¹ [Auct.] ad Her. 2.34; Cic, . Inv. 1.91Google Scholar. Cicero has perhaps misremembered the line; caesae accidissent should probably be caesa accidisset (as at Varr, . LL 7.33Google Scholar and Prise, . Inst. 2.320.15ff.Google Scholar): Jocelyn, H.D., The Tragedies of Ennius (Cambridge 1967) 352Google Scholar.

² In a pedagogical context pundits might regard such arguments as tenuous (Cic. loc. cit.; cf. Quint. 5.10.84), although Cicero seems, like others, to have been much taken with that particular strategy – and, in oratorical practice (as noted above in the text), such a feint was obviously found useful (both Crassus and Cicero using it in their defences of Caelius in 56); Crassus, Cael. frag. 12 Male. = Cic, . Cael. 18Google Scholar. For a discussion, Austin, R.G., M. Tulli Ciceronis pro M. Caelio oratio (Oxford 1960) 68Google Scholar. For Cicero's fondness for Ennius: Gell, . NA 12.2.6Google Scholar.

St Jerome was also taken with the line (again, in the context of discussing prominent women): Ep. 127.5.

³ Cornelia as icon: Plut. Mor. 145F (= ‘Advice to the Bride and Groom’, 48); Jer, . Ep. 107.4.6Google Scholar. Cf. Hemelrijk, E., Matrona docta. Educated Women in the Roman Élite from Cornelia to Julia Domna (London and New York 1999) 64-8Google Scholar. Roller (cited in introductory note) also discusses this political exchange (discussed below), observing the extraordinary fact of a politician's mother being drawn in this particular fashion into deliberative or forensic oratory.

⁴ The information that Gaius' salvo was produced in contiene is provided by Seneca (Dial. 12.16.6); cf. Plut, . CG 4.3Google Scholar:

⁵ For these fragments, see also Häpke, N., C. Semproni Gracchi Oratoris Romani Fragmenta Collecta et Illustrata (Munich 1915) 97-8Google Scholar.

⁶ Plut, . CG 13.2Google Scholar; cf. Diod. Sic. 34/35.25.2; Plut, . CG 4.2Google Scholar (on her influence).

This might have gone even further. Some implicated her in the death of Aemilianus: App, . BC 1.20.83Google Scholar; cf. Cic, . Rep. 6.12Google Scholar (propinquorum manus). It is difficult to know if this was mooted at the time or was part of subsequent speculation. According to Stockton, D. (The Gracchi [Oxford 1979] 224)Google Scholar, this may have been the reason she came under fire – and may explain C. Gracchus' public defence.

⁷ That she was ‘slandered for having adulterous relations’ is the assumption of Corbelli, A., Controlling Laughter. Political Humor in the Late Roman Republic (Princeton NJ 1996) 148Google Scholar; ‘Dining Deviants in Roman Political Invective’, in Hallett, J.P. and Skinner, M.B. (eds), Roman Sexualities (Princeton NJ 1997) 99–128, esp. 110-11Google Scholar.

Cf. Leonhard Burckhardt and Jürgen von Ungern-Sternberg's full discussion of this incident: Burckhardt, L. and von Ungern-Sternberg, J., ‘Cornelia, Mutter der Gracchen’, in Dettenhofer, M.H. (ed.), Reine Männersache? Frauen in Männerdomänen der antiken Welt (Köln 1994) 97–132, at 116-19Google Scholar.

⁸ The apophthegm may not have been original; cf. Plut, . Phoc. 19.3Google Scholar; Mor. 241D (= ‘Sayings of Spartan Women’, 9). Indeed, the parallel prompts Burckhardt and von Ungern-Sternberg, ‘Cornelia’ (n. 7) 106 to dismiss the item altogether. For them it is an ‘urban myth’ (Wanderanekdote), possibly fabricated by Gaius Gracchus.

⁹ So too was memory of the care which she lavished on their education: Cic, . Brut. 104, 211Google Scholar; Quint, . Inst 1.1.6Google Scholar; Tac, . Dial. 28.4–6Google Scholar; Plut, . TG 1.4–5Google Scholar; Jer, . Ep. 107.4.6Google Scholar. Roller (cited in introductory note) also discusses the positive nature of the ‘jewel’ anecdote and the tradition of Cornelia's strong interest in her sons’ education.

¹⁰ Plut, . TG 8.5Google Scholar.

¹¹ We observe that Peter Green had also noticed this in passing: Juvenal. The Sixteen Satires (Harmondsworth 1967) 155 nn. 13-14Google Scholar; cf. Roller (introductory note), offering further references.

¹² There is, perhaps, something of a formula at work here. A scholiast on Medea 264 reports that Parmeniskos preserved a tradition that claimed that Medea had seven sons and seven daughters who were killed by the Corinthians: Schwartz, E. (ed.), Scholia in Euripidem 2 (Berlin 1891) ad loc. [p. 159]Google Scholar.

¹³ Such cynicism with regard to aristocratic pride seems better to belong to Juvenal's epoch; cf. Henderson, John, Figuring Out Roman Nobility. Juvenal's Eighth Satire (Exeter 1997)Google Scholar. From surviving examples of first-century oratory, it is clear that the rhetoric of diminution might embrace a ridiculing of an opponent's pride in, and deluded dependence upon, the imagines (e.g. Cic, . Leg. Agr. 2.1Google Scholar; Pis. 1; cf. Hölkeskamp, Karl-Joachim, ‘Self-Serving Sermons: Oratory and the Self-Construction of the Republican Aristocrat’, in Smith, C. and Covino, R. [eds], Praise and Blame in Roman Republican Rhetoric [Swansea 2010] 17–34, esp. 17-19)CrossRefGoogle Scholar. And, it is true, Sallust puts that type of derision into a second-century mouth [Iug. 85.4-6, 10, 29,38]). But, again, this would not explain the personal nature of Gaius’ (comparative) attack on his mother's detractor. (As observed above, Cornelia's pride in her sons was part also of a friendly tradition – whatever critics of the Gracchi might say. It did not require this type of vilification of those who observed it in an unfriendly way.)

¹⁴ Plut, . CG 13.2 (cited above)Google Scholar. The ‘letter of Cornelia’ preserved by Nepos would have been evidence adduced to refute the allegation. The vexed issue of its authenticity is not one that needs to be addressed here. These ancient allegations are viewed with reserve by Burckhardt, and von Ungern-Sternberg, , ‘Cornelia’ (n. 7) 109-20Google Scholar.

¹⁵ References cited above, n. 6. Burckhardt, and von Ungem-Sternberg, , ‘Cornelia’ (n. 7) 100Google Scholar dismiss the allegation itself as an absurdity. Eventually, they observe, someone will believe anything (or, rather, everything will find a believer).

¹⁶ For no less than Carneades as perhaps one of those guests, see Jer, . Comm. in Sophoniam prophetam, Prologus, PL 25.1337c (= 655 Adriaen)Google Scholar.

¹⁷ The area's reputation for luxury and excess is too well established to require extensive reference here. Cicero referred to Baiae, or perhaps the whole Bay of Naples, as cratera ilium delicatum (Att. 2.8.2). A mere reference to ilia vicinitas was sufficient: Cic, . Cael. 47Google Scholar; cf. 27 (for an example). For Cornelia's villa (and steady stream of guests), Plut, . CG 19.1–2Google Scholar; cf. Oros. 5.12.9 (dating her withdrawal to Misenum to after, and – implicitly – in response to, Tiberius' death. Cornelia would thus have been at Misenum from 133/132).

¹⁸ Tony Corbeill, as we have noted (n. 7), suggests that she was ‘slandered for having adulterous relations'. That is not impossible – but there is certainly no evidence for it (either that she was slandered in that way or that she had had illicit affairs).

We do, indeed, explore the possibility that the hostile insinuation to which Gaius responded was of a highly personal nature (and with regard to Cornelia's intimate life), but it is of an altogether different kind. See below.

¹⁹ Cf. Corradi, G., Cornelia e Sempronio (Rome 1946) 8Google Scholar, underlining the significance of this item being preserved; and Guarino, A., ‘Minima de Gracchis’, Studi in onore di Arnaldo Biscardi (Milan 1982) 53–67Google Scholar, underlining, as we do, Cicero's detachment from such futuristic reasoning. See, on the other hand, Petrocelli, C., ‘Cornelia the Matron’, in Fraschetti, A. (ed.), Roman Women (1994Google Scholar, Eng. trans. Lappin, L., Chicago and London 2001) 34–65, at 37Google Scholar, suggesting – implicitly – that Cicero subscribed to the line of thought.

²⁰ We need not dwell for the moment on the clear synkrisis of the maledictor and Cornelia in Gaius’ sharp riposte (and the implication therein that the synkrisis had been invited by his opponent), though we shall return to that (as it is of the essence). It is unlikely that the man, as a critic of Cornelia , had made a direct comparison between his own qualities and those of his target; it must have been implicit.

²¹ , as Valgiglio, E. notes (Plutarco. Vita dei Gracchi [Rome 1957] 124)Google Scholar – and both words pick up on parere (in the Senecan fragment).

²² In the forceful repetition was a twist; at the personal level, ‘Do you dare insult my mother?’; and then, at a level of communal resonances, ‘Do you dare insult the mother of Tiberius Gracchus?’. Bauman, R.A., Women and Politics in Ancient Rome (London and New York 1992) 44CrossRefGoogle Scholar, taking his cue from Plutarch's report, suggests that, by using Cornelia's name rather than saying ‘my mother’, her son ‘has almost elevated her to a concept’. But Gaius could have had it both ways. There is no need to abandon either fragment.

²³ And by 123/122, only Gaius and, it seems, one ‘boy’ (often taken as the one surviving son of Tiberius) were left alive: C. Gracchus, ORF³ frag. 47 Malc. (= Schol. Bob, . in Cic. Sull. 81 Stangl)Google Scholar, where Gaius, speaking at the time of his tribunician office, claims that he and an unspecified puer were the only descendants left to the houses of P. Africanus and Ti. Gracchus (presumably the Elder). This, of course, may not take into account girls. Cf. Sempronius Asellio, ap. Gell, . NA 2.13.1–2Google Scholar (on Tiberius' sole surviving son [in 133]); Val. Max. 9.7.2 (elaborating on the fate of Tiberius' three sons); Plut, . CG 15Google Scholar (on Gaius' ‘child’ [paidion] in 121). Our colleague, Patrick Tansey, would strongly argue that Gaius' reference above is to his own son. For a discussion (and references to earlier discussions), see Dixon, S., Cornelia. Mother of the Gracchi (London and New York 2007) 25, 70 n. 32, 71 n. 53Google Scholar (and see below, n. 60). Whatever the answer, a tenuous link to posterity.

The fruit of the original marriage (that of Gracchus the Elder and Cornelia), as remembered by Jerome, (Ep. 54.4)Google Scholar, who honoured the latter, was one that brought her anything but laetitia.

²⁴ App, . BC 1.20.83Google Scholar.

²⁵ The question remains vexed; cf. Burckhardt and von Ungern-Sternberg, ‘Cornelia’ (n. 7) 102-3, for references to earlier discussions (adding to that bibliography, the discussion by Petrocelli, , ‘Cornelia the Matron’ [n. 19] 38–43)Google Scholar; Dixon, , Cornelia (n. 23) 7 (for despair)Google Scholar.

²⁶ These dates are by way of necessarily rough calculations based on Polyb. 31.27 [32.13].1-3 (the date of Scipio Africanus' death, 183 or 182 BC, at which point we believe the best evidence suggests Cornelia was still unmarried), the likely birth-date of Tiberius, Cornelia's putatively eldest son (between July 163 and July 162: Sumner, G.V., The Orators in Cicero's Brutus: Prosopography and Chronology [Toronto 1973] 58)Google Scholar, and the presumption that Gaius (bom in 154 or 153 [Sumner, loc. cit. 70]) was Cornelia's last or second last child, or possibly, following the intriguing suggestion of Carcopino, J. (Autour de Gracques [Paris 1967] 81)Google Scholar, and of Burckhardt, and von Ungern-Sternberg, , ‘Cornelia’ (n. 7) 104 n. 27Google Scholar, in utero at the time of his father's death. On this complicated issue, we suggest readers have recourse to an Appendix attached to the longer version of this paper to be found on our Macquarie staff websites, and to the forthcoming treatment by Patrick Tansey, who will deal with the earlier discussions of Mommsen, Th., ‘Die Scipionenprozesse’, Hermes 1.3 (1866) 161–216, at 204-6Google Scholar (= Römisches Forschungen [Berlin 1864-1879] II, 489-91)Google Scholar, Carcopino, K. Moir, ‘Pliny UN 7.57 and the Marriage of Tiberius Gracchus’, CQ 33 (1983) 136-45, at 144Google Scholar, and others, and who will argue, inter alia, that we cannot be sure how many boys and how many girls there were amongst Cornelia's twelve children. That argument will challenge our presumption above (i.e. that Gaius was Cornelia's last or second last child), but not mie out the datum as a possibility. We reject the scepticism of Dixon, , Cornelia (n. 23) 7Google Scholar concerning the number of Cornelia's children.

²⁷ For references to Tiberius' military service, see MRR 1.388, 393, 395-6, 398, 401, 402Google Scholar. It will be clear from the preceding note that we are not prepared to offer any simulations of certitude with regard to Cornelia's birth-date. But in 180, three years after her father's death (by which time we would think of Cornelia being married), Cornelia may have been around fifteen years old (or younger). Tiberius Gracchus will have been around forty. (For his birth-date, probably 220, see Sumner, , Orators in Cicero's Brutus [n. 26] 38-9.Google Scholar) If the marriage is to be dated later, Cornelia is still likely to have been in her early teens and Gracchus will have been in his mid-forties. In itself, this is no argument for the marriage remaining unconsummated (see below, n. 51 on Mia and Pompey).

²⁸ Many scholars familiar with the shortcomings of Solinus will assume that these two citations represent a single piece of evidence (which they probably do), with Solinus drawing directly from Pliny. In a separate paper, Hillard will argue that so much of the prosopographia to be found in both these sources was derived from a common source from which both authors were prepared to draw in a quasi-verbatim manner. The source is likely to have been Late Republican, possibly Varro. If so, the item gains in stature.

²⁹ For what it is worth, Pliny, for instance, does not record that this was rumoured of Cornelia. (See, by way of example, the use of such formulae as ferunt at 7.79 or even in acta … relatum est as at 2.147.) He offers Cornelia as the proof that women are bom with this problem - and that it is a bad omen. The fact, in her case, is not in doubt for Pliny. (It has to be acknowledged that such certitude regarding the data that come his way is characteristic of Pliny.) Roller (cited in the introductory note) also refers to the ominous nature of Cornelia's medical complaint and interprets the likely Roman reaction in much the same way that we do, registering other discussions not cited here.

³⁰ It is rejected, as noted above, by Burckhardt and von Ungern-Sternberg, ‘Cornelia’ (n. 7) 101, who suspect that the item (the plausibility of which they acknowledge) is the product of hostile fabrication.

³¹ On the Scipiones and Sempronii in this regard, see Burckhardt, and von Ungem-Sternberg, , ‘Cornelia’ (n. 7) 105 n. 31Google Scholar. The apparently extraordinary focus might, of course, be the result of the unusually close attention this family grouping received in the surviving source tradition.

³² Sor. Gyn. 1. [III] 17; cf. Hippoc. Gynaikeion (‘Diseases of Women’) 1.20; Arist, . Gen. An. 4.4.773aGoogle Scholar; Hist. An. 10.4.636bGoogle Scholar; Sor, . Gyn. 3. [I] 7, 9Google Scholar. This is complicated by the denial, in some quarters, of the normal existence of a hymenial membrane. For detailed discussions of this topic, see Sissa, G., ‘Une virginité sans hymen: le corps féminin en Grèce ancienne’, Annales ESC 39 (1984) 1119-39Google Scholar (= ‘Maidenhood without Maidenhead: The Female Body in Ancient Greece’, in Halperin, D.et al. [eds], Before Sexuality. The Construction of Erotic Experience in the Ancient Greek World [Princeton 1990] 339-64), esp. 1130 (= 354f)Google Scholar, correcting Rouselle, A., ‘Observation féminin et idéologique masculine: le corps de la femme d'après les médecins grecs’, Annales ESC 35 (1980) 1089-115, esp. 1104Google Scholar; Porneia (Paris 1983) 41, 48Google Scholar (= Porneia. On Desire and the Body in Antiquity [Eng. trans. 1988] 23, 27, 33Google Scholar); Hanson, A., “The Medical Writers' Woman’, in Halperin et al., op. cit. 309-38, esp. 324-30Google Scholar; Dean-Jones, Lesley Ann, “The Cultural Construct of the Female Body in Classical Greek Science’, in Pomeroy, S.B. (ed.), Women's History and Ancient History (Chapel Hill 1991) 111-37, esp. 121 and 135 n. 48Google Scholar; Women's Bodies in Classical Science (Oxford 1994) 54, 130Google Scholar.

³³ Aristotle, (Hist. An. 10.4.636b)Google Scholar believed this to be in some cases congenital and in others a pathological condition where the os uteri had grown together; cf. Hanson, A.E., ‘Talking Recipes in the Gynaecological Texts of the Hippocratic Corpus’, in Wyke, M. (ed.), Parchments of Gender. Deciphering the Bodies of Antiquity (Oxford 1998) 71–94Google Scholar, 88 for references and a discussion of the remedies thought potent against closed or deviant uterine mouths.

³⁴ Arist, . Gen. An. 4.4.773aGoogle Scholar; cf. Sor, . Gyn. 3. [I] 9Google Scholar.

³⁵ Hist. An. 10.4.636b.

³⁶ Gen. An. 4.4.773a.

³⁷ See Schilling, R., Pline l'Ancien. Histoire Naturelle Livre VIII (Paris 1977) 157Google Scholar: ‘Là où Aristote se borne à un constat d'ordre médical, Pline voit matière à présage.’

One of the anonymous readers, taking a cue from modern studies exploring the gender-roles that might be inferred in the famous ‘letter’ of Cornelia (we leave aside here the much debated question of the authenticity of that document and the accompanying bibliography of pertinent items), suggested that Cornelia was perhaps criticised for appropriating masculine modes of conduct (and that this criticism might have been served by claims that she had been bom deformed and ‘unwomanly’); on the gender issue, cf. Hallet, J.P., ‘Matriot Games? Cornelia, Mother of the Gracchi, and the Forging of Family-Oriented Political Values’, in McHardy, F. and Marshall, E. (eds), Women's Influence on Classical Civilization (London and New York 2004) 26-39, esp. 304, 36-8Google Scholar; and ‘Absent Roman Fathers in the Writings of their Daughters: Cornelia and Sulpicia’, in Hübner, F.R. and Ratzan, D.M. (eds), Growing Up Fatherless in Antiquity (Cambridge 2009) 175-91, esp. 181-2CrossRefGoogle Scholar. We take a very different tack, and leave that as an article for another to write.

³⁸ We have contemplated an even greater range of possibilities, including endometriosis (which is typically a late-onset complaint [i.e., post-pubertal]) and hermaphroditism (from which the new bom child would have been unlikely to survive the demands of the haruspices in terms of expiation; cf. Brisson, L., Sexual Ambivalence. Androgyny and Hermaphroditism in Graeco-Roman Antiquity [1997, Eng. trans. Lloyd, J., Berkeley and Los Angeles 2002] 7–40)Google Scholar.

³⁹ It is unlikely, however, that the omen (if an imperforate hymen or suchlike obstruction) was recognised at birth. As our gynaecological consultants inform us, with an imperforate hymen, the external genitalia usually have a normal appearance. Likewise, as far as vaginal atresia is concerned, these women also are likely to have normal external genitalia. The Romans might conduct attentive post-natal inspections (see e.g. Obseq. 40, with regard to an examination of newborn twins in 108 BC). But unless such inspections were extremely thorough, these were not problems that were likely to be detected until puberty. Even today, such defects are rarely recognised until the onset of menses, when the retention of mucus and blood will lead to amenorrhea, pelvic pain and/or a palpable abdominopelvic mass.

Much of the surviving evidence indicates that close inspections had once been made at the age of puberty with regard to marriageability – though even in this regard the physical inspections practised in more remote times were giving way, ‘with regard to modesty’, in favour of set ages being fixed for eligibility; cf. Corbett, P.E., The Roman Law of Marriage (Oxford 1930/1969) 51Google Scholar (with a discussion of Servius' interesting remarks ad Virg. Aen. 7.53 on the pubescent Lavinia); Watson, A., The Law of Persons in the Later Roman Republic (Oxford 1967) 39–40Google Scholar; Treggiari, S., Roman Marriage. lusti Coniuges from the Time of Cicero to the Time of Ulpian (Oxford 1991) 39Google Scholar.

⁴⁰ All the more so, if we accept the hypothesis - put forward by Carcopino, , Autour de Gracques (n. 26) 82Google Scholar; Burckhardt, and von Ungern-Sternberg, , ‘Cornelia’ (n. 7) 104Google Scholar – that Gracchus the Elder, in observing that Cornelia might still bear a child (parere adhuc potest : see n. 67 below), was not enjoining his wife to marry again (and have children to a second husband) but alluding to the fact that she was carrying a child - and that Gaius was the one in utero at the time of his father's pronouncement and self-sacrificial offer.

⁴¹ For a discussion of ‘perhaps the most outrageous of Martial's obscene epigrams’, Watson, P., ‘Non tristis torus et tarnen pudicus: The Sexuality of the Matrona in Martial’, Mnemosyne 58.1 (2005) 62–87, at 63-70CrossRefGoogle Scholar, underlining that the behaviour attributed to the historical matronae here and the equestrian position linked with Andromache were more usually associated with meretrices, the key distinction being, of course, that the sexuality of a matrona was to be ‘channelled into procreation’ (81).

⁴² The language is familiar to readers of graffiti – and of Catullus (but hardly in a domestic context, as here). Martial is using language that many others would have avoided. On the use of pedicare, see the cautious statement of Adams, J.N., The Latin Sexual Vocabulary (Baltimore 1982) 2Google Scholar; cf. on the word generally, id. 123-4. See also Richlin's, A. discussion of ‘four-letter words’ in The Garden of Priapus. Sexuality and Aggression in Roman Humor (New Haven 1983) 18–26Google Scholar.

⁴³ On the porticus Octaviae replacing the earlier portico, Vell. Pat. 1.11.3. On the statue (and its transfer to the porticus Octaviae), Plin, . NH 34.31Google Scholar; cf. Chioffi, L., ‘Statua Corneliae’, LTUR 4 (1999) 357-59Google Scholar, to the bibliography of which add Kajava, M., ‘Cornelia Africani f. Gracchorum’, Arctos 23 (1989) 119-31Google Scholar; Petrocelli, , ‘Cornelia the Matron’ (n. 19) 61-4Google Scholar; Ruck, B., ‘Das Denkmal der Cornelia in Rom’, MDAI(R) 111 (2004) 477–494Google Scholar; Dixon, , Cornelia (n. 23) 29-30, 56-9Google Scholar. Bettini, Maurizio, Nascere: Storie di donne, donnole, madri ed eroi (Turin 1998) 109-12, 125-7 nn. 25-44Google Scholar, makes the interesting suggestion that the unlaced sandals of Cornelia's statue were a symbol of her fecundity. Was, then, the statue itself a celebration of her frequent child-bearing?

⁴⁴ Ep. 1.42.

⁴⁵ Dover, K.J., Greek Homosexuality (Cambridge MA 1978) 101Google Scholar; cf. Kay, N.M., Martial Book XI. A Commentary (London 1985) 281Google Scholar.

⁴⁶ On the popular feelings for Julia, Plut, . Pomp. 53.4Google Scholar, Caes. 23,4.

⁴⁷ Seneca does not explicitly register Lucretia as the first Brutus' wife – but as his equal. (She was, of course, not the wife of Brutus.)

⁴⁸ We owe a debt of gratitude to Kay, Martial Book XI (n. 45) 281-2, for first drawing this Senecan reference to our attention – though Kay does not note the telling absence of Julia, seeing Seneca's trio as something of a parallel. Kay merely suggests that although the two ‘modern’ exempla at Mart. 11.104.18 were perhaps not as revered as Cornelia, their stature allowed them to be ‘included for the same irreverently humorous reasons’. (That is the usual assumption; cf. Obermayer, H.P., Martial und der Diskurs über männliche ‘Homosexualität in der Literatur der frühen Kaiserzeit [Classica Monacensia 18, Tübingen 1998] 25Google Scholar, where the women are seen as models of rectitude, also citing Seneca.) It is the anomalous inclusion of Julia that is of the essence here.

Jerome also grouped Lucretia, Cornelia and Porcia (perhaps in the debt of Seneca); in Iovinianum 1.49 (320) C; cf. Petrocelli, , ‘Cornelia the Matron’ (n. 19) 61Google Scholar; Dixon, , Cornelia (n. 23) 61-2Google Scholar. It is Julia's absence from these repertoires that prompts us to a different reading of Martial's reference to her in 11.104.

⁴⁹ All three are included in Valerius Maximus’ section de amore coniugali (4.6), examples calling for maxima veneratio (4.6 praef). In Cornelia's case (4.6.1), this affection is only testified to implicitly – and inversely (in the willing self-sacrifice of her husband to preserve her life); but this is enough to have her enshrined, paradoxically, alongside Alcestis, by Ael, . VH 14.45.1Google Scholar. Valerius Maximus draws attention to the same paradox; cf. Petrocelli, , ‘Cornelia the Matron’ (n. 19) 41Google Scholar. Of particular interest here is that Julia is remembered for her vulnerability (4.6.4), Porcia for her courage, her spiritus virilis and fortitude demanding the admiratio of all centuries (4.6.5).

⁵⁰ Compliance: Dixon, , Reading Roman Women (London 2001) 40Google Scholar (a discussion of male fantasy, ‘dream lovers’, and ‘female compliance with male preferences’). Demeaning: Martial's ‘wife’ clearly finds it so (Mart. 11.104. 21-22); cf. Parker, H.N., “The Teratogenic Grid’, in Hallett, and Skinner, (eds), Roman Sexualities (n. 7) 47–65, esp. 53Google Scholar: “The Roman created what we may call “the scale of humiliation“: vagina, anus, mouth.'

⁵¹ R.J. Seager (‘Iulia’ [2], OCD ³ 777), correcting G.E.F. Chilver (‘Julia’ [1], OCD² 566) gives her birth-date as c. 73. Münzer, F., RE 10, col. 894Google Scholar (‘Julius’ 547) had dated her birth to 83; this was queried by Gelzer, M., Caesar. Politician and Statesman (1921, Eng. trans. Needham, P., Oxford 1968) 21 n. 5Google Scholar (preferring 76); countered by Syme, R., ‘No Son for Caesar?’, Historia 29 (1980) 422-37, at 422-3Google Scholar (= Roman Papers 3 [Oxford 1984] 1236-50, at 1237Google Scholar) opting for a date between 77 and 75, but favouring 76 or 75. Marshall, B.A., “The Engagement of Faustus Sulla and Pompeia’, Ancient Society 18 (1987) 91–101, at 91-2CrossRefGoogle Scholar, suggests a birth-date between 81 and 79, but that is not generally followed; cf. Meier, Chr., Caesar (1982, Eng. trans. McLxntock, D., London 1995) 216Google Scholar; and Pelling, C.B., Plutarch. Caesar (Oxford 2011) 153Google Scholar (opting for 76 or 75).

⁵² Richlin, , ‘Not Before Homosexuality: The Materiality of the Cinaedus and the Roman Law against Love Between Men’, Journal of the History of Sexuality 3 (1993) 523-73, at 523Google Scholar; Edwards, C., The Politics of Immorality in Ancient Rome (Cambridge 1993) 85CrossRefGoogle Scholar; Williams, C., Roman Homosexuality (Oxford 1999/2010) 407 n. 294Google Scholar (refining the previous two readings).

⁵³ Cf. Plut. De cap. ex inim. utilitate 6 (= Mor. 89E), where it is made clear that such careful parting of the hair with a single finger was taken as a sign that a man was of a womanish nature and licentious (which, Plutarch feels the need to emphasise, was far from a reliable guide in Pompey's case). For the timing of one such abusive demonstration, 7 February 56 BC, see Cic, . Q.fr. 2.3.2Google Scholar (where Cicero sanitises the hostile script for his brother's consumption); cf. Dio 39.19.1 (reporting that Pompey's personal peculiarities were targeted). The account of the chanting in Cicero's letter to Quintes follows a very different script. M. Licinius Calvus (ap. Sen. Controv. TAJ) composed a similar taunt aimed at Pompey. For the intent of such a gibe, see Williams, , Roman Homosexuality (n. 52) 237Google Scholar.

⁵⁴ On the familiar ‘Biblical’ euphemism for carnal knowledge (in koine Greek at least), see Bauer, W., Arndt, W.F. and Gingrich, F.W., A Greek-English Lexicon of the New Testament (Chicago 1957) 160bGoogle Scholar (s.v. γνγνώσκω [5]). Such a metaphor might have been lifted from a Latin source; cf. Adams, , Latin Sexual Vocabulary (n. 42) 190Google Scholar.

⁵⁵ Plut, . Brut. 13.3–11Google Scholar; cf. 13.2 (on the sharing of a bed).

⁵⁶ H. Rackham (Loeb) for the former; H. Le Bonniec (Budé) and Blake, K. Jex (in Blake and Sellers, E., The Elder Pliny's Chapters on the History of Art [Chicago 1896/1976)] 68) for the latterGoogle Scholar.

⁵⁷ On Strongylion, see Gross, W.H., ‘Strongylion’, DKP IV 398, 2Google Scholar; Sellers, E.[-Strong], in Blake and Sellers, Elder Pliny (n. 56) 68Google Scholar; cf. de Santerre, H. Gallet, Pline l'Ancien. Histoire Naturelle Livre XXXLV (Paris 1953) 81, 262-3Google Scholar.

⁵⁸ We cannot rule out that the item was a malicious invention of subsequent hostile historical tradition; see above, n. 30. But, as we have argued above, there is also no reason to assume that such an item was not broadcast at the time.

⁵⁹ E.g. that of Ptolemy Physkon, Plut, . TG 1.4Google Scholar. Dixon, , The Roman Mother (London and Sydney 1988) 22Google Scholar, plausibly assumes that there were others. (No evidence exists.) For the ideology of the univira, see Dixon, loc.cit. From this ideal there were probably dissentients at the time; divorce and remarriage were too common. It was certainly challenged in the Augustan period: Herbert-Brown, G., Ovid and the Fasti (Oxford 1994) 13, 147-8Google Scholar. But in Cornelia's day (and later: see Propert. 4.11), the sentiment no doubt won the approbation of many.

⁶⁰ … nec quisquam de P. Africani et Tiberi Gracchi familia nisi ego et puer restaremus (Gaius Gracchus, de legibus promulgates frag. 47 Malc. [= Schol. Bob. in Cic. Sull. p. 81 Stangl]): ‘nor does anyone remain to the houses of Publius Africanus and Tiberius Sempronius (i.e. the Elder) except me and [this] boy.’ (We are assuming that the lament focussed on male issue – but see above n. 23.) Gaius conveniently omits consideration of the Scipiones Nasicae. This is rhetoric. Gaius also ignores, for the purpose of argument, his sister Sempronia, as indeed does ‘Cornelia’ or whoever composed the letter published under her name: Cornelia P.f, epistulae ad Caium filium 4 [= codd. Corn. Nep. in fin.] (Cugusi, P., Epistolographii Latini Minores [Turin 1970] 1, 112Google Scholar; cf. 2.67-8 [for a succinct commentary on the lines]), referring to Gaius as her last surviving child.

⁶¹ This would have been in addition to those claims which suggested that Cornelia actually had a hand in his rumoured murder.

A notion of the ‘Gracchan blight’ might underlie the highly negative portrait of Sempronia at App, . BC 1.20Google Scholar – unless it was simply historical observation that Sempronia was ‘unlovable, misshapen and childless’ (a highly unusual picture): . The childlessness of Scipio Aemilianus and Sempronia had, it seems, been a cause for concern in the 130s. Such concern possibly lay behind the incident (if it was historical) reported at Numantia in 134-3, from which Marius emerged with such kudos. Who would succeed to Scipio's position was a question asked by his entourage. To those invited to share his table, Scipio recommended Marius as a possible worthy heir (Val. Max. 8.15.7; Plut, . Mar. 3.3)Google Scholar. His succession was not to be along bloodlines.

⁶² Liv. 38.57.5-8; cf. the emendation suggested by Konrad, C.F., ‘Livy on the Betrothal of Cornelia Gracchi (38.57.7)’, Philologus 133 (1989) 155-7CrossRefGoogle Scholar, and accepted in J. Briscoe's Teubner edition (Stuttgart 1991) 602.

⁶³ Livy places the anecdote amongst those which he describes as the greatly variant stories circulating with regard to Scipio – but which are all worthy of presentation (proponenda): 38.57.8, cf. 56.1 (for the context of unstable and untrustworthy traditions). Plutarch (TG 4.23) reports the doublet, observing that ‘some’ hold to the former story, but that ‘most historians’ hold to the latter, as he does. He furthermore cites Polybius as confirmation that Cornelia was betrothed after Scipio's death and that the selection of her husband was made by Scipio's relatives , Scipio having left Cornelia ‘unespoused and unbetrothed’ . (In the surviving text of Polybius, no explicit statement to that effect remains; that does not rule out Plutarch having available a fuller text; cf. Polyb. 31.27 [32.13]. 1-3.) Cf. Valgiglio, , Plutarco. Vita dei Gracchi (n. 21) 34-5Google Scholar.

⁶⁴ And any suggestion that Scipio Africanus died envisaging that his younger daughter would remain unmarried flies in the face of Polybius' understanding (loc. cit. n. 63) of the financial arrangements that Africanus had set in place for his daughters' dowries; cf. Dixon, , ‘Polybius on Roman Women and Property’, AJP 106 (1985) 147-70Google Scholar.

⁶⁵ A possible parallel has already perhaps sprung to the mind of readers. In 6th-century Athens, Peisistratos found a political union with the Alkmaionidai irresistible but the danger of religious contagion (i.e. the Alkmaionid ‘curse’) troubling. His solution is well known; he coupled with his wife κατὰ νόμον (Hdt. 1.60.2-3; 61.1-2). That is as far as the musing mind can go. Tiberius Gracchus did not take that path. He procreated – perhaps after hesitation.

⁶⁶ Indeed, the very statue erected in Cornelia's honour may have symbolically represented a celebration of her fecundity. See above n. 43.

⁶⁷ Cic, . Div. 1.36, 2.62Google Scholar, citing a letter written by C. Gracchus to M. Pomponius; Val. Max. 4.6.1; Plin. NH (offering the only surviving Latin version of Gracchus' utterance: immo vero, inquit, meum necasse, Cornelia enim iuvenis est et parere adhuc potest); Plut, . TG 1.2–3Google Scholar; de vir. ill. 57.4. Note, in passing, the emphasis in this context on Cornelia's child-bearing. Cicero's source (Gaius' friend Pomponius) obviously suggests that the incident was susceptible to a reading friendly to the Gracchan household. It is also clear that the surviving tradition (concentrating, as it does, on the elder Gracchus’ selflessness) is heir to that friendly interpretation. On the other hand, it seems a safe presumption that enemies of the Gracchi would have focussed not upon the heroic self-sacrifice of Gracchus but upon the inescapable severity of the divination. It is noteworthy that this tradition did not prevail - but that need not detract from the logic of a hostile interpretation of the omen.

Köves-Zularf, Th., Reden und Schweigen. Römische Religion bei Plinius Maior (Munich 1972) 273-4Google Scholar, ingeniously points to the relevance in this regard of the dire omen which preceded the death of Gracchus the Elder; cf. Burckhardt and von Ungern-Sternberg, ‘Cornelia’ (n. 7) 103-5. The haruspices, summoned to interpret the sign, announced that one of the serpents must be killed and that the sex of that reptile would correspond to the imminent death of either Gracchus or Cornelia, upon which Gracchus ordered that the male snake be killed on the grounds that his wife was young and might still bear children. At Div. 2.62, Cicero professes a certain degree of bafflement; why not let both snakes go free? Plutarch answers that question: the soothsayers had ordained that one snake was to be killed. The haruspical response is startling; the two reptiles most aptly, one would think, signify the genius of Gracchus and the iuno of Cornelia (cf. Corradi, , Cornelia e Sempronia [n. 19] 14Google Scholar; Burckhardt, and Ungern-Sternberg, von, ‘Cornelia’ [n. 7] 105, esp. n. 30)Google Scholar and therefore marital union. Köves-Zularf, loc. cit. 273 n. 515, highlights the fact that the haruspices’ finding presupposes the end of the married couple's coexistence, and that the marriage was not meant to be, not sanctioned, and must be ended (‘daß die Ehe des Sempronius Gracchus mit Cornelia gar nicht als richtige Ehe galt und deswegen nicht geschützt, sondern vernichtet werden sollte’). For a discussion of the practicalities of the haruspical response, see Köves-Zularf, loc. cit. 270-8 (cf. 283, 287-8 n. 556); he cogently, in our opinion, links this to Cornelia's birth-omen (n. 515; for his earlier discussion of which, see 222-4) – though Burckhardt and von Ungern-Sternberg, loc. cit. 105 n. 31, remain wary.

⁶⁸ Enemies: reconciliation with Scipio Africanus, dramatic and unexpected, no doubt brought new enmities. Gracchus' career had involved controversial decisions (Plut, . Marc. 5.1–3Google Scholar; for further references, see MRR 1.442Google Scholar); and his eldest son is reported to have had deadly enemies at the very outset of his senatorial career (Plut, . TG 6.2Google Scholar and the reference to the young manἐχθροί). The careers of Gracchus' two last surviving sons added exponentially to those ranks.