^{page 182 note 2} Since the lamented death in 1927 of (Professor Flamstead Walters, who had been my partner 1914), in editing the text of Livy since 1901, I have enjoyed the co-operation of Professor Stephen K. Johnson, now of Newcastle-on-Tyne, formerly my colleague at Manchester, in the long task. To him I am greatly indebted for indispensable criticism of the first draft of this paper, as well as for his own addition to it.

^{page 182 note 3} Titi Liui Ab Urbe Condita Libri a Vicesimo Sexto ad Tricesimum. Recensuit A. Luchs, Berlin, 1879.

^{page 182 note 4} See e.g. Walters, C. F., C.R. xviii (1904), p. 392Google Scholar, and the Praefatio to Livy, , Vol. i (Oxford, 1914), § 29Google Scholar.

^{page 182 note 5} They are so few and far between, and so slight in character (e.g. xxii. 39. 1; eo sed N, esset N ⁴), that they seemed best unmentioned in our Notes to xxi–xxv lest it should be supposed that we attributed to them any connexion with Spirensis.

^{page 184 note 1} It is rightly numbered 161 at the bottom; but erroneously numbered 162 at the top, a mistake perhaps made by some scribe who was familiar with the practice of numbering pages rather than folia, and who, since he found the last folio numbered 160 on its recto, concluded that the number to put on the recto on his added folio was 162.

^{page 187 note 1} One other person who occasionally appears, but too rarely to be worth a separate name, uses just the same ink and a hand which Professor Rostagno attributes to the same period. His distinguishing feature is that he puts little curved strokes underneath letters to be deleted; for instance, on Fol. 101 v. 1. 6 from below (xxvii. 11. 15), a passage which shows that he is intermediate between the first hand and the Spirensian corrector, who has improved upon his deletion, The inaccuracy involved in calling him N ² seems a less evil than to burden the list with another number for a man who so rarely intervenes, His only merit is that like the other N ² he draws our attention to what he did not understand.

^{page 188 note 1} Almost certainly, but it is just conceivable that N ³ and N ⁿ drew from a Spirensian source other than that used by N ⁴; see Professor Johnson's remarks below, p. 198. In that case N ⁿ might have been later than N ⁴.

^{page 188 note 2} The term N ^x we reserve to denote a corrector of whom we feel completely uncertain; and therefore it may cover any number oi persons. It must not, therefore, be used for a corrector whom we can distinguish from every other and whose place we can nearly (if not quite) fix in the chronological succession. For this such a name as N ⁿ seems, by algebraic analogies, to be better fitted.

^{page 189 note 1} See Plate C.

^{page 189 note 2} See Plate B.

^{page 190 note 1} See Plate A.

^{page 190 note 2} I hope I have been consistent in this; but it is possible that here and there in the earlier part of ten years' work I may have omitted the qualifying phrase.

^{page 190 note 3} See Plate C.

^{page 194 note 1} Plate B.

^{page 194 note 2} Cf. Ter, . Andr. 921Google Scholar; Cic. Verr. Act. II. i, § 116 i; Off. 3. 19. 76; Att. 1. 12. 1; Liv. xxxv, 13. 5.

^{page 194 note 3} It is written by an abbreviation which Luchs read as alid, but which Rostagno confirmed me in interpreting as aliunde.