The chronology of the various extant pedigrees of Powys has always been a problem for those who have attempted to analyze them. One guidepost in time is the virtually certain knowledge that Vortigern, as overking of the Britons in 425 and for perhaps a quarter of a century thereafter, must have been born late in the fourth century. An estimate of 375/380 should be close. The next man in the pedigrees whose birthdate can be closely estimated is Selyf Sarff Cadan, who fell at the Battle of Chester in 616.[1] Already the king by that date, Selyf could scarcely been born later than about 590[2] nor earlier than perhaps 560 or he would have been too old for the battlefield. Assigning him a birthdate near 575 would be chronologically consistent with what is know of his descendants.[3]

        One should expect to find six, at most seven, generations to span the 200 years from Vortigern to Selyf.  But the medieval genealogists were faced by pedigrees which list 10 generations from one man to the other.[4] Their solution was to delete 3 names near the top. What had read "Cadell Ddyrnllwg ap Pasgen ap Brydw ap Rhuddfedel Frych ap Cyndeyrn ap Gwrtheyrn" was emended to "Cadell Ddyrnllwg ap Cyndeyrn ap Gwrtheyrn" with the curious justification that 3 of the four names which follow Gwrtheyrn were also names of his sons which must have been listed vertically. No one would dispute that Cyndeyrn (or Cadeyrn or Cattegern, however spelled) and Brydw and Pasgen were, in fact, names of his sons.  But not exclusively; they were fairly common male names of that era. They had no idea who Rhuddfedel Frych was, but decided he had to go as well. Voila! We now have a "chronologically stable" pedigree which makes Cadell Ddyrnllwg the grandson of Gwrtheyrn (Vortigern).

      Or do we? Turning to other ancient pedigrees, it could be seen that 8 generations were listed from Cadell Ddyrnllwg down to Selyf.[5] If in fact Cadell was two generations removed from Vortigern, his birthdate should fall somewhere near 435.  Thus, we should expect only five generations in a "stable" pedigree to reach Selyf. Once again the medieval shears came out to right things.  What had been "Selyf ap Cynan Garwyn ap Brochwel Ysgithrog ap Cyngen ap Maucant (or Mawgan or Manogan) ap Pascent ap Cattegern ap Cadell" was recast to omit "Maucant ap Pascent ap Cattigern". The net result of these emendations can best be seen in a chart:

                   385 Vortigern
                   415 Cadeyrn
                   445 Cadell Ddyrnllwg
                   475 Cyngen Glodrydd
                   510 Brochwel Ysgithrog
                   545 Cynan Garwyn
                   575 Selyf

          Our own inquiry into the matter was prompted primarily by the assignment of what seemed to be an unreasonably late floruit for Cadell Ddyrnllwg. The fanciful tale related by Ninnius, while typical in its assignment of heavenly powers to men canonized as saints, contains at least some historical basis. If we accept the chronology of Ninnius, the story of Cadell was set during the first visit of Bishop Germanus to Britain in 429. But if Cadell was born in 445, there must be some mistake; it must have been during his second visit in 447. This works if we assume Cadell was about 2 years old at the time or at most a teenager. But Ninnius tells us Cadell already was the father of 9 children at the time. Common sense tells us his age must have been nearer to 40 than to his teens. And if the setting was, in fact, the 429 visit of St Germanus[6], the Cadell in his report must have been born before the turn of the century, perhaps even as early as 380.  Even the renowned Peter Bartrum concedes that Cadell was contemporary with Vortigern.[7]

        So even after deleting six generations (about 200 years) from the various pedigrees of this family, we still have failed to achieve "chronological stability". Anyone can make a jigsaw puzzle "fit" by trimming the pieces, but this method will never reproduce the original uncut picture. Perhaps it is time to question the boundries which we have used as absolutes. Is it possible that one or more of the men have been confused with same-named men of an earlier era?

       The earliest known pedigree of the Gwrtheyrn called Vortigern, that from Ninnius, makes him "son of Vitalis, son of Vitalinus, son of Gloiu". These Latin names were rendered by the Welsh as "ap Gwidol ap Gwdoleu ap Gloyw Gwallt Hir".[8] But the pedigrees cast for the Powysian dynasty cite the ancestry of their Gwrtheyrn as "ap Rydeyrn ap Deheuwaint ap Endicant ap Endeyrn ap Enied ap Endos ap Enddolen ap Afallach ap Affleth ap Beli Mawr".[9] That same former Catuvellauni family appears in the cited pedigrees of Cunedda and Coel Hen.[10] When birthdate estimates are applied to the various branches which are consistent with the known floruit of those two men, the Rydeyrn ap Deheuwaint cited as father to Gwrtheyrn belongs to the middle of the second century or about 155 AD. If we were to assign a birthdate near 185 for that Gwrtheyrn, we would find it consistent with those pedigrees previously considered too long. With no other changes, Cadell Ddyrnllwg would occur five generations later around 350.  But we are persuaded by the preponderance of citations which say the father of Cadell was named Cadeyrn, not Pasgen as in the pedigrees first cited in note[4].

         The earliest known pedigrees[11] begin this family with Cadell and say nothing of his ancestry. They do, however, cite a son named Cadeyrn as they descend down to Selyf and beyond. If we were to simply transpose the two names, it would both move the birth of Cadell to near 380 and give him a father named Cadeyrn. The resulting pedigree would look like this:

           155  Rydeyrn
           185  Gwrtheyrn
           220  Cadeyrn
           250  Rhuddfedel Frych
           285  Brydw
           315  Pasgen
           350  Cadeyrn
           380  Cadell Ddyrnllwg

           By simply repeating a common error found in many pedigrees, that of rolling two men named Cadeyrn into one and omitting the intervening names, we would effectively repeat the medieval emendation that removed all the "sons of Vortigern" (except Cadeyrn) plus Rhuddfedel Frych.

          This brings us to the evidence gleaned from the Pillar of Eliseg. Whatever may have been carved there early in the ninth century, we only know what was legible when transcribed by Edward Lluyd in 1696. That transcription contains no connected pedigree beyond Eliseg of the late seventh to mid-eighth centuries. It appears to conclude with the names Cyngen, Pasgen, Maun and Annan, then Brydw son of Vortigern. Whatever relationship existed between Annan and Brydw is left to conjecture.

       We know a young lady named Annun was a member of the household of Madrun, daughter of Vortimer the Blessed, and is cited as her handmaiden. If the Annan of the Pillar were a female, she may have been the wife of Maun and a daughter of Brydw. Those pedigrees which include "Cyngen ap Maucann"[12] in the Powys dynasty might be no more than a scribal attempt at rendering "Cyngen, son of Maun and Annan". Certainly there is no Maucann or Mawgan in Lluyd's list of Pillar names. It is not necessary to conjecture whether the Pillar Annan was identical to the handmaiden of Modren vz Vortimer[13], nor even whether Vortimer was identical to Brydw.[14] But no one would argue that the Powys dynasty proudly proclaimed it descended from a mere handmaiden of a granddaughter of Vortigern.

            The possibility that Annan was female and formed the connection between Vortigern and the Powys dynasty would place virtually all the ancient pedigrees into chronological alignment, something none of the other emendations have done.  The chart would look like this:
                 155  Rydeyrn[15]
                 185  Gwrtheyrn
                 220  Cadeyrn
                 250  Rhuddfedel Frych
                 285  Brydw                                   Gloyw Gwallt Hir 280
                 315  Pasgen                                  Gwidolin 310
                 350  Cadeyrn                                 Gwidol 350
                 380  Cadell Ddyrnllwy                      Gwrtheyrn (Vortigern) 385
                 410  Pasgen                                  Brydw 415
                 440  Maun====================Annan 450
                                   475 Cyngen
                                   510  Brochwel Ysgithrog
                                   545  Cynan Garwyn
                                   575  Selyf, obit 616

Notes

[1] Annales Cambriae lists the date as 613; the actual year is debatable.
[2] Celtic tradition never included child-kings. Selyf must be presumed to have been in his 20's or older at the Battle of Chester. 
[3] Six generations after Selyf's brother, Eiludd, we find Cyngen ap Cadell who was born c. 775; this would be the expected generational span for 200 years.
[4] ABT 20 and HLG 2 from Bartrum, P.C. (1966): Early Welsh Genealogical Tracts, (Cardiff).
[5] Harleian Ms 3859, 22 & 27. 
[6] The year 447 for the second visit has received wide popularity, but it may have been as early as 435.  See http://www.vortigernstudies.org.uk/artsou/constant.htm.
[7] Bartrum, P.C. (1966): Early Welsh Genealogical Tracts, (Cardiff), pp. 129 but inexplicicably suggests that Cadell could have been born c. 404 and still be a grandson of Vortigern
[8] No earlier pedigree of Gloyw is extant but modern genealogists make him a son of the Rhodri (Rydeyrn?) ap Euddigan (Endigant) who occurs in the pedigrees of Coel Hen, and whose birth early in the second century makes such a connection untenable.
[9] ABT 9 and Jesus College Ms 119 "Buchedd Bueno" in Bartrum, P.C. (1966): Early Welsh Genealogical Tracts, (Cardiff). 

[10] Harleian Ms 3859 1 and 10

[11] Harlean Ms 3859, 22 and 27,
[12] ibid and Jesus College Ms 20, 18 which render variant spellings "Maucant" and "Manogan". 
[13] The handmaiden was an offical member of the court, not a mere servant.  Nothing in known customs would preclude a young girl from serving in that capacity for an older sister.
[14] Ninnius does not list Brydw among the sons of Vortigern but does name Vortimer.  They may refer to the same person. 
^[15] ABT 1; ABT 9; and JC 20,5 extend this line back to Beli Mawr.