¹ The article by E. Swoboda, Petillius No. 8, RE XIX.I (1937), columns 1138-1146, is the standard treatment. But the question of Tacitus's attitude to the man is not there examined. For the correct spelling of the nomen, with double L, see Münzer, F., RE XIX.I (1937), col. 1137.Google Scholar

² The full name appears only in CIL XVI 20.

³ See Münzer, F., art. Petillius No. 11, RE XIX.I (1937), columns 1150-1152Google Scholar; Broughton, T. R. S., The Magistrates of the Roman Republic (1951), i, p. 400Google Scholar.

⁴ Livy xli 18.11.

⁵ Tacitus, , Histories iii 79.1Google Scholar (incautum), iv 76.3 (temeritatis), iv 77.2 (temeritate), iv 78.2 (incuria), v 20.1 (neque satis cautum). It is certain that Tacitus was influenced by Livy in the Histories: see Syme, R., Tacitus (1958), p. 685 f.Google Scholar

⁶ Münzer, F., RE XIX.I, columns 1136 f.Google Scholar; Broughton, T. R. S., op. cit. (note 3), p. 369.Google Scholar

⁷ Münzer, F., art. Petillius Nos. 5, 6, 7, RE XIX.I (1937), column 1138.Google Scholar

⁸ Thus Townend, G., ‘Some Flavian connections’, JRS li (1961), pp. 54–61Google Scholar, at p. 59. Swoboda, E., art. Petillius Nos. 9-10, RE XIX.I (1937)Google Scholar, col. 1150, is inconsistent, suggesting that Petillius Rufus is the father of Gerialis, but that Q. Petillius Rufus is brother of Cerialis and grandson of Petillius Rufus.

⁹ Cf. Syme, R., Tacitus (1958), pp. 256Google Scholar ff. on the structure of the Annals. Note that in vi 4.1, the man is called Lucanius Latiaris; probably Latinius, earlier, is a slip, see Syme, R., op. cit., p. 277 n. 1Google Scholar and Ten Studies in Tacitus (1970), p. 70.Google Scholar

¹⁰ On the normal consular age, see Syme, R., Tacitus (1958), pp. 652 ff.Google Scholar

¹¹ That is, if he had been praetor at 30, the normal age (Syme, R., op. cit., p. 652 f.Google Scholar), before his legionary command, as was customary, although he could have held that post some years before the praetorship at this period, as pointed out by Fabia, P., ‘Le premier consulat de Petilius Cerialis’, Revue de Philologie xxxiv (1910), pp. 5–42, at p. 32.Google Scholar

¹² A good parallel for this style of nomenclature would be M. Macrinius Avitus Catonius Vindex, clearly a close kinsman of M. Macrinius Vindex: see H. G. Pflaum, Les carrières procuratoriennes équestres sous le Haut-Empire romain (i960), Nos. 161 and 168. I think it likely, also, that C. Quinctius Certus Poblicius Marcellus (cos. suff. 120) was a son of the delator Publicius Certus (although R. Hanslik, art. Publicius No. 36, RE XXIII.2 (1959), column 1903 f, assumes without question that he was a C. Poblicius Marcellus adopted by a Quinctius Certus): Syme, R., ‘People in Pliny’, JRS lviii (1968), pp. 135-51, at p. 150.Google Scholar

¹³ Over twenty qualify for inclusion in PIR ² ii, pp. 38 ff.

¹⁴ The best parallels from this period are the brothers Scribonii (E. Groag, art. Scribonius Nos. 25 and 28, RE IIA.I (1921) columns 888-891) and Domitii (PIR ² D 152 and 167).

¹⁵ The Umbrian Caesii are Sex. Caesius Propertianus (PIR ² G 204)—who was, ironically enough, in the service of Vitellius in 69; C. Caesius Sabinus (Ibid. 205), a friend of the poet Martial; and C. Caesius Aper (Ibid. 191), prefect of a cohort in Pannonia in 60, later made a senator, presumably by Vespasian. After completing this article I discovered that Sir Ronald Syme had postulated Sabine origin for the Petillii: see ‘Antonine Relatives: Ceionii and Vettuleni’, Athenaeum xxxv (1957), pp. 306–15, at pp. 313 ff.Google Scholar

¹⁶ For the date, see Syme, R., Tacitus (1958), p. 765 f.Google ScholarBurn, A. R., in Dorey, T. A. (ed.), Tacitus (1969), p. 60Google Scholar, n. 6, who argues for A.D. 61, seems, in my view, to overlook the main argument adduced by Syme, the impossibility of fitting all the events in Annals xiv 29-39 into one and the same year, 61.

¹⁷ Syme, R., op. cit. (note 16), p. 175Google Scholar, hints at this; and it has apparently been argued by H. Allgeier, Studien zur Kriegsdarstellung des Tacitus (unpublished dissertation, Heidelberg, 1957), p. 181, known to me only from Streng, M., Agricola: Das Vorbild römischer Statthalterschaft nach dem Urteil des Tacitus (1970), p. 122Google Scholar, n. 155 (who does not herself accept his arguments): in Allgeier's view, Tacitus intended ‘die “Darstellung vom egregius dux Cerialis” zu korrigieren’. It is certainly conceivable that Tacitus's attitude may have changed, but I find enough signs of disapproval already present in the Agricola (see p. 188, n. 44, below) to make this doubtful.

¹⁸ See Townend, G., JRS ii (1961), p. 58 f.Google Scholar The deification of Vespasian's daughter Domitilla is recorded by Statius, Silvae i, 97 f. and by inscriptions and coins which probably belong to the 80s A.D., cf. PIR ² F 417: for her own daughter (in Townend's view the child of her marriage with Cerialis) Ibid. 418.

¹⁹ One wonders if Tacitus knew that a Petillius had (apparently) once claimed some special connection with the Capitol, the mintmaster Petillius Capitolinus, cf. F. Münzer, art. Petillius No. 7, RE XIX.I (1937), column 1138.

²⁰ See generally Zancan, P., La crisi dell’ Impero nell’ anno 69Google Scholard.C. (1939); also Syme, R., Tacitus (1958) esp. p. 208Google Scholar. Numerous passages could be cited; cf. for example Histories i 56.1; i 60; ii 82.2; iii 49. 1–2.

²¹ On the rebellion see e.g. Syme, R., op. cit. (note 20), pp. 172 ff.Google Scholar; Cambridge Ancient History x (1934), pp. 842 ff.Google Scholar There is a problem over the status of Cerialis when appointed: had he already held the consul-ship? He certainly ought to have, before commanding an army or more than one legion (ultimately some eleven legions in fact). Fabia, P., Rev. de Philologie xxxiv (1910), pp. 5 ff.Google Scholar argued that he was adlectus inter consulares (for the precedents, from 29 B.C., see Cassius Dio Iii 42.4). See also Vulic, N., ‘Petilius Cerialis’, Klio vii (1907), pp. 457–8Google Scholar, who tries to extract some sense from the muddled and tendentious version in Josephus, , Bellum Judaicum vii 4.2Google Scholar (on which see below, p. 187, n. 41). Syme, R., ‘Consulates in absence’, JRS xlviii (1958), pp. 1–9Google Scholar, at p. 6, suggests that he may have been designated consul on his departure and have held office in absentia (not before April, for Vespasian and Titus were still consuls in March).

²² Many e.g. Briessmann, A., Tacitus und das flavische Geschichtsbild (1955), p. 102Google Scholar and Streng, M., Agricola (1970), pp. 53 ff.Google Scholar, take it as representing Tacitus's own views. Even if that were the case, it would give no clue to his attitude towards the man, cf. his use of the much hated Eprius Marcellus as a mouthpiece for his own political credo (Histories iv 8, cf. Syme, R., op. cit. (note 20), pp. 25 f., 509, 547)Google Scholar. But cf. Martin, R. H., ‘Tacitus and his predecessors’, in Dorey, T. A. (ed.), Tacitus (1969), pp. 117-47Google Scholar, at p. 130 who points out that ‘the Sallustian allusion at the outset warns us not to take at its face value Cerialis's claim to be speaking simple, unvarnished truth’. And note the comments of P. A. Brunt, ‘The revolt of Vindex and the fall of Nero’, Latomus xviii (1959), pp. 531-59Google Scholar, at p. 553 f. It is instructive to consider the footnote in Ramsay's, G. G. translation of The Histories (1915), p. 382Google Scholar: ‘Nowhere in the classics do we find the Roman claim to Empire… more powerfully, more simply and on the whole more justly put than in this speech… Much of the speech of Cerialis, with additions perhaps, rather than corrections, might be applied to our own Indian Empire’.

²³ It would take me too far afield to comment here on the arguments adduced by R. Urban, Historische Untersuchungen zum Domitianbild des Tacitus (1971), esp. pp. 76 ff. I hope to tackle this question shortly elsewhere.

²⁴ Cf. Syme, R., Tacitus (1958), pp. 175, 190 n.Google Scholar 7, 452, on the significance of the inclusion of the Claudia Sacrata story for an understanding of Tacitus's sources.

²⁵ Syme, R., op. cit. (note 24), p. 302 f.Google Scholar

²⁶ Seneca, , de beneficiis ii 21.5Google Scholar.

²⁷ It is hard to see how Dudley, D. R. and Webster, G., The Rebellion of Boudicca (1962), p. 97Google Scholar, can conclude, on the basis of Tacitus's account, that Cerialis's ‘military distinction, knowledge of the province, and his personal qualities made him an ideal choice as the first Flavian governor [of Britain]’.

²⁸ I hope to tackle the perennial problem of the chronology of Agricola's governorship shortly, with reinforced arguments in favour of the earlier dating.

²⁹ Degrassi, A., I Fasti consolari (1952), p. 25.Google Scholar

³⁰ PIR ² F 398.

³¹ Thus Eck, W., Senatoren von Vespasian bis Hadrian (1970), p. 60Google Scholar f. Dr. Eck kindly informs me that in his view AE 1952, 168 (Aries) may be restored to show that M. Pompeius Silvanus was cos. iii des.; and he believes that he died before taking office for A.D. 83 [see now W. Eck, ‘M. Pompeius Silvanus, consul designatus tertium—Ein Vertrauter Vespasians und Domitians’, Zeitschrift für Papyrologie und Epigraphik 9-3 (1972), PP- 259-76].

³² Cf. p. 180, n. 8 above. It is hard to believe that a younger brother of Cerialis could have been called Q. Petillius Rufus; they might perhaps have been cousins.

³³ See Syme, R., Tacitus (1958), pp. 652 ff.Google Scholar

³⁴ See previous note.

³⁵ CIL XVI 20.

³⁶ CIG II 3173, lines 22-3.

³⁷ Since L. Vitellius in A.D. 47: Degrassi, A., I Fasti consolari (1952), p. 13.Google Scholar

³⁸ See Monumenta Germaniae historica, Auctorum Antiquissimorum tom, ix, ed. Mommsen, T., Chronica minora saec. IV, V, VI, VIII, vol. i (1892), pp. 57Google Scholar, 222, 284, 417 ff., where Rufo or Tito Rufo appears, variously with or without iteration as colleague of Domitian. But the latter, actually holding his ninth consulship, is labelled variously as II, V, VI; and the Rufus of 83 is in most cases conflated with the other Rufi of the 70s, 80s and 90s, the coss. ord. 78, 83, 85, 88 and 97 being regarded as the same man holding a series of five consulships. Jerome-Eusebius (Migne, J.-P., Patrologia Latina t.xxviii (1866), p. 205Google Scholar) has Tito Rufo with no iteration. It is just possible that if Cerialis had indeed been adlectus inter consulares (see p. 183, n. 21 above) in 70, there might have been some doubt whether he should really be called cos. ii in 74 (although this is in fact recorded on CIL XVI 20), and thus if he were cos. again in 83, there could have been hesitation between II and III.

³⁹ See e.g. Eck, W., Senatoren (1970), p. 61Google Scholar, who opts for A.D. 83 for these men, on the grounds that as Q,. Petillius Rufus was cos. ii ord. and a kinsman of Domitian, there would have been no loss of prestige for Crispus or Veiento: but it would have been still more satisfactory if they were suffecti to the emperor and to a cos. iii ord. [Cf. now W. Eck Ztschr. f. Pap. u. Epigr. 9.3 (1972), p. 271, n. 40, who points out that Dr. J. Morris had already postulated that Cerialis was the cos. ord. 83 in his unpublished London dissertation (1955). Dr. Eck regards this as ‘recht wahrscheinlich’.]

⁴⁰ Most notably in Josephus, BJ vii 4.2; also Silius Italicus, Punica iii 607 ff.; Martial ii 2.4; Frontinus, Strat. iv 3.14; and cf. Fabia, P., Rev. de Philologie xxxiv (1910), pp. 21 ff., 26 ff.Google Scholar

⁴¹ Josephus, , loc. cit. (note 40)Google Scholar, states that Cerialis was given consular rank and sent to govern Britain, having previously governed Germany; and that on his way there he dealt with the German rebels; but, he goes on, it was really Domitian who deserved the credit, his mere arrival resulting in their instant submission. This confused account clearly reflects the official version current under the Flavians.

⁴² The work by R. Urban (cited above, p. 184, n. 23) is only the latest of many to deal with this.

⁴³ Dessau, H., Inscriptiones Latinae Selectae (1892-1916), 1025, 9200Google Scholar.

⁴⁴ Note the direct or implied criticism of Agricola's predecessors (unnamed) in 18.5-6, 19.1, 20.1 (incuria… priorum), 22.3, 22.4 (which contrasts with 8.3). Burn, A. R. in T. A. Dorey (ed.), Tacitus (1969), p. 50Google Scholar, attributes the growth of the abuses stamped out by Agricola (19.1–20.1) to the governorship of Frontinus: but the plural priorum does not accord with this, and incuria looks like a reference to Cerialis, in the light of what has been adduced earlier. Burn himself (p. 48) notes that 22.3, contrasting the way in which Agricola gave the enemy no rest with their previous ability ‘to make good their losses of the summer by successes in winter… casts a bleak light on some events of the winters between Cerialis’ main campaigns’.

⁴⁵ See, for example, Webster, G., ‘The military situations in Britain between A.D. 43 and 71’, Britannia i (1970), pp. 179–97CrossRefGoogle Scholar, at p. 191 f.; Wenham, L. P., ‘The beginnings of Roman York’, in Butler, R. M. (ed.), Soldier and Civilian in Roman Yorkshire (1971), pp. 45-53Google Scholar, esp. 48 ff.; B. R. Hartley, ‘Roman York and the northern military command to the third century A.D.’, Ibid., pp. 55–69, at p. 56 f.; Jones, G. D. B., ‘The Romans in the North-West’, Northern History iii (1968), pp. 1–20, at p. 2.CrossRefGoogle Scholar

⁴⁶ S. Frere, ‘Introduction’, in R. M. Butler (ed.), Soldier and Civilian in Roman Yorkshire (1971), pp. 15–19, at p. 16 f.; B. R. Hartley, Ibid., p. 57.

⁴⁷ See p. 179 and n. 4 5 above; further J. S. Wacher, Excavations at Brough-on-Humber 1958-1961 (1969), p. 3, who places Period I ‘before A.D. 70’, but finds no structural evidence to suggest Roman occupation before that date, and Period IIA c. A.D. 70, a temporary camp of unknown size; and on p. 5 cautions that it is impossible to distinguish between occupation under Bolanus and under Cerialis.

⁴⁸ Cf. Wheeler, R. E. M., The Stanwick Fortifications (1954), pp. 23Google Scholar ff. Wheeler was of course writing before the existence of the third-century civitas of the Carvetii (JRS lv (1965), p. 244) was known. Is it possible that the perennially puzzling passage in Pausanias, viii 43.4, stating that ‘the Brigantes had made an armed attack on the Genunian district whose inhabitants were Roman subjects’ and as a result were deprived of some of their territory, in the reign of Antoninus Pius (138-61) may be explained in terms of the Carvetii? Stevens, C. E., ‘Hadrian and Hadrian's Wall’, Latomus xiv (1955), p. 392Google Scholar and n. 3, suggested that the mysterious Γενουνιαν might be emended to Ούενουτ;ίαν—and, remembering that Celtic Gw- might easily be rendered as υ- in Latin, this is perfectly reasonable—making it the Venutian district. Stevens, following E. Birley, Roman Britain and the Roman Army (1953), pp. 31-47, places this district north of the western end of Hadrian's Wall, but Edenside and the Solway might equally be possible–and this is, of course, the Carvetian area, which, geographically, extends precisely to the vicinity of Stanwick.

⁴⁹ Wheeler, R. E. M., loc. cit. (note 48).Google Scholar

⁵⁰ Dobson, B., ‘Roman Durham’, Trans, Archit. & Archaeol. Soc. Durham & Northumberland, N.S., ii (1970), pp. 31–43, at p. 39 f.Google Scholar

⁵¹ Dobson, B., op. cit. (note 50), p. 40Google Scholar. Hartley, B. R., Northern History i (1966), p. 11Google Scholar, comments that ‘We may almost certainly accept Wheeler's thesis that the earthworks at Stanwick were then stormed and that they mark the last stand of Venutius, who, ironically enough, in view of Tacitus's estimate of his abilities, showed himself without a grasp of the realities of warfare. To put it at its lowest, he was extremely sanguine, if he believed that he could successfully defend a perimeter enclosing over 700 acres against the disciplined assault of Roman troops’.

⁵² Mr. C. E. Stevens points out to me, however, that the slighting of the defences of'Phase II’ (Wheeler, p. 18) could be the work of Bolanus, and that the phrase of Statius, cinxitque haec moenia fossa (Silvae v. 2.146) could refer to a siege of Stanwick by him. Bolanus would not have had sufficient forces to knock Venutius out permanently and ‘Phase III’ could therefore have been begun when he was obliged to withdraw. Certainly, as E. Birley pointed out twenty years ago (Roman Britain and the Roman Army (1953), p. 13) the language of Statius ‘would have been out of place if it bore no resemblance whatever to the facts’.

⁵³ Rey Cross, Crackenthorpe and Plumpton Head, CW² xxxiv (1934), 50 ff.Google Scholar; cf. Hartley, B. R., Northern History i (1966), p. 11, with references.Google Scholar

⁵⁴ Cf. for differing views Birley, E., Roman Britain and the Roman Army (1953), pp. 31 ff.Google Scholar; Ogilvie, R. M. and Richmond, I. A., vita Agricolae (1967), pp. 55, 206Google Scholar; Webster, G., The Roman Imperial Army (1969), p. 73Google Scholar; Frere, S., Britannia (1967), p. 100Google Scholar; Wenham, L. P., in R. M. Butler (ed.), Soldier andCivilian in Roman Yorkshire (1971), p. 50Google Scholar; B. R. Hartley, Ibid., p. 58, now tentatively assigns Carlisle to Frontinus.

⁵⁵ CIL XVI 20.

⁵⁶ PIR² J 322. Frontinus's consulship must have come in A.D. 73, cf. R. Syme's review of A. Degrassi, I Fasti consolari (1952) in JRS xliii (1953), p. 151Google Scholar, evidently overlooked by S. Frere, Britannia (1967), p. 101. I do not think the comments by Torelli, M., ‘The cursus honorum of M. Hirrius Fronto Neratius Pansa’, JRS lviii (1968), pp. 170–5Google Scholar, at p. 174, n. 22 alter th e situation: the consul…ON…of Inscr. Ital. xiii, 1, tab. ix, frag, VIII need not be Frontinus, even if not Fronto.

⁵⁷ Though if A. Dederich's emendation were accepted, Frontinus himself may refer to the campaign against the Silures in Strat. i 5.26.

⁵⁸ Syme, R., in Cambridge Ancient History xi (1936), p. 152.Google Scholar Recently some support has been forthcoming for this view, cf. B. R. Hartley, loc. cit. (note 54) and Jones, G. D. B., Northern History iii (1968), p. 6Google Scholar, although neither refers to Syme.

⁵⁹ Cf. note 28 above.

⁶⁰ Their previous contact must therefore have been during the term of office of one or more of Agricola's predecessors. Birley, E., Roman Britain and the Roman Army (1953), p. 15Google Scholar, takes ex aequo egerant (20.3) to refer to previous ‘dealings with Romans’. Ogilvie, R. M. and Richmond, I. A., vita Agricolae (1967), p. 219Google Scholar, translate ‘had lived on an equal footing with others (i.e. independent)’ which simply reproduces the note by H. Furneaux and J. G. C. Anderson in the edition of 1922, p. 100. But the Latin does not have a word which could be translated as ‘others’ and the context shows that it implies ‘with Rome’, not with other tribes.

⁶¹ Ogilvie and Richmond do not comment on multae civitates. R. Syme in Cambridge Ancient History xi (1936), p. 153, interprets it as ‘members of the Brigantian confederation’—yet in 17.1 Tacitus calls the Brigantes a civitas in the singular.

⁶² Birley, E., Roman Britain and the Roman Army (1953), p. 45 f.Google Scholar