¹ Norman, P. and Reader, F.W., Archaeologia lxiii (1911–1912), 269Google Scholar f., 342, pl. LXV, 2; V.C.H., London (1909), 86, fig. 35; Royal Commission on Historical Monuments (England), London iii. Roman London (1928), pl. 13 (below).

² J.M.C. Toynbee, Art in Britain under the Romans (1964), 164 note 2. The reference is à propos of a relief from High Rochester, which certainly represents a group of three water-nymphs.

³ On the legend see Severyns, A., Le cycle épique dans l'école d'Aristarque, Bibliothèque de la Faculté de Philosophic et Lettres de l'Université di Liège 40 (1928), 266–71Google Scholar; Moreau, 15; Gury, F. in LIMC iii, I (1986), 609Google Scholar; Kahil, L. in LIMC iv, I (1988), 498, 562.Google Scholar

⁴ Generally for the artistic representations see Kahil, L. in LIMC iv (1988), 503–4Google Scholar, 562, pls 291-3; cf. LIMC iii (1986), 583 nos 185–6, pl. 471Google Scholar. For the Greek vase-paintings see also Chapouthier, F., ‘Léda devant l'oeuf de Némesis’, Bulletin de correspondence hellénique lxvi–lxvii (1942–1943), 1–21Google Scholar; J.D. Beazley, Etruscan Vase Painting (1947), 39–42.

⁵ Apollodorus, , Bibliotheca III. 10. 7Google Scholar (also quotes the Attic version of the legend); cf. Hyginus, Fabulae 77; Scriptores rerum mythicarum Latini tres Romae nuper reperti, ed. Bode, G.H. (1894), pp. 119 f. (Mythogr. II, 132).Google Scholar

⁶ Scriptores rerum mythicarum, op. cit. (note 5), p. 64 (Mythogr. 1, 204), ll.28–30; cf. Horace, Satirae II. I. 26; idem, Ars Poetica 147.

⁷ Ausonius, Epigrammata 66; idem, Griphus ternarii numeri 10; Servius, , Commentarii in Vergilii Aeneidos III. 328Google Scholar; Schol. Callimachus, In Dianam 232; Schol. QTV Odyssey xi. 298Google Scholar; Tzetzes ad Lycophron, Alexandra 88; Eustathius, , Commentarii ad Homeri lliadem XXIII. 638Google Scholar; idem, Commentarii ad Homeri Odisseam xi. 297–302; Scriptores rerum mythicarum, op. cit. (note 5), pp. 27 (Mythogr. 1,78), 163 (Mythogr. 111,3. 6).

⁸ Schefold, 1098, 1100 f., pl. 90; Moreau, 17, pl. 22 below; H. Mielsch, Römische Stuckreliefs, Mitteilungen des Deutschen Archäologischen Instituts, Römische Abteilung, Ergänzungsheft 21 (1975), 66Google Scholar, 151 (K 64a), pl. 65, 1; K. Schefold, Die Göttersage in der klassischen und hellenistischen Kunst (1981), 246, fig. 344; LIMC iii (1986), 626 no. 146Google Scholar; iv (1988), 504 no. 12, pl. 293.

⁹ Robert, C., Die antiken Sarkophagreliefs ii (1980), 6 f. no. 2, pl. 11Google Scholar; Espéiandieu, E., Recueil général des bas-reliefs, statues et bustes de la Gaule romaine i (1907), 78 f. no. 96Google Scholar; Moreau, 16, pl. 19 below; Koch, G. and Sichtermann, H., Römische Sarkophage, Handbuch der Archäologie (1982), 156 f.Google Scholar; LIMC iii (1988), 504 no. 11, pl. 293.Google Scholar

¹⁰ Metropolitan Museum of Art, inv. no. 13.225.7 (Rogers Fund 1913): Moreau, 16, pl. 13; M. Bell in K. Weitzmann (ed.), Age of Splrituality. Late Antique and Early Christian Art, Third to Seventh Century (1979), 239 f. no. 215; LIMC iii (1986), 626 no. 146aGoogle Scholar; iv (1988), 505 no. 13, pl. 293.

¹¹ Eiden, H., ‘Spätrömisches Figurenmosaik am Kornmarkt in Trier’, Trierer Zeitschrift xix (1950), 52–71Google Scholar; Parlasca, K., ‘Das Trierer Mysterienmosaik und das ägyptische Ur-Ei’, Trierer Zeitschrift xx (1951), 109–25Google Scholar; K. Parlasca, Die römischen Mosaiken in Deutschland (1959), 56 f., pls 54, 55 (1); Moreau, 10, 15 ff., colour pl. I.; R. Schindler, Landesmuseum Trier. Führer durch die vorgeschichtliche und römische Abteilung (2nd edn, 1970), 68, fig. 212; LIMC iii (1986), 626 no. 145, pl. 501; J. Schwarz and J.-J. Hatt, ‘Une interprétation nouvelle de la mosäique dite du “Kornmarkt” de Trèves’, Bulletin de la Société Nationale des Antiquaires de France (1985), 37–45.Google Scholar

¹² Moreau (10) interprets the torrent of liquid as a ‘column’; cf. Parlasca, in Trierer Zeitschrift xx (1951), 112Google Scholar, 115 f. (arguing, however, that the mosaicist has misunderstood what was in earlier works a stream of liquid).

¹³ Isiaca, Aula: Rizzo, G.E., Le pitture dell'Aula Isiaca di Caligola, Monumenti della plttura antica scoperti in Italia iii, Roma ii (1936), 12–14, figs 9, 10, pls iv, vGoogle Scholar; Schefold; Moreau, 16 f., pls 22 above, 23; Schefold, op. cit. (note 8), 245, fig. 343. Hannover gem: Antike Gemmen in deutschen Sammlungen iv. M. Schlüter, G. Platz-Horster and P. Zazoff, Hannover, Kestner-Museum; Hamburg, Museum für Kunst und Gewerbe (1975), 189 f. no. 942, pl. 124; LIMC iv (1988), 504 no. 10, pl. 293.Google Scholar

¹⁴ I am grateful to Dr Wilson for drawing my attention to this point. The simplest explanation for the difference (and the one which I prefer) is that the Romano-British sculptor devised or copied a variant form of the standard Roman iconography for the triple birth; but Dr Wilson suggests as an alternative possibility that the piece was based on a somewhat different iconographic model in which the Dioscuri were spectators rather than new-born infants. He is not convinced that the side-figures on the London relief are necessarily children.

¹⁵ On these bastions see R. Merrifield, The Roman City of London (1965), 68–72; P. Marsden, Roman London (1980), 170–3. For finds in Duke Street Merrifield, op. cit., 304, 321 f. Our stone was evidently found on the south-west side of the street (the excavation area is described as ‘between no. 7 and the synagogue at the corner of Church Passage, extending to Mitre Street on the west’: Archaeologia lxiii (1912), 269)Google Scholar; but the most likely source remains the city-wall, and specifically the nearby bastion B₇. It may well have been re-used in one of the medieval walls which criss-crossed the site, ‘chiefly on the north portion adjoining Duke Street’ (Archaeologia, loc. cit.).

¹⁶ London: Merrifield, op. cit. (note 15), 68, 112, 320 (B2). Chester: R.P. Wright and I.A. Richmond, Catalogue of the Roman Inscribed and Sculptured Stones in the Grosvenor Museum, Chester (1955). Cities of Gaul: see e.g. Espérandieu, É., Recueil général des bas-reliefs, statues et bustes de la Gaule romaine iv (1911), 3–54, nos 2756-2853 (Sens)Google Scholar; v (1913), 211–84, nos 4012–4125 (Arlon); vi (1915), 317–406, nos 5140–5224 (Neumagen).

¹⁷ For this and the remaining reliefs from the tomb see Mielsch, op. cit. (note 8), 66f., 151–4, pls 65–8.

¹⁸ See e.g. R. Ling, Roman Painting (1991), 135–41. Even in displays of statuary, where programmatic arrangements and juxtapositions can certainly be identified, choices were often dictated by other factors, formal or aesthetic: see e.g. Bartman, E., ‘Decor et duplicatio: pendants in Roman sculptural display’, American Journ. Arch, xcii (1988), 211–25CrossRefGoogle Scholar; eadem, ‘Sculptural collecting and display in the private realm’, in E.K. Gazda (ed.), Roman Art in the Private Sphere (1991), 71–88.

¹⁹ For recent surveys of the subject-matter employed in early-imperial funerary sculpture see Boschung, D., Antike Grabaltäre aus den Nekropolen Roms, Acta Bernensia x (1987), 47–52Google Scholar; Sinn, F., Stadtrömische Marmorurnen, Beiträge zur Erschliessung hellenistischer und kaiserzeitlicher Skulptur und Architektur 8 (1987), 54–83.Google Scholar It is worth quoting Sinn's comments on mythical themes on Roman sarcophagi (ibid., 81): ‘Man stellte immer wieder fest, dass nur wenige unmittelbar auf den Tod zu beziehen sind und nur ein Teil der Darstellungen als mythische Exempla gelten können, die zu Jenseitshoffnungen berechtigten oder Trost spenden konnten durch den Hinweis, dass auch Helden sterblich sind. Eine ganze Reihe der zur Darstellung gebrachten Themen sind jedoch von solchem Grauen geprägt, dass der Anlass für ihre Anwendung im Grabbereich entweder so abstrakt war, dass er nicht mehr erschlossen werden kann, möglicherweise auch in zeitgenössichen Ausdeutungen zu suchen ist, oder im Verhältnis der Römer zur griechischen Kunst begründet liegt, die per se als ehrwürdig und nobel gait und daher zum Grabschmuck herangezogen wurde ohne Rücksicht auf die Themen’. For a detailed and sensitive analysis of the various levels of meaning in sarcophagus reliefs see H. Sichtermann in Koch and Sichtermann, op. cit. (note 9), 583–617.

²⁰ Relevant here is P. Zanker's thesis that many motifs of funerary art in early-imperial times were borrowings from public art which gradually took on ‘a broad and indistinct range of meanings, embodying personal values and aspirations, equally applicable to the dead and their survivors’: The Power of Images in the Age of Augustus (1988), 276–8. For the image of the wolf and twins he suggests in one case that it is symbolic of a mother's ‘love and selfless devotion’, in another that it alludes to the dead man's ‘virtues as a Roman citizen, his correct attitude toward the new age’, in a third that it simply refers to the fact that the deceased was the twin brother of the man who commissioned the relief (but actually the inscription reveals only that they were brothers, not twins): idem in J. Huskinson, M. Beard and J. Reynolds (eds), Image and Mystery in the Roman World (1988), 8 f. On grave-altars, as Glenys Davies points out to me, the image of the wolf and twins is more usually paired with that of the doe and Telephus, an apt combination on both formal and thematic grounds.

²¹ Stupperich, R., ‘A reconsideration of some fourth-century British mosaics’, Britannia xi (1980), 289–301CrossRefGoogle Scholar; Ling, R., ‘Brading, Brantingham and York: a new look at some fourth-century mosaics’, Britannia xxii (1991), 147–57.CrossRefGoogle Scholar

²² It is hardly worth saying that our relief has nothing to do with the well-known Mithraic sculpture from Housesteads, which appears to show the birth of Mithras from an egg (instead of a rock), perhaps by assimilation to (or confusion with) the birth of the Orphic deity Phanes from the cosmic egg: see e.g. Levi, D., Hesperia xiii (1944), 299–301.Google Scholar