^{page 312 note 1} F. B. Marsh, The Reign of Tiberius, 14.

^{page 312 note 2} R. M. Ogilvie, Commentary on Livy, 463.

^{page 312 note 3} The Romans and their Gods, i8.

^{page 312 note 4} Tacitus: Greece and Rome, New Surveys, no. 4 (1970), 33. See now the same author's Annals of Tacitus (I-VI), i. 37–40.Google Scholar

^{page 312 note 5} Thuc. 3. 64. 4.

^{page 312 note 6} This Latin terminology is discussed by Hoffmann, H. in Gymnasium, lxxv (1968), 240 fr.Google Scholar

^{page 312 note 7} Thuc. 1. 86. 2.

^{page 313 note 1} Lysias 14. 23.

^{page 313 note 2} Pind. Pyth. 2. 72.

^{page 313 note 3} Ad Her. 2. 5; Cic. inv. 2. 32–4 (for more on the Greek background see Classen, C. J., Rh. Mus. cviii [1965], 122).Google Scholar

^{page 313 note 4} Cic. Cluent. 70.

^{page 313 note 5} Ad Her., loc. cit. Or ‘negare oportebit de vita eius et de moribus quaeri, sed de eo crimine, quo de arguatur; quare, ante factis omissis, illud, quod instet, id agi oportere’ (inv. 2. 27).

^{page 313 note 6} Or ‘ilium ante occultasse sua flagitia; se planum facturum ab eo maleficium non abesse’ (auct. ad Her., loc. cit.).

^{page 313 note 7} Inv. 2. 34. Compare Quint. 7. 2. 33 ‘dicens neminem non aliquando coepisse peccare’.

^{page 314 note 1} Cie. Sulk: 6g.

^{page 314 note 2} Isoc. 7. 38.

^{page 314 note 3} Ann. 3. 65.

^{page 314 note 4} As to this, we may note that even the would-be orator is bidden to avoid obviously unreliable stories about the accused: ‘satins est omni se ante actae vitae abstinere convicio quam levibus aut frivolis aut manifesto falsis reum incessere, quia fidesceteris detrahitur’ (Quint. 7. 2. 24)—much more, presumably, the historian who hoped to be believed.

^{page 314 note 5} Ann. 6. 48, where the mere report of Arruntius' opinion does not, surely, involve Tacitus in any inconsistency (contra R. Syme, Tacitus, 422).

^{page 314 note 6} Koestermann, E., Cornelius Tacitus (1- III), i. 38.Google Scholar By contrast Goodyear, F. R. D., Annals of Tacitus (I-VI), i. 40Google Scholar, doubts whether Tacitus even saw the contradiction between his own analysis and that of Arruntius. He recognizes, however, from passages such as Ann. 6. 48 and Hist. 1. 50. 4 that Tacitus was ‘not entirely unacquainted with the idea of change in character’ but is among a few classical writers who have ‘an occasional inkling of change and development of character’ (ibid. 38). But if our author, with his admitted ‘skill in indirect characterization’, makes little attempt to present the process of change, this should be taken as an indication of his choice not to do so rather than to his limited powers of perception—because the exploitation of this skill was more appropriate to biography rather than to history? See Momigliano, A., The development of Greek biography, 1971, 40,Google Scholar 99–100: we may compare Plutarch's readiness to allow a change in the character of Pericles which most of his predecessors had seen as ‘steadily and sometimes rather unpleasantly democratic’ (Connor, W. R., Theopompus and Fifth Century Athens, 114, adding, however, that even in this biography the change is really a reversion to Pericles' naturally aristocratic disposition).Google Scholar

^{page 315 note 1} R. SyMC, op. cit. 421.

^{page 315 note 2} Op. 61. 419.

^{page 315 note 3} 5. To. 28.

^{page 315 note 4} As Hoffmann, loc. cit. 231 n. 54 and 246, allows.

^{page 316 note 1} As in Plato, Symp. 205 E.

^{page 316 note 2} See Plato, Rep. 476 ff.; Robinson, R., Plato's earlier Dialectic, 194;Google Scholar Cross, R. C. and Woozley, A. D., Plato's Republic, 166.Google Scholar

^{page 316 note 3} Long, A. A., Problems in Stoicism, 185 and 184; note also, 175 ‘freedom to act out of character is a concept generally denied or ignored in Greek philosophy’ (my italics).Google Scholar

^{page 316 note 4} Hoffmann favours F. Klingner's theory that Tacitus transfers to his analysis of Tiberius the idea of the harmful effect of the removal of metus hostilis, an idea applied by earlier writers to the analysis of the decline of the state or to a whole class within the state. If Tacitus is consciously doing this, then he is adapting a theory which goes back long before the Stoics, though it is sometimes ascribed to Posidonius (Earl, D. C., The Political Thought of Sallust, 51).Google Scholar It may be added that other historians, by whom Tacitus is likely to have been influenced not less than by philosophers, used the idea to explain change, rather than to explain it away, and indeed to suggest that it is inevitable (see Walbank, F. W., A Historical Commentary on Polybius, i. 697, 744–5).Google Scholar

^{page 316 note 5} Quint. 9. 2. 66; see Cousin, J., R.E.A. xxix (1951), 244.Google Scholar

^{page 317 note 1} See R. Syme, op. cit. 420.

^{page 317 note 2} It also seems distinct from most of the cases of inconsistency of behaviour discussed by Dutoit, E., Mus. Helm ii (1945), 39, which involve deliberate pretence or constant vacillation, the latter quite out of place in any Roman nobilis, trained or not in Greek philosophy. Syme's verdict allows for a greater element of artificiality, though it is not specifically related to the rhetorical exigencies of the law-courts, to the indirect influence of which I would now attribute more than I did (J.R.S. xlix [1959], 56) in the case of Sallust's allegations of dissimulatio.Google Scholar

^{page 317 note 3} I wish to thank Professor A. Momigliano for stimulating discussion in advance of the writing of this article, which he also made time to read on the eve of departure for Italy; but, since detailed comment was then impossible, all heresy is my own.