Hostname: page-component-586b7cd67f-2plfb Total loading time: 0 Render date: 2024-11-27T15:52:51.329Z Has data issue: false hasContentIssue false
  • English
  • Français

The Military Situations in Britain Between A.D. 43 and 71

Published online by Cambridge University Press:  09 November 2011

Graham Webster
Get access Rights & Permissions [Opens in a new window]

Abstract

In 1960 I published an account of the campaigns of Aulus Plautius and Ostorius Scapula in Britain, and attempted to show that the Fosse Way was part of a frontier system established by the first governor. There have been subsequent modifications as new information has come to light. A number of new sites has been discovered and the basic pattern of the Plautian frontier, as suggested in the original paper, now requires further consideration. Our conception of the military situations between 43 and 71 has become more complicated with increased knowledge of the tide of battle sweeping forward to the west and then the ebb bringing troops back into the rearward areas subsequent to the Boudiccan revolt. Several sites such as The Lunt at Baginton, Metchley and Usk are being carefully and extensively excavated. These are producing not only interesting sequences of buildings but quantities of pottery and coins which make closer dating possible. The amount of pottery in particular allows detailed studies of shapes and fabrics, enabling the changes to be more narrowly assessed than previously. This present paper is an attempt to consider this new information and to clarify the position as it is now seen. But the situation is still far too fluid for anything but an interim statement; as more work proceeds further adjustments will become necessary.

Type
Articles
Information
Britannia , Volume 1 , November 1970 , pp. 179 - 197
Copyright
Copyright © Graham Webster 1970. Exclusive Licence to Publish: The Society for the Promotion of Roman Studies

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

1 ‘The Roman Military Advance under Ostorius Scapula’, Arch. Journ., cxv for 1958 (1960), pp. 49–98.Google Scholar

2 The Roman Conquest of Britain, 1965 jointly with D. R. Dudley; and ‘The Claudian Frontiers in Britain,’ Studien zu den Militärgrenzen Roms, 1967, pp. 42–53, a summary up to 1964.

3 Dudley, and Webster, , The Roman Conquest of Britain, 1965, p. 89 ff.Google Scholar

4 Suetonius, Claudius, 17; Vespasian, 4.

5 Antiq.Journ., xlix (1969), p. 38.Google Scholar

6 Richmond, Sir Ian, Hod Hill, ii, 1968, pp. 32–33 and fig. 14.Google Scholar

7 Proc. Soc. Antiquaries, iv for 1856–1859Google Scholar; Arch. Journ., xcvi (1939), pp. 114–131.Google Scholar

8 Current Archaeology, No. 14 (1969), p. 80.Google Scholar

9 Interim reports in Antiq. Journ., in particular xlv (1965), pp. 2–3Google Scholar, and Antiquity, xxxix (1965), p. 178.Google Scholar

10 As at Maiden Castle and Hod Hill.

11 The distribution of Claudian coins within the colonia suggests the possibility of an early site in the western part of this area. (Miss Dunnett, B. R. K., Trans. Essex Archaeol. Soc, 3rd ser. ii (1968), pp. 137–142Google Scholar; see also JRS, lvii (1967), p. 189Google Scholar, a mid first cent, building with military bronzes in Insula 11, and JRS, lvii (1968), p. 196 for pits with military bronzes in Insula 25.Google Scholar

12 RIB., 200.

13 RIB., 201; there is also a fragment of a first-cent, tombstone to a soldier of Coh. I Vangionum (?), but he could have been a veteran, RIB., 205.

14 V.C.H., Essex, v, 1966, p. 4.Google Scholar

15 Fifth Report on the Excavations of the Roman Fort at Richborough, ed. by B. W. Cunliffe, 1968.

16 Ibid., fig. 27.

17 Period Ib, preceding the building of the palace c. 75.

18 The Roman Conquest of Britain, 1965, Appendix V, p. 192.

19 Based on the evidence of material found below the forum and street (Sussex Archaeol. Collect., lxxvi (1935), pp. 163–65;Google Scholar also J. Holmes, Chichester: The Roman Town, 1965, pp. 17–18). The implication here is that the town did not develop on Roman urban lines until after the death of Cogidubnus. There must, however, have been a town here during the lifetime of the king and there is evidence that it was provided with at least a temple and a statue of Nero. This can be derived from the two inscriptions; RIB, 92 is a fulsome dedication to Nero in Purbeck marble which is now lost but dated to 58. Its size suggests that it probably came from the base of an equestrian statue. It appears to have been the result of a general thanksgiving probably following the fall of Artaxata when ‘statues and arches were voted’, (Annals, xiii, 41); the chronology of the campaign of Corbulo is confused since Tacitus in his usual manner deals with several seasons in a single passage (34-42). It would have presented an opportunity for Cogidubnus to demonstrate his loyalty. The other sculpture raises issues too involved to be treated here in detail. RIB, 91 is the dedication of a temple by a guild, presumably of shipwrights since Neptune and Minerva are invoked. A guild wealthy enough to provide for such a substantial monument may indicate the presence of a naval base in Bosham Harbour, where local firms could have obtained lucrative contracts. Cogidubnus appears here as a king and imperial legate and the problem is to date the inscription. There is no evidence from the stone itself except that Domus Divinae excludes the reign of Vespasian (69–79), and the king may well have been alive for some years after this. There must have been exceptional circumstances under which the title of Legatus Augusti would have been given. If it had been Claudius, the absence of any reference in the satirical skit attributed to Seneca (The Apocolocyntosis) is remarkable. Nero did almost nothing to advance the political aspirations of the provincials and there appears to have been no occasion to have made such an exception unless Cogidubnus performed some remarkable service to Rome during the Boudiccan revolt: but surely Tacitus or Dio would have remarked upon it. One is left with Vespasian who had memories of the king in 43, when there must have been a close collaboration. The state of Britain during the Civil War was one of some confusion and Cogidubnus may well have used his influence on behalf of the Flavian cause which Vespasian could have acknowledged in this remarkable way. In that case the erection of this temple belongs to the reign of Titus or early in that of Domitian. It may even be possible to link th e occasion with the passage in the Agricola (21, ut templa fora domos extruerent) the effect of which is seen at Verulamium (Antiq. Journ., xxxvi (1956), pp. 8–12Google Scholar). Professor Frere has put forward an alternative suggestion that the status of legate may have been conferred by Claudius after Cogidubnus had supplied the Romans with a force of his own troops when the Iceni revolted in 48/49 (Britannia, p. 83). The words used by Tacitus, socialis copias, are in contrast to legionaries and the succeeding passage—et distributis cohortibus turmas quoque pedilum ad munia accingit—makes it clear that regular Roman auxiliary troops are involved. Another use of the same word by Tacitus to refer to auxiliaries occurs earlier in the Annals (iv, 73 turmas socialis equitesque legionum). The theme is developed by Mr. Boon, G. C. in his paper, ‘Belgic and Roman Silchester: excavations of 1954–58’, (Archaeologia, cii (1969), pp. 37Google Scholar and 38). But not only does he suggest, as Frere has done, that Cogidubnus had given the Romans military aid but that his kingdom was extended and that at this point h e was given the title legatus. It is a neat idea that the loyal ally should be given extended power in order that Scapula could concentrate on Caratacus. It depends on a passage of Tacitus—quaedam civitates Cogidumno regi donatae (Agricola, 14). Richmond interpretes this as the gift of ‘at least two other tribes as subjects’ (Roman Britain, 2nd. ed., 1963, p. 23). This is developed in the edition of the Agricola b y R. M. Ogilvie and Ian Richmond (1967, p. 189) as ‘certain cantons were given to King Cogidubnus’ as if he was already king before the gift. On this slender evidence, Mr. Boon postulates the idea that the extra tribes were governed by Cogidubnus as legate in addition to his kingdom.

20 JRS, xliii (1953), p. 82; an aerial photograph by Dr. St. Joseph.Google Scholar

21 V.C.H., Essex, v, 1966, p. 5; as foreseen by Fox, The Archaeology of the Cambridge Region, 1923, p. 174.

22 Dio, lx, 20, 2, as suggested by S. S. Frere, Britannia, p. 64.

23 Archaeol. Cantiana, lxxxiii (1968), p. 75.Google Scholar

24 As demonstrated with a list in J. S. Wacher (Ed.), The Civitas Capitals of Roman Britain, 1966, pp. 31–45. The only addition is Chelmsford where an early ditch and rampart is recorded (V.C.H., Essex, v, 1966, p. 5). Confirmation of a fort at Alcester, Warwicks., has come with the discovery of military equipment in a recent excavation directed by Mr. Steven Taylor.

25 As at Camulodunum and Verulamium (Antiq. Journ., xxxvii (1957), p. 5); at Bagendon, on the other hand, the Roman commander preferred the site at Corinium where the ground was more open and level.Google Scholar

26 Archaeologia, cii (1969), fig. 5Google Scholar. Some of this material was found in a pit in Insula XXIII in the northern part of the city. A fort in this area might be related to the length of street at the East Gate which is at an angle to the general town grid, on the analogy of a similar arrangement at Verulamium. The presence of a fort here in the Plautian phase would upset the chronology offered by Mr. Boon of the sequence of earthworks (ibid., p. 14).

27 Antiq. Journ., xlviii (1968), pp. 210–230Google Scholar; JRS, lvii (1967), p. 190.Google Scholar

28 JRS, lv (1965), p. 74–76 and pl. ix; lviii (1968), p. 189.Google Scholar

29 Frere suggests Newton-on-Trent (Britannia, p. 70; JRS, lv (1965), pp. 74–76 and pl. x (1)), but such an advanced position would seem to fit a later situation.Google Scholar

30 S. S. Frere, Britannia, p. 70. Mr. Hartley, B. R. has noted ‘a striking lack of Claudian pottery’: Northern History, i (1966), p. 10.Google Scholar

31 This could be supported by the presence of a cemetery in this area (JRS, xxxix (1949), p. 58 and fig. 11, no. 20), but this includes a standard-bearer of IX (RIB, 257).Google Scholar

32 Frere, Britannia, p. 75, fn. 1; although Richmond retained Collingwood's appraisal of 1930 in the revised edition of The Archaeology of Roman Britain, 1969, p. 72.

33 First noted from the air by Dr. St. Joseph and excavated by Miss E. Blank, of Leicester City Museum, to whom I am indebted for this information. The site has not yielded any dating evidence or equipment but the position would be suitable for that of a fort.

34 Antiq, Journ., xlii (1962), pp. 3–5; xliii (1963), pp. 15–16; xlv (1965), pp. 97–101.Google Scholar

35 By Todd, Malcolm (JRS, lvii (1967) pp. 183; lviii (1968), pp. 184–85 and fig. 11); this suggests that Dr. Felix Oswald failed to find the fort but worked mainly in a large military compound associated with metal working.Google Scholar

36 By Wacher, John (JRS, liv (1964), p. 159; lvi (1966), p. 203 and fig. 10).Google Scholar

37 JRS, lviii (1968), p. 186. This area near the river, I am informed by Mr. B. R. Hartley, has produced a quantity of Claudian samian.Google Scholar

38 JRS, lvi (1966), p. 203; lvii (1967), p. 182; the presumed military ditch was found to cut through two earlier pits, one of which contained Roman pottery.Google Scholar

39 Malcolm Todd, The Roman Fort at Great Casterton, Rutland, 1968.

40 JRS, lix (1969), p. 227.Google Scholar

41 Barry Cunliffe, Roman Bath, 1969, p. 2.

42 Annals, xii, 31. The words of Tacitus detrahere arma suspectis parat would seem to indicate that not all of the tribes were subject to a search for arms but only those whose loyalty was suspect. By Roman law, the Briton was not allowed to carry arms except for hunting or self-defence on a journey. But it is not necessary to imagine the existence of hidden arms although they may well have existed in a Celtic society so recently subdued. Swords and spears could easily be made by local blacksmiths; thus Scapula's act would seem to be one of terrorization.

43 The discharge of veterans, often the prelude to serious campaigning, at least created a reserve at the new colonia at Camulodunum.

44 castrisque legionum indicates that more than one legion was involved. The action was now against the Silures, and Gloucester seems to be the best position for a base of operations.

45 Summarized in Exeter and its Region, 1969, pp. 145–48, with an additional possible fortlet at Broadbury (fig. 24).

46 Fox, Aileen and Ravenhill, W. L. D., ‘Early Roman Outposts on the North Devon Coast, Old Burrow and Martinhoe’, Proc. Devon Archaeol. Explor. Soc, No. 24 (1966), pp. 3–39Google Scholar; see also Antiquity, xxxix (1965), PP. 253–57.Google Scholar

47 By the same authors, Cornish Archaeology, v (1966), pp. 28–30; vi (1967), pp. 32–33; vii (1968), pp. 40–42.Google Scholar

48 The excavators suggest that as Martinhoe is 300 ft. lower than Old Burrow, it may have had better visibility when there were hill mists.

49 The fifty-four coins allow for a close assessment. There is nothing later than Claudius and his issues are hardly worn. The samian also does not include any Neronian sherds (Richmond, Sir Ian, Hod Hill, ii, 1968, pp. 117–123).Google Scholar

50 Interim reports, Proc. Dorset Natur. Hist. Archaeol. Soc, lxxxii (1961), pp. 88–108; lxxxvi (1965), pp. 135–49; final report forthcoming.Google Scholar

51 There are forty-two coins of Claudius and twenty-three of earlier issues, while at Hod Hill there were only twelve of Claudius and fifty-two earlier issues, including thirty-one under Gaius.

52 The absence of any coins of Nero makes it difficult to put this beyond 64 when this emperor started large issues of base metal, Brit. Mus.: Coins of the Roman Empire, i, p. clxvi.

53 The equipment seems to suggest a mixed garrison.

54 JRS, lix (1969), p. 228. The area investigated has revealed water tanks built against the back of a rampart. The pottery includes Claudian and Neronian sherds (Information from Mr. Norman Field).Google Scholar

55 There is a fine signal station built into the west end of the Iron Age Hillfort at Abbotsbury Castle with a typical Roman military ditch cutting down the slope as at Waddon (R.C.H.M., Dorset, I. West, 1952, p. 11 and plan on p. 10). There should be other stations on headlands along the coast if they have survived erosion. The signal station at Stoke Hill near Exeter, at one time thought to be later (Trans. Devonshire Ass., xci (1959), pp. 71–80Google Scholar), is because of its similarity in plan to those on the North Devon coast, now thought by Lady Fox to be part of the first-cent, system. There is another possible site at High Peak on the coast (Trans. Devonshire Ass., lxii (1930), pp. 119–20).Google Scholar

56 V. E. Nash-Williams, The Roman Frontier in Wales, 2nd edn. edited by Michael G. Jarrett, 1969, pp. 77–80. Dr. Jarrett suggests that it may belong to this period, ‘Early Campaigns in Wales’, Arch, journ., cxxi (1965), p. 26; the samian from this site is in a wretched condition which makes dating difficult.Google Scholar

57 JRS, lix (1967), p. 123.Google Scholar

58 The Roman Frontier in Wales, pp. 124–26.

59 A coin of 64 was found in the clay wall of a barrack-block; another coin of 66 was sealed by a street of the fortress: this information has been kindly supplied by Mr. Henry Hurst in advance of publication (H. R. Hurst, Excavations in Gloucester, 1969, privately circulated by the Gloucester City Museum).

60 Reliquae-Brittannico-Romanae, ii, pls. xi to xv.

61 A new discovery may have a bearing on the situation in Central Wales at this period. This is a 10 acre fort at Caersws discovered by Dr. Joseph, St. (JRS, lix (1969), pp. 121–22).Google Scholar

62 Annals, xii, 40, Silures…lateque persultabant.

63 V. E. Nash-Williams, The Roman Frontier in Wales, 2nd revised edition, 1969, p. 5 and fig. 3.

64 Boon, G. C., ‘Remarks on Roman Usk’, Monmouthshire Antiquary, I, ii (1962), pp. 28–33Google Scholar; there is doubt about an early samian bowl in the Caerleon Collection and the possibility, as Mr. Boon indicates, is that this is all that is left of forty boxes of pottery sent from Usk to Caerleon 1893–95, Monmouthshire Antiquary II, i (1965), pp. 42–51.Google Scholar

65 JRS, lviii (1968), p. 177.Google Scholar

66 V. E. Nash-Williams, The Roman Frontier in Wales, 2nd revised edition, 1969, p. 46; JRS, lvi (1966), p. 198Google Scholar. See now Monmouthshire Antiquary, II, iv (1970), pp. 163–98.Google Scholar

67 S. S. Frere, Britannia, p. 84. V. E. Nash-Williams, op. cit., p. 81.

68 Possibly the Jay Lane site where the date c. 47–61 has been postulated by Mr. Stanford, S. C., Trans. Woolhope Field Club, xxxix (1968), p. 237.Google Scholar

69 Graham Webster and Donald R. Dudley, The Roman Conquest of Britain, 1965, pl. 40.

70 Ibid., fig. 39.

71 Trenches by Dr. A. W. J. Houghton failed to produce a single sherd or any indications of internal buildings.

72 Found in a trench cut by Dr. Joseph, St., Trans. Birmingham Archaeological Soc., lxix for 1951 (1953), p. 54, samian forms 24/25, 29, 27 and a mortarium rim probably of Flavian date.Google Scholar

73 Trans. Shropshire Archaeol. Soc., lviii (1967), pp. 8–18.Google Scholar

74 JRS, lv (1965), pp. 76–77.Google Scholar

75 Trans. Lichfield S. Staffs. Archaeol. Hist. Soc., 8 for 1966–1967 (1968), pp. 1–38. There are now twentyfour pre-Neronian coins.Google Scholar

76 Although Tacitus says of him in a later context that he had been loyal to Rome (fidusque diu) and only after his break with Cartimandua did he take up arms against Rome.

77 Arch. Journ., cxxv for 1968 (1969), pp. 198 and 212–13.Google Scholar

78 V.C.H., Staffs., i, p. 189: Antiq. Journ., xiv (1934), p. 183. Sections were cut through the defences by the Newcastle-under-Lyme Archaeological Society in 1969, to whom I am grateful for the opportunity of examining the sections and obtaining information in advance of publication.Google Scholar

79 N. Staffordshire J. Field Studies, viii (1968), pp. 19–38.Google Scholar

80 Finds of coins and pottery on a modern housing estate; information kindly given by Mr. Roy Hughes of the Derby Museum.

81 Trans. Birmingham Archaeological Soc., lviii (1937), pp. 68–83.Google Scholar

82 Information kindly supplied by Mr. T. Rowley in advance of publication.

83 Trans. Birmingham Archaeol. Soc., Ixxii (1956), pp. 1–4.Google Scholar

84 Antiquity, xl (1966), pp. 300–4 and fig. 1;Google ScholarTrans. Birmingham Archaeol. Soc, lxxx (1965), pp. 82–83, fig. 4.Google Scholar

85 The northernmost fort was trenched by boys from Wolverhampton Grammar School and produced at least one Claudian vessel (William Salt Arch. Soc. Proc, for 1927 (1929), pp. 185–206); the other fort has been explored by the Kidderminster Arch. Soc. and produced 14 coins of Claudius (W. Midlands Annual Archaeol. News Sheet, No. 11 (1968)).Google Scholar

86 JRS, lix (1969), P. 103.Google Scholar

87 Trans. Birmingham Archaeol. Soc., lxxii (1943), pp. 27–31Google Scholar; Ixiv (1946), pp. 39-40: Trans. Worcestershire Archaeol. Soc., xxxix (1963), pp. 55–58.Google Scholar

88 Epigraphische Studien, iv (1967), p. 102, note 1.Google Scholar

89 Annals, xii, 40, senectute gravis et multa copia honorum, per ministros agere et arcere hostem satis habebat.

90 Agricola, 14, paucis admodum castellis in ulteriora promotis…

91 Annals, xii, 40, fratrem ac propinquos Venutii.

92 T. May, The Roman Forts of Templeborough near Rotherham, 1922, pl. xxii; redrawn by Dr. Grace Simpson in Britons and the Roman Army, 1964, p. 11 and fig. 1.

93 JRS, lix (1969), p. 103 and pl. ii (1).Google Scholar

94 J. Wacher in Excavations at Brough-on-Humber 1958–1961, 1969, pp. 5–8.

95 Antiq. Journ., xviii (1938), pp. 262–77. This is a site which would well repay further investigation; it is possible that much of the pottery is not as early as stated in the report.Google Scholar

96 There is a surprising number of Claudian coins for York (it is not certain that these are genuine York finds) which can hardly be accounted for as ‘old savings-deposits brought from Lincoln’ (Eburacum, Roman York, 1962, p. xxx) since they are bronze issues. Pre-Flavian pottery has been noted (JRS, lv (1965), p. 204Google Scholar) and Mr. B. R Hartley has suggested that the rebuilding of the Agricolan rampart may imply an earlier small fort (Northern Hist., i (1966), pp. 10–11).Google Scholar

97 It may be significant that when Seneca, or whoever it was, was writing the Apocolocyntosis after the death of Claudius, it was the caeruleos scuta Brigantas (12) which came to mind.

98 E. Birley, Roman Britain and the Roman Army, 1953, p. 7. I t could, however, be argued that it was a momentary panic decision by Nero as news of the disaster of 60 came to Rome. The procurator may well have given the impression that all was lost. (B. H. Warmington, Nero: Reality and Legend, 1969, pp. 78–79.)

99 So far Dr. Manning has found no buildings other than granaries and store-buildings. The brief account in Current Archaeology, No. 11 (1968), pp. 257–59, is now out-dated by a further season of extensive work in 1969. In spite of the great effort in stripping this area it is still only a small part of the whole.Google Scholar

100 Bull. Board of Celtic Stud., xiii (1950), p. 239; Dr. Grace Simpson has identified a few pre-Flavian samian sherds (V. E. Nash-Williams, The Roman Frontier in Wales, 2nd revised edition, 1969, pp. 99, fn. 3).Google Scholar

101 This fort has produced a few pre-Flavian sherds (The Roman Frontier in Wales, p. 96).

102 The Belgic-type pottery found there (Jarrett, M. G., ‘Early Roman Campaigns in Wales’, Arch. Journ., cxxi (1965), p. 29), is more likely to have been introduced by the Roman Army than by the Silures.Google Scholar

103 No one has suggested that the occupation of this fort could be earlier than the traditional date of 75.

104 The Neronian cremation burial below the later legionary fortress must belong to this period (J. Chester N. Wales Architect. Archaeol. Hist. Soc., xxxv (1942), pp. 49–51; xxxviii (1950), p. 18).Google Scholar

105 Agricola, 5.

106 Annals, xiv, 38, contractus deinde omnis exercitus subpellibus.

107 These include a cuirass-hinge (Acc. No. B 12.152.929); part of a cuirass-hook (Acc. No. 6.8.152.929) and a lunate pendant (Acc. No. B. 15.152.929).

108 East Anglia, Thames and Hudson, 1960, p. 114.

109 JRS xliii (1953), p. 82Google Scholar; first discovered by Mr. Riley, D. N., JRS, xxxv (1945), p. 82; lix (1969), pl. ii(2).Google Scholar

110 Antiq. Journ., xxxvi (1956), pp. 73–75. The fragment of fretted bronze is part of a bronze plate (opus interrasile) which weighted down the saddle cloth, J. Werner in Berträge zur älteren europäischen Kulturgeschichte, Band I, p. 423.Google Scholar

111 Antiquity, xiii (1939), 189–190Google Scholar. S. S. Frere, Britannia, 1967, fig. 2, p. 72. The evidence is inconclusive as Frere himself has stated (Antiq. Journ., xxi (1941), p. 54Google Scholar) and consists of ditches thought to have been in the form of a square, with some iron spearheads and a first-cent, brooch (Proc. Suffolk Inst. Archaeol., xxiii (1939), pp. 173–74.Google Scholar

112 Current Archaeology, no. 16 (1969), p. 134. This could be Plautian.Google Scholar

113 Trans. Birmingham Archaeol. Soc, lxxxiii (1969), pp. 63–129.Google Scholar

114 Further extensive work was carried out in 1969 (report forthcoming).

115 D. R. Dudley and G. Webster, The Rebellion of Boudicca, 1962, p. 74.

116 Trans. Birmingham Archaeol. Soc., lxxix (1964), pp. 117–20. The absence of later coins in this hoard of sixteen, mostly Claudian imitation, strongly suggest that it was buried before 64.Google Scholar

117 Ibid., lxxiiii (1956), p. 36.

118 Work carried out by Mr. Keith Scott who has kindly allowed me to anticipate his report.

119 Discovered by Dr. Joseph, St. (Trans. Birmingham Archaeol. Soc., lxix (1953), p. 52)Google Scholar. A small excavation produced a piece of Neronian samian (Ibid., lxxiii (1957), p. 102, fig. 2).

120 JRS, xlviii (1958), pp. 94–95.Google Scholar

121 Annals, xiv, 38.

122 Annals, xiv, 32, et inerat modica militum manus may imply no more than clerks or technicians in a store base.

123 S. S. Frere, Britannia, 1967, p. 93.

124 Aileen Fox, Roman Exeter, 1952, p. 17.

125 Information kindly supplied by Mr. Alec Down.

126 Seep. 182, fn. 19 above.

127 Antiq.Journ., xlii (1962), p. 5.Google Scholar

128 Hist., iii, 45, cohortes alaeque.

129 The poet Statius credits him with building forts and watchtowers speculas castellaque longe and with the dramatic episode of wrenching a breastplate from a British king, regi rapuit thoraca Britanno (Silvae, v, 2. 143).

130 Information kindly supplied by Mr. Round, A. A.; see also Trans. Lichfield and S. Staffs. Archaeol. Hist. Soc., viii (1968) p. 11.Google Scholar

131 Information kindly supplied by Mr. Trevor Rowley.

132 There are two coins of Vespasian, Cos. IIII (A.D. 72–73) (nos. 7–8 in the report) and a sherd of samian which ‘can scarcely be much earlier than 75’ (Report, p. 87).

133 Report, p. 39. One sherd of samian is dated c. 80–105, but the presence of the nearby town makes it difficult to be certain of the context of unstratified pieces on the site of this fort.

134 This is not, however, a view put forward by Professor Frere (Britannia, p. 117, fn. 2). He argues that XX was withdrawn from Inchtuthil while II Adiutrix was still at Chester. All that we know for certain is that XX was probably in Scotland up to 87, and possibly later, and that II Adiutrix was in Moesia by 92. There could have been a slight over-lap, but surely the decision to withdraw from Scotland must be linked with the need for troops in Dacia in 87 and 88, to avenge the Romandefeat in 87? On this also hangs the argument about the unfinished bath-house at Wroxeter. Frere suggests that XX set about rebuilding the fortress on the basis of a four legionary garrison in Britain. But had this been so, Wroxeter would not seem to have been a very suitable place for a permanent legionary base, with another so near at Chester. In broad terms of strategy, Carlisle would have been a much better site, since the legions would then have been equally divided, two facing the north, and two the west. This leaves the problem of the Wroxeter bath-house unsolved. The masonry is certainly similar to that at Chester, and it is not impossible that XX may have built it as a ‘parting gift’ to the new city where many of their veterans had already settled. The abandonment of the work could have been due to the troubles in the north necessitating the presence of the legion there at the end of the first century, or to a veto on the project from a high authority.