¹ S 72–4. Unless otherwise indicated, S = Suetonius, Tiberius; T = Tacitus, Annales; D = Cassius Dio.

² Gantz, T., Early Greek Myth. A Guide to the Literary and Artistic Sources (1993), 318–28Google Scholar.

³ Iliad 3.237.

⁴ ILLRP 1271a.

⁵ For details and references to the temple and the cult of the twins at Rome: Sihvola, J. in Steinby, E. M. (ed.), Lacus Iuturnae I (1989), 76–109Google Scholar; Poulsen, B., ‘Cult, myth and politics’, in Nielsen, I. and Poulsen, B. (eds), The Temple of Castor and Pollux. The Pre-Augustan Temple Phases with Related Decorative Elements (1992), 46–53Google Scholar; LTUR I (1993), 242–5, ‘Castor, Aedes, Templum’ (I. Nielsen); La Rocca, E., ‘“Memorie di Castore”: principi come Dioscuri’, in Nista, L. (ed.), Castores. L'immagine dei Dioscuri a Roma (1994), 73–90Google Scholar. R. M. Ogilvie argued that the main temple of the cult of Castor and Pollux in Latium was at Tusculum, in the territory of which Lake Regillus lay, and that the vow of a temple to the twins at Rome rewarded not merely their assistance to the Romans but their desertion of the Latins, analogous to an evocatio, the ceremonial request to the enemy's deity to change sides: Ogilvie, R. M., A Commentary on Livy (1965), 288–9, 781Google Scholar. See further below.

⁶ The classic (and only) description of the parade is that of Dionysius of Halicarnassus, at 6.13.4. Cobblers, sutores, near the temple: Pliny, NH 10.121–2. Bankers or moneylenders, argentarii, ‘behind the Temple of Castor’ (‘post aedem Castoris’): CIL VI.9177, 30748, cf. 9393 (= ILS 7696). A cloakmaker, sagarius, likewise ‘behind the Temple of Castor’, CIL VI.9872. Slave-dealers, ‘mancipia ementes vendentesque, ad Castoris’: Seneca, De Constantia Sapientis 13.4, cf. Plautus, Curculio 481. Taverns: Catullus 37, cf. Appian, BC 1.54. The Severan marble plan of the city appears to show a row of shops on the far side of a small piazza at the rear of the temple: http://formaurbis.stanford.edu/fragment.php?record=85 (T. Najbjerg, J. Trimble).

⁷ Briefly: Ginge, B., Becker, M. and Guldager, P., ‘Of Roman extraction’, Archaeology 42.4 (1989), 34–7Google Scholar. Full publication: Bilde, P. Guldager and Poulsen, B., The Temple of Castor and Pollux II.1* The Finds (2008), 253–322Google Scholar. The rich commercial life around the Aedes Castoris and the adjacent Basilica Iulia and Scalae Graecae is vividly evoked by R. Neudecker, ‘Ein göttliches Vergnügen. Zum Einkauf an sakralen Stätten im kaiserzeitlichen Rom’, in Neudecker, R. and Zanker, P. (eds), Lebenswelten. Bilder und Räumer in der römischen Stadt der Kaiserzeit (2005), 81–100Google Scholar.

⁸ References at LTUR, loc. cit. (n. 5).

⁹ The association was first discussed by Scott, K., ‘Drusus, nicknamed “Castor”’, CP 25 (1930), 155–61Google Scholar, and ‘The Dioscuri and the imperial cult’, CP 25 (1930), 379–80Google Scholar; brought up to date by Poulsen, B., ‘The Dioscuri and ruler ideology’, SO 66 (1991), 119–46CrossRefGoogle Scholar (an influential article: cf. the Appendix, below), and La Rocca, op. cit. (n. 5). Most recently, see Suspène, A., ‘Tiberius Claudianus contre Agrippa Postumus: autour de la dédicace du temple des Dioscures’, RPh 75 (2004), 99–124Google Scholar, and Sumi, G. S., ‘Monuments and memory: the Aedes Castoris in the formation of Augustan ideology’, CQ 59 (2009), 167–89CrossRefGoogle Scholar.

¹⁰ Suetonius, Claudius 1.3, cf. Porphyrio ad Hor., Carm. 4.4.27–8, morbo; D 55.1.4, nosoi, cf. 55.2.1; Pliny, NH 7.84, aegrotum; Seneca, Consolatio ad Marciam 3.1, aegrum; Valerius Maximus 5.5.3, gravis et periculosa valitudo.

¹¹ The main accounts of Drusus' death vary somewhat. D 55.1 has him turned back by the female apparition and dying ‘from some disease’ on the other side of the Rhine. (Strabo, a contemporary, mentions in passing, at 7.291, that he died across the Rhine, but gives no cause.) Suetonius, Claudius 1.2–3, seems to date the apparition earlier, to 11 b.c., and has Drusus die ‘from disease in his summer camp’ at a place thereafter called ‘Scelerata’, ‘Accursed’. Livy, Periochae 142 has him die, apparently across the Rhine, thirty days after his leg was broken by his horse falling on it.

The significance of Drusus' death is emphasized by the female apparition. She represents a whole range of such figures who warn conquerors and explorers that certain boundaries are set and if those boundaries are crossed, such hubris arouses the envy of supernatural powers: Krappe, A. H., ‘Der Tod des Drusus’, Zeitschrift für Deutsches Altertum und Deutsche Literatur 75 (1930), 290–6Google Scholar.

¹² Valerius Maximus 5.5.3 (Loeb translation by D. R. Shackleton Bailey, slightly modified). Valuable analysis at Wardle, D., ‘The heroism and heroisation of Tiberius: Valerius Maximus and his emperor’, in Defosse, P. (ed.), Hommages à Carl Deroux 2 (2002), 433–40Google Scholar.

¹³ Various details at Livy, Per. 142; Valerius Maximus 5.5.3; Pliny, NH 7.84; D 55.2.1; Consolatio ad Liviam 89–94; Seneca, Consolatio ad Polybium 15.5. The first four of these emphasize the haste of Tiberius' journey: beyond that, how much of the story is fact and how much embroidery is unknown.

¹⁴ S insists at 7.3 that Tiberius walked all the way with the body. But Mainz to Rome is about 800 miles by modern highways with bridges, viaducts, and tunnels, none of which was available in 9 b.c. Tiberius had also to reckon with an additional unknown distance beyond the Rhine, the heights of the Alps, and an extremely harsh winter (T 3.5.1). An epic journey indeed.

‘Municipiorum coloniarumque primores’: Suetonius, Claudius 1.3; D 55.2.1. Cf. Seneca, Consolatio ad Marciam 3.1: ‘ingens civium provinciarumque et totius Italiae desiderium, per quam effusis in officium lugubre municipiis coloniisque usque in urbem ductum erat funus triumpho simillimum.’ And Consolatio ad Liviam 169ff., at 173: ‘funera ducuntur Romana per oppida Drusi.’ At Rome the order of the scribes received the body and brought it to the Forum where Tiberius delivered the eulogy; thence it was conveyed to the Circus Flaminius, where Augustus delivered a second eulogy. From there the knights carried the corpse to the Campus Martius, where it was burned at the Ustrinum, and the ashes were buried in the Mausoleum. Drusus was posthumously given the name Germanicus, and awarded statues, an arch, and a cenotaph on the Rhine. On all of this, see the thorough commentary of Swan, P. M., The Augustan Succession: An Historical Commentary on Cassius Dio's Roman History Books 55–56 (9 B.C.–A.D. 14) (2004), 44–7Google Scholar, with references.

¹⁵ References (more have accumulated for each): PIR ² C 857 (Drusus), A 885 (Antonia), I 221 (Germanicus), C 942 (Claudius). Nero Claudius Drusus Germanicus Imp. appears often on the coins of his son Claudius: RIC Claudius 69–74, 93, 98, et al. Livia as mother of Drusus Germanicus: named Livia Drusi f. Augusti (before a.d. 14), CIL IX.3304 (Superaequum); and Iulia Augusta (after a.d. 14), CIL II.2038 (Anticaria: ‘mater Ti. Caesaris principis et conservatoris et Drusi Germanici genetrix orbis’) and XI.7416 (Ferentium). In one inscription, CIL XI.1165 (Veleia) she is, tortuously, Iulia Augusta, daughter of the divine Augustus, mother of Ti. Caesar Augustus, son of the divine Augustus, and of Nero Claudius Drusus (no mention of Germanicus).

¹⁶ Consolatio ad Marciam 3.2, Loeb translation by J. W. Basore: ‘Non desiit denique Drusi sui celebrare nomen, ubique illum sibi privatim publiceque representare, libentissime de illo loqui, de illo audire; cum memoria illius vixit.’ Probably to be dated to the years a.d. 33/37: Bellemore, J., ‘The dating of Seneca's Ad Marciam de Consolatione’, CQ 42 (1992), 219–34CrossRefGoogle Scholar.

¹⁷ Valerius Maximus 4.3.3; Velleius Paterculus 2.97.2–3; Seneca, Consolatio ad Marciam 3.1, cf. Consolatio ad Polybium 15.5 (Loeb translations by D. R. Shackleton Bailey, F. W. Shipley, and J. W. Basore, respectively).

¹⁸ Consolatio ad Liviam: whatever the origins of this poem, there is no reason to doubt that it is exactly what it claims to be, and that it was produced sometime between the dedication of the Temple of Castor and Pollux in a.d. 6 and Livia's death in a.d. 29; cf. NP, s.v. Compare the poetic effusions of Clutorius Priscus, one on the death of Germanicus, for which he was rewarded by Tiberius, and one anticipating the death of Drusus Caesar, for which the Senate had him executed: T 3.49–51; D 57.20.3.

Pliny's dream: Pliny, Epp. 3.5.4, with the remarks of Flower, H. I., The Art of Forgetting. Disgrace and Oblivion in Roman Political Culture (2006), 3–5Google Scholar. Tacitus perhaps recalls Pliny's Bella Germania when he refers to the German deeds of ‘Drusus Germanicus’ in his own Germania (34.3, on Drusus' daring) and his Historiae (5.19).

¹⁹ Drusus Germanicus on public inscriptions: notably CIL VI.40329 = ILS 148 (Rome, from the Campus Martius); 40330 (Rome: his elogium in the Forum of Augustus); 40337; 40339 (Rome: the dedication of the Temple of Castor and Pollux, on which see below); 40424 (Rome, possibly from the Ara Pietatis Augustae); AE 1962.37 (Saepinum); and the fragment of fasti at AE 1981.316 (Hispellum). The honours of Germanicus: the third mention of his father noted in the text above appears in the Tabula Hebana, recording the Lex Valeria Aurelia of a.d. 20, the first two in the partially overlapping Tabula Siarensis, which records the decree of the Senate that preceded that law. The bibliography on these and several related fragments is enormous, their reconstruction fiendishly complicated: easily the best place to start is Crawford, M. H. (ed.), Roman Statutes (1996), 1.507–47Google Scholar.

The memory of Drusus may also dominate the scenes on two well-known silver cups from Boscoreale: de Caprariis, F., ‘Druso, Giove Feretrio e le coppe “imperiale” di Boscoreale’, MEFR 114 (2002), 713–37CrossRefGoogle Scholar, arguing (vs. A. L. Kuttner, Dynasty and Empire in the Age of Augustus. The Case of the Boscoreale Cups (1995)) that they reflect well-known images generated around the dead Drusus rather than a specific (and otherwise unattested) public monument erected to commemorate Tiberius' victory.

For some of the many posthumous portraits of Drusus: Boschung, D., ‘Die Bildnistypen der iulisch-claudische Kaiserfamilie’, JRA 6 (1993), 51Google Scholar; Rose, C. B., Dynastic Commemoration and Imperial Portraiture in the Julio-Claudian Period (1997), 83, 90, 100, 108, 110, 153Google Scholar; Megow, W. R., Kameen von Augustus bis Alexander Severus (1987), 180, 204, 276Google Scholar. Most intriguing is a veiled bust ‘Found in Capri’ and acquired by the British Museum in the nineteenth century, whose face shows ‘a certain boyishness, in spite of his obvious maturity’ (Pollini, J., ‘Gnaeus Domitius Ahenobarbus and the Ravenna Relief’, RM 88 (1981), 129Google Scholar, with Tafeln 38 and 39, dating the portrait to ‘the later Julio-Claudian period’, 130), and who is clearly the same person as a togate figure from Caere (ibid., Tafeln 37 and 38). The man's identity is contested, but a convincing case for Drusus is offered at Rose, 63–4 with n. 75.

²⁰ T 1.33.2, translated by A. J. Woodman; D 55.27.3. Tacitus reverts to the theme of Drusus' popularity at 2.41.3 and 6.51.1. The obvious bears re-stating, that all of his posthumous honours were for a man who did not actually conquer Germany, although everyone agreed that it was inevitable: cf. Strabo 7.1.3; Porphyrio on Horace, Epp. 1.3 pr (‘Drusus qui subactis Germanis Germanicus dictus est’).

²¹ Consolatio 169–72; Seneca, Consolatio ad Polybium 15.5: ‘totum exercitum non solum maestum sed etiam attonitum corpus Drusi sui (sc. Tiberii) sibi vindicantem.’ Cenotaph: D 55.2.3; Suetonius, Claudius 1.3, ‘honorarium tumulum’. The tumulus was almost surely mentioned in the Tabula Siarensis a 26–8, in which, if we accept W. D. Lebek's ingenious restoration of the Latin, we can see the compromise as the Senate decreed in honour of Germanicus ‘that a third arch either [be built onto or be placed near that burial mound] which [the army had first begun to construct on its own initiative] for Drusus, the brother of Tiberius Caesar Augustus, and then [completed] with the consent of the Divine Augustus’: Crawford, op. cit. (n. 19), 1.515. Claudius ‘a Germanicus’: Josephus, AJ 19.217.

²² Claudius 1.3, Tiberius 50.1. Suetonius uses the letter to introduce a grossly unfair section of the biography of Tiberius devoted to his supposed hatred of his close relatives, suggesting that Tiberius betrayed his brother's words. But this is the only instance given of such antipathy, against many displays of fraternal affection; the incident is not dated, and the Republican Tiberius might well have produced the letter after the death of Augustus, as part of his own resistance to taking up the principate; and Suetonius contradicts himself, for it can hardly have been a betrayal if Drusus himself made no secret of his intentions.

²³ Suetonius, Claudius 1.5. The fragmentary inscription from the Forum Augustum, published in 1933, is now CIL VI.40330: ‘[Nero] Cl[a]udiu[s] Ti(beri) f(ilius) / [Dru]sus German[i]cus / [co(n)s(ul)] pr(aetor) urb(anus) q(uaestor) aug(ur) imp(erator) / [app]ellatus in Germania’. Pliny the Elder informs us that Augustus himself composed the elogia in his Forum: NH 22.13.

²⁴ Memorials: inscriptions in Mausoleum and Forum Augustum; eulogy; memoir; permission for tumulus on the Rhine (Tabula Siarensis). Eulogy overshadows that delivered by Tiberius: Suetonius recalls Augustus' emotional plea to the gods, but does not mention the speech of Tiberius; while in its narrative of events, the Consolatio, whose author claims to have been there, likewise recalls Caesar's tearful laudation and his dramatic plea to the gods for a similar death (209–12, cf. 464–5), and completely ignores Tiberius' speech.

²⁵ T 3.5.1: ‘circumfusas lecto Claudiorum Iuliorumque imagines.’

Burial in the Mausoleum: von Hesberg, H. and Panciera, S., Das Mausoleum von Augustus. Der Bau und seine Inschriften (1994), 74–5Google Scholar. None of the earlier burials in the tomb was in fact a member of the Julian family, but Agrippa was at least the natural father of Augustus' two sons by adoption, Marcellus was his nephew, and Octavia was his sister; whereas Drusus was only his stepson and the husband of his niece. Tumulus Iuliorum: T 16.2; cf. Livy, Periochae 142, Drusus buried ‘in tumulo C. Iuli’. Consolatio 161–3: ‘Quod licet, hoc certe: tumulo ponemur in uno, / Druse, neque ad veteres conditus ibis avos. / Miscebor cinerique cinis atque ossibus ossa.’ Livia comforts herself: ‘This at least is possible — in this tomb shall we be laid together, Drusus, nor buried shalt thou go to the sires of old; I shall be mingled with thee, ashes with ashes, bone with bone.’ (Loeb translation by J. H. Mozley.)

The position of the statue in the Forum Augusti is assured by the location of the fragments of its inscription: Spannagel, M., Exemplaria Principis. Untersuchungen zu Enstehung und Ausstattung des Augustusforums (1999), 288–91Google Scholar.

Ovid, Fasti 1.707: ‘fratres de gente deorum.’ Valerius Maximus 5.5.3: ‘fraternum iugum, Claudiae prius, nunc etiam Iuliae gentis … decus.’

²⁶ D 55.8.1–2 (Loeb translation by E. Cary, modified): Τιβέριος δὲ ἐν τῇ νουμηνίᾳ ἐν ᾗ ὑπατεύειν μετὰ Γναίου Πίσωνος ἤρξατο, ἔς τε τὸ Ὀκταουίειον τὴν βουλὴν ἤθροισε διὰ τὸ ἔξω τοῦ πωμηρίου αὐτὸ εἶναι, καὶ τὸ Ὁμονόειον αὐτὸς ἑαυτῷ ἐπισκευάσαι προστάξας, ὅπως τό τε ἴδιον καὶ τὸ τοῦ Δρούσου ὄνομα αὐτῷ ἐπιγράψῃ, τά τε νικητήρια ἤγαγε καὶ τὸ τεμένισμα τὸ Λίουιον ὠνομασμένον καθιέρωσε μετὰ τῆς μητρός.

The invaluable commentary of P. M. Swan makes annotation superfluous, op. cit. (n. 14), 71–5. On the complex connotations of Concord at Rome, see especially Levick, B., ‘Concordia at Rome’, in Carson, R. A. G. and Kraay, C. M. (eds), Scripta Nummaria Romana, Essays Presented to Humphrey Sutherland (1978), 217–33Google Scholar.

²⁷ 55.9.6, with Swan, op. cit. (n. 26). This act of imperial brutality sits ill with Tiberius' later professed concern for the provincials, and he would be the first to recognize the irony of its connection with ‘concord’. However we explain it, two banal observations are valid: retired or not, he had not forgotten the Temple of Concord; and its importance to him trumped common morality.

²⁸ Castor and Pollux: D 55.27.3–4, Loeb translation by E. Cary, modified; full Greek text below at n. 56. Dio gives the year; the day appears in the Fasti Praenestini and at Ovid, Fasti 1.705.

Concord: D 56.25.1: τῷ δὲ δευτέρῳ τά τε ἄλλα τὰ προειρημένα ἐγένετο, καὶ τὸ Ὁμονόειον ὑπὸ τοῦ Τιβερίου καθιερώθη, καὶ αὐτῷ τό τε ἐκείνου ὄνομα καὶ τὸ τοῦ Δρούσου τοῦ ἀδελφοῦ αὐτοῦ καὶ τεθνηκότος ἐπεγράφη. Dio and the FP give the year; the FP, the Fasti Verulani, and Ovid the day, Ovid at Fasti 1.637ff.

²⁹ Good brief introductions to the two temples at LTUR I (1993), 316–20, ‘Concordia, Aedes’ (A. M. Ferroni), and 242–5, ‘Castor, Aedes, Templum’ (I. Nielsen). A replacement for the standard monograph on Concordia (C. Gasparri, Aedes Concordiae Augustae (1979)), announced in LTUR I as in preparation, has not yet appeared; the Augustan Temple of Castor on the other hand is now thoroughly treated in the sumptuous volumes of P. Guldager Bilde and B. Poulsen, op. cit. (n. 7), in four parts, and of S. Sande and J. Zahle (eds), The Temple of Castor and Pollux III. The Augustan Temple (2008).

Gleaming: Ovid on the niveum templum of candida Concordia, Fasti 1.637. So much is known of the artwork in the Temple of Concord (mainly from Pliny) that a programme has been discerned, no mere museum collection but a symbolic paean to the values proclaimed by the dynasty: Kellum, B., ‘The city adorned: programmatic display at the Aedes Concordiae Augustae’, in Raaflaub, K. A. and Toher, M. (eds), Between Republic and Empire: Interpretations of Augustus and his Principate (1993), 276–307Google Scholar; Bravi, A., ‘Tiberio e la collezione di opere d'arte dell’ Aedes Concordiae Augustae', Xenia Antiqua 7 (1998), 41–82Google Scholar (approved by Hölscher, T., ‘Greek styles and Greek art in Augustan Rome: issues of the present versus records of the past’, in Porter, J. I. (ed.), Classical Pasts. The Classical Traditions of Greece and Rome (2006), 253–4Google Scholar, promising to return to the subject soon). These interpretations may seem over-determined to some readers, and curiously neither mentions the posthumous equestrian statues in the Temple of Concord of Gaius and Lucius Caesar, of Germanicus, and probably of Drusus Caesar, which would have been hard to miss: Tabula Siarensis b 1–12, interpreted by Heinemann, A., ‘Eine Archäologie des Störfalls. Die toten Söhne des Kaisers in der Öffentlichkeit des frühen Prinzipats’, in Hölscher, F. and Hölscher, T. (eds), Römische Bilderwelten. Von der Wirklichkeit zum Bild und zurück (2007), 90–3Google Scholar. German spoils: S 20, discussing Tiberius' German triumph in a.d. 12, adds retroactively, ‘He also dedicated the Temple of Concord and that of Pollux and Castor in his own name and that of his brother, from the spoils’, ‘dedicavit et Concordiae aedem, item Pollucis et Castoris suo fratrisque nomine de manubiis’. Ovid confirms that ‘munera triumphatae gentis’ paid for Concord: Fasti 1.647–8.

³⁰ On the date of the vows, see below.

³¹ Fires: LTUR I (1993) as above n. 29, 317 (Concordia), 244 (Castor: correctly adding ‘although such a destruction is not specifically mentioned in the sources’); cf. Poulsen, op. cit. (n. 9), 121. Repeated without question at Haselberger, L. (ed.), Mapping Augustan Rome (2002), 97, 83Google Scholar.

The only source for these fires, and another in 12 b.c., is Dio, who gives examples of buildings damaged in each case. 14 b.c., Basilica Paulli and Temple of Vesta, burned in the same fire, 54.24.2. 12 b.c., portents of the death of Agrippa: ‘many (buildings)’ burned, including the Hut of Romulus, which was set alight by crows dropping on it burning meat from an altar: 54.29.8. 9 b.c., portents of the death of Drusus: ‘many (buildings)’ ruined or destroyed by a storm and lightning, and many temples, and the Temple of Jupiter Capitolinus and associated deities harmed: 55.1.1. How could Dio have failed to name Concord and Castor in such reports; and how could he have failed to mention such damage as the reason for Tiberius' rebuilding of them?

Even more striking, the Consolation to Livia carefully reminds Livia at lines 401–4 of the damage done by lightning to Capitoline Jupiter, Juno, and Minerva (along with the sancta domus Caesaris, on the Palatine) as portending the death of Drusus. How could the poet, who mentions the Temples of Concord and Castor at lines 283–90, possibly have ignored any damage done to them in 9 b.c. and repaired by Drusus' brother?

³² CIL VI.40339 (Rome). Brackets [] indicate portions that have been lost; parentheses () the expansion of standard epigraphic abbreviations.

³³ Alföldy, G., Studi sull'epigrafia augustea e tiberiana di Roma, Vetera 8 (1992), 39–58Google Scholar.

³⁴ Sande and Zahle, op. cit. (n. 29), 179–80, 183.

³⁵ As in notes 37 and 41 below.

³⁶ The following sketch is heavily dependent on Zanker, P., Forum Romanum: die Neugestaltung unter Augustus (1972)Google Scholar. For the arch of Gaius Caesar, see Rose, C. B., ‘The Parthians in Augustan Rome’, AJA 109 (2005), 58–64CrossRefGoogle Scholar.

³⁷ It is inconceivable that the filiation of Tiberius (‘son of Augustus’ in this case) would be omitted from a public monument. That Drusus also was named as privignus Augusti, ‘stepson of Augustus’ must remain a hypothesis, but is extremely probable. Alföldy, op. cit. (n. 33), 51, adduces ILS 148 (Rome), an exact parallel for the name and relationship restored here, and AE 1981.316 (Hispellum): both were highly visible public documents. The inclusion of Augustus, who loved Drusus as a son, would be a gracious gesture by Tiberius, his exclusion hard to imagine. For what it is worth, the double appearance of Augustus in the two filiations is nicely balanced visually, and the lettering fits perfectly within Alföldy's careful and elegant reconstruction of the text.

³⁸ Tiberius might even claim an ancestral connection with the twins, for his distant ancestor Appius Claudius Sabinus, consul in 495 b.c., had immigrated from the otherwise unknown town of Regillum, which was probably in the territory of Tusculum, the main site of the cult of Castor and Pollux in Latium, and presumably near the equally unknown site of the Battle of Lake Regillus. Note also that as Princeps Tiberius had a villa at Tusculum (CIL XV.7814), where he certainly stayed in a.d. 34 and 36, and that the imperial cult there became entwined with that of Castor and Pollux (CIL XIV.2620, 2630).

³⁹ Horace 4.4.73–5: ‘nil Claudiae non perficiunt manus, / quas et benigno numine Iuppiter defendit’. 37–8: ‘quid debeas, o Roma, Neronibus, / testis Metaurum flumen.’

⁴⁰ S 20: text in n. 29 above.

⁴¹ By sheer chance, or divine intervention, one of the six fragments used by Alföldy in the reconstruction at CIL VI.40339 bears the remains of two letters which seemed to confirm that the dedicatory inscription did indeed identify the building as the Temple of Pollux and Castor. Viz., frag. d, (certainly from the temple) to be restored: ‘[aedem Pollucis e]t C[astoris]’. This was first observed by G. Tomasetti in 1890: Alföldy, op. cit. (n. 33), 48–9, with discussion. The size of the letters assigns them to the third line of the text in his version. There is an interpunct on the stone, a small triangle marking the division between words. The first word ended with T, the second, only the top of which is preserved, began with a C or possibly G, O, or Q. From this point, the possibilities for words, combinations of words, and abbreviations on such conventional public inscriptions is severely limited (Alföldy duly cites parallels): the words aedem, Castoris, et, Pollucis, de and manubiis are all but assured in some order, and [r]ef(ecerunt) is preserved.

However, Sande sees traces of relief on the left side of the fragment (her ARC 19), two little drops that are man-made, and asserts that they represent decoration appropriate to the end of the architrave, hence that the letters belong not to the middle of the line but to its beginning. ‘T.’ will then represent not ‘[e]t.’ but ‘T(emplum) C[astorum/is]’. As reconstructed, this is (apparently) the beginning of the dedicatory inscription. Alföldy's two shallow columns above this, his third line, are thus rejected along with the blocks on which their fragmentary letters appear.

Readers need not be warned that ‘T. Castoris/um’ cannot be. To begin the dedication of a major monument with a one-letter abbreviation would be unnecessary and inelegant, indeed grotesque, and no parallels are cited. To squeeze the proper names, let alone the inevitable titulatures of the two dedicators into what Sande assumes to be space for approximately forty-one letters between ‘C[astorum/is]’ and ‘ref(ecerunt)’ would be impossible however abbreviated, and no parallels are cited. The temple, commonly known as ‘Aedes’, was indeed called ‘Templum’ on occasion, as Sande points out, but such occasions are all literary: on inscriptions it is invariably ‘Aedes’. The existence and significance of Sande's traces of decoration on the stone will have to be decided by experts. For the present: non liquet. If the preserved letters ‘t.c(?)’ do not represent the words ‘et Castoris’, which we may deduce from Suetonius to have been used in the text, it is impossible to say what they signify.

The historian Florus, Suetonius' contemporary, has ‘youths’ turn up with laurelled letters announcing victory over the Cimbri in 101 b.c., which they deliver to the ‘praetor pro aede Pollucis et Castoris’ (1.38.20). This, the only other reference to the Temple of Pollux and Castor as such, surely reflects the Tiberian inscription. Poets might invert the normal order of the pair, but it is hard to explain why the prosaic Suetonius and Florus would do so — unless they saw it on the temple.

⁴² Suetonius, Divus Iulius 10.1; D 37.8.2.

⁴³ G. D. Hadzsits' careful assembly and analysis of the complex evidence is essential on all of this: ‘History of the name of the Temple of Castor in the Forum’, in Hadzsits, G. D. (ed.), Classical Studies in Honor of John. C. Rolfe (1931), 101–14Google Scholar. Though not much regarded by subsequent scholarship, it nevertheless strikes me as correct, however improbable the results may first appear. Hadzsits concluded (113): ‘that the temple was at first and for long, Castor's; that, once Greek mythologies were widely diffused, it was possible in popular parlance to think of it and speak of it as the shrine of Castor and Pollux; that Pollux did become associated with Castor in worship – to what extent, exactly, we cannot tell, nor precisely when, though it would seem that this was an established fact in the second century b.c.; that the Greeks inevitably called it the shrine of Castor and Pollux, regardless of dates; that it was officially rechristened by Tiberius (before he became Princeps) as “the temple of Castor and Pollux,” or as the “temple of the Castors”.’

The crucial witness is Livy. The standard legend of the battle of Lake Regillus, as recounted above (and conveyed by such authors as Cicero, ND 2.6, 3.11–13; Dionysius 6.13.1; Florus 1.5.4; Plutarch, Coriolanus 3.4; Valerius Maximus 1.8.1), is a much later invention, as all would agree. But in his account of the battle (2.20.12), Livy says only that at a critical stage the dictator vowed a temple to Castor, which was later dedicated by his son (2.42.5): no divine epiphany, no Pollux, cf. Hadzsits, 101–5. (Ogilvie, op. cit. (n. 5), 289, avoids the problem.) Cults and sites dedicated to one or other of the brothers alone: E. Bethe, PW 5.1090, s.v. Dioskuren; note especially Pausanias 3.13.1, 20.1. A rock near Lake Regillus was said to preserve the hoof mark of Castor's horse: Cicero, ND 3.11–12.

Greek authors refer to ton Dioskouron hieron, Dioskoreion, and naos ton Dioskouron. These last designations are clearly anachronistic when referring to the Republic, and they uniformly mislead, since the term ‘Dioskouroi’, so natural to the Greeks to designate the inseparable Castor-and-Pollux, does not translate any Latin equivalent: the word ‘Dioscuri’ seems never to appear in classical Latin, whether literary or epigraphical, certainly not in an alternate name for the Temple of Castor and Pollux. Cf. Hadzsits, 105–6, 110–11. Note particularly Cicero, ND 3.53, where the word is left in Greek, and Augustus, RG 20.3, where the Temple of Castor, aedem Castorem, of the Latin original is translated in the Greek version as the Temple of the Dioscuri, τοῦ ναοῦ τῶν Διοσκόρων.

The Roman plural for the brothers, hardly equivalent to the neutral ‘Dioskuroi’, was the unbalanced ‘Castores’. References to the twins as such are rare in literature, the earliest being the elder Pliny (NH 1.2a, 7.86, 10.121 (temple), 34.23 (temple), 35.27). Their temple is called aedes Castorum only by Pliny (twice) and by the fourth-century Historia Augusta (twice) and Notitia of the city; and although the Castores turn up in inscriptions, their temple does not.

As far as I am aware, aedes Castoris and aedes Castorum refer only to the temple in Rome, that is, shrines to the twin gods elsewhere always mention Castor and Pollux.

⁴⁴ For ancient references to the temple, see Hadzsits, op. cit. (n. 43) and, conveniently, LTUR I (1993), 242–5. Add, for aedes Castoris, the important I. de Delos 1511 (a senatusconsultum of 58 b.c.); and for aedes Castoris et Pollucis, CIL VI.2203. For the widespread phenomenon of twins being designated by one name: Harris, J. R., The Cult of the Heavenly Twins (1906), 58–62Google Scholar.

⁴⁵ Propertius 3.14.17; Ovid, Amores 2.16.13. Thereafter at Seneca, NQ 1.1.13 and Pliny, NH 2.101, both referring, as does Ovid, to the brothers in their rôle as stars. And thereafter only in Suetonius and Florus (as above), referring to the temple.

⁴⁶ CIL IX.2443 = ILS 147. The inscription has had a remarkably confusing history: Bernecker, A., ‘Zur Tiberius-Inschrift von Saepinum’, Chiron 6 (1976), 185–92Google Scholar; Stylow, A. U., ‘Noch einmal zu der Tiberius-Inschrift von Saepinum’, Chiron 7 (1977), 487–91Google Scholar. The composite text presented here is that of Stylow's definitive reconstruction. Note also AE 1991.530, a dedication to Tiberius from a local magistrate at Saepinum in 3/2 b.c.

⁴⁷ The fantasy of the dead Drusus engaged in public works with his living brother, thus attested on the Temple of Pollux and Castor and the gates of Saepinum, reappears on CIL VI.40337, the dedication of an unknown building between a.d. 4 and 14: the verb is lost, but again there seems to be no indication that Drusus was not alive. It was also presumably repeated on the inscription on the Temple of Concord in a.d. 10.

Up until Tiberius, temples vowed by one Republican nobleman, normally because of military victories, were often completed after a lapse of time and dedicated by another, usually a son or other relative: e.g. Honos et Virtus, vowed by the great Claudius Marcellus and dedicated by his son in 205 b.c., Livy 29.11.13; or the original Temple of Castor, dedicated by the dictator's son, Livy as above; or the Temple of the Lares Permarini, vowed by L. Aemilius Regillus in 190 b.c. and dedicated by his clansman M. Aemilius Lepidus in 170 b.c., complete with a long eulogy on a tablet over the door listing the man's deeds and his vow, Livy 40.52.4. But I can find no instance before Drusus of a temple at Rome erected by a dead man, and no case where the dedicator's death is ignored.

⁴⁸ Polydeukes/Pollux the elder brother: Theocritus, Idylls 22.176, 183. It would be pedantic to observe that Tiberius and Drusus were not actually twins, for Tiberius was comparing, not identifying, the two pairs. On the dissimilarity of Castor and Pollux: Harris, op. cit. (n. 43), 45–8.

⁴⁹ AE 1963.104. Bibliography reviewed and argument presented: Alföldy, G., ‘Pontius Pilatus und das Tiberieum von Caesarea Maritima’, SCI 18 (1999), 85–108Google Scholar (whence AE 1999.1681). Further arguments: idem, ‘Nochmals: Pontius Pilatus und das Tiberieum von Caesarea Maritima’, SCI 21 (2002), 133–48Google Scholar (AE 2002.1556). Conclusions summarized: idem, ‘Zwei römische Statthalter im Evangelium: die epigraphischen Quellen’, in dal Covolo, E. and Fusco, R. (eds), Il contributo delle scienze storiche allo studio del Nuovo Testamento (2005), 226–36Google Scholar (AE 2005.1583).

⁵⁰ Josephus, BJ 1.412 (Loeb translation by H. St. J. Thackeray): τοῦτο δὲ πύργοις τε διείληπται μεγίστοις, ὧν ὁ προύχων καὶ περικαλλέστατος ἀπὸ τοῦ Καίσαρος προγόνου Δρούσιον κέκληται. AJ 15.336 (Loeb translation by R. Marcus): τὸ δὲ λοιπὸν περιεῖχεν λίθινον τεῖχος πύργοις διειλημμένον, ὧν ὁ μέγιστος Δρούσιον ὀνομάζεται, πάνυ καλόν τι χρῆμα, τὴν προσηγορίαν εἰληφὼς ἀπὸ Δρούσου τοῦ Καίσαρος προγόνου τελευτήσαντος νέου.

⁵¹ Josephus, BJ 1.414, with Alföldy, op. cit. (n. 49, 1999), 96–101. Must be correct: the words ‘Tiberieum’ on the inscription and ‘Druseum’ in Josephus look to be unique in literature and epigraphy, both Greek and Latin; they are structures recalling two prominent and famously close brothers; they are essentially contemporary (note that Pilate's work was a refurbishment); and of all the cities in the Roman world, they come from the same one. Coincidence is unthinkable: how could the Tiberieum be anything but a pendant to the Druseum?

⁵² Alföldy, op. cit. (n. 49, 2002), 148, citing Strabo 17.1.6 and Lucian, Quomodo 62. Note also a ship from Alexandria named for the Dioscuri: Acts 28.11.

⁵³ Work on Herod's harbour had certainly begun long before, and the pendant towers of the Druseum and the Tiberieum might be an afterthought, but 4 b.c. appears to be the latest possible date. It is true that Josephus mentions only the Druseum, and the Tiberieum is only certified as existing by a.d. 26/36, but again it is hard to conceive the one being built without the other and, again, Pilate's task in a.d. 26/36 was to restore a previously existing structure.

⁵⁴ It has been argued, and is commonly believed, that Augustus' sons Gaius and Lucius Caesar preceded Tiberius and Drusus as Castor and Pollux. There is no evidence for this: see the Appendix below.

⁵⁵ Considered further below.

⁵⁶ Ovid, Fasti 1.705–8, Penguin translation by A. J. Boyle and R. D. Woodard; Consolatio ad Liviam 283–90, Loeb translation by J. H. Mozley; Valerius Maximus 5.5.3 (translation above).

⁵⁷ Despite the forbidding reputation passed on by our three major sources, there is good evidence for Tiberius' exceptional popularity with the people of Rome, some of which is considered in Champlin, E., ‘Tiberius the Wise’, Historia 57 (2008), 408–25Google Scholar.

⁵⁸ 55.27.3–4. Loeb translation by E. Cary, modified: μέχρις οὗ ἥ τε σιτοδεία ἐπαύσατο, καὶ μονομαχίας ἀγῶνες ἐπὶ τῷ Δρούσῳ πρός τε τοῦ Γερμανικοῦ τοῦ Καίσαρος καὶ πρὸς Τιβερίου Κλαυδίου Νέρωνος, τῶν υἱέων αὐτοῦ, ἐγένοντο. τοῦτό τε γὰρ αὐτοὺς ἐπὶ τῇ τοῦ Δρούσου μνήμῃ παρεμυθήσατο, καὶ ὅτι τὸ Διοσκόρειον ὁ Τιβέριος καθιερώσας οὐ τὸ ἑαυτοῦ μόνον ὄνομα αὐτῷ, Κλαυδιανὸν ἑαυτὸν ἀντὶ τοῦ Κλαυδίου διὰ τὴν ἐς τὸ τοῦ Αὐγούστου γένος ἐκποίησιν ὀνομάσας, ἀλλὰ καὶ τὸ ἐκείνου ἐπέγραψε.

⁵⁹ Germanicus and Drusus Caesar almost the same age: Sumner, G. V., ‘Germanicus and Drusus Caesar’, Latomus 26 (1967), 413–35Google Scholar, convincingly argued for birthdates of 24 May 15 b.c. and 7 October 14 b.c. Ovid, Tristia 2.167; Ex Ponto 2.81–4: ‘Quem pia vobiscum proles comitavit euntem, / digna parente suo nominibusque datis, / fratribus adsimiles, quos proxima templa tenentis / divus ab excelsa Iulius aede videt.’ The implied emphasis here, hard to catch in translation, is on four generations of Caesars.

A few years later, during his consulship in a.d. 15, Drusus Caesar was given the nickname of Castor after he struck a distinguished member of the equestrian order, almost certainly the praetorian prefect Sejanus himself: D 57.14.9, 22.1; T 4.3.2: on which see Scott, op. cit. (n. 9). Who gave him that nickname and how widespread it was is not known, but it is very curious, and not immediately explicable: Pollux was the boxer, not Castor.

I find no evidence that Drusus was Princeps iuventutis, as claimed at Poulsen, op. cit. (n. 9), 128. While Germanicus and Drusus were depicted as loving brothers, there seems to be no official representation of them as Castor and Pollux: but see below. One of the most attractive portrayals of their affection is a fine coin issued after their deaths by the koinon of Asia, showing them sitting, togate, in curule chairs, and calling them ‘Drusus and Germanicus Caesar the new gods of brotherly love’, neoi theoi philadelphoi. The reference is to a cult of the old Attalid dynasty of Pergamum, paid to dead kings or their dead relatives: Levy, B., ‘The date of Asinius Pollio's proconsulship’, JNG 44 (1994), 79–89Google Scholar. Its application to Tiberius' dead sons is particularly neat.

⁶⁰ The birth: I.It. 13, 1.216; T 2.84.1 (dating it to a.d. 19, and adding the usual nasty Tacitean comment, ‘for he turned everything, however chance, to glory’). The coins: RIC 32 (Rome). The complex of symbols — crossed cornucopiae, heads (the children assure the future), and caduceus (the wand of Mercury) — came to represent felicitas temporum, the good fortune of the age: for date, parallels, and precedents, see Meise, E., ‘Der Sesterz des Drusus mit den Zwillingen und die Nachfolgepläne des Tiberius’, JNG 16 (1966), 12–14Google Scholar.

⁶¹ Coins: RPC 946 (Cyrenaica: ‘Tiberius and Germanicus Caesar’, with portraits on the reverse, ‘Drusus Caesar son of (Tiberius) Augustus’, with portrait, on the obverse); RPC 1171 (Corinth: ‘Twin Caesars’, with facing busts). Near Salamis: IGRR 3.997. Ephesus: IK Ephesos VII.2.4337. All of this material is presented at Poulsen, op. cit. (n. 9), 128–9. (Pace Poulsen, portraits on glass medallions from the north-western provinces do not appear to represent the infant twins: Boschung, D., ‘Römische Glasphalerae mit Porträtbüsten’, BJ 187 (1987), 193–258Google Scholar.)

⁶² Death of Germanicus: T 4.5.1. Tiberius' nickname of The Twin seems to be attested only at Josephus, AJ 18.206 (explicitly) and on the papyrus BGU 156.6.

⁶³ Pliny NH 10.121–2, Penguin translation, J. F. Healey, substantially modified.

Pliny appends his usual moralizing comment, 10.123: ‘The Roman people considered the bird's talent a sufficiently good reason for a funeral procession and for the punishment of a Roman citizen. Yet in Rome many leading men had no funeral rites at all, while no one avenged the death of Scipio Aemilianus after he had destroyed Carthage and Numantia.’

⁶⁴ Compare the popular reaction to a false rumour spread at Rome in a.d. 19 that Germanicus had recovered from what was to prove his final illness at Antioch: ‘a general rush was made from every side to the Capitol with torches and victims, and the temple gates were all but torn off, that nothing might hinder them in their eagerness to pay their vows. Tiberius was roused from sleep by the cries of the rejoicing throng, who all united in singing: safe is Rome, safe too our country, for Germanicus is safe.’ (Suetonius, Caligula 6.1, Loeb translation by J. C. Rolfe).

Whether the raven incident is fact or folklore is moot. Macrobius relates two anecdotes about ravens trained to salute Augustus as imperator, the second of them by a poor shoemaker, indeed Augustus has a houseful of such avian salutatores: Saturnalia 2.4.30. There are also echoes of our tale at Plutarch, Moralia 973B–D.

⁶⁵ S 74.

⁶⁶ The standard work remains Thiersch, H., Pharos. Antike Islam und Occident. Ein Beitrag zur Architekturgeschichte (1909)Google Scholar.

⁶⁷ C. Krause, Villa Jovis: die Residenz des Tiberius auf Capri (2003), a sumptuously illustrated work in the popular series Zaberns Bildbänder zur Archäologie: the Pharus is the subject of an appendix, ‘Anmerkungen zur Gesamtanlage’ (92–7). In a particularly sceptical review of this book (GFA 7 (2004), 1063–9Google Scholar), U. Wulf-Rheidt drew attention to both errors and fragile speculations. At p. 1067 Wulf-Rheidt agrees with Krause's argument that the ‘Torre del Faro’ cannot be the Pharus, but she finds the essentially unexcavated remains of the ‘Loggia della Marina’ too exiguous to support his elaborate reconstruction of the Pharus there. In the only other serious review of Krause, Gros, P. (JRA 17 (2004), 593–8)Google Scholar appears to accept the identification of the ‘Loggia’ with the Pharus. Cf. in slightly more detail the version in Krause's full-scale reconstruction of the villa: Krause, C., Villa Jovis. L'edificio residenziale (2005), 251–8Google Scholar. This work has thus far escaped the notice of both reviewers and standard bibliographies.

⁶⁸ Providing enough material for an entire dissertation: Jaisle, K., Die Dioskuren als Retter zur See bei Griechen und Römern und ihr Fortleben in christlichen Legenden (1907)Google Scholar; cf. Harris, op. cit. (n. 43).

⁶⁹ Lucian, Quomodo 62; Strabo 17.1.6. There is a large, complex, and contentious bibliography on the construction and dedication of the Pharos in the third century b.c., well summarized by Bing, P., ‘Between literature and the monuments’, in Harder, M. A., Regtuit, R. F. and Walker, G. C. (eds), Genre in Hellenistic Poetry (1998), 21–43Google Scholar = (revised) Bing, P., The Scroll and the Marble. Studies in Reading and Reception in Hellenistic Poetry (2009), 194–216Google Scholar. The only matter relevant here is that it is established beyond reasonable doubt that a statue of Zeus Soter stood atop the enormous structure (thus Bing, building on Fraser, P. M., Ptolemaic Alexandria (1972), II.47–8Google Scholar; et al.): that is clear from ancient representations and from an epigram by the contemporary Posidippus (Gow-Page XI). Where then are the Theoi Soteres, the Dioskouroi? They appear only in Lucian, writing almost 500 years after the construction of the Pharos, and the problem is compounded by the two versions of the dedicatory inscription as presented by Strabo and by Lucian. P. M. Fraser translates these as follows: ‘Sostratos the Cnidian, friend of the sovereigns, dedicated this for the safety of those who sail the seas, as the inscription says’; and ‘Sostratos, the son of Dexiphanes, the Cnidian, dedicated this to the Saviour Gods on behalf of those who sail the seas’. The two texts do not fully overlap in either form or content, but they do not actually disagree with each other and, as Fraser noted, Strabo's may be read as a paraphrase.

We cannot dismiss Lucian as a late source, or as erroneously altering Dis Soter to Theoi Soteres. Such arguments are library-bound, and they stumble on the test of autopsy. Fraser (I.19), following an older scholarly tradition, reasonably remarked that ‘the explanation may simply be that the dedication was in fact to all the deities who protect seafarers, and that Posidippus singled out Zeus, pre-eminent among such, because his statue crowned the lighthouse’.

It might be noted in this regard both that the lighthouses inferred by Alföldy at Herod's harbour at Caesarea were probably dedicated ‘to the sailors’ (Josephus), just as the Pharos was at Alexandria (Strabo and Lucian), and that they (the Druseum and the Tiberieum) would have been dominated by the great temple of Rome and Augustus, the Sebasteion or Caesareum. In that temple at Caesarea stood a colossal statue of Augustus, allegedly equal in quality to and modelled on the statue of Zeus at Olympia. That is to say, if Herod at Caesarea was indeed recalling the sailors' safe haven at Alexandria, Augustus, Tiberius, and Drusus at Caesarea would have nicely complemented Zeus and the Dioskouroi at Alexandria.

⁷⁰ Lucian, Quomodo 62, as translated by Costa, C. D. N., Lucian, Selected Dialogues (2005)Google Scholar.

⁷¹ Readers need not be reminded of the uncertainties involved, not only those noted already, but the assumption that the inscription in Lucian's Alexandria read the same in Tiberius' day; that Strabo is paraphrasing, not quoting, it; that Tiberius was responsible for constructing the Pharus on Capri; that the tower could accommodate both a statue of Zeus and a dedicatory inscription to Kastor and Polydeukes; that the Tiberieum and Druseum at Caesarea were lighthouses; et al.

⁷² S 19: ‘Proelia, quamvis minimum fortunae casibusque permitteret, aliquanto constantius inibat, quotiens lucubrante se subito ac nullo propellente decideret lumen et exstingueretur, confidens, ut aiebat, ostento sibi a maioribus suis in omni ducatu expertissimo.’

⁷³ The link is noted by Lindsay, H., Suetonius, Tiberius (1995), 98, 185Google Scholar. Vigourt, A., Les présages impériaux d'Auguste à Domitien (2001), 335–8Google Scholar, is disappointing on the death of Tiberius. The contrary nature of the omen is noteworthy, in that a lamp going out is good for the Claudii, just as the appearance of the Dioscuri was bad for one of them. Tiberius' special relationship with fire and the sun, indeed his unique mastery of flame, will be pursued elsewhere.

⁷⁴ Poulsen, op. cit. (n. 9), 122–6, Cf. the contributions of La Rocca, op. cit. (n. 5); Spannagel, op. cit. (n. 25), 28–34; Heinemann, op. cit. (n. 29), 45–8, 75–6; Sumi, op. cit. (n. 9), 179–81.

⁷⁵ See Swan, op. cit. (n. 26), 88–91.

⁷⁶ Rose, op. cit. (n. 36), 45.

⁷⁷ Mitford, T. B., ‘A note from Salamis’, in Bradeen, D. W. and McGregor, M. F. (eds), ΦOPOΣ (1974), 110–16Google Scholar.

⁷⁸ J. Pollini, The Portraiture of Gaius and Lucius Caesar (1987): they ‘recall classical figures like the Dioskouroi’ (p. 19, and especially n. 4); they are ‘in a sense like the Dioskuroi’ (p. 20).

⁷⁹ Buxton, B. and Hannah, R., ‘OGIS 458, the Augustan calendar, and the succession’, in Deroux, C. (ed.), Studies in Latin Literature and Roman History XII (2005), 290–306Google Scholar.