^{page 123 note 1} ProfessorBrink, G. O. reminds me that Sörbom, Variatio Sermonis Tacitei, p. 4, among examples of variatio in the use of proper names, quotes Ann. 3. 56 Tiberium Neronem—Neronem—Tiberius. Similarly Ann. 1. 4–5 have Tiberium Neronem—Tiberius—Neronem. Since in the former case there is nothing more than a literary variation, itGoogle Scholarcannot be assumed without argument that Neronem in 1. 5 is anything more. But in 3. 56 the reference is to the tribunician power first conferred on Tiberius in 6 B.C.; there is thus no anachronism in the use of Neronem there. Similarly in 1. 3 Nero, used of Tiberius immediately prior to his adoption by Augustus, is legitimate. Neronem in 1. 5 is different in being an anachronism (an unparalleled one, I think), and it is accordingly reasonable to believe that its employment in that passage is deliberate.

^{page 123 note 2} Livia, and Tanaquil, : C.R. xli (1927), P. 55 a.Google Scholar

^{page 123 note 3} The letters (a), (c), (d) should be ignored for the present. They do not mark Tacitean parallels but represent correspondence between the versions of Tacitus and one or both of Suetonius and Dio Cassius; their significance is discussed later.

^{page 124 note 1} Tiberius, and the death of Augustus: A. J. Ph. xliv (1923), 145 f.Google Scholar

^{page 124 note 2} Dio also, it will be seen, has the same basic story; it follows that the factual parallelism is not the creation of Tacitus, but goes back to the source that both he and Dio used.

^{page 124 note 3} There are some differences of detail in their account of the time and manner in which the poison was given to Claudius; for their significance cf. Momigliano, A., Rendiconti. R. Accad. Naz. d. Lincei, Serie VI, vol. viii (1932), 293 f. But since Tacitus gives no details of the alleged poisoning of Augustus, this portion of the story does not concern us here.Google Scholar

^{page 125 note 1} Since it is quite unlikely that the correspondence already existed in Tacitus' source(s) for the two reigns, there are two alternatives: (i) that the apparent correspondence is illusory, (ii) that the correspondence is of Tacitus' making. Exact parallelism between the two occasions is precluded by the fact that, whereas Tiberius' accession was effective from the time of the announcement from the house at Nola, the significant moment in Nero's case was his proclamation as imperator in the praetorian camp (Ann. 12. 69; cf. Ramsay's translation (vol. ii, p. 115, footnote 3) for the increasing importance of the military in the appointment of a new emperor). In view of the difference in the circumstances, it may be argued that the degree of verbal correspondence is too great to be accidental. If Neronem in 1. 5 is deliberate (cf. p. 123, n. 1), an intentional correspondence between the two passages is perhaps made more likely. However, even if a resemblance between themes (h) and (i) in Ann. 1 and 12 is denied, it does not invalidate the other conclusions suggested in this paper.

^{page 127 note 1} Whereas (h–i), which, it has been suggested (vide § 2 supra), have a different origin, are continuous with (f) and (g) in Ann. I, but separated from them in Ann. 12.

^{page 127 note 2} The parallel is already noted in Aurelius Victor, de Caesaribus 4. 15: ‘ceterum funus [sc. Claudii], uti quondam in Prisco Tarquinio, diu occultatum, dum arte mulieris corrupti custodes aegrum simulant atque abeo mandatam interim priuigno, quern paulo ante in liberos asciuerat, curam reipublicae.’

^{page 127 note 3} The phrase is taken from Walker, B., The Annals of Tacitus (Manchester, 1952); see chapter iv, especially pp. 33–34, and chapter viii, p. 158.Google Scholar

^{page 127 note 4} Pippidi, D. M., Autour de Tibère, Bucharest, 1944Google Scholar (see the important review by Balsdon, J. P. V. D. in J.R.S. xxxvi (1946), 168–73) tne three devices are (i) general psychological affirmations—this does not concern us here, (ii) ‘glosses’ (gloses, eclaircissements), (iii) ‘rumores’.Google Scholar