, 5o8, 514, 519, 527, &c.).
69. Hist. Reg. Brit. XI. c. 5 ; Oxford Brut, 233.
70. Rhys's Celtic Britain, 3rd ed. 107.
71. Excid. Brit. c. 25 (Chr. Min. III. 38, 40).
72. In addition to the remarks of Gildas in the Epistola, chs. 33-6 (Chr. Min. III. 44-8) and of the author of the Historia Brittonum, ch. 62 (ibid. III. 205), see the traditions of Maelgwn as supreme king (Anc. Laws II. 48-50, 584) and his exploits in different parts of Wales as recorded in the Vitae Sanctorum (Rees's Cambro-British SS.). As to Rhun, see Anc. Laws I. 104-5 and the Vita S. Cadoci (Cambro-Brit. SS. 52-5).
73. Skene's Four Ancient Books of Wales II. 431-5, where the exploits of Cadwallon in different parts of Wales are referred to.
74. Skene's Four Ancient Books of Wales II. 173, 447 ; Cambro-Brit. SS. 79; Owen's Pembrokeshire I. 222, note 2 ; III. 281.
75. 'Annus CLXIX. Gueith Cairlegion et ibi cecidit Selim filius Cinan ' (Y Cymm. IX. 156 ; Bede's H. E. II. 2 ; Owen's Pembrokeshire III. 282, note i). The above annal is to be reckoned from the false Bedan date of the Saxon Advent, viz. 449 + 168 = A.D. 617. Cf. Plummer's Bede II. 77.
76. Rhys's Celtic Britain, 3rd ed. 130.
77. lolo MSS. 86. The same fragment contains the equally fictitious patria of Fferyllwg ' between Wye and Severn ' (Owen's Pem. III. 257, note 3).
78. Y Cymm. VII. 119, note 3 ; IX. 179, note 6.
79. In 685 Ecgfrid gave St. Cuthbert 'terramquae vocatur Cartmel et omnes Britannos cum eo '. Hist. de S. Cuthberto (Symeonis Dunel. Opera I. 141, 231. Surtees Society).
80. Hist. Reg. Brit. XII. cc. 14-18.
81. ' Dum ipse [Osguid filius Eadlfrid] regnabat venit mortalitas hominum Catgualart regnante apud Brittones post patrem suum et in ea periit.' Hist. Britt. c. 64 (Chr. Min. III. 208). Oswy reigned from 642 to 670, and the plague referred to raged in 664-5 (Bede's H. E. III. 27). The Ann. Camb. places the obit of Cadwaladr oppo- site Annus CCXXXVIII, which if calculated from 428, the true year of the Saxon Advent, gives 428 + 237 = A.D. 665. According to Geoffrey, Cadwaladr died in 689 (XII. 18), which historically is the year of the obit of Ceadwalla of Wessex in Rome (Bede's H. E. V. 7). Allowing one year for Geoffrey's aliquantulum temporis (XII. 17) and adding the eleven years of adversity (XII. 16), and also the twelve years of prosperity (XII. 14), we obtain 1 + 11 + 12 = 24 years as the length of Cadwaladr's reign, which brings us to the true date of Cadwaladr's death, viz. 689 - 24 = A.D. 665. As Cadwaladr succeeds his father Cadwallo[n] immediately, and as the latter is made to die on November 17, after a reign of forty-eight years (XII. 13), we obtain 665 -48 = A.D. 617 as the first year of Cadwallo[n]'s reign according to Geoffrey, which is historically the date of the Battle of Chester. Geoffrey, therefore, has clearly confounded the three kings, Cadwallon, Cadwaladr, and Ceadwalla ; and by making Cadwaladr die in the year of Ceadwalla's death, he has almost certainly given us the true deathday of Cadwaladr as that of Cadwallon, in which case Cadwaladr died on November 17, 665.
82. Cynan's obit is placed opposite Annus CCCLXXII, which in the era of the Ann. Camb. gives 445 + 371 = 816. For the Pedi- grees see Y Cymm. IX. 169, 172 (Ped. I and IV) ; VIII. 87 (Peds. XVII and XIX). Owen's Pembrokeshire III. 209.
83. Anc. Laws I. 342. ' Rrodri vab Kamwri ' (from MS. Z). The same idea is implied in what Asser says of certain South Welsh kings seeking Alfred's protection, being forced thereto filiorum Rotri vi. The vis or camwri denotes the aggressive policy of the kings of Gwynedd (Stevenson's Asser, p. 66).
84. Jesus Coll. MS. 20, Peds. XVIII, XX, XXI. For Seisyllwg, see Oxford Mabinogion, p. 25, at the end of the Mabinogi of Pwyll. It is so called from Seisyll (Ped. XXVI, Y Cymm. IX. 180), King of Ceredigion sometime in the eighth century, who deprived Dyved of the cantrevs which together were afterwards known as the gwlad of Ystrad Tywi. Before this deprivation the kings of Dyved had come into possession of Brycheiniog through Ceindrech, a lady of the line of Brychan. Brycheiniog afterwards, however, appears to have had a line of its own, represented in Asser's day by Helised ap Teudubr. De rebus gestis Aelfredi, c. 80 (Stevenson's Asser, p. 66).
85. Annus CCCCXXXIII in the Ann. Camb., which in the era of the Annales gives 445 + 432 = 877.
86. Such is the tradition of the tripartite division as given by Gerald in his Descriptio Kambriae I. 2 (Girald. Camb. Opera VI. 166).
87. Peds. I and II in Y Cymm. IX. 169, 171 ; Ped, VIII in ibid. VIII. 85.
88. Ex omni natione certainly stands for examinatione and goes with diligenti as in the prefaces of the other Latin texts ; likewise for medio circiter read mediocriter.