(b) ix. 36, 1, ‘Silva erat Ciminia magis tum invia atque horrenda quam nuper fuere Germanici saltus.’ In this Niebuhr found an allusion to the campaigns of Drusus, B.C. 12-9, and accordingly assumed that the first decade was not published till B.C. 9. But the passage may equally well refer to earlier campaigns, e.g. of Julius Caesar. Nor can it be shown that the history of Dionysius of Halicarnassus, published B.C. 7, was used by Livy for Books viii.-x. Book ix. must have been written before B.C. 20, or Livy would have mentioned the recovery of the standards from the Parthians in ix. 18, 9.
(c) xxviii. 12, 12, ‘Hispania prima Romanis inita provinciarum, quae quidem continentis sint, postrema omnium nostra demum aetate ductu auspicioque Augusti Caesaris perdomita.’ This was written not earlier than B.C. 19, if it refers to Agrippa’s victory over the Cantabrians.
(d) Book lix. mentioned the lex de maritandis ordinibus, and consequently cannot have been earlier than B.C. 18.
(e) The books in which Pompeius figured were composed in the lifetime of Augustus (Tac. Ann. iv. 34, above).
(f) Book cxxi., according to the oldest MS. of the Periochae, was published after the death of Augustus; so doubtless were the remaining Books (A.D. 14-17).
A work of such compass, and occupying so many years of the author’s life, would naturally be published in sections. This a priori view is corroborated by several considerations: (a) There are separate prefaces to various sections (vi. 1; xxi. 1; xxxi. 1); (b) Livy’s style was censured[70] by Asinius Pollio, who died A.D. 5; (c) Augustus was acquainted with Livy’s sympathetic treatment of Pompeius (see above); (d) Livy had great fame in his lifetime: Pliny, Ep. ii. 3, 8, ‘Numquamne legisti Gaditanum quemdam T. Livi nomine gloriaque commotum ad visendum eum ab ultimo terrarum orbe venisse statimque ut viderat abisse?’
The historians from whom Livy derived his materials, and whom he himself mentions are: Fabius Pictor (i. 44, 2, etc.). Livy refers to him six times, but it may be questioned whether he used him at first-hand. More probably he took his opinions on the authority of later annalists like Macer, Antias, and Tubero. Cincius Alimentus (xxi. 38, 3): the Cincius quoted in vii. 3, 7, may be the same, or an antiquarian of the Ciceronian or Augustan age; Cato (xxxiv. 15, 9); Calpurnius Piso (xxv. 39, 15); Coelius Antipater (xxix. 25, 3); Claudius Quadrigarius (vi. 42, 5, etc.); Valerius Antias, quoted thirty-five times—far more frequently than any other authority; Licinius Macer; Aelius Tubero (iv. 23, 1); Clodius Licinus (xxix. 22, 10); Rutilius (xxxix. 52, 1); Polybius; Silenus (xxvi. 49, 3), a Greek, whose account of the Second Punic War was favourable to the Carthaginians.
A criticism of Livy’s use of these sources is impossible, except in the case of Polybius, all the others having perished. His tone in alluding to the Greek historian is remarkable for its coldness: xxx. 45, 5, ‘Polybius haudquaquam spernendus auctor’; cf. xxxiii. 10, 8. Although Polybius is not mentioned till Book xxx., he was undoubtedly used throughout the third decade, as well as in the fourth and fifth. Livy follows him very closely. Where Livy differs from Polybius he is probably following the account of Coelius Antipater, who is his leading authority for the Second Punic War.
Livy is not careful to reconcile his sources, and so frequently contradicts himself. His way of explaining a discrepancy between his authorities is by striking an average (xxvi. 49, 6, ‘si aliquis adsentiri necesse est, media simillima veris sunt’). His irresolution was noted by Quintilian, ii. 4, 19, ‘saepe quaeri solet de tempore, de loco, quo gesta res dicitur, nonnumquam de persona quoque, sicut Livius frequentissime dubitat.’ This of course has its good side: it saves him from dogmatizing on uncertain points, and he has a hearty appreciation of the confusion in his authorities: xxxvii. 34, 5, ‘is ubi et quando et quo casu captus sit, sicut pleraque alia, parum inter auctores constat.’ He recognizes the value of contemporary evidence: xxii. 7, 4, ‘Fabium aequalem temporibus huiusce belli potissimum auctorem habui’; xxi. 38, 3, ‘L. Cincius Alimentus, qui captum se ab Hannibale scribit, maxime auctor moveret.’ Criticism of his authorities is most conspicuous in the case of Valerius Antias, whom at first he followed in good faith; he condemns him again and again for exaggeration and credulity, e.g. xxxiii. 10, 8, ‘si Valerio qui credat, omnium rerum immodice numerum augenti’; xxxix. 43, 1, ‘Valerius Antias, ut qui nec orationem Catonis legisset et fabulae tantum sine auctore editae credidisset.’ He also recognizes the bias of Licinius Macer: vii. 9, 5, ‘quaesita ea propriae familiae laus leviorem auctorem Licinium facit.’ For the untrustworthiness of family records, cf. viii. 40, 4, ‘vitiatam memoriam funebribus laudibus reor falsisque imaginum titulis, dum familiae ad se quaeque famam rerum gestarum honorumque fallenti mendacio trahunt.’
Livy often refers to authorities whom he does not name: ‘invenio apud quosdam,’ ‘satis constat’; and to tradition: ‘fama est,’ ‘dicitur,’ ‘fertur,’ ‘traditur.’ Tradition was the sole source for events prior to the sack of Rome by the Gauls, cf. vi. 12, 2 sqq.