444 NOTES AND QUERIES. [12 s. x. JUNE 10, 1022. reject the identification quoted. Moreover, about nine miles south of Ancaster and twenty-six miles from Lincoln lies Keisby, and that, as I pointed out in my note on ' Lindum Colonia and its Neighbours ' ('N. & Q.,' 12 S. ix. 524), represents the
- ' Chiesebi " of Domesday Book. " Chiesebi "
is O.E. Ciesanby, " the habitations of Ciesa." " Ciesa " postulates *Clesa > *Ceasi, and these are O.E. adaptations of Alemannic " Caus-i " the shifted form of Germanic Gaus-, O.E. *Geas < Ges, which we get in
- ' Gesecg " for Gesseg in the pedigree of the
Kings of Essex. In my note on ' Ancaster and Annhun Rex Grsecorum ' (< N. & Q.,' 12 S. vii. 227), I gave reasons for equating " Ane-" in the thirteenth- century form " Anecaster," with O.E. Andan, the owning case of Anda ; and I identified the person thus named with the grandson of Seaxneat, who is miscalled " Andsecg," son of " Gesecg," in the pedigree just now referred to. The scribal forms, as we know them, date from about A.D. 875. They are evidently corrupt and we must amend them to Andaeg, son of Gesaeg. These names, like O.E. " Bseld-seg," that of the ancestor of the Balthungs, present a Germanic ending of names of men (viz. -ag), in its O.E. form. Now, granted an O.E. And-aeg (with which cp. And-hun, And-raed, And- scoh), let us ask what form that name would assume in Brythonic writings of the eighth century and the ninth. In the Nomina Civitatum in the ' His- toria Brittonum ' (ed. T. Mommsen, 1894, p. 210) we find a town-name " Cair Guor- anegon " recorded. This Romano-British city has never yet been located correctly. Crude attempts have been made to identify the name with " Worcester." But the O.E. for Worcester was " Wiogoraceaster " and there cannot be any nominal connexion between that and Cairg uoranegon. " Wio- gora " is found in the ninth-century Bodleian MS., Hatton, No. 20, which pre- serves King Alfred the Great's translation of Pope Gregory's ' Cura Pastoralis.' It is the owning case of *Wigwaru, and the town- name means the Chester of the Wigfolk. King Alfred was writing in A.D. 890. Nen- nius's compilation was made in A.D. 837. The headword of GuorAnegon is the Old Welsh preposition guor, guar, which means " upon," " over," and is corrupted into gor. WTien set before names of men it has titular meaning and is equivalent to prceses, a protector or ruler ; cp. " praeses pro- vincise," the governor of a province. Nen- nius assigns this title to each of three Romano- British kings who, so he believed, were ruling in the fourth century, viz., GuorThegirn, GuorThemir and GuorAnogon. The last is made king of Cantium and is said to have been unaware that his kingdom had been granted to Hengist. " Anogon " is one form. Other variants in the ' Historia Brittonum ' (cap. 37 and 66a) are Anegon, Ancgon, Ancguon, Amgon. In every case these scribal variants are compounded with Guor, and they preserve the same name and the same title as that presented in the Nomina Civitatum, viz., Cair GuorAnegon the City of Anegon the Governor. The ending -on is an addition frequently made, and sometimes quite erroneously, by Nennius to O.E. personal names : cp. Alus-on, Titin-on, Gueay-on. " Aneg-" is an early Welsh form which points to a still earlier * Ant -eg. In the middle of a Welsh word nt lost its t by phonetic rule and be- came nn ; cp. Welsh dant, " a tooth," dannedd, "teeth," and see ' An Introduc- tion to Early Welsh,' by John Strachan (1909), p. 8, f. Consequently *Guor Ant eg would become *Guor Anneg, vocally, and scribally *Guor Aneg. I have previously explained that " Anto " is the Alemannic shifted form of Anda, and that the Brythons elected to equate " Anto " with Latin Anton-ius (Ant + on again), and to express it as Annhun or Annwn. This Anto, Anda, Annhun, the prince of the " Greeks " or Creacas of Lincolnshire, had passed away at a great age some fifty years before Hengist and Horsa arrived in the consulship of Felix and Taurus, i.e., in A.D. 427-428. Why then did Nennius synchron- ize him with GuorThegirn and Hengist ? A scribal error in the ' Historia Brittonum ' deceived Nennius, and was the cause of the anafhronism. In cap. xxxi. the Saxon ad- vent is ascribed to " Anno cccxlvii. post passionem Christi " (Mommsen's edition). The oldest form of this datum occurs in a twelfth-century Corpus MS., which yields " cccxlviiii. a passione Christi." This in- dicates A.D. 29 + 348 = A.D. 377, and that coincides approximately with the demise of Antis and the coming-in hither of the Alemannic king named Fraomarius by Am- mianus Marcellinus. It cannot, however, fit GuorThegirn and Hengist. Well, an Anglo-Saxon form of C was C, and C, F and L are easily confounded. So, if we