Why Ossa was not Osla Gyllellfawr

The identification of Osla Gyllellfawr with Ossa, grandfather of Ida, king of Bernicia, as suggested by corrupted late entries in the Bonedd y Saint, shown below, is incorrect:

Gen. ByS 70 ByS 71 JC20 17 seg.
4 Eda Glynuawr Tegyth
3 Osswallt Oswydd aelwyn Gwynbei drahavvc Ceit
2 Mwc Mawr Drevydd Douc
1 Ydolorec vrenin Offa kyllellvawr Llewarch hen

The names given in the above chart are in the manuscript form. The first two individuals in ByS 70 are Oswald and his brother Oswiu, wrongly shown as his father but corrected in the above table. The third name would then be their grandfather, the Bernician Æthelric, wrongly shown as Oswiu’s father and this has also been corrected for. An alternative interpretation, which would be correct in terms of parentage, would be Oswine son of Osric son of the Deiran Æthelric.

Bartrum maintained the first name in ByS 71, Eda Glinfawr, is Æthelric’s father, Ida. However, that is impossible since Eda can safely be placed in gen. 4, more than a century after Ida. This conclusion is arrived at by noting Eda was the grandson of Mwng Mawr Drefydd who was in conflict with Mechydd ap Llywarch Hen. Llywarch can safely be placed in gen. 1. Furthermore, as Ossa was Ida’s grandfather he would have been far too early to be a contemporary of Arthur.

It would seem that the author of ByS 71 added cognomens that did not actually apply to Eda and Offa but instead belonged to other individuals with similar names. So, Glinfawr came from the name of the father of Ecgbert of York, Eata glinmawr, mentioned in HB 61. Likewise, Offa’s cognomen derived from that of Esla, see Dating the Wessex generations.

Melville Richards’ identification of the 8th C Offa of Mercia is unlikely for the chronological reason.[1] Also, the dissimilarity in the names suggests it is doubtful that Osla was the Kentish Ochta as suggested by Idris Llewelyn Foster.[2]

[1] Richards, M., 1948, 46.
[2] Foster, I. L., 1961, 42.

Dating the Wessex generations

Below is a table that dates the Wessex genealogies up to Alfred’s time.

Gen. Reign Obit Gen. obit Deviation
18 Alfred 899 889 10
17 Æthelwulf 858 862 -4
16 Egbert 839 836 3
12 Ine 728 729 -1
11 Cædwalla 688 702 -3
10 Centwine 685 676 -14
9 Cynegils 642 649 9
8 Ceolwulf 611 622 -11
7 Ceawlin 588 596 -8
6 Cynric 581 569 12
5 Creoda/Cerdic 554 542 12
4 Elesa/Esla 516
3 Giwis/Wig 489
2 Brond 462
1 Bældæg 436
0 Woden 409

The generations have been arrived at by using the following structure:

Yorke, B., 2003

Each generation has been dated using the obit of the individual ruler who died last in that particular generation. So, for example Alfred’s obit has been used rather than those of his brothers. With Saxon pedigrees, unlike Welsh ones where the crown generally passed down to the next generation, it was not unusual for it to go to a sibling. There are gaps in the generations, for example between Egbert of gen. 16 and Ine of gen. 12 as we do not have the obits of the intervening rulers. Note, the obits of Ceawlin, Cynric and Cerdic are not those indicated by the Anglo-Saxon Chronicle but instead the values obtained by Dumville’s correction. The zero generation is that ascribed to Woden, unlike the tables for the British kings which uses Arthur’s generation. The two sets cannot be synchronised since they use differing step-sizes.

The analysis shows that for Wessex the generation step-size was 26.7 years. The deviation column shows how far the individuals obit differs from that ascribed to the generation to which he belongs. Most deviations should be less than half a step-size and indeed this is the case.

Cotton MS Tiberius B V 1 British Library

As Sisam noted Elesa/Esla and Giwis/Wig are alliterative pairs for the same individuals. To this list has been added another pair, namely Creoda/Cerdic. The generation obit values for Elesa to Woden are estimated values using the step-size given above. It will be noted that Esla was a contemporary of Arthur. This together with the similarity of names allows us to be fairly confident that he was Osla Gyllellfawr. Sisam points out that the name Esla is unknown in English.[1] This may indicate its British origin. He says Elesa is also unknown unless it is etymologically the same as Elsa in Widsith.[2]

The postulate that there were two Wessex dynasties resolves the paradox that although the ASC tells us Wessex originated from the South coast with Cerdic, the oldest Saxon sites are around the upper Thames valley and the founder was a Giwis.

[1] Stanley, E. G., 1990, 164.
[2] Sisam, K., 1953, 302.

Arthurian connections with Ewyas and Ergyng

In Culhwch ac Olwen,during his chase, Twrch Trwyth killed Llygadrudd Emys and Gwrfoddw, Arthur’s uncles, his mother’s brothers. The latter name appears in that of Gwrfoddw Hen, king of Ergyng, but he appears to be a later ruler. However, Welsh dynasties often preserved the same name, so Gwrfoddw Hen may have been a descendant. This suggests Eigr, Gwrfoddw’s sister, could have come from that region.

Ergyng may have covered parts of Herefordshire, Monmouthshire and Gloucestershire. The Brut y Brenhinedd calls Eudaf, an ancestor of Arthur, as “Eudaf yarll ergig ac euas”, that is Earl of Ewyas and Ergyng. However, Geoffrey refers to him anachronistically as “Octavius dux Wisseorum”, presumably the territorial name being derived from Welsh Ewyas.

Magnus Maximus had a daughter, named Sevira, by Elen, daugther of Eudaf. It was through Gwrtheyrn’s marriage with Sevira that he gained control of the territory that was to become known as Ewyas. Geoffrey referred to him as the “Consol Gewissiorum”. He invited Germanic warriors to settle in the Abingdon area to help defend attacks on his territory in Ergyng. The ASC confuses this event with the later settlement in Kent. The name for Gwent is easily confused with that for Kent. Gwrtheyrn locating the Gewisse, a Saxon tribe, in the upper Thames valley made logistical sense, as his opponent, Emrys Wledig i.e. Aurelius Ambrosius, was based in the Wiltshire area. Located in that county is the village of Amesbury, formerly known as “Ambres byrig” in the Cartularium Saxonicum.[1] It is likely that the East Wansdyke earthwork was built by the Britons as a defense against attack from the north.

Cerdic is attributed in the Anglo-Saxon Chronicle as the founder of the Wessex dynasty. However, as Dumville has pointed out, his reign started later than the chronicle suggests. In fact, he belonged to a second Wessex dynasty. The first dynasty was founded by the eponymous ruler, Wig/Giwis, the two names being, as Sisam explains, alliterative pairs. His reign was followed by that of Esla/Elesa. The latter is known as Osla Gyllellfawr, whose defeat by Arthur brought the first Wessex dynasty to an end. For obvious reasons, this disaster goes unmentioned in the ASC.

[1] Birch, W. de G., 1887, 178.


The name Cawrnur occurs in the Kadeir Teyrnon, The Chair of the Prince, which speaks of pale horses under saddle being led from him. In the Marwnat vthyr pen, Uthr Pen[dragon]’s Elegy, there is a reference to an attack on the sons of someone named Cawrnur. Sims-Williams wrote:

“Presumably the fact that Cawrnur and Arthur rhyme partly explains their collocation, but both poems may allude to some lost Arthurian story.”[1]

If we speculate that for the sake of rhyming Cawrnur is a variant of the individuals actual name than a reasonable candidate would be Cawrdaf ap Caradog Freichfras who was of gen 0, see St. Collen. According to triad 13 he was one of the Chief Officers of the Island of Britain. He appears as one of Arthur’s counselors in Breuddwyd Rhonabwy when Osla Gyllellfawr asked for a limited truce.

Gen. ByS 51 ByS J 51 ByS Y(S) 88 ByS Y(S) 89
2 St. Dyfnog St. Dyfnog
1 Medrod Medrod Gwenhwyach Iddew Corn Brydain St. Cathen
0 Cawrdaf Cawrdaf Gocuran Gawr Cawrdaf Cawrdaf
-1 Caradog Freichfras Caradog Freichfras Caradog Freichfras Caradog Freichfras
-2 Llŷr Marini Llŷr Marini

Gwenhwyach was the wife of Medrod. TYP 53 indicates a dispute between her and  Gwenhwyfar led to Camlan. Iddog Cordd Prydain, the Embroiler of Britain, appears in Rhonabwy‘s Dream as one of the messengers between Arthur and Medrod. However, he twisted Arthur’s words when reporting them as he was keen for the battle to occur. These hostilities may be what is alluded to in the references to Cawrnur. The Pen. 51 version of triad 51 tells us that Idawc ap Nyniaw was called  Idawc Korn Prydyn from which we can conclude Iddog Cordd Prydain is the same person as Iddew Corn Brydain.

However, as Gwenhwyfar would have belonged to gen. 0, the Gwenhwyach of ByS J 51 could not have been her sister. Furthermore, Medrod ap Llew has been conflated with Medrod ap Cawrdaf. The existence of two Medrods would explain why different personalities have been ascribed to the name Medrod.

[1] Bromwich, R., Jarman, A.O.H., Roberts, B. F., 1991, 53.