¹ To these accretions, I think, belongs also the quies sents Patricii at 457 This is of course the dating of the ‘Annales Cambriae’; however, since the date 458 (Kl iiii) in Chron. Scot, belongs to a set of three entries that are postdated by one year (as is proved by the Cottonian Annals with their corresponding entries under the same years as A.U.), the immediate source of this entry in A.U. would appear to have been their ‘Tigernach’exemplar, which in turn derived this detail from some (Irish) ancestor of our ‘Annales Cambriae’.

² In the fifth century section alone I note the following: Both Ann. Inisf. and Chron. Scot, give the Bedean A.M. for 429 as 4481 instead of 4381; the synchronism at 432 in Ann. Inisf. (a morte Conculainn herois 434 [anni], a morte Conchobuir meicc Nessa 413) is contradicted by the obits of these persons in the same annals (Cuchulainn d.1 A.D., Conchobur 20 A.D.), which do, however, agree with the years in the corresponding synchronism of Chron, Scot. (431 and 412).

³ A.U. is quoted from a photostat of T.C.D, MS H1 1.8, Chron. Scot, from a photostat of T.C.D, MS H1. 18, both in the National Library; Ann. Inisf. from the facsimile published by the Royal Irish Academy; the other annals from the printed editions in Revue Celtique. Unless otherwise stated, A.U entries are by the first hand.

⁴ O’Rahilly, , Early Irish history and mythology, p. 242.Google Scholar

⁵ Correct: 433, 435, 437, 438, 439, 440; wrong: 455.

⁶ Cath. Bull, xix (1939). 681–2.

⁷ A similar suggestion has been made recently by Grosjean, Père P, in Anal. Bolland., 63 (1945). 110, n.2.Google Scholar

⁸ The obit of Patrick in Ann. Inisf. may for the moment be disregarded. It is entered in an uncertain place round about 490, but with a passion date implying 461.

⁹ Here for once A.U. has a wrong epact : u instead of xii as correctly in Ann. Inisf.

¹⁰ See O’Rahilly, , The two Patricks, p. 64, f.Google Scholar; Early Irish history and mythology, p. 241 ; and here below, p. 260.

¹¹ Suprascript.

¹² The age of the moon is wrong; it should be xxuii.

¹³ These dates are entirely hypothetical ; they are calculated by counting the KI’s from the last one (impliedly) dated, viz. 442. The same applies to the following tables.

¹⁴ The eclipse was identified with that of 445 by Charles O’Conor in his edition of Ann. Inisf., in Rerum Hibernicarum Scriptores ii (1825). 2, n.3. There was certainly no solar eclipse in 444. However, Mr J. D. Schove, B.sc, has pointed out recently (in a paper read to the Irish Historical Society) that O’Conor’s identification cannot be correct. The text of the entry: eclypsis solis in nona hora clearly indicates that the eclipse recorded is that of December 23, 447 This eclipse could not have been observed in the north-west of Europe; the entry must be derived from a mediterranean source. The sole reference which I can find is in the Continuatio chronicoruin Hieronymianorum by the Spanish chronicler Hydatius (compiled ca 468 A.D.), who notes under 447: Solis facta defectio die X kal. lanuarias qui fuit tertia feria. The ninth hour is not mentioned in the original chronicle, but in the ‘Excerpta Montepessulana’: ipsorum (i.e. Valentiniani et Theodosii) quoque anno xxiiii, anno ab incarnatione domini cccclli et a mundi conditione [m^°], anno quinquies milaesimo et deli, olimp. cecuii, eclipsis solis facta X kal iunii, luna xu, quod fuit feria iii, ab hora iiii usque in sextam et ab hora uiii usque in nonam (M. G. H., Chronica minora ed. Th. Mommsen, vol. 11, p. 25 [1894]). The chronology of this note is very confused. The 24th year of Valentinian and Theodosius = 307th olympiad is 448 ; the year of the world (Jerome’s ‘years of Abraham’) 5651 is 452; May 23 was not a Tuesday in either 447, 448, or 452, thus iunii is a scribal blunder for ianuarii (December 23 was a Tuesday in 447 and again in 452); luna xu also does not apply to any of these years, no matter whether the 23rd of May or the 23rd of December is meant (it was the age of the moon on December 23, 451, which is the year 452 of the Dionysian paschal cycle). That the note in question was intended for 447 is suggested by its continuation: in anno quoque sequenti ab incarnatione domini ccccxl<u>iii ini non. iunii luna uiiii in Hi feria occiduo fuscatur tertia hora noctis—where it is implied that the preceding year was 447. This or a similar note probably underlies the Ann. Inisf., where, accordingly, we should read: eclypsis solis in nona<m> hora<m>. Since the shifting back over two years (447–445) has no parallel in the early section of these annals, the mistake must have been made already in their source. Although this eclipse is not recorded in any other body of Irish annals that I have examined, it may be safely assumed that the entry was present in the ‘Ulster Chronicle’, and probably with a fuller wording than what we read now; Ann. Inisf., in their usual way, give a mere extract of their exemplar.

¹⁵ To this set might belong also the entry referring to 455, which stands in the place of 454 (three Kl’s after ‘Council of Chalcedon’).

¹⁶ A.U. has both entries against 452, the defeat of Leinster alone (in a second hand) once more against 453; but the latter date is implied also in Chron. Scot, (feria iiii, to be corrected to u after the neglected leap-year) and in Cott. Ann. Perhaps the Ulster Chronicle gave as the alternative birthdate of Brigit 452, and placed the defeat of Leinster in 453.

¹⁷ In A.U. the order of these latter dates has been inverted:

Note that the disturbance starts again immediately after 450.

¹⁸ Conaill .m̄. cremthaine MS.

¹⁹ The feriáis are as follows:

N.B.‘Thrasamund’s accession to the Vandal kingdom of Africa was as late as 496; 492 and 497 (‘496’in the numbering of A.U.) have the same feria.

²⁰ The synchronism (A.M. 5660, A.D. 448) may be corrected as referring to A.D. 458, the (misplaced) obit of senex Patricius in Chron. Scot.—Against the Kalend next following stands the obit of Ibar (d. 501).

²¹ Here is a comparison of A.U and Ann. Inisf.:

²² From Macalister’s ‘Sources of the preface to the “Tigernach” Annals’(I.H.S., vol. iv, no. 13, pp. 38–57) the corollary at once occurs to the reader that the rule for adding 1 to the A.U. dates after 486 does not apply to the secondary entries. But, and this shows how warily we must walk here, if the error was already there in the ‘Ulster Chronicle’ there is a further sub-corollary, i.e., that the rule does apply to such secondary entries as were derived from the ‘Ulster Chronicle’.

²³ On this date see O’Rahilly, T. F., Early Irish history and mythology, pp. 505–6.Google Scholar

²⁴ ‘The ancient list of the coarbs of Patrick’ ed. Lawlor, H. J. and Best, R. I., in R.I.A Proc, xxxv, sect. C (1919), pp. 316–62.Google Scholar

²⁵ Ibid., p. 335 A.

²⁶ 491 as the obit of Patrick and 496 as that of Cormac in this list are further proof of the antiquity of the misdating in A.U.

²⁷ O’Rahilly, T. F, The two Patricks, p. 64 fGoogle Scholar. The a.u.c. date is wrong; correct mclxxxxiiii to mclxxxxuii.