¹ The Celtic Church in Britain and Ireland (translated by Meyer, A.), p. 34Google Scholar ; Juv. II. 159, XV. 112.

² Eng. Hist. Rev., 1913, 1 sqq.; cf. P.-W. VIII. 1389 s.v. Hibernia, where he is less emphatic.

³ E.g. IV. 17, XIV. 240, XV. 13 sqq., and especially X. 174 sqq.

⁴ No one else condescends to particulars even formulates a charge. Duff, Mr. (Juv. p. xxxiii)Google Scholar says that ‘J.'s idea of historical perspective is very peculiar’ but that is irrelevant here, and in any case his chief example, that Gallicus (XIII. 157), may rather be the generic use of a name.

⁵ Dessau, in Hermes 46, p. 156Google Scholar ; E.E. IX., p. 560 Haverfield.

⁶ Statius, , Silv. IV. 4. 62Google Scholar, ‘nigrae litora Thyles,’ within the Roman sphere; V. 1, 91; V. 2, 55; Silius III. 597, XVII. 416.

⁷ DrMacdonald, in J.R.S. IX., pp. 136–8.Google Scholar Juv. XV. was written perhaps in 127; Ptolemy's account of Britain is mainly pre-Hadrianic.

⁸ Haverfield, , I.c., p. 6Google Scholar and n. 8, citing Mr. Curie.

⁹ If ‘rhetor’ = professor of rhetoric, as it usually does, and as the context rather suggests, the hyperbole is perhaps rather more evident.

¹⁰ Juv. XIV. 275.80. ‘Plus hominum … is a clearer instance of hyperbole than our line.

¹¹ , Tac.Agr. 10, 38Google Scholar; Valmaggi, , Circumnavigazione della Britannia? in Atti, Torino, 53 (1918), PP. 579–586Google Scholar.

¹² C.I.L. X. 5382 = Dessau 2926.

¹³ Cichorius, s.v. Cohors in P.-W. IV. 280–3.

¹⁴ Domaszewski, , Rangordnung d. röm. Heeres, pp. 122, 129Google Scholar : ‘Wo in dieser Zeit (Claudius to Trajan) nur die praefectura cohortis bekleidet wird, liegt bloss der Anfang der Carriere vor.’ But Juvenal among others of the twenty relevant instances cited, ib. n. 8, probably held no further command. A single command, followed or preceded by municipal office, was not unusual. (C.I.L. IX. 5841, X. 4619.) The ‘tribunatus semestris’ often held by equites at this time as their first office (Juv. VII. 88–9 with Mayor's n.; , Plin.Ep. III. 8Google Scholar, IV. 4; Domaszewski, I.c.; Mommsen, , Staatsr. Fr. tr. V., p. 124Google Scholar, n. 1, VI. 2, p. 156, n. 2; Hirschfeld, , K.V. ², p. 422Google Scholar, n. 1) was legionary, and largely formal, not serving as a stepping-stone to higher office; neither it nor the centurionate nor the ‘praefectura fabrum’ need be assumed as a preliminary in Juvenal's case. If Martial became an ‘eques’ through this tribunate (III. 95, 9) we need have no doubt of Juvenal's rank.

¹⁵ Next to Egypt, Juvenal seems best acquainted with Britain (II. 159, 160; see also IV. 126, 7, XIV. 147, VIII. 51, 2; XIV. 196 is neutral).

¹⁶ MrDuff, (Juv., p. xx)Google Scholar is among the dissentients; he concludes, with hesitation, that ‘the local magnate of Aquinum was not the poet himself, but a kinsman by blood or adoption.’ Schanz, , Gesch. d. röm. Litteratur II.³418Google Scholar, takes a similar view; see Addenda.

¹⁷ Appendix 58 to Sittengeschichte, Vol. IV., pp. 310 sqq. of the English translation; Preface to Juv., pp. 14, 16. His argument indeed may be converted into one for 59 ( Nettleship, , Journ. of Phil. 16, p. 45)Google Scholar.

¹⁸ In Dürr, , das Leben Juvenals, p. 28Google Scholar ; accepted by Hardy, (Preface to Juv., p. xi sqq.)Google Scholar, whose discussion of the lives of Juvenal is the best in English. Schanz, I.c.; see Addenda.

¹⁹ Mommsen, , op. cit., VI. 2, p. 153Google Scholar : ‘II n'y a pas de limites d'âge auxquelles ces nominations soient subordonnées.’ Cf. Vita Hadriani, 10, 6; C.I.L. XI. 1437.

²⁰ Agricola's governorship is reckoned from 77 (J.R.S. X. pp. 68 sqq.). Friedländer, (Juv., p. 17)Google Scholar says: ‘Sehr wohl kann J. die Feldzüge des Agricola mitgemacht haben.’ So Hardy, p. xxxiv, who also fixes on 81.

²¹ This motive is clearly seen in the last discovered life, Dürr, p. 28: ‘Sub honoris specie in Aegyptum cohortis praefectus mittitur in exilium, ubi duas edidit alias [satiras].’

²² Vitae V., VI. in Jahn's edition of 1851, combined as IV. by Dürr, p. 25; Hardy, p. xiv; Schol. ad VII. 93 in Br. Museum MS. 15 B. XVIII. (not in Jahn's edition of scholia): ‘Hie autem tangit Traianum … quod Traianus sic ultus est, sic uindicauit, ut ipsi Iuuenali daret preposituram belli Scotici et numquam dedit ei successorem qui eum adiuuaret ubi cottidie incursus Scottorum est …’ For ‘belli’ the MS. has ‘galli’ clearly enough, though it is rubbed here; ‘belli’ (Dürr, p. 26) is more probable than ‘valli.’ For my information I have to thank the Keeper of the MSS. in the British Museum.

²³ Nettleship, , I.c., p. 42Google Scholar : ‘Another tradition relegates him to Scotland, or the Scottish border.’ Cf. pp. 43–4. So Hardy, pp. xxx, xxxiv, and Duff, p. xiii. Friedlander is silent, ignoring all the lives but the first. Only Strong, H. A., so far as I know (Class. Rev. V., p. 279)Google Scholar, notes the true meaning of Scoti, but he misses the full inference.

²⁴ Skene, , Celtic Scotland, I. 1–5Google Scholar with authorities there cited; Brown, Hume, Hist, of Scotland, I., p. 51Google Scholar. I have found no instances of the name applied to Ireland after 1000.

²⁵ I quote from a letter of Dr. H. H. E. Craster, sub-Librarian of the Bodleian (where there are two copies of Ognibene's commentary), to whom I owe many thanks. Dürr, p. 25, enumerates the MSS.

²⁶ So the Keeper of MSS. in the British Museum informs me.

²⁷ Tac. Agr.: ‘In spem magis quam ob formidinem’ ib. ‘si … velut e conspectu libertas tolleretur.’ Incerti Panegyricus Constantio Caesari XI. (of Britain 55 B.C.). ‘Natio etiam tune rudis et solis Pictis et Hibemis assueta hostibus, adhuc seminudis, facile Romanis armis signisque cessit, prope ut hoc uno Caesar in ilia expeditione gloriari debuerit quod nauigasset oceanum.’ Haverfield, , I.c., p. 4Google Scholar : ‘The Irish Channel was never probably quite free from pirates. The first record of raids is about 270, when the Dessi “fled” from Meath to South Wales.’ But Joyce, , Social Hist, of Ancient Ireland, I., p. 73Google Scholar, records a bardic legend of an expedition ‘beyond the sea’ by Crimthann Nia Nair, King of Ireland A.D. 74–90, and of his return to Howth with much treasure. Hist. Brittonum in Mommsen's, Chron. Minora, III., p. 165Google Scholar, mentions Picts and Scots as foes in Severus' reign.

²⁸ Agr. 22–25.

²⁹ Appian, Praef. 5: οί Ρωμαȋοι τò κράτιστον αύτς (τς ΒρεττανίϨος) ξΧoυσιν ύπρ ἢμισυ, ούϨν τς ἂλλης ού γρ εüϕορος αύτοȋς έστίν ούδ ἢν έΧοσιν Cf. Strabo IV. 5, 3. But Tacitus, disapproved: ‘perdomita Britannia et statim missa,’ Hist. 1. 1Google Scholar (referring only to the Highlands after Mons Graupius, as DrMacdonald, shows, J.R.S. IX., p. IIIGoogle Scholar sqq., especially 132–3). Incerti, Pan. Constantio Caes. XX.Google Scholar indicates a protectorate about 300 A.D.

³⁰ By DrMacdonald, , I.c., p. 134Google Scholar; so Gaheis, , P.-W. X. 132Google Scholar, ‘zunächst als provisorisches Ziel.’

³¹ ‘Pateretur’ is probably sub-oblique = ‘patietur’ of Agr.'s thought, though it may be Tac.'s retrospect. ‘Brit, ipsa’ might also = ‘citra finem Britanniae,’ without excluding our general view.

³² Cf. n. 27 supra. Incerti Pan. Constantino Aug. Chap. VII.: ‘Neque enim ille … non dico Caledonum, Pictorum aliorumque silvas ac paludes, sed nec Hiberniam proximam nec Thylen ultimam … dignabatur acquirere.’ If Zimmer's theory of a copious trade direct between Ireland and France could be proven (Sitzungsb. Berlin. 1909, pp. 363 sqq.), we might suppose that Agr.'s previous government of Aquitania had suggested this policy. But Haverfield, I.c., rightly objects that there is no evidence and no internal probability in favour of this view, which involves a mistranslation of ‘miscuerit’ in Agr. 24. A ‘lively trade’ would surely have been traceable by Roman coins or other remains in the south of Ireland: but there is no such evidence. Rhys', suggestion (Proc. of Br. Acad., 1903–1904, p. 55)Google Scholar that Gallic Druids were trained in Ireland, not in Britain, seems to lack confirmation. (See Addenda.)

³³ Gudeman on Agr. I.c.

³⁴ In Chap. 7, 5, the reference to Britain becomes clear only from Chap. 8. The advance of the praetorium to Eburacum—a first-rate event—is ignored by Tacitus.

³⁵ Haverfield proposes to take ‘prima’ as neut p1. = ‘the first part of his advance’ Gudeman as = ‘primum,’ ‘for the first time (to Ireland).’ None of the many emendations is successful. See n. 50 infra.

³⁶ Chap. 10, 5, compared with 24, 1.

³⁷ Skene, , op. cit., p. 43Google Scholar, speaks of camps from Dumfries to Wigtonshire, but there is no evidence that these are Roman. DrMacdonald, in J.R.S. IX., p. 134Google Scholar.

³⁸ Agr. 20, 2, ‘nihil … apud hostes quietum pati, quominus subitis excursibus popularetur.’

³⁹ Gaheis, in P.-W. X. 133Google Scholar : ‘Im J. 81 drang A. über die Linie Forth-Clyde hinaus … und zwar, wie es scheint, im Westen über die Clota.’

⁴⁰ Neilson, Annals of the Solway to 1307; Scot. Hist. Rev. 18, pp. 80–1 (where the reference to Camden's, Britannia, ed. Gibson, , p. 1195Google Scholar, wrong); Macdonald, in J.R.S. IX., p. 134Google Scholar ; SirRamsay, J. H., Foundations of England, I. 69Google Scholar.

⁴¹ Here the ‘Roll of Caerlaverock’ is interesting—a list of 60 ships, mostly from the Cinque Ports, mustered there in 1300, under one of the earliest English ‘admirals.’ Neilson, pp. 54 sqq.

⁴² Agr. 20, ‘aestuaria ac silvas ipse praetemptare.’

⁴³ Life of St. Patrick, p. 288 n.

⁴⁴ Hist, of Greece, IV. (ed. 1888), chap. 47, p. 492, n. 3.

⁴⁵ Roberts-Gardner, , Inscr. of Attica, p. 49Google Scholar ; Hicks-Hill, , Gk. Historical Inscr., p. 121Google Scholar ; Holm, , Hist, of Greece, II., p. 372Google Scholar.

⁴⁶ IV. 5, 3.

⁴⁷ , Caes.B.G. IV. 20Google Scholar sqq. As Tacitus, says (Agr. 13, 2)Google Scholar of Caesar in Britain, ‘potest uideri ostendisse posteris.’

⁴⁸ Neilson, p. 27; S. Adamnan tried to recover Irish captives from the King of Northumberland.

⁴⁹ Cambrensis, Giraldus, Itinerarium Cambriae, II. 1Google Scholar.

⁵⁰ So the disputed phrase is best taken: cf. , Caes.B.G. IV. 23Google Scholar with Stock's n.

⁵¹ I have hitherto failed to trace this coin, which was reported by me to Professor Haverfield; but I have reason to believe that the find is genuine.

⁵² Gaheis, , I.c. 135Google Scholar, makes Domitian's second ‘acclamatio’ a record of Agr.'s work in 81; but then (ib. 132) he wrongly assigns Titus' last two ‘acclamationes’ to the British campaign of 80. MrAnderson's, objection (Class. Rev. XXXIV., p. 160Google Scholar) is thus easily met.

⁵³ Tac. Agr. 13, 2; Lucan II. 571–2; Strabo IV. 5, 3; Hor. Epodes 7, 7; n. 27 supra.

⁵⁴ Neglected British History in Proc. of the Br. Acad., 1917–8, pp. 251 sqq. Windisch in Das keltische Britannien had previously advanced a similar view, rejected by Professor, Haverfield, J.R.S. III., pp. 147–8Google Scholar.

⁵⁵ Tysilio IV. 16 (tr. by San Marte, Halle, 1854) [Peace was made with Vespasian in the West through Gweyridd's Queen, a daughter of Claudius] ‘worauf sie zusammen nach London gingen, von wo sie Abtheilungen und Haufen beider Heere zur Unterwerfung von Irland absandten.’ Geoffrey IV. 16 [after an indecisive fight at Exonia] ‘mane facto mediante Genuissa regina concordes effecti sunt duces, et commilitones suos in Hyberniam ablegaverunt.’

⁵⁶ The same error is found elsewhere, e.g. in Eusebius and Eutropius. Professor Flinders Petrie's attempt to refer the name to Selsey Bill is unconvincing.

⁵⁷ Hegesippus V. 15 in Migne, Patrol., vol. 15; cf. , Jos.B.J. V. 361 sqq.Google Scholar, especially 367–8, which gave the cue for amplification.

⁵⁸ Gaheis, , I.c. 135Google Scholar, ranges himself with the majority. ‘Dagegen [against the invasion] spricht aber die klare Darstellung bei Tacitus. Die zum Beweis herangezogene Iuvenalstelle besagt nichts.’