¹ See, for example, H. Dragendorff, ‘Die Amtstracht der Vestalinnen’, RhM 51 (1896), 281–302; I. Santinelli, ‘Alcune questioni attinenti ai riti delle vergini Vestali: “Vesta aperit” (Cal. Philoc, al 7 giugno)’, Riv Fil 30 (1902), 255–62; G. Dumézil, ‘QII 18, “Te, amata, capio”’ REL 41 (1963), 89–91. Several more general studies of the priesthood, of course, exist. The most helpful reference works are Giannelli, G., Il Sacerdozio delle Vestali Romane (1913)Google Scholar and Guizzi, F., Aspetti Giuridici del Sacerdozio Romano: Il Sacerdozio di Vesta (1968)Google Scholar. Brief introductory discussions may be found in Balsdon, J.P.V.D., Roman Women: Their History and Habits² (1974)Google Scholar and Pomeroy, S. B., Goddesses, Whores, Wives and Slaves (1976)Google Scholar.

My references to modern studies in the notes do not aim to provide a full bibliography of each particular aspect under consideration; rather they aim, individually, to provide brief attestation for each point and, together, give some idea of the range and quality of modern works on the Vestals.

² The term ‘sexual’ will be used throughout this paper to differentiate not only between male and female, but also between different developmental stages of a woman’s career, virginal, matronal, post-menopausal and so forth.

³ A. D. Nock, ‘A Diis Electa: A Chapter in the Religious History of the Third Century’, HTR 23 (1930), 251–74 (reprinted in Nock, Arthur Darby, Essays on Religion and the Ancient World, ed. Stewart, Zeph (1972) 1, 252–70Google Scholar) tries to draw some potentially important conclusions from an apparent upsurge of interest in the Vestals and the Vesta cult in the mid-third century A.D. This, however, places excessive reliance on negative evidence, especially archaeological. It is surely unwarranted, for example, to draw detailed comparisons between different surviving phases of the Atrium Vestae, when that building suffered so badly from fire during its history. See Van Deman, E. B., The Atrium Vestae (1909)Google Scholar.

⁴ See, for example, H. J. Rose, ‘De Virginibus Vestalibus’, Mnem. N.S. 54 (1926), 442–3.

⁵ Sokolowski, F., Lois sacrées des Cités Grecques, Supplément (1962)Google Scholar, nos. 54 and 91. More generally, see A. D. Nock, ‘Eunuchs in Ancient Religion’, ARW 23 (1925), 25–33 (reprinted in Essays 1, 7–15) and Fehrle, E., Die kultische Keuschheit im Altertum (1910)Google Scholar. The latter also deals with the notion of the sacred marriage of the priestess to the deity, surely inapplicable in this case (but see pp. 215–17).

⁶ One might also note in the context of ‘holy virginity’ the implicit or explicit comparison of Vestals with Christian nuns (e.g. SirCato Worsfold, T. Bt., The History of the Vestal Virgins of Rome (1932). 11Google Scholar)—or even with dons at Oxbridge women's colleges (Balsdon, op. cit., 242).

⁷ ‘Order’ and ‘College’ are used loosely to refer to the Vestals throughout this article. Strictly speaking they were, if not a subsection of, at least associated with the Pontifical College.

⁸ This is not accepted by A. Brelich, who regards the Vesta cult as ‘public’ from its inception. See Vesta (1949), esp. 9.

⁹ This is the correct term for what is commonly called the Temple of Vesta, for the building was not in fact an augurated templum in the technical sense. See Varro ap. Aul. Gell. XIV, 7, 7.

¹⁰ See, for example, G. Wissowa in Roscher, Myth. Lex. VI, 260 and Guizzi, op. cit. (n. 1), 109 (though it will be seen that Guizzi's final solution to the problem of the status of the priestesses is somewhat more subtle than this stress on their marital aspect might suggest).

¹¹ For further discussion of these festivals and a collection of the ancient evidence, see Warde Fowler, W., The Roman Festivals of the Period of the Republic (1899), 71–2Google Scholar (Fordicidia), and 115 and 206–9 (Consualia).

¹² Wissowa, op. cit., 266–8. The cult title of the goddess herself, ‘Vesta Mater’ (see, e.g., Cic., Har. Resp. 12; ILS 2942, 4930, 4937), and the presence of a phallos in the temple (Pliny, NH 28, 39) have also been adduced in support of this fertile matronal aspect of the priestesses.

¹³ See Dragendorff, op. cit. (n. 1); G. Wissowa, RK² (1912), 509 n. 5; F. Guizzi, op. cit. (n. 1), 111–12; and Jordan, H., Der Tempel der Vesta und das Haus der Vestalinnen (1886), 43–56Google Scholar.

¹⁴ The fullest ancient description of -the captio is that of Aulus Gellius (1, 12, 10–14, on which the following account is based.

¹⁵ For the ‘capture’ of the bride, see Festus p. 364/5L. The comparison between this ceremony and the Vestal captio is made, for example, by I. Santinelli, ‘La condizione giuridica delle Vestali’, Riv Fil 32 (1904), 63.

¹⁶ Wissowa, RK ² 510 n. 4.

¹⁷ Aron, G., ‘Etudes sur la condition juridique des prêtres à Rome: les Vestales et le Flamine de Jupiter’, Nouvelle Revue Historique de Droit Français et Etranger 28 (1904), 33;Google Scholar Santinelli, art. cit. (n. 1).

¹⁸ See Guizzi, op. cit. (n. 1), 113 and Fehrle, op. cit. (n. 5), 206–10. It might be adduced in favour of this view that a lack of soberness in dress and manners could itself lead to suspicion of a Vestal's unchastity. See Livy IV, 44, 11, where Postumia comes under suspicion ‘propter cultum amoeniorem ingeniumque liberius quam virginem decet’ (cf. Val. Max. 11, 1).

¹⁹ There were six priestesses throughout the historical period. Earlier there had been fewer, though the exact number is debated. See Dion. Hal., Ant. Rom. II, 67, 1 (four) and Plutarch, , Numa X, 1Google Scholar (two).

²⁰ Wissowa, Myth. Lex. 264. Alternatively the Virgo Vestalis Maxima alone might be seen in the role of materfamilias (Balsdon, op. cit. (n. 1), 235, but note that he (323 n. 3) wrongly quotes Mommsen as a supporter of this view; Röm. Strafr. (1899), 18 clearly states that Vestals are to be seen as daughters of the early kings).

²¹ Throughout the historical period Vestals were allowed to leave the order after thirty years’ service and even marry, if they wished. Apparently few did so and there was a tradition of bad luck associated with such marriages (Dion. Hal., Ant. Rom. II, 67, 2; Plut., Numa X, 1–2. For long periods of service within the order, see Tac, Ann. II, 86 and ILS 4923). H. J. Rose (Mnem. N.s. 54 (1926), 446–8) lays great stress on a passage of Dionysius (Ant. Rom. I, 76, 3) which states, in the context of the unjust treatment of Rhea Silvia, that the length of service in earliest times was only five years. R. regards this as almost conclusive proof that the Vestals should be seen, in origin, as the daughters of the early kings, serving the cult before their marriage. However, while rightly drawing attention to a neglected passage, R. seems to lay too much stress on such isolated testimony, which seems contradicted by the popular tradition about the Vestals and could well be the product of later rationalizing—on the lines of, ‘Originally the Vestals were both fewer in number and served for a shorter period of time’. I would prefer to see the thirty year term not as an extension of an original five years, but as related to women's child-bearing capacity; for it is often found in traditional societies that women undergo a marked change in their religious roles after their years of fertility. (See, e.g., A., and Strathern, M., Self-Decoration in Mount Hagen (1971), 104.Google Scholar) Whatever the exact age of menopause in antiquity—on which there is some debate (see, e.g., Frommer, D. J., ‘Changing age of the menopause’, British Medical Journal 2 (1964), 349–51;CrossRefGoogle ScholarPubMedAmundsen, D. W. and Diers, C. J., ‘The age of menopause in Classical Greece and Rome’, Human Biology 42 (1970), 79–86Google ScholarPubMed—it is certain that a woman's fertility would have declined markedly by her fortieth year (see in general Hawthorn, G., The Sociology of Fertility (1970)Google Scholar and Petersen, W., Population³ (1975), 190–219)Google Scholar; thus it is likely that the child-bearing years would be perceived as ending around that time.

²² Plut., QR 85, with H. J. Rose, ‘Iterum de Virginibus Vestalibus’, Mnem. N.s. 56 (1928), 79–80 and H. Hommel, ‘Vesta und die frührömische Religion’, ANRW 1, 2 (1972), 404.

²³ Rose, Mnem. N.s. 54 (1926), 445–6 and ‘The Bride of Hades’, CPh 20 (1925), 240–2.

²⁴ For the red flammeum of the bride, see Festus p. 79 and 82L; for the white Vestal suffibulum, Festus p. 474/5L and Suda s.v. Νουμᾶς Πομπίλιος.

²⁵ This is the view of, for example, C. Koch, RE VIII A 2, 1743.

²⁶ See Rose, Mnem. N.s. 54 (1926), 444.

²⁷ Aul. Gell. 1, 19.

²⁸ A. vom Blumenthal, ‘Zur römischen Religion der archaischen Zeit—1. Zur captio der Vestalinnen’, RhM 87 (1938), 268–9. Other solutions include those of G. May, ‘Le Flamen Dialis et la Virgo Vestalis’, REA 7 (1905), 14–15 (that amata is the past participle of emere, in the sense of ‘to acquire’) and G. Dumézil, ‘QII 18, “Te, amata, capio”’, REL 41 (1963), 89–91 (that it may be related to similar titles of affection found in other areas of Indo-European fire cult). For a full bibliography, see Guizzi, op. cit. (n. 1), 130–7.

²⁹ Serv. ad. Aen. x, 228.

³⁰ See Volterra, ‘Il preteso tribunale domestico in diritto romano’, RISG 2 (1948), 103 ff. V. argues that while the existence of a consilium necessariorum is well attested in the ancient sources, there was no obligation on the paterfamilias to consult it, even in the exercise of his powers of life and death over his family. Thus it did not operate as a iudicium or court, in the manner of the pontifical college in cases of incest. But cf. Kunkel, ‘Das Konsilium im Hausgericht’, ZSS 83 (1966), 219–51.

³¹ For an interesting discussion of the stress placed by the ancients on these separate categories, representing successive stages in a woman's life, see Detienne's, M. treatment of the Aristaeus and Orpheus myth (‘Orphée au miel’ in Faire de l'histoire III, ed. Le Goff, J. and Nora, P. (1974), 56–75)Google Scholar.

³² Tuccia, for example, carried water in a sieve, Dion. Hal., Ant. Rom. II, 69; Tertullian, Apol. XXII, 12; Augustine, CD x, 16.

³³ This seems to have been the case in 216 (Livy XXII, 57, 2–5; Plut., Fab. XVIII, 3) and 114–113 B.C. (Asc. p. 45–460; Obs. 37; Mac, Sat. 1, 10, 5; Dio XXVI, fr. 87; Val. Max. III, 7, 9). The best discussion of the later incident is by North, J. A. (The Interrelation of State Religion and Politics from the Second Punic War to the time of Sulla, Unpublished DPhil. thesis, Oxford 1968)Google Scholar who relates it convincingly to the situation of increasing anxiety about the Cimbri and Teutones, who invaded Italy in 113 B.C.

³⁴ See, especially, Pliny, Ep. IV, 11. There has been much debate as to whether the burial alive of the guilty priestess should be seen as a punishment for crime or as the removal of a prodigium. For the opposing views see Wissowa, Myth. Lex. VI, 260–2 and Koch, RE VIII A2, 1747–52.

³⁵ The relief sculpture depicting Vestals may conveniently be found in I. Scott Ryberg, ‘Rites of State Religion in Roman Art’, MAAR 22 (1955). The statues from the Atrium Vestae are illustrated by E. B. Van Deman, ‘The Value of the Vestal Statues as originals’, AJA 12 (1908), 324–42. She gives a critical evaluation of their value as evidence in the study of the Vestal order.

³⁶ This is clearly represented in sculpture, for example, on the so-called Sorrento and Palermo bases, Scott-Ryberg, op. cit., figs. 26 and 27.

³⁷ See Ovid, , Fasti VI, 457;Google Scholar Prudentius, Contra Symm. II, 1095; and Daremberg-Saglio, , Diction naire des Antiquités Grecques et Romaines IX, 2, 950Google Scholar.

³⁸ Tib. 1, 6, 67; Plaut., Miles 792.

³⁹ Festus p. 454L: ‘Senis crinibus nubentes ornantur, quod [h]is ornatus vetustissimus fuit. Quidam quod eo Vestales virgines ornentur, quarum castitatem viris suis † sponoe *** a ceteris’. It is unclear exactly how the hair was arranged in this style. Presumably it was tied together into six locks and then in some way intertwined. It is possible that it is represented in a stylized way on at least one of the statues of priestesses from the Atrium Vestae. See Jordan, op. cit. (n. 13), tab. x, fig. 11. Note also the illustrations in Wüscher-Becchi, E., ‘Die Kopftracht der Vestalinnen und das Velum der “gottgeweihten Jungfrauen”’, Röm. Quartalschrift 16 (1902), 313–25Google Scholar.

⁴⁰ Firth, R., We, the Tikopia² (1957), 507Google Scholar; idem, Symbols Public and Private (1973), 288–90.

⁴¹ It is possible that the bridal parallels extend further than this. For example the cutting of the Vestals' hair has been compared to the touching of the bride's head with the so-called hasta caelibaris, though this perhaps is more tenuous. For further discussion and references, see Guizzi, op, cit. (n. 1), 110–11.

⁴² For the matronal character of these rites, see, esp., Plut., Caes. IX. Maidservants, flutegirls and the like were, of course, also present, in attendance, and it was through corrupting one of these that Clodius, disguised as a woman, gained admittance in 61. The leading role of the Vestals at the ceremony is indicated by Cic, Har. Resp. 37 and Ad. Att. 1, 13, 1. For an interesting discussion of the significance of the whole cult, see G. Piccaluga, ‘Bona Dea’, SMSR 35 (1964). 195–237.

⁴³ Pighi, J. B., De Ludis Saecularibus (1965), 189Google Scholar (with 151, ll. 35–6). It is also suggested by Wissowa (Myth. Lex. VI, 267) that they should be seen as associated with the matronae in the procession that went out from Rome to meet Octavian on his return from Actium (Dio LI, 19, 2).

⁴⁴ Dio LVI, 10, 2.

⁴⁵ Of course, a rejection of these admittedly tentative arguments concerning the masculine status of Vestals does not involve the devaluation of the clear ambiguity 1 have established between their virginal and matronal roles.

⁴⁶ Plut., Numa x, 3; Dio XLVII, 19, 4. Dio dates the grant of a lictor to the Vestals to 42 B.C., while Plutarch ascribes it to their ancient privileges. Mommsen's reasons for preferring the account of Dio (because of the unlikelihood of a woman having a lictor in early Rome) seem inadequate in this context (Röm. Staatsr.³ I, 391, n. 3).

⁴⁷ Tac., Arm. 1, 14, 3; Dio LVI, 46, 2.

⁴⁸ Aul. Gell. VII, 7, 2; Plut., Publ. VIII, 4.

⁴⁹ Evidence given by women is attested at Cic., 2 Verr. I, 37, 93–4; Suet., Claud. XV, 2 and XL, 2. The whole issue is complicated by Tac., Ann. II, 34, a passage concerning Urgulania, the friend of Livia, whose influence in the state was great: ‘Ceterum Urgulaniae potentia adeo nimia civitati erat, ut testis in causa quadam, quae apud senatum tractabatur, venire dedignaretur: missus est praetor qui domi interrogaret, cum virgines Vestales in foro et iudicio audiri, quotiens testimonium dicerent, vetus mos fuerit’. If it can be assumed that Urgulania was, at least, an ex-Vestal, then this passage would support the views of Aulus Gellius and Plutarch (n. 48). If, however, she never had any connection with the Vestal order (and there is no evidence for this) the sense of the Latin would have to be, ‘although it is the custom even for Vestal virgins (sc. who are very special, holy people) to give their evidence openly in court’. This, of course, does not accord well with the position of Gellius and Plutarch.

⁵⁰ Aul. Gell. I, 12, 9.

⁵¹ Gaius, Inst. I, 145.

⁵² Cic., Top. IV, 18 concerns the invalidity of a woman's will made without capitis deminutio. of course women did manage to fulfil all the requirements and bequeath property during the Late Republic. See, e.g., Cic., Ad. Att. VII, 8, 3; Pro Cluentio XIV, 41; 2 Verr. I, 43, 111. For further discussion of the testamentary restrictions on women, see Watson, A., The Law of Succession in the Later Roman Republic (1971), 22–3Google Scholar and Buckland, W. W., A Textbook of Roman Lam (rev. P. Stein, 1963), 288Google Scholar.

⁵³ Gaius, Inst. I, 157 and 171. From then on any woman was free to change her tutor if she was inconvenienced by him.

⁵⁴ Gaius, Inst. II, 112–13.

⁵⁵ It is alluded to by Dumézil, who points out (Archaic Roman Religion (English ed., with revisions, 1970), 587) that among primitive peoples virginity is often conceived ‘as an intermediate stage between femininity and masculinity’ and refers in this context to some of the Vestals' male characteristics.

⁵⁶ RE VIII A², 1717–76. Koch's attempt to demonstrate a fundamental relationship but consequent crucial opposition between Vestals and Flamen Dialis is, of course, of great interest, especially in the context of recent anthropological work on the role such oppositions play in the religion of many traditional societies. However, while the broad similarity of some of the privileges of these two early priesthoods has long been recognized, it seems to me dangerous to overstress the relationship—as K. tends to. For example, while it is correct to point out that both Vestal and Flamen came out of patria potestas on entry to their priesthood, the significance of this privilege was surely mu h greater for a girl (under the age of ten) than for a male (and adult) Flamen. Similarly, although the praetor's edict united the two priesthoods in the following way, ‘sacerdotem Vestalem et flaminem Dialem in omni mea iurisdictione iurare non cogam’ (Aul. Gell. x, 15), as K. himself admits (1736), there were occasions on which the Vestals were compelled to take an oath (Plut., Numa x), while the Flamen never could. Doubts must also exist over some of the postulated ritual connections between the two priesthoods. Dumézil, for example, has shown that the role of the Vestal Virgin along with the Flamen Dialis at the ceremony of the Equus October is little more than wishful thinking (Archaic Roman Religion, 220–4).

⁵⁷ Brelich, op. cit. (n. 8), esp. 58.

⁵⁸ See, e.g., th e reviews by Weinstock, S. (JRS 40 (1950), 150–1)Google Scholar, J. Bayet (REL 28 (1950), 452–4) and H. J. Rose (CR N.s. 1 (1951), 107–8).

⁵⁹ Brelich, op. cit., esp. chap. 6., ‘Mutterschaft und Jungräulichkeit’, 57–67.

⁶⁰ These ideas are most fully discussed in chapters 5–7 (‘Die “Elemente”’, ‘Mutterschaft und Jungfräulichkeit’ and ‘Das Männliche’), 48–85.

⁶¹ Guizzi, op. cit. (n. 1), 106–8. See also G. May, art. cit. (n 28), 4–14.

⁶² Hommel, art. cit. (n. 22).

⁶³ See, esp., 406–16.

⁶⁴ Esp. 403–5. The two aspects are conveniently summarized at 416–19.

⁶⁵ I have not here considered discussions of the Vestals which do not relate their position to the traditional family structure of Rome. One might note in passing, however, the views of Dumézil and Lambrechts. D. (op. cit. (n. 1), 585–7), while mentioning the intermediate status of the virgin, lays greatest stress on the connection of virginity to royal power. L. (‘Vesta’, Latomus 5 (1946), 321–9)Google Scholar recognizes the apparent contradictions inherent in their dress and privileges, but regards them as brides of the god of the underworld. (Similarly, Kristensen, W. B., ‘De Antieke Opvatting van Dienstbaarheid’, Mededeelingen D. K. Acad. van Wetenschappen—Letterkunde 78 (1934), 83–114Google Scholar.) For Dumézil's view of the position of Vesta herself, see Tarpeia (1947), 100–09.

⁶⁶ Purity and Danger: An Analysis of the Concepts of Pollution and Taboo ² (1969).

⁶⁷ Purity and Danger, 55.

⁶⁸ ibid., 41–57. See also, Rogerson, J. W., Anthropology and the Old Testament (1978), 112–14Google Scholar. The traditional theological position on the Abominations of Leviticus and other such prohibitions may be found conveniently in Encyclopedia Judaica (1971) s.v. Dietary Laws.

⁶⁹ ‘Animals in Lele Religious Symbolism’, Implicit Meanings: Essays in Anthropology (1975) (reprinted from Africa 27 (1957), 46–57)CrossRefGoogle Scholar.

⁷⁰ Bulmer, R., ‘Why the cassowary is not a bird’, Rules and Meanings: The Anthropology of Everyday Knowledge, ed. Douglas, M. (1973), 167–93Google Scholar (reprinted from Man, N.s. 2, 1 (1967), 5–25); Tambiah, S. J., ‘Animals are good to think and good to prohibit’, Ethnology 8, 4 (1969), 423–59CrossRefGoogle Scholar (partially reprinted in Rules and Meanings, 127–66). Douglas discusses these contributions in ‘Self Evidence’, Implicit Meanings, 276–318.

⁷¹ Of course, the frequent association of the anomalous with the sacred may well mean that the former is, in certain cases, generally perceived to cause the latter. Indeed the complex nature of the layers of causality in individual instances is perhaps not adequately treated by any of the writers referred to.

⁷² Rules and Meanings, 191–2 ( = Man, N.S. 2, 1 (1967), 21). He regards the explicit statements in Leviticus that the taboo on certain animals results from aspects of ambiguity as ‘rationalization’ by ‘sophisticated professional rationalizers’. While accepting the validity of his basic position I feel that the real intricacies of cause and effect are perhaps being oversimplified.

⁷³ The advantages of this modified approach are evident in several areas of the study of taxonomy. For example, it enables mythical hybrids to be analysed in much the same way as the interstitial pig or pangolin. These were problematic on Douglas' original theory, which considered that the protection of the taxonomic system was the ultimate cause of the special treatment of the anomalous. For, as Sperber, Dan has pointed out (‘Pourquoi les animaux parfaits, les hybrides et les monstres sont-ils bons a penser symboliquement?’ L'Homme 15, 2 (1975), 5–34CrossRefGoogle Scholar. Rev. version to appear in translation in French Anthropology since Lévi-Strauss, ed. J. Pouillon (Cambridge, forthcoming)), why, in that case, would men have invented new hybrids (the centaur or the sphinx) ‘qui ne font que compliquer la tâche’?

Sperber provides a useful bibliography on animal classification (to which add Soler, J., ‘La nourriture dans la Bible’, Annales ESC 28, 4 (1973), 943–55)Google Scholar. He himself tries to institute a yet more rigorous enquiry into the whole field of primitive taxonomy and outline a possible explanation for symbolic animals on the basis of the relationship between taxonomy and norm.

⁷⁴ Aul. Gell. I, 12, 9.

⁷⁵ ibid.

⁷⁶ idem, I, 12, 18.

⁷⁷ Pliny, Ep. VII, 19, 1–2. In this case the matron was an affinis, a relation by marriage, but the situation is still clearly to be distinguished from the return of the Vestal to her agnatic family.

⁷⁸ For the nuances of this useful term ‘evocation’, see Sperber, Dan, Rethinking Symbolism (English ed. 1975)Google Scholar.

⁷⁹ For an interesting series of papers on women's position in various religious and social groups, see Ardener, Shirley (ed.), Perceiving Women (1975)Google Scholar.

⁸⁰ See Parke, H. W. and Wormell, D. E., The Delphic Oracle (1956), I, 34–41Google Scholar.

⁸¹ Diod. Sic. XVI, 26, 2. For the Pythia portrayed as old, see Aesch., Eum. 38.

⁸² Diod. Sic. XVI, 26, 2.

⁸³ Paus. VIII, 5, 11–12.

⁸⁴ Soph., El. 1145–7.

⁸⁵ ibid., 997.

⁸⁶ Vernant, J.-P., Mythe et Pensée chez les Grecs (1965), 110–11Google Scholar.

⁸⁷ In Structural Anthropology (English Ed., 1968), 224. A clear discussion of this, based on the principles of Lévi-Strauss, may be found in Leach, E., ‘Genesis as Myth’ (in Genesis as Myth and other essays (1969)Google Scholar, reprinted from Discovery vol. 23, May 1962)Google Scholar: ‘In every myth system we will find a persistent sequence of binary discriminations as between human/superhuman, mortal/immortal, male/female, legitimate/illegitimate, good/bad… followed by a “mediation” of the paired categories thus distinguished. “Mediation” (in this sense) is always achieved by introducing a third category which is “abnormal” or “anomalous” in terms of ordinary “rational” categories. Thus myths are full of fabulous monsters, incarnate gods, virgin mothers. This middle ground is abnormal, non-natural, holy. It is typically the focus of all taboo and ritual observance’ (p. 11). See also the discussion of the structuralist analysis of Greek myths by Detienne, M. in Dionysos mis à mort (1977, English trans. Dionysos Slain, 1979)Google Scholar, chap. 1, ‘Les Grecs ne sont pas comme les autres’, 17–47. His criticism of Leach (among others), however, seems extreme. It is surely not true to say that ‘Leach en a conclu abusivement que l'aspect mediateur du mythe était sa seule fonction’ (p. 20, my italics). More wide-ranging criticisms are made by Kirk, G. S. in Myth, Its Meaning and Function in Ancient and Other Cultures (1970)Google Scholar, esp. chap. II, ‘Lévi-Strauss and the Structuralist approach’, 42–83.

⁸⁸ See, esp., Implicit Meanings, 282–7 and 297–302.

⁸⁹ M. Douglas seems to veer between overschematization and confusion in her discussion, Implicit Meanings, 287–9 (and see n. 64 above).

⁹⁰ See e.g. Purity and Danger, 2. In the course of an interesting discussion of ‘dirt’, which deals with its opposed definitions, based, on the one hand, on considerations of hygiene and, on the other, on the concept of matter out of place—or the confusion of categories—she adduces the example of an ambiguous bathroom. Thi s had been formed out of a piece of corridor and, while it included all the correct bathroom fittings and was perfectly clean, it still remained partially in its old use with its line of gumboots, stacked up gardening tools and old prints. In fact it fell between the categories of bathroom and corridor, partaking of each, and for this reason could not be used, at least by the author, without a feeling of great discomfiture.

⁹¹ See above n. 57.

⁹² See above p. 18. with n. 58.

⁹³ Prop, II, 29, 27; Ov, ., Fasti III, 417Google Scholar.

⁹⁴ Cic., Font. 47; Dam. 144; Har. Resp. 12; ILS 2942, 4930, 4937.

⁹⁵ Ov, ., Fasti VI, 291–2Google Scholar.

⁹⁶ For the cultic connection of Vesta and the ass, see Ov, ., Fasti VI, 319–48Google Scholar. Asses were decorated with garlands of loaves during the Vestalia.

For the recognition of the ass' blatant sexuality, see SHA, Commodus Antoninus x, 9;Google ScholarAntoninus Elagabalus VIII, 7; Apuleius, Met. passim and esp. x, 19–23. Also Brelich, op. cit., chap. 5, ‘Der Esel’, 85–95. (Esp. 86–8. Much of the rest of the chapter is concerned with a cosmic analogy, a comparison of the turning of the ass around the millstone and the rotation of the heavens.)

⁹⁷ Dion. Hal., Ant. Rom. II, 66, 2.

⁹⁸ LL, 61.

⁹⁹ Dion. Hal., Ant. Rom. IV, 2; Pliny, NH XXXVI, 70, 204; Plut., Fort. Rom. 323. On this and all the related stories, see Euing, L., Die Sage von Tanaquil (Frankfurter Studien zur Religion und Kultur der Antike, Band 11, Frankfurt 1933)Google Scholar, chap, III, ‘Die Stellung der Tanaquil in den Geburtsmythen des Königs Servius Tullius’, 20–40.

¹⁰⁰ Plut., Rom. II, 3–5.

¹⁰¹ Serv., ad. Aen. VII, 678.

¹⁰² Plut., Rom. III, 2–3.

¹⁰³ Lucr. II, 593; VI, 639–702; Cic., 2 Verr. IV, 106; Pliny, NH II, 193.

¹⁰⁴ Lucr. II, 214–15; Cic., Rep. VI, 15; Nat. Deor. II, 118; Verg., Aen. I, 42; Pliny, NH XVIII, 277.

¹⁰⁵ Cic., Nat. Dear, II, 41.

¹⁰⁶ Cato, Ag. 81; Ov, ., Fasti V, 515–16;Google Scholar VI, 381–2.

¹⁰⁷ Lucr. VI, 968; Ov, ., Fasti IV, 785–6;Google Scholar Sil. It. 1, 429–30.

¹⁰⁸ Lucr. V, 244–9; Cic., Ac. II, 61.

¹⁰⁹ Pliny, NH XXXVI, 202–3; Prop. I, 1, 27.

¹¹⁰ Lucr. V, 1283–5; Verg., Aen. II, 664–5; Caes., Bell. Alex, XIV; as signal, Caes., Gall. II, 33; Liv. XXV, 9, 10.

¹¹¹ Cic., 2 Verr. v, 6, 14; Top. xx, 74; Tac., Ann. xv. 57.

¹¹² Ter., Andr. 129; Ov., Am. I, 15, 41.

¹¹³ For the absence of a cult image, see Ov, ., Fasti VI, 295–8Google Scholar. The primacy of the element of fire is supported by a wide range of comparative evidence. See Dumézil, G., Archaic Roman Religion, 311–26Google Scholar (with references).

¹¹⁴ It might be argued that these oppositions are already accommodated in ancient thought by the opposition of Vulcan and Vesta and thus our cult would merely concern the civilizing aspects of fire. This is hardly plausible in the light of, for example, Vulcan's metallurgical function. However, even if it were so, I would argue that it was impossible for one side of the element to be evoked without the other. Thus, I am in opposition to Dumézil, G. (Fêtes romaines d'été et d'automne (1975), 61–77Google Scholar), who attempts to draw a sharp distinction between the fields of Vesta and Vulcan. In general terms, however, he demonstrates convincingly the ambiguity inherent in the basic element of fire.

¹¹⁵ Scott Ryberg, op. cit. (n. 35), 49–53 with figs. 26 and 27.

¹¹⁶ See, for example, the treatment of Ovid, (Fasti VI, 249–460,Google Scholar on the Vestalia) and Dion. Hal. (Ant. Rom. II, 64, 5–69, 3).

¹¹⁷ As with the Vestals, one might suggest that there is a further trigger to ambiguity, in this case in the figure of the virgin mother of Christ.

Schilling, R., in an article entitled, ‘Vestales et vierges chrétiennes dans la Rome antique’ (RSR 35 (1961), 113–29CrossRefGoogle Scholar, repr. in Rites, Cultes, Dieux de Rome (Etudes et Commentaires 92, 1979), 166–82) has attempted to draw a sharp distinction between Vestals and nuns. Some points of difference that he has demonstrated between the two categories of women seem perfectly valid, but do not directly challenge the notion of a deeper level of similarity. I would not, of course, accept his view that, while nuns were in a sense Brides of Christ, Vestals should be seen simply in the context of ritual purity.

¹¹⁸ op. cit. (n. 87), 224–7.

¹¹⁹ For an interesting study of this type, see Vernant, op. cit. (n. 86), 97–143. He is concerned with the hearth cult in Greece and discusses such aspects as the relationship of Hestia to marital exchange and the position of women in their role both as moving ‘commodities’ and as symbols of the permanence of the house.

¹²⁰ G. Dumézil, Fêtes romaines d'été et d'automne.

¹²¹ O. Dix, Unpublished M.A. thesis, University of East Anglia, 1978. I am very grateful to Ms Dix for allowing me to consult an early version of this dissertation.