John Rhys and Thomas Jones suggested many emendations in order to regularise syllable counts. Some of their emendations may be excessive, since variation may have been tolerated, as in the saga englynion analysed by Jenny Rowland, which have many stylistic similarities to Englynion y Beddau. That said, it is clear that the scribes sometimes disregarded metrics and tinkered with the texts in order to clarify their prose sense (e.g. I.35c and III.2c).

Nearly all the englynion in Englynion y Beddau are milwr or penfyr, the two three-line types already found in the Juvencus manuscript in the early tenth century. These were not used by the court poets of the twelfth century onwards – or, rather, not in the official poetry attributed to them by name. There are also some examples of four-line englynion, either the englyn gwastad (first attested in manuscript at Llanbadarn in the late eleventh century) or the englyn cyrch, which chiefly occurred in popular verse in the later Middle Ages. Four-line englynion of the more elaborate types used in the court poetry of the twelfth century onwards are absent, and the syllable counting and ornamentation is less strict than in that corpus.

Without ruling out a much earlier date, it seems reasonable to suppose that some parts of Englynion y Beddau were composed at least as early as the first half of the twelfth century. Once their format was established, however, one can imagine later poets adding further englynion in the same metres, as has been suggested in the case of the englynion referring to Norman kings in prophetic englynion such as the Cyfoesi. Metrics are therefore an unreliable dating guide. III.17, an englyn penfyr, is unlikely to be earlier than the fourteenth century (see commentary), but it does not stand out metrically.

Englynion milwr

Rowland shows that 7-7-8 and 8-7-7 were probably acceptable in englynion milwr as well as the canonical 7-7-7. In the Beddau there is one example (37) of 7-7-8, and although its 8 in line c could be reduced to 7 by deleting the optional adjective plural ending -ion, that is probably not mandatory.

The anonymous Englynion y Beddau name and describe the burial places of early medieval heroes, providing a key to a narrative literature which is largely lost. In this they resemble Trioedd Ynys Prydain (‘The Triads of the Island of Britain’). While the Triads’ mnemonic function is accomplished by grouping information in threes, Englynion y Beddau are suitable for memorisation because they are metrical and use poetic formulae.

Two series of Englynion y Beddau survive, Series I in the Black Book of Carmarthen (NLW Peniarth 1), c. 1250, and derivatives of it, and Series III in various derivatives of a lost copy, with partially modernised orthography, in a manuscript belonging to William Salesbury (c. 1510–c. 1580). The so-called ‘Series II’ and ‘Series IV’ are not true ‘series’ (see below, p. 18). A reference such as I.10 or III.10 refers to the number of the stanza (usually in englyn metre) within the relevant series, and a reference such as I.10c refers to the line (here the third) within the stanza.

On 32r of the Black Book a late medieval scribe added the title Englynnionn y Beddev. This is the first attestation of the now customary title, translated as ‘The Stanzas of the Graves’ by Thomas Jones in his 1967 edition. He also included the alternative title Beddau Milwyr Ynys Prydain, ‘The Graves of the Warriors of the Island of Britain’. Although not attested before the early seventeenth century, this title is quite apt. All stanzas, except one (I.70) which was added to the Black Book of Carmarthen by a second thirteenth-century scribe, refer to men rather than women; even Branwen, whose grave is described in the Second Branch of the Mabinogi, is missing. With the exception of Beli ab Benlli Gawr – another addition (I.73) – the warriors seem to be humans rather than giants, although that distinction was fluid and could change over time. Typically the men are heroes praised for their ferocity in battle. They come from the whole of Britain, including Mercia and present-day Scotland. Warriors from the ‘Old North’ include Cynon ap Clydno Eidyn (I.8–11), Owain ab Urien, and Rhydderch Hael (I.13–14), Einion ap Cunedda (I.72), Llofan (III.5–8), and Elidir Mwynfawr (III.15). From Wales come Cynddylan of Powys (I.15), Aergol of Dyfed, and Maelgwn Gwynedd’s son, wife, and bard (I.70–71), among many others. From Mercia comes Panna ap Pyd (III.9/I.10).

It might be supposed that the extant series of Englynion y Beddau are the result of combining shorter, exclusively regional series comparable to the Irish poem on the graves of the Leinstermen. Good evidence to the contrary is provided by the stanzas which group more than one grave place within the indivisible unit of the englyn (I.4, 7, 8, 10, 12, 13, 26, 36, 39(?), 44, 46, 50, 71; III.3(?), 7, 9, 13). While some of the paired places are very close together (I.4/III.13, I.46, III.7) or at least in the same part of Wales (I.7, I.12(?), I.36, III.3(?)), others of them are widely separated, notably I.10/III.9 (graves in England and Wales), I.26 (Ardudwy and Brycheiniog), and I.71 (Dinorben and Dyfed), and quite likely I.8, I.13, and I.50 as well. Evidently their authors were not composing englynion for series organised on a regional basis.

A work not limited to a narrow geographical area can nevertheless show a regional bias, owing to the limitations of its author’s knowledge or interests. An example is the Old English tract on ‘The Saints who Rest in England’, which reached its current form c. 1031. Despite its title, the first half is biased towards Northumbria and Mercia, while the second half, which was evidently added later, ignores Northumbria and is biased towards Wessex. Another good example is the eleventh-century(?) Irish tract on the ‘Wonders of Ireland’. In its fullest form, in the Book of Ballymote, it has 34 entries, of which less than a third refer to Munster and Leinster, indicating a northern or central Irish place of origin for the tract. The eleventh-century Latin version lacks items 15–20, 25, 27–28, 30 and 33, nearly all of which refer to places in Meath. This indicates that the vernacular ‘Wonders of Ireland’ was originally a regionally biased northern compilation, subsequently conflated with a strongly regional Meath collection of marvels. It is interesting from our point of view to note that the latter (Meath) collection is not dispersed through the expanded text; on the contrary, its entries tend to survive in clumps, between item 15 and the end of the ‘Wonders’. A similar tendency, and various regional biases (to be discussed later), can also be detected in other compilations, such as Bonedd y Saint and Englynion y Beddau.

(i) The manuscripts and stemma

Since all other manuscripts of Series I derive from the Black Book of Carmarthen (NLW Peniarth 1, 32r–35r), their variant readings are ignored in the edition below, although occasional reference is made in the commentary to significant variants or marginal comments. The englynion, set out as prose in the Black Book, are printed as verse, and word division and capitalisation are introduced. The exact layout on the manuscript page can be seen on the National Library of Wales’s website.

In roughly chronological order, the manuscripts of Series I which have been consulted are as follows:

Although there were presumably a number of pre-Reformation copies of Englynion y Beddau in addition to the Black Book of Carmarthen, none survives, and there are no certain allusions to the text, even in the Triads. It may be significant that Gruffudd Hiraethog (d. 1564), despite his keen interest in old stories and traditions about places, never seems to draw on Englynion y Beddau.

The first extant post-medieval manuscripts are Thomas Wiliems’s possibly direct copy, in BL Add. 31055 (c. 1591–96), of William Salesbury’s now lost manuscript of Series III and Jaspar Gryffyth’s modernised copy of Series I in Llanstephan 120, made sometime between c. 1597 and 1607 as part of his transcription of the Black Book. Neither of these scholars had any evident interest in the contents, to judge merely by the scarcity of marginalia, except that Thomas Wiliems notes the name dylan beside III.8c and correctly identifies ‘Ân ap lhian’ in III.17a as ‘Merdhin Embrys’. The same is true of the texts of Series III and Series I in Peniarth 111, transcribed c. 1607 by John Jones of Gellilyfdy from a lost manuscript by Roger Morris of Coedytalwrn, Llanfair Dyffryn Clwyd, Den. (fl. 1582–97). Lack of annotation does not prove lack of interest, however, for John Jones ‘corrects’ Mabon vap Madron (III.15c) to Mabon vap Mydron in order to agree with Mabon am Mydron in the Black Book of Carmarthen poem Pa gur yv y porthaur? And some years later he would quote I.73 in connection with the alleged grave of Beli ap Benlli Gawr (see commentary). All these scribes were connected with the Denbighshire–Flintshire region, where the humanistic spirit of the Renaissance flourished, and no doubt their priority was to preserve texts rather than to comment on them. Another copyist in the same area, apparently involved in transmitting Englynion y Beddau, was Salesbury’s contemporary Richard Longford (or Langford), but his role remains unclear.

In Peniarth 98ii (not before 1617), Dr John Davies of Mallwyd, copying Series I from the Black Book on pp. 18–25 and Series III from John Jones’s Peniarth 111 text on pp. 48–50, displays a different kind of interest from John Jones. Unlike Jones, who was an exact, careful copyist, Dr Davies silently emends, improves the metre, and underlines problematic words for further consideration or for inclusion in his Dictionary.

Series I

Series I is found in the Black Book of Carmarthen (NLW Peniarth 1). This was written c. 1250, and certainly after 1212, possibly at the Augustinian priory at Carmarthen. It has been known by its present name at least since the sixteenth century. Until recently it was believed that an early-fourteenth-century Gwynedd(?) scribe who added lines to the Afallennau prophecies at the foot of 24v referred to it as ‘y Llyfr du o Gaerfyrddin’ in an ‘exact copy’ of it, mentioned by Humphrey Humphreys (d. 1712), bishop of Bangor and Hereford, but Gruffudd Antur has shown that the bishop was probably misled by a seventeenth-century semi-facsimile on parchment made by Robert Vaughan (now Peniarth 107). The Black Book scribe was evidently a southerner, judging by his language, and many of the poems he included point to a connection with the south-west, perhaps with Carmarthen itself or with one of the nearby monasteries patronised by the family of the Lord Rhys (d. 1197), such as the Cistercian house at Hendy-gwyn (Whitland, Crm.) or the Premonstratensian house at Talyllychau (Talley, Crm.). While the Lord Rhys and his father-in-law Madog ap Maredudd (d. 1160) are prominent in the Black Book praise poems, none of his sons appear; this suggests that the mid-thirteenth-century Black Book scribe drew on a small collection of panegyrics from the previous century.

Englynion y Beddau appear on fols 32r–35r. I.1–69 are written by the main scribe and end with what looks like a concluding englyn (I.69). Afterwards another thirteenth- century scribe inserted I.70–73 in the remaining space at the foot of 35r. He might perhaps have added more if there had been room. His addition is the clearest indication that the text grew by accretion. It seems the main scribe did not copy I.1–69 at a single sitting. Within his part there are a number of changes of ink or pen, sometimes within stanzas, but once (33v, line 8) at the start of a new stanza (I.32). This stanza is one of a pair of linked stanzas, the second of which (I.33) corresponds to the first stanza of Series III. Does this indicate that he was moving to a new exemplar connected somehow to Series III? More likely, this semi-correspondence is a coincidence.

Dating Englynion y Beddau is difficult, owing to the lack of precisely dated material with which to compare them and the probability that the series grew by accretion, like other medieval compilations such as the Triads, the genealogies, and Bonedd y Saint. That is clearly the case with some other poems in englyn metre, including Ymddiddan Arthur a’r Eryr and Enwev Meibon Llywarch Hen in the Black Book itself, where the stanza ‘Tri meib Llywarch (recte Kynwarch)’ is an interpolation inspired by Brut y Brenhinedd. A comparable Middle Irish example of accretion is provided by the above-mentioned poem on the deaths (and graves) of ancient heroes, ‘Fíanna bátar i nEmain’, by Cináed úa hArtacáin (d. 975). Two stanzas before the end, in the Book of Leinster copy, Find, a mid-twelfth-century bishop of Kildare, interpolated eleven stanzas of his own composition, bringing the poem up to date. In that case we know the names and dates of the two poets. In the case of Englynion y Beddau we are in the dark, but the parallel suggests that the date of one stanza or group of stanzas cannot automatically be applied to the rest.

Citing features such as their language, style, metre, ‘heroic mood’, and textual corruption, Thomas Jones dated the composition of Englynion y Beddau to ‘a period considerably earlier than … the Black Book of Carmarthen …, probably as early as the ninth or tenth century’. Comparison with the Old Welsh englynion added to the Juvencus Manuscript in the early tenth century suggests that so early a date is quite possible. Yet it cannot be proved without making unwarranted assumptions about when the features in question became obsolete. The ‘heroic mood’ is an example.

The ‘heroic mood’

At one time, scholars used to assign pure expressions of the ‘heroic ideal’, such as we see in Englynion y Beddau, to an early date and a later date to more realistic or even cynical opinions about the value of fighting to the death and so on. Later they came round to the view that the contrast was less a matter of date than of genre: panegyric versus narrative. In fact the militaristic praise poetry of the twelfth- and thirteenth-century court poets has a purer, more unquestioning heroic ethos than the undoubtedly earlier ‘saga englynion’ of Canu Llywarch Hen. Englynion y Beddau may even cultivate an archaic tone.

One page of Lhwyd’s copy of Series I and Series III from the lost manuscript of Roger Morris survives as Llanstephan 145, fol. 68. This runs from I.62 (misnumbered ‘63’) to I.72/73 (treated as one stanza, owing to a lacuna, and numbered 73) and is followed by the complete text of Series III (our MS B), numbered from 74 to 91 (but what should be numbered ‘83’ is numbered ‘82’ so that the sequence is … 82, 82, 84 …). Later bound next to it (probably by Moses Williams) as Llanstephan 145, fol. 69, is the index to both series which Lhwyd sent to William Baxter, probably in early 1708. The index was started by Lhwyd (up to H-) and was completed by Williams, his assistant. The latter’s contribution, in a neat cursive hand, is printed in italics below.

Lhwyd’s own orthography is employed in the index, with Greek letters used as follows: χ = ch; δ = dd; ƛ/Λ = ll/Ll; ρ/P = rh/Rh; τ = th. An undotted i may sometimes be iota representing ‘dark y’, and sometimes be the result of forgetting to write a dot; whichever the case, it is transcribed as ı below. There is one example of ʒ for g (s.n. Λuχ) and one ε for e (s.n. Koel).

Alongside the following transcription I give, to the right, the standard numbering of stanzas in Series I (plus the relevant Black Book text) and in Series III. Strikethrough indicates crossings out in the manuscript. On the use of bold and underlining see below.

At least two systems of numbering can be detected behind the index. The Series I numbers in the part mostly written by Lhwyd, from Aber Dyar 55 in col. 1 to Hennyn henben 72 early in col. 2, are mostly one higher than the modern numbering (marked in bold without underlining in the corresponding Black Book citations), with the exceptions of Gurien 2, Guén 3, Aren 4, Auen ab Tede 4, and Bryn Aren 4. Lhwyd originally numbered Dılan Bd. at Λan Beyno (I.4) as ‘5’. The obvious explanation for the ‘5’ is that he originally counted I.4ab and I.4d as two separate items – 4 Tede at Bryn Aren and 5 Dılan – resulting in the subsequent englynion being numbered +1; for example, Aber Guenoli and Guàƛog hír (I.7) are numbered 8 rather than 7.

(i) The manuscripts and stemma

S All extant manuscripts seem to derive from William Salesbury’s lost manuscript ‘S’. The primary witnesses to this (W, N, T, P, and B) are transcribed in full in the edition below, with their exact lineation and indentation. B’s numbering of the englynion is included, as are the relevant entries in the index by Edward Lhwyd and Moses Williams, now bound as fol. 69 of NLW Llanstephan 145, after the B text on fol. 68. Williams’s entries are in italics. Occasional reference is made in the commentary to significant variants or marginal comments in secondary witnesses, including the well-known Peniarth 98ii by Dr John Davies, Mallwyd, the manuscript printed by Thomas Jones (his ‘Peniarth 98B’). The englynion are divided and numbered as in Thomas Jones’s edition, including the problematic stanzas 11 and 12, even though it will be suggested in the commentary that III.11d may really be line a of III.12.

For a stemma, see Figure 4 below, p. 280, and for further discussion of the secondary witnesses see above, pp. 82–93.

In the absence of S the primary witnesses to Series III are as follows: