A book on rhetoric was known to Quintilian and Seneca the elder, apparently in the form of a letter addressed to the author’s son (Quint. x. 1, 39, above).
Quint. ii. 5, 20, ‘quemadmodum Livius praecipit’ (on models of style); Sen. Contr. ix. 2, 26, ‘Livius de oratoribus ... aiebat’ (on obscurity of expression); Sen. Contr. ix. 1, 14, ‘T. Livius tam iniquus Sallustio fuit ut hanc ipsam sententiam ... obiceret Sallustio.’
These minor works have perished, and of his great history only a portion survives.
Its title, according to the oldest MSS., the summaries of the lost Books, and the grammarians, was Ab urbe condita libri; and this is corroborated by Livy’s own language: i. praef. 1, ‘si a primordio urbis res populi Romani perscripserim’; and by Pliny, N.H. praef. 16, ‘T. Livium ... in historiarum suarum, quas repetit ab origine urbis, quodam volumine.’ Livy refers to it loosely as meos annales (xliii. 13, 2). Separate parts may have had special titles: thus Books cix-cxvi. were known as Civilis belli libri viii. (Codex Nazarenus of the Periochae).
The number of Books now extant is thirty-five, viz., i.-x., which carry the history down to B.C. 293, and xxi.-xlv., covering the period B.C. 218-167. Of these xli. and xliii. are incomplete. But we possess summaries (Periochae or Argumenta) of Books i.-cxlii., except cxxxvi. and cxxxvii., which show that the narrative was continued to the death of Drusus in B.C. 9. There is no evidence that it actually went further; but as the death of Drusus is hardly an event of sufficient importance to form the conclusion of so great a work, it has been thought that Livy may have intended to finish with the death of Augustus—the point from which Tacitus starts. The total number of Books would then have been probably one hundred and fifty.
The division into Books (libri or volumina) is due to the author: vi. 1, 1, ‘quae ab condita urbe Romani gessere quinque libris exposui.’ The division into decades (i.e. sets of ten Books) is first mentioned towards the end of the fifth century; it is merely a conventional arrangement, the subject-matter falling naturally into sets of fifteen Books, which again sometimes embrace three sub-divisions each a half-decade, or two, a half-decade and a decade.
An epitome was known to Martial, xiv. 190,
‘Pellibus exiguis artatur Livius ingens,
quem mea non totum bibliotheca capit.’
The evidence of the date of composition is as follows:
(a) i. 19, 3, ‘Bis deinde post Numae regnum [Ianus] clausus fuit, semel T. Manlio consule post Punicum primum perfectum bellum, iterum, quod nostrae aetati dei dederunt ut videremus, post bellum Actiacum ab imperatore Caesare Augusto pace terra marique parta.’ Now, as the first closing of the temple of Janus by Augustus was in B.C. 29, and as Livy is silent as to the second closing after the Cantabrian war in 25, it follows that this passage was written B.C. 29-25. The use of the title Augustus, conferred on Octavian in 27, puts the earliest possible date two years later. The history therefore was not begun before B.C. 27.