^{page lxvii note 1} Bull of Alexander III (1160).

^{page lxvii note 2} C ³ inserts Littere quas after no. 32, and repeats Ad gerendam, which was given before as Ad regendam.

^{page lxviii note 1} Omit, C ² and C ³.

^{page lxviii note 2} Omit, C ¹ and C ².

^{page lxviii note 3} Omit and insert after no. 56, D.

^{page lxviii note 4} Letter of Henry, archbishop of Reims (1174).

^{page lxviii note 5} Omit, H ¹.

^{page lxviii note 6} Glossed vacat in A.

^{page lxviii note 7} Letter repeated at no. 51 and after no. 59.

^{page lxviii note 8} Omit, H ¹ and H ².

^{page lxviii note 9} Transposed from no. 45.

^{page lxix note 1} Letter repeated at no. 51 and after no. 59.

^{page lxix note 2} Omit, H ¹.

^{page lxix note 3} Letter of Gilbert Foliot.

^{page lxix note 4} Bull of Urban III.

^{page lxix note 5} A gathering is lost, but the index supplies the information.

^{page lxix note 6} An addition to Personam domini Lond.

^{page lxix note 7} F ¹ gives the verse before Magnum michi, and the sermons at the end of the corpus; E, E ^l and F ² give the verse and the sermons after no. 78; H ² omits the sermons.

^{page lxix note 8} Omit, H ¹.

^{page lxix note 9} Miscellaneous document.

^{page lxx note 1} Bull of Alexander III.

^{page lxx note 2} Blank reverse to the letter, and folio excised.

^{page lxx note 3} Letter of Giles, bishop of Évreux.

^{page lxx note 4} An anonymous excuse.

^{page lxxi note 1} Except the unique selection of letters in MS. Regin., lat. 189 (B ⁵).

^{page lxxi note 2} Except MS. Digby 209 (E ¹), which is a copy of the earlier MS. Auct. F.1.8 (E).

^{page lxxi note 3} See the plan of the arrangement of the MSS.

^{page lxxi note 4} 45 if a bull of Pope Alexander III (Litteras a tua), which occurs in all the extant MSS., is counted.

^{page lxxi note 5} MSS. Corpus Christi College, Cambridge, 273 (C ²) and Vat., lat. 6024 (C ³) collect the sermons at the end of the forty-four letters.

^{page lxxi note 6} See below, p. 2.

^{page lxxi note 7} Marcel Bouteron, ‘Arnoul, Évêque de Lisieux, Études’, (École nat. des Charles, Positions des thèses, 1905), p. 2.Google Scholar

^{page lxxi note 8} See above, p. lxi.

^{page lxxi note 9} Cf. the fate of MS. Vat., lat. 6024; see below, pp. lxxxii, seq.

^{page lxxii note 1} At the end of the MS., on fo. 59V, in a new hand, is written, “Liber iste est Sancti Andree Vercell”, etc.

^{page lxxii note 2} Below, pp. lxxiv, seqq.

^{page lxxii note 3} See below, p. 45, n. 6.

^{page lxxii note 4} Keener eyes have read, ‘Sancte Marie de V…’, which has been interpreted as the Cistercian monastery of Ourscamp in the diocese of Noyon. This is certainly a likely attribution, for Arnulf was intimately connected with its daughter house of Mortemer (see above, p. xvi, n. 5).

^{page lxxiii note 1} They are independent MSS., but share some peculiarities.

^{page lxxiii note 2} For detailed descriptions of the volume, see Neues Archiv, (1878) iii, 150–1Google Scholar; Poupardin, ‘Dix-huit lettres inédites’ in Bibl. de l'École des Chartes, lxiii (1902)Google Scholar; and Brooke, Z. N., ‘The Register of Mr. David’ in Essays in history presented to R. Lane Poole (Oxford, 1927).Google Scholar

^{page lxxiii note 3} Pp. lxxxii, seq.

^{page lxxiv note 1} See Delisle, L. V., Le cabinet des manuscrits (Paris, 1868–1881), ii, 228.Google Scholar

^{page lxxiv note 2} In A this increment occupies a distinct portion in the MS. The poems end with half a page left blank. Then these letters begin on a new gathering in a fresh hand, and have no titles.

^{page lxxvi note 1} Unless we postulate this order of copying, it is difficult to account for the truncation of the letter Causam que inter Hermerium at the bottom of a page within a gathering in D.

^{page lxxvi note 2} Notice the characteristic example in diagrams 3 and 4—ierusolimam.

^{page lxxvi note 3} The spelling neglegens is very common in A. D usually repeats it, but it is frequently emended by expunction. The scribe of D usually turns o in A into cum, e.g. cummendo, cummodius; this too, has been almost invariably altered later. A peculiarity of D's scribe is his weakness for the double ss, e.g. gauissus, asscribi.

^{page lxxvi note 4} See below, p. lxxxii.

^{page lxxvii note 1} Above, p. lxxvi, and n. 1.

^{page lxxviii note 1} This letter, Quam utilis, ep. no. 29, provides a problem. Written in 1160 (see above, p. xl), it contains a long description of the disputed election of Alexander III to the papacy. The ist Edition MSS. address it to the bishops Gilbert of London, Hilary of Chichester and William of Norwich, and, when cardinals are mentioned, they are referred to in the 3rd person. The unique Vatican codex (C ³) has the same title, but refers to the cardinals in the 2nd person. D ¹ had the 3rd person readings, but they are altered to the 2nd person in D ², and the letter is addressed to the Roman cardinals. This is characteristic of the 2nd Edition. From the sense there can be no doubt that the letter was originally sent to the cardinals. If it were a mere confusion of title there would be no problem. But rubrication followed engrossing, and it is in any case unlikely that the scribe would change the readings to agree with the title. We must suppose then that there were two forms of the letter, and, although primarily written to the cardinals, it was also used by Arnulf to acquaint the English church with the events at Rome. The ist Edition took the ad hoc version because it was the more accessible, and the original form was restored in C3 because of the special nature of that codex; but C ³ received the 1st Edition address because the rubricator was following a normal copy. We learn otherwise that this rubricator was working mechanically, for he did not realize that there was a disorder in the MS. (see above, p. lxvii, n. 2), and gave the title of Littere quas to Sustulit uirum. This error, however, was corrected.

^{page lxxix note 1} See below, p. 130, note 8. The second possessionis is in the next line practically under the first in D, making such a slip easy.

^{page lxxxii note 1} For instance, he did not realize that a gathering was missing in the MS., and he printed without hiatus or correction. This has been rectified in some reprints. He failed entirely to see the significance of the excision of 4 words from the end of the Pro Simone letter (see above, p. lxxv), and, although he took the marginal correction there, he did not notice the vacat when the words occurred again at the end of Causam que inter Hermerium, although the cat of the cancellation made him change the last word from meretur to mereat. The reprinters, and Giles in the latest edition, reproduce this confusion uncorrected.

^{page lxxxii note 2} Above, p. lxxiii.

^{page lxxxii note 3} Ep. no. 82.

^{page lxxxii note 4} There is a blank verso to fo. 64, and the later letters include one of Giles, bishop of Evreux.

^{page lxxxii note 5} The first part of the codex is not without its peculiarities; see above, p. lxxviii, n. 1. This suggests that Arnulf sent the whole collection, making this one correction in the first part.

^{page lxxxii note 6} Thirty-four of these letters have an address instead of a title; ep. no. 54 is divided into two separate letters, and there are increments to epp. nos. 63, 76, 79, 89 and 91. Also D's address to ep. no. 77 is corrected.

^{page lxxxiii note 1} Pp. lxxxv, seq.

^{page lxxxiii note 2} The addition of the g pieces and of the increment occurred at the same time. There is no break of any kind between them. There is, however, a decisive break before the 9 pieces, for the letter Magnam michi begins on a new gathering in a fresh hand.

^{page lxxxiii note 3} See above, p. lxxi.

^{page lxxxv note 1} Ascription on fo. 151^v. For the priory, see V.C.H. Hampshire, ii, 164Google Scholar, seqq.

^{page lxxxv note 2} The rubricated introduction to the florilegium (fo. 79) is as follows: ‘Incipit conpendiosa defloratio ex libro epistolarum magistri lohannis Sares-beriensis, qui postmodum fuit episcopus Carnotensis, super causa beati Thome martyris a Guidone priore Suwicense diligenter excerpta. Que studiose dictare uolentibus turn pro modo scribendi, turn pro sententiarum elegantia, turn pro causa prescripti martyris declaranda non erit inutilis. Predictus autem liber epistolarum a prescripto magistro lohanne post prefati martyris passionem in unum corpus diligenter et breuiter est collectus, et in iiij^or uolumina luculenter ac studiose diuisus. Porro in eiusdem libri collectione et diuisione ac emendatione iam dictus prior, tune canonicus Mereton, memo-rato magistro lohanni comes indiuiduus ac familiaris adhesit’. It must be remembered that John's brother, Richard, became a canon at Merton (Webb, , John of Salisbury, p. 2).Google Scholar

^{page lxxxvi note 1} Inserted in the text of F ¹, before the verse (see plan on p. lxix above), are two letters touching the bishop of London. These would be more readily obtained at Merton than at Southwick.

^{page lxxxvi note 2} Becket had received schooling at Merton, and the importance of the priory in this period is shown by several references in the Becket correspondence.

^{page lxxxvi note 3} For a description of these MSS., see Robertson's introduction to vol. v of the Materials.

^{page lxxxvi note 4} Ep. no. 72.