^{page 142 note 1} Gardner, P., J.H.S. ix, 1888, p. 47Google Scholar ff.

^{page 142 note 2} e.g. the famous so-called ‘Thusnelda’ in the Loggia dei Lanzi at Florence, most probably a personification of Germania (Strong, Roman Sculpture, pl. lxviii); Armenia personified as a male Armenian captive on coins of Augustus with the legend CAESAR DIVI F. ARMENIA CAPTA (Mattingly, Coins of the Roman Empire in the British Museum, vol. 1, pl. 2⁴) etc.

^{page 143 note 1} Vita 5, 2. ‘Britanni teneri sub Romana ditione non poterant.’

^{page 143 note 2} ibid. ii, 2, ‘Britanniam petit, in qua multa correxit murumque per octoginta milia passuum primus duxit, qui barbaros Romanosque divideret’ 12, 1. ‘Compositis in Britannia rebus transgressus in Galliam …’

^{page 143 note 3} The Correspondence of Marcus Cornelius Fronto, Loeb., edit. vol. 2, p. 22Google Scholar. ‘Quid ? avo vestro Hadriano imperium obtinente quantum militum ab Britannis caesum.’

^{page 144 note 1} Mr. Mattingly of the British Museum has now established the following chronological sequence for the earliest group of ‘cos. III’ coins, namely those on which the Emperor is given the title of Pontifex Maximus:—

Our Britannia type, which in its earlier issue bears the legend PONT. MAX. on the reverse, thus belongs to the year 119 A.D.

^{page 144 note 2} Cohen, , La monnaies de l'empire romain², ii, p. 213Google Scholar, no. 1285 ƹ.

^{page 144 note 3}

Cohen, op. cit.² ii, p. 121Google Scholar, no. 197 Ʋ (Plate xxiv, 1).

^{page 144 note 4} G. Macdonald, Roman Wall in Scotland, p. 6.

^{page 145 note 1}

Cohen, , op. cit. ² ii, p. 121Google Scholar, nos. 194-196, 198, 199. Æ^1,2. Macdonald, G., op. cit. pl. i, A. 2Google Scholar. Haverfield, Roman Occupation of Britain, fig. 4, 2 (Plate XXIV, 2).

^{page 145 note 2} G. Macdonald, op. cit. p. 7.

^{page 146 note 1} Mr. Mattingly has pointed out to me that Britannia's attitude bears a distinct resemblance to that to Securitas.

^{page 146 note 2} Cf. the cloak with a fringed border worn by the captive woman on the so-called ‘Trophy of Marius’ (Bieńkowski, De simulacris barbararum gentium apud Romanos, p. 39, fig. 19).

^{page 146 note 3} Never a sceptre, as Cohen sometimes describes it.

^{page 146 note 4} This remarkable type of shield, for which I can find no parallel, seems to have been peculiarly British, though neither Caesar nor Tacitus nor any other of our ancient authorities mention it. The popular notion that the British chariot wheels were equipped with spikes or scythes appears to rest on very doubtful authority (Daremberg et Saglio, s.v. Essedum, p. 815, and Déchelette, Manuel d'archéologie, iii, 1183)Google Scholar.

^{page 146 note 5} Mr. Mattingly has made the interesting suggestion that these are not rocks but the actual Wall itself. The regularity with which they are arranged certainly favours the idea that they represent courses of stone, but at the same time it should be noticed that in Hadrian's Dacia type, where there is no question of a wall, the rocks on which the province is seated are even more symmetrical.

^{page 146 note 6} Bieńkowski op. cit. p. 79, fig. 75; Lucas, , Jahrbuch des kaiserlichen deutschen archäologischen Institute, Band xv, 1900, fig. 13Google Scholar.

^{page 147 note 1} Cohen, , op. cit.² ii, p. 153Google Scholar, nos. 555, 556. Macdonald, , op cit. pl. i, A. 1Google Scholar; Haverfield, op. cit. fig. 4, 1.

^{page 147 note 2}

Cohen, , op. cit.² ii, p. 109Google Scholar, no. 28. ƹ (Plate XXIV, 3).

^{page 147 note 3} Birmingham Post, Tuesday, July 15, 1924, p. 7; Antiquaries Journal, October, 1924; Year's Work, 1923-4, p. 82; Classical Review, vol. xxxviii, NoV.-DeC, 1924, p. 146Google Scholar.

^{page 147 note 4} Haverfield, , Victoria County History of Shropshire, i, pp. 216, 244, 245Google Scholar.

^{page 148 note 1} Cohen, , op. cit.² ii, pp. 314–315Google Scholar, nos. 463-472.

^{page 148 note 2}

Cohen, , op. cit.² ii, p. 281Google Scholar, no. 115. ƹ (Plate XXIV, 4).

^{page 148 note 3} Here, however, it is just possible to suppose that the seat is concealed behind the shield, whereas the design on Hadrian's coin does not admit of this supposition.

^{page 148 note 4} The British Museum does not possess a specimen of this coin, and the bronze cast in the Ashmolean Museum is taken from a poorly preserved specimen, and is therefore not very reliable as evidence for details. But after examining the latter, I am inclined to decide in favour of Britannia being here represented with a helmet. Possibly the combination of long chiton and himation indicates an attempt to assimilate the Britannia to the Roma type. It is not possible to come to any definite conclusion about the treatment of the hair of this Britannia from the Ashmolean cast.

^{page 149 note 1}

^{page 149 note 2} I cannot discover anything to account for the use of the legionary standard in this and in the third type of Pius 140-143 Britannia issue. In all Hadrianic types and in all the other types of Pius where a standard is an attribute of a personified province it is the vexillum which is represented. The introduction of the legionary standard in these two types seems to have been an experiment which was tried and then abandoned, for the vexillum reappears in the Britannia type of Pius' 155 issue.

^{page 149 note 3} Cohen, , op. cit.² ii, p. 213, no. 1285Google Scholar.

^{page 149 note 4} Cohen, , op. cit.² ii, p. 37, no. 179, pp. 51–52. nos. 326, 327Google Scholar.

^{page 149 note 5} Cohen, , op. cit.² ii, pp. 314–315, nos. 463-472Google Scholar.

^{page 149 note 6} Cohen, , op. cit.² ii, p. 281, no. 113Google Scholar.

^{page 149 note 7} e.g. Roscher, , Lexicon, Band iii, p. 335, Abb.14-Daremberg et Saglio, figs. 7466 and 7467Google Scholar.

^{page 149 note 8} Vergil, , Ecl. i, 66Google Scholar.

^{page 150 note 1}

Cohen, , op. cit.² ii, p. 282, no. 119Google Scholar.

s16_inline7

Cohen, , op. cit.² ii, p. 281Google Scholar, no. 116; Macdonald, G., op. cit. pl. ii, 1Google Scholar (Plate XXIV, 6).

^{page 150 note 2} Cohen makes this remark only in connexion with his no. 116, but presumably he intended it also to apply to his no. 119, the types of both coins being identical. Jatta (Le rappresentanze figurate delle provincie romane, p. 15), following Cohen, makes the mistake of placing this coin under the heading ‘Tipo della provincia capta.’

^{page 150 note 3} In the British Museum specimen the spear and the spike on the shield are practically obliterated. It is, however, just possible to distinguish their traces.

^{page 150 note 4} Atkinson, D., The Governors of Britain from Claudius to Diocletian, J.R.S. xii, 1922, p. 66Google Scholar.

^{page 150 note 5} Iuli Capitolini Antoninus Pius 5, 4. ‘Britannos per Lollium Urbicum vicit legatum alio muro cespiticio summotis barbaris ducto.

^{page 150 note 6} G. Macdonald, op. cit. p. 8.

^{page 150 note 7} It must have been for this purpose that the coins with Victory and the legend BRITAN were struck, cf. supra p. 149, note 6, and G. Macdonald op. cit. Plate i, A 3; Haverfield, op. cit. fig. 4, 3. It is natural to suppose that these were issued immediately after the successes of Lollius and probably before the Britannia types.

^{page 150 note 8} I suggest the following chronology for the cos. III Britannia types. The activities of Lollius occasioned three experiments in ‘Britannias’ early in 143 A.D.—(i) Britannia in quasi-seated attitude with helmet (?), long chiton and spear = Cohen 115; (ii) Britannia seated on globe; (iii) Britannia seated to l. on rock = Cohen 119. All these coins have the legend IMPERATOR II. (i), from its close resemblance to the Hadrianic type, I should place first in date—(iii) would certainly seem to come last. It is decidedly the best design of the three, and was evidently popular, since Commodus' medallion (vide infra) is an almost exact copy of it. (ii) may be dated between (i) and (iii); as far as style and composition go, it holds an intermediate position between the other two. I would also suggest that the cos. III coins with the same type as (iii) but without IMPERATOR II, ( = Cohen 116) were struck later in 143, when the interest in the ‘salutatio’ had died down, to commemorate the progress of the work on the wall, the latest and most successful of the three IMPERATOR II types being repeated for this purpose.

^{page 151 note 1} Pausanias viii, 43, 4. Ἀπϵτέμϵτο δὲ καὶ τῶν ἐν Βρϵττανία Βριγάντων τὴν πολλήν, ὅτι ἐπϵσβαίνϵιν καὶ οὖτοι σὺν ὅπλοις ἦρξαν ὲς τὴν Γϵνουνίαν μοῖραν, ὑπηκόους Ρωμαίων.

^{page 151 note 2} Haverfield, Roman Occupation of Britain, pp. 120, 121; Macdonald, op. cit. pp. 9, 10.

^{page 151 note 3} Haverfield, op. cit. fig. 6, p. 122.

^{page 151 note 4} Haverfield, op. cit. fig. 5, p. 121; Macdonald, op. cit. plate 1, 13, p. 8.

^{page 151 note 5} Ritterling, , Westdeutche Zeitschrift, Korrespondenz-blatt, xxii (1903), p. 217Google Scholar.

^{page 151 note 6}

Cohen, , op. cit. ², ii, p. 282Google Scholar, nos. 117 118, Ʋ. Macdonald, G., op. cit. pl. ii, 2Google Scholar; Num. Chron. 1907, pl. xi, 5-8 (Plate XXIV, 7).

^{page 152 note 1} This is not, I think, an umbo, as Bieńkowski (op. cit. p. 34) describes it.

^{page 152 note 2} I am convinced by an examination of the coins that this object is a vexillum and not ‘a sceptre surmounted by an eagle’ as it is called by Cohen, (op. cit.² ii, p. 282Google Scholar) and his follower Jatta (op. cit. p. 15).

^{page 152 note 3} Gnecchi, , I medaglioni romani, i, pl. 211Google Scholar; Grueber, Roman Medallions in the British Museum, pl. i.

^{page 153 note 1}

Cohen, , op. cit.² iii, p 232, no. 37Google Scholar; Gnecchi, , op. cit. ii pl. 78²Google Scholar; Grueber, op. cit. pl. xxix; Macdonald, G., op. cit. pl. ii, 3Google Scholar; Bieńkowski, op. cit. fig. 46; Jatta (op. cit. p. 15) makes the curious mistake of assigning Commodus' medallion to the reign of Hadrian. The origin of his error is probably to be found in his misreading of the following sentence in Bieńkowski (op. cit. p. 56), where the medallion is mentioned : ‘Noch wilder und barbarenhafter sieht Britannia (fig 46) auf den Münzen Hadrians (Cohen, n 194-9), Pius (ebd. n. 113-119) und Commodus' (ebd. n. 37-38) aus.’ Fig. 46 is the medallion of Commodus, but Jatta, who seems to have taken over Bieńkowski's sentence without further verification, concludes that it reproduces a Hadrianic type, as Bieńkowski's wording which is certainly misleading, might at first sight suggest. As a result of this confusion, Jatta does not include in his collection the real Britannia type of Hadrian, a serious omission.

^{page 153 note 2} Cassius Dio LXXII, 8. Ἐγένϵτο δὲ καὶ πόλϵμοί τινϵς αὐτῷ … μέγιστος δὲ ὁ Βρϵταννικός. τῶν γὰρ ἐν τῇ νήσῳ ἐθνῶν ὑπϵρ βϵβηκότων τὸ τϵῖχος τὸ διόριζον αὐτούς τϵ καὶ τὰ τῶν 'Ρωμαίων στρατόπϵδα, καὶ πολλὰ κακουργούντων, στρατηγόν τέ τινα μϵτὰ τῶν στρατιωτῶν οὓς ϵἶχϵ κατακοψάντων, φοβηθϵὶς ὁ Κόμμοδος, Μάρκϵλλον Οὕλπιον ὲπ' αὐτοὺς ἔπϵμψϵν … Μάρκϵλλος μὲν δὴ τοιοῦτος ὢν τούς τϵ βαρβάρους τοὺς ἐν Βρϵταννίᾳ δϵινῶς ἐκάκωσϵ. Æ Lampridii Commodus Antoninus 13, ‘in Britannia … imperium eius recusantibus provincialibus, quae omnia ista per duces sedata sunt.’ Cf. 8, ‘appellatus est Commodus etiam Britannicus ab adulatoribus, cum Britanni etiam imperatorem contra eum deligere voluerint.’

^{page 153 note 3} The only point in which Commodus' medallion really differs from the sestertius of Pius is in a detail of the standard. The standard held by the Britannia of Commodus does not show the half-moon decoration which appears on the standard in both the second and third types of 143 A.D.

^{page 153 note 4} Cf. Rostovtzeff, , ‘Commodus-Hercules in Britain,’ J.R.S. xiii, p. 96Google Scholar. This revival of the peaceful Britannia type does, however, take on a new significance if we accept Mr. Collingwood's theory (in J.R.S. xiii, 69 ff.) that between 181, the year of the great catastrophe, and 185 Ulpius Marcellus carried out extensive repairs on the Wall. The type would then commemorate the restoration to Britain of her line of frontier defence—the Wall and the watch thereon.

^{page 153 note 5} Cohen, , op. cit.² iii, p. 231Google Scholar, no. 35. Obv. M. COMMODVS ANTONINVS AVG. PIVS. Rev. BRITT. (exergue) P.M. TR.P. VIIII IMP. VII COS IIII P.P. (circumf.) s.c. ƹ.

^{page 154 note 1} Cohen, , op. cit.² iv, p. 6Google Scholar, nos. 25-31. I am following Cohen's dating of the silver coins on which Severus is simply described as PIVS AVG. The bronze Africa coins with the same types can be dated as belonging to the years 194 and 195, owing to the presence of IMP. III or IMP. IV on the obverse.

^{page 154 note 2} ibid. iv, p. 27, no. 228, p. 153, no. 102.

^{page 154 note 3} Africa (ibid. vi, pp. 500, 503, 504; vii, pp. 62, 105, 235); Hispania (ibid. vi, p. 66); Gallia (ibid. v, p. 528; vi, pp. 49-50, 80); Francia (ibid. vii, pp. 249, 349); Alamannia (ibid. vii, pp. 248, 346, 377); Pannonia (ibid. v, pp. 193-4, 217 226; vi, 171, 192); the Pannonian city of Siscia (ibid. vi, p. 316); Illyricum (ibid. vi, p. 304); Dacia (ibid. v, pp. 187-188, 269, 361; vi, p. 136, 184) and Sarmatia (ibid. vii, p. 377). Colonial coins are not here included.

^{page 154 note 4} Cohen, , op. cit.² iv, 75-77, 209-210, 275–277Google Scholar.

^{page 154 note 5} Macdonald, G., op. cit. pl. ii, 4–10Google Scholar.

^{page 154 note 6} Cohen, , op. cit.² iv, p. 77Google Scholar, no. 733. Obv. L. SEPT. SEVERVS PIVS AVG. Rev. VICTORIAE BRITANNICAE S.C.

^{page 156 note 1} Cohen, , op. cit.² iv, p. 195Google Scholar, no. 495. Obv. M. AVREL. ANTONINVS PIVS AVG. Rev. PONTIF. TR. p. XIII cos III s.c.; p. 210, no. 639. Obv. M. AVREL ANTONINVS PIVS AVG. Rev. VICTORIAE BRITANNICAE s.c.; p. 277, no. 223, Obv. p. SEPTIMIVS GETA PIVS AVG. BRIT. Rev. VICTORIAE BRITANNICAE. (Plate XXIV, 9, 10).

^{page 156 note 2} It is noteworthy that in this Severus, Caracalla and Geta type the figure of Britannia, if Britannia she be, is on a small scale and is strictly subordinate to the composition as a whole, being no longer the central object of interest as in the second-century types. For the mural crown, cf. the Britannia medallion in a mosaic at Berlin (Jatta, op. cit. fig. 1).

^{page 156 note 3} Cohen, , op. cit.² vii, p. 8Google Scholar, nos. 54-61. Num. Chron. 1907, pp. 305, 306, pl. i, 9 (Plate XXIV, 11). The caduceus which Cohen suggests as the attribute of Britannia on one coin (no. 56) would be an odd attribute for Britain, and it seems probable that it is a standard indifferently rendered. Cohen also describes her as holding a trident in some cases; but in all the British Museum specimens, at any rate, the attribute is certainly a standard.

^{page 156 note 4} Arethuse, Jan. 1924, pl. viii, 5, 6. I am able to reproduce here these medallions from Aretbuse through the kindness of M. Jules Florange.

^{page 156 note 5} Arethuse, loc. cit. pl. vii; Gordon Home, Roman York, p. 79.

^{page 157 note 1} A similar object appears to be worn by Constantius' kneeling Britannia, but it cannot be distinguished very clearly on these smaller medallions.

^{page 157 note 2} The legend round the circumference of the reverse reads REDDITOR LVCIS AETERNAE.