1. Apart from a brief period of captivity in Persia (614-631), the relic of the Wood of the Cross was kept in Jerusalem during the three centuries from its invention until it was removed to Constantinople, probably just prior to 638 when Jerusalem surrendered to the Arabs. During these three centuries, numerous fragments were obtained by favoured pilgrims and ecclesiastics. These tiny splinters, often housed in large and elaborate cruciform reliquaries, were regarded as possessing the same powers as the main piece of the Wood of the Cross in Jerusalem, because the Cross was seen as one indivisible whole (see esp. Paulinus of Nola Ep. 32, 11 (PI 61, 336; tr. Goldschmidt, R.C., Paulinas’ Churches at Nola [Amsterdam 1940])Google Scholar. In this paper, references to the Wood of the Cross indicate specifically the main relic located in the Holy Sepulchre church in Jerusalem, but with the rider that all the other fragments distributed across Christendom should not be regarded as independent entities — still less as mere ‘symbols’: all fragments, whatever their size, were regarded as the real thing. The Jerusalem relic had of course the advantage not only of distinctive size but also of being kept and venerated at or near its locus, the rock of Golgotha. Whether or not this main piece was cruciform in shape, or even (in the literal sense) complete, its representation in visual imagery always takes the geometric form of the original cross on which Christ was crucified, because the relic was wholly identified with the original cross. Fourth- to seventh-century sources give the relic various titles: Wood of the Cross, True Cross, Tree of Life, Lignum Vitae, Lignum Sanctum Crucis, Crux Domini, Crux Christi; for the purposes of this discussion, I have treated all these titles as synonyms emphasising one or other aspect of the cross:

2. Most notably Weitzmann, K., ‘“Loca Sancta” and the Representational Arts of Palestine’, DOP 28 (1974) 40 Google Scholar; and Kühnel, B., From the Earthly to the Heavenly Jerusalem: Representations of the Holy City in Christian Art of the First Millennium. Römische Quartalschrift für christliche Altertumskunde und Kirchengeschichte, Supplement 42 (Rome 1987)66 Google Scholar (hereafter E to HJ). See also Corbo, V.C., ll Santo Sepolcro di Gerusalemme: Aspetti archeologici dalle origini al periodo crociato, Part II (Jerusalem 1981-2 Google Scholar), English text accompanying Pl.3: ‘… the hillock of Calvary … with a cross on the top’; and Pl.40: ‘… at the time of Constantine the rocky hillock of Calvary was simply crowned by a cross’.

3. Kühnel, , E to HJ 66 Google Scholar; Weitzmann, , ‘Loca Sancta’, 40 Google Scholar.

4. For a discussion of the dating of the various Constantinian buildings on the site, see Couasnon, C., The Holy Sepulchre in Jerusalem, tr. , J.-P.B. and Ross, C. (London 1974)Google Scholar esp. 14ff; and Corbo, V., Il Santo Sepolcro, vol. 1, esp. 223–228 Google Scholar.

5. The Bordeaux Pilgrim, ed. Geyer, P., Itinera Hierosolymitana, Saec IIII-VIII (CSEL, Vienna 1898) vol. 58, 21ff Google Scholar.

6. Eusebius of Caesarea, Vita Constantini, 3:28ff., tr. Bagster, S., Greek Ecclesiastical Historians, I (London 1845)Google Scholar (hereafter GEH).

7. Cyril of Jerusalem, Catechetical Lectures, 1.1, 4.14, etc. (PG 33); referred to by Walker, P. W.L., Holy City, Holy Places? (Oxford 1990) 254, n. 48 Google Scholar (hereafter HCHP).

8. Eusebius’ dedicatory oration ‘On Christ’s Sepulchre’ is dated to 335 by Walker, , HCHP, 251 Google Scholar; see also Rubin, Z., ‘The Church of the Holy Sepulchre and the conflict between the sees of Caesarea and Jerusalem’, in Levine, L.I. (ed.) Jerusalem Cathedra 2 (Jerusalem and Detroit 1982) 86 Google Scholar.

9. The Bordeaux Pilgrim (as in note 5); tr. Wilkinson, J., Egeria’s Travels to the Holy Land (revised edition) (Jerusalem 1981) 158 Google Scholar.

10. Eusebius, Vita Constantini, 3.40; tr. Bagster, GEH I.

11. Walker, , HCHP 128-129 Google Scholar; see also 252ff. Noting Eusebius’ ‘strategic silences’ on other matters not to his liking, Walker suggests that as a diligent historian, Eusebius may well have doubted the authenticity of thè Wood of the Cross; that, as metropolitan of Palestine, he may have wished to play down the ‘discovery’ because of its political value to the Jerusalem church in its attempts to achieve ecclesiastical autonomy; that, as a theologian in the Origenist tradition, he would have disliked any cult of physical relics; and that, as one whose theological emphasis was on the resurrection rather than the crucifixion, he may well have been uneasy about such a strong focus on Golgotha.

12. Walker, ibid., esp. 116ff; 254ff.

13. Cyril, Catech., 1.1, 4.14, etc. (ten occasions).

14. Cyril, ibid., 4.10, 10.19, etc. (five occasions).

15. Cyril, Epistola ad Constantium Imperatorem (PG 33) 1168B.

16. Itinerarium Egeriae, ed. Franceschini, E. and Weber, R. (CCSL 175) 27-90 Google Scholar. For a discussion of the date of her journey, see Wilkinson, , Egeria, 237-9, 330-1 Google Scholar.

17. Jerome, , Epistola ad Eustochium 108, ed. Hilberg, I. (CSEL 55).Google Scholar

18. Ibid., 9.2, tr. Wilkinson, J., Jerusalem Pilgrims before the Crusades (Warminster 1977) 49 (hereafter JPBQ.Google Scholar

19. E.g. Wilkinson, , JPBC 177(6)Google Scholar; but for an interpretation of this passage as a reference to the Wood of the Cross, see Frolow, A., La relique de la Vraie Croix: recherches sur le développement d’un culte (Paris 1961) 165.Google Scholar

20. E.g. Itin.Eg., 24.7; 25.1,8,11; 37.1,4,5,8; etc.

21. Itin.Eg., 37.1.

22. The fact that Egeria employs this standing cross as a landmark might suggest that it was a permanent fixture, although as she invariably mentions it in the context of liturgy and festivals, a moveable ceremonial cross is equally possible.

23. Itin.Eg., 25.8; see also 49.3.

24. Eusebius, , Vita Constantini, 3.40 (quoted in translation above).Google Scholar

25. Itin.Eg., 37.1

26. Itin.Eg., 48.1-2. The early 7th century Chronicon Paschale suggests that the Feast of the Invention of the Cross had begun in 334, at the same time as the inauguration of the ‘Church of the Holy Cross’: that is, Constantine’s basilica at Golgotha. See the discussion of this passage in Whitby, M. and Whitby, M. (eds.), Chronicon Paschale 284-628 AD (Liverpool 1989) 20 Google Scholar, n. 62. See also n. 49 below.

27. E.g. Grabar, A., L’Iconoclasme byzantin: dossier archéologique (Paris 1957) 28 Google Scholar, n. 2; Barag, D., ‘Glass Pilgrim Vessels from Jerusalem’, part I, Journal of Glass Studies 12 (1970) 40 Google Scholar; Wilkinson, JPBC 177.6.Google Scholar

28. Theophanes, , Chronographia, (ed. de Boor, ) 86. (AM 5920).Google Scholar

29. Kedrenos, George, Compendium Historiarum (PG 121) 644.Google Scholar

30. Xanthopoulos, Nikephoros Kallistos, Eccl. Hist. 14.9 (PG 146) 1084.Google Scholar

31. Holum, K., ‘Pulcheria’s Crusade AD 421-22 and the Ideology of Imperial Victory’, GRBS 18 (1977) 163, n. 46Google Scholar. The earlier date would however make no difference to the present argument.

32. The only scholar, as far as I am aware, who raises doubts (other than about the dating of Theophanes’ account) is Wortley, John, ‘The Trier Ivory Reconsidered’, GRBS21 (1980) 381ffGoogle Scholar. Wortley does not specifically question the gift of a jewelled cross to Jerusalem, but regards the translation of relics of Stephen to Constantinople in 420/1 as ‘a fictitious event’ (p.382).

33. Wilkinson, , JPBC 177.6Google Scholar, assumes this to be the case. Barag, , ‘Glass Vessels’, 40 Google Scholar, says that if ‘Aetheria’s’ account is not later than the early 5th century, then Theodosius II must have replaced the cross, of which she wrote, with the ‘crux gemmata’; see also Barag, ibid., n. 21. Kühnel, E to HJ 66, claims that it was Theodosius I (sic) who renewed the ‘crux gemmata’.

34. Holum, , ‘Pulcheria’s Crusade’, 164 Google Scholar; Holum, , Theodosian Empresses: Women and Imperial Dominion in Late Antiquity (Berkeley 1982) 108 Google Scholar; Holumand, K.G., Vikan, G., ‘The Trier Ivory, “Adventus” Ceremonial, and the Relics of St. Stephen’, DOP 33 (1979) 131-2.Google Scholar

35. Wortley, , ‘Trier Ivory Reconsidered’, 383ff, esp. 389, n.30.Google Scholar

36. The suggestion that part, or all, of this story has been made up would be quite plausible for Theophanes’ early 9th century context of post-iconoclasm emphasis on images and relics, and in particular, of iconodule attempts to reclaim the cross as an orthodox imperial symbol after its virtual hijacking by the iconoclast emperors. Even the image of a jewelled cross on Theodosian coinage of 421 -2 does not really provide confirmation of Theophanes’ story, as it appears to be a processional cross (held by a figure of Victory). Holum (‘Pulcheria’s Crusade’, 166-7) suggests that this cross is most likely patterned on a ceremonial cross in the palace at Constantinople, which he proposes as the ‘common model’ for the solidi of Theodosius II and the ‘Golgotha cross’. (The fact that one image is patterned or modelled after the form of another image does not of course indicate that the second image necessarily represents the first: it may still, in its primary role, be an image of the prototype — the most obvious example being the icon of a holy person). Holum’s theory may be supported by the fact that the jewelled cross on the Theodosian coins is not pictured in conjunction with a hill, or even on a stepped base: the cross on a stepped base does not appear until the late 6th century, on the coinage of Tiberius II. It is noteworthy, however, that when these stepped crosses do appear on coins, they are not depicted as being jewelled.

37. See e.g. Grabar, , L’Iconoclasme byzantin, 28, n. 2Google Scholar; and Barag, , ‘Glass Vessels’, 40, n. 27Google Scholar. For the claim that the mosaic depicts an earlier imperial gift, see Kühnel, , E to HJ, 66 Google Scholar (here, ‘Theodosius I’, with no supporting evidence); and Conant, K.J., ‘The Original Buildings at the Holy Sepulchre in Jerusalem’, Speculum 31 (1956) No. 1, 5-7 Google Scholar, who identifies the S.Pudenziana cross with an ‘imperial gift’ on the Golgotha rock, dating from the mid-fourth century or ‘soon’ after (again, without supporting evidence).

38. The S. Pudenziana apse mosaic originally included a dedicatory inscription on its lower edge, dating the decoration of the church to the time of Pope Innocent I (402-417); see Dassmann, E., ‘Das Apsismosaik von S.Pudentiana in Rom’, Römische Quartalschrift für christliche Altertumskunde und Kirchengeschichte 65 (1970) 67-68.Google Scholar

39. Barag, , ‘Glass Vessels’, 40, n. 27.Google Scholar

40. Breviarius de Hierosolyma (ed. Weber, R., CCSL 175) 105-112 Google Scholar; tr. Wilkinson, , JPBC, 59-61 Google Scholar. For a discussion of the dating, see Wilkinson, ibid., 4-5, 183.

41. Ibid., 5.

42. See Appendix A.

43. See, e.g. Wilkinson, , JPBC, 177.6.Google Scholar

44. A and B lines 9-10; A lines 41-42; B line 42: ‘crux Christi’.

45. B lines 40-43: ‘crux Christi … et ipsa crux …’.

46. Paulinus of Nola, Ep. 31 (PL 61) 325f.

47. B lines 40-42.

48. Form A: ‘caelum desuper patente’. Form B has: ‘celum desuper aureum’. The apparent contradiction is discussed by Wilkinson, JPBC 59 n. 2; 182 n. 6a, 6b. See also note 56 below.

49 Theodosius, , De situ terrae sanctae, 31 (CCSL 175) 124 Google Scholar; for a discussion of the dating of this work, see Wilkinson, , JPBC, 5, 185 Google Scholar. The September Feast of the Holy Cross was timed to coincide with the date of the founding of Solomon’s temple as recorded in the Old Testament, and was also (in reference to this) called the Encaenia (Itin.Eg., 48.1-49.3). This festival celebrated not only the Invention of the Cross but also the founding of Constantine’s basilica at Golgotha (by Egeria’s time, claimed to have been built over the spot where the relic was found) and the Anastasis rotunda built over the Tomb of Christ, which for Christians symbolically replaced the Jewish temple. This indicates that at least by the 380s, and probably earlier, the Wood of the Cross was not associated only with Easter, but had become virtually an independent entity, validated almost as much by its locus (which it in turn validated) as by its history.

50. See also Barag, , ‘Glass Vessels’, 47 and n. 69Google Scholar, who notes that ‘Gregory of Tours (d.593/4) states that this ceremony (i.e. the veneration of the Cross) takes place on Ash Wednesday’ — whereas Egeria had noted it only on Good Friday.

51. Itin.Eg., 37.1

52. On documented fragments of the cross, see Frolow, Vraie Croix.

53. The reliquaries in the papal oratories are recorded in the Liber Pontificalis, ed. Duchesne, vol. I, 242, 261; see Frolow, , Vraie Croix, 173 (20, 23)Google Scholar. For further contemporary examples of gold, jewelled and cruciform reliquaries of True Cross fragments, see ibid., 176 (26), 177 (29), etc.

54. Frolow, , Vraie Croix, 180-1 (34)Google Scholar; no doubt, this circle of twelve stones around the True Cross fragment was symbolic of Christ surrounded by his twelve apostles, as in the iconography of some Early Christian church decoration: for example, the apse mosaic in Paulinus’ church at Nola, where the apostles were represented as a circle of doves around the cross ( Paulinus, , Ep. 32, ed. Goldschmidt, , Churches, 38)Google Scholar. But the point about the design of this reliquary, in the present context, is the close association between jewels and the Wood of the Cross — as also in the Breviarius.

55. The text, in both forms A and B, first mentions ‘silver screen(s)’ around the rock itself (lines 35-6). Only A mentions the ‘silver doors’, in an ambiguous sentence: ‘Habet ostia argentea ubi fuit crux domini exposita’ (lines 38-42A), which could refer either to the screen around the rock, or to one around, or in front of, the ‘crux domini’. Both A and B then describe what is apparently another screen (lines 45-6), associated with the ‘crux domini’ rather than the rock; in A, this second screen is ‘auro et argento multum ornatae’.

56. cf. Itin.Eg., 25.8, where reference is made to the rich silk and gold textiles used at festivals. The existence of a temporary baldachin might explain the apparent contradiction at this point in the Breviarius (1.44) between version A and version B: the cross, if displayed under a silk and gold canopy out in the court, would still have been ‘under open sky’ (A) and yet ‘under a golden sky/canopy’ (B).

57. Itin.Eg., 37.3.

58. This process of transferral has been discussed by a number of scholars, e.g. Wilkinson, , Egeria, 302ffGoogle Scholar; and Ousterhout, R., ‘The Temple, the Sepulchre, and the Martyrion of the Saviour’, Gesta 29/1 (1990) 44-53.CrossRefGoogle Scholar

59. lines 50-54. A mentions both; B the Horn of Anointing, but not the Ring of Solomon.

60. lines 46-49.

61. See Appendix A.

62. It is even possible that Form A, if taken on its own, does not refer, on the second occasion, to anything other than the rock, with a historical reference to the crucifixion: ‘ubi fuit crux domini exposita’ (A lines 41-42): ‘where the cross of the Lord was (once) set up’. The adjective ‘ornata’ would then have to qualify the neuter plural ‘ostia’ instead of the feminine singular ‘crux’. Considered in conjunction with Form B, however, the reference appears more likely to be to the relic, not just the rock.

63. Antonini Placentini Itinerarium (ed. Geyer, P., CCSL 175) 127-174.Google Scholar

64. E.g. Sophronius of Jerusalem, Anacreonticon 20, ed. Gigante, M. (Rome 1957)Google Scholar; Theodosius, , De situ terrae sanctae, 31 (The Feast of the Holy Cross) (ed. Geyer, , CCSL 175), 124.Google Scholar

65. For illustrations and discussion of most of the extant ampullae, see Grabar, A., Ampoules de Terre Sainte (Monza-Bobbio) (Paris 1958).Google Scholar

66. The relatively small size of the ampullae (approximately six inches in diameter in most cases) would not preclude a fine detail such as jewels being indicated, as there is a surprising amount of detail included on them: for example, a starry sky behind one of the cross-images, or the metal grille of the tomb aedicule.

67. E.g. Monza 5, Grabar, Ampoules, Plate XIa.

68. E.g. Monza 4, ibid., Plate Xb; this particular image depicts the hill or rock of Golgotha as a set of steps beneath the cross — a common alternative scheme of representation, referring to the steps cut into the rock. A small proportion of these flared and knobbed ampulla crosses (e.g. Monza 13, ibid., Plate XXV) are depicted without either the hill or the steps, but under a stylised arch comprising two lateral columns joined at the top by an arc in the form of a garland; the arch may refer to an actual shrine or to the chapel in which the Wood was kept near the rock, or it may simply be an artistic convention for depicting something to which great honour was due: cf. Monza 4 (PI. Xb) which has a similarly shaped arch (plus base) consisting entirely of a garland surrounding both cross and steps.

69. E.g. Bobbio 1, 2, ibid., Plates XXXII, XXXIII; Monza 10, Plate XVI.

70. Translated, this form of inscription reads: ‘oil of the Wood of Life from the holy places of Christ’, in reference to the fact that these ampullae contained oil sanctified by contact with the Wood of the Cross.

71. E.g. Weitzmann (as in note 2).

72. Grabar, , Ampoules, 55-56.Google Scholar

73. Ibid., 56.

74. Vikan, G., Byzantine Pilgrimage Art (Washington D.C. 1982) 22-24 Google Scholar (hereafter BPA); and Vikan, G., ‘Pilgrims in Magi’s Clothing: The Impact of Mimesis on Early Byzantine Pilgrimage Art’, ed. Ousterhout, R., The Blessings of Pilgrimage (Illinois Byzantine Studies I. Urbana and Chicago 1990) esp. 102-3 Google Scholar. See also Grabar, , Ampoules, 65-66 Google Scholar. However, Kühnel, E to HJ, 95-6, interprets these ampulla crosses as the relic adored by the pilgrims, and also as the monument. Vikan identifies the two kneeling figures as pilgrims, from the style of their hair and beards, and their foreign, exotic clothing — such as trousers (BPA 24). He also suggests that their hands, reaching forward as if to touch the cross, display ‘a characteristic gesture of veneration’ (ibid., 23-24). If so, then this particular detail is further evidence that the iconography is symbolic rather than literal, because although this gesture was indeed characteristic of much relic veneration (and could therefore be ‘read’ as ‘veneration’), it was specifically forbidden in the case of the Wood of the Cross (Itin.Eg., 37.3). It is possible that the practice had changed since Egeria’s time; but the Piacenza Pilgrim’s account, which is contemporary with the ampullae, still appears to make a distinction between specific modes of veneration for the Wood and the Title, which were kept together: he kissed the Wood, but handled and kissed the Title (Itin., 20. The ‘title’ was a trilingual superscription originally attached to the top of the cross, and is mentioned in all four gospels: see, esp. John 19:19-22). But the iconography is clearly not intended as a literal representation of the veneration ceremony, any more than as a narrative depiction of the historical crucifixion; and Vikan’s parallel between the story of doubting Thomas (who appears on some other ampullae) and the seeing and touching experience of the pilgrim (BPA 24-25) is an appropriate one, supported by the 4th century writings of Cyril (Catech., 13.39). (For a discussion of Cyril’s teaching about Thomas, and Cyril’s distortion of the meaning of the gospel account, see Walker, HCHP, 331-2).

75. For discussion of this point, see e.g. Dassmann, , ‘Das Apsismosaik’, 77-8 Google Scholar; Conant, , ‘Original Buildings’, 5-9 Google Scholar; Kühnel, , E to HJ 66ffGoogle Scholar. There is some disagreement as to just which buildings are included in the background of the mosaic; this is particularly so of the right hand side (see Kühnel, ibid., 66ff).

76. E.g. Kühnel, ibid., 66ff; Conant, , ‘Original Buildings’, 5-7.Google Scholar

77. Conant, ibid., esp. 6ff, and Plate X.

78. Ibid., 6.

79. Ibid., 6: ‘Architecturally the solution is a good one. The large platform would be very handsome’.

80. See Corbo, Il Santo Sepolcro, II, Plate 3; III, Plates 87-89.

81. See e.g. Grabar, , Ampoules, 55-56 Google Scholar; Vikan, , BPA 22-23 Google Scholar. For a literary parallel, see note 94 below.

82. I.e. the ‘real’ heaven, not the one in the mosaic.

83. This was recorded in a drawing by Ciacconio in 1595 (Cod.Vat.Lat.5407, f81); see Dassmann, , ‘Das Apsismosaik’, Plate 3a; Kühnel, E to HJ 187 n. 34.Google Scholar

84. Paulinus of Nola, Ep. 32: ‘… the Cross and the Lamb testifying him as the sacred sacrificed One …’ (tr. Goldschmidt, Churches, 39).

85. The link between apse mosaic and high altar is made explicit in Paulinus’ iconography and in his accounts of the decoration of his churches at Nola and Fundi (Ep. 32).

86. For a discussion of the dating of the glass vessels, see Barag, , ‘Glass Vessels’, esp. 36, 48.Google Scholar

87. On their place of origin, see Barag, ibid., 48.

88. Ibid., esp. 41-44.

89. cf. Cyril, , Catech. 13.28 (PG 33, c.805)Google Scholar, in which Golgotha is called the centre of the earth.

90. cf. the ampulla images discussed above, in which the Four Rivers of Paradise flow down the rock of Golgotha from the foot of the cross.

91. Ibid., 39-41. Whilst Barag does not claim that the cross on the glass vessels is necessarily a Theodosian monumental cross, he does state that the cross on a stepped base is ‘identified with the cross actually standing on Golgotha until 614’, adding that this and the other two variants — the cross on the omphalos, and the leaved cross — ‘represent three aspects of a single element, i.e., the Cross of Golgotha’ (ibid., 44); and it is clear from the context that he is here referring specifically to a supposed monument and not to the relic of the True Cross. Elsewhere, he states that ‘The form of Theodosius’ cross [i.e. the supposed monumental ‘crux gemmata’] greatly influenced early Christian art even after the destruction of the Church of the Holy Sepulchre in 614, when this treasured symbol fell into the hands of the conquerors’ (ibid., 40). Whilst a great number of ceremonial artefacts were looted at this time, including crosses, the ‘treasured symbol’ which mattered most to the Christians was the relic of the True Cross, carried off by the Persians (who recognised its religious and also its political value as a symbol of victory), and later restored to Jerusalem by Heraclius as the crowning triumph of his campaign.

92. See e.g. the accounts of the Piacenza Pilgrim, Itin., 19; and Theodosius, De situ, 7. These are the two sources cited by Barag (‘Glass Vessels’, 40) to support his claim that pilgrims mention the steps ‘leading to the cross’ (ibid., 40, 41).

93. The Piacenza Pilgrim, Itin., 19.

94. Sophronius, Anacreonticon 20, lines 30-32; tr. Wilkinson, JPBC 91.

95. For contemporary beliefs about the ‘lignum vitae’, see e.g. Fortunatus, Venantius, Hymns 33, 34, ed. Walpole, A.S., Early Latin Hymns (Cambridge 1922) 165-177 Google Scholar. On this point, Peter Brown has suggested to me that Queen Radegund’s devotion to the True Cross (the reception of a fragment of which at her convent at Poitiers was the occasion for the writing of these two hymns), and her vows taken before this relic, indicate that ‘choosing the cross’ — in a life of religious seclusion — was identified with ‘choosing life’. See also Talbot Rice, D., ‘The Leaved Cross’, BS 11 (1950) 72-81 Google Scholar; Underwood, P.A., ‘The Fountain of Life in Manuscripts of the Gospels’, DOP 5 (1950) 95ff.Google Scholar

96. The whole question of the reliability of such texts is too complex to be dealt with here. Whether or not the texts are reliable, however, those who argue for the existence of the ‘crux gemmata’ have invariably employed these same texts to support their claims. If the texts are in fact unreliable, then the case for the ‘crux gemmata’ is even more shaky than I have suggested in this paper.

97. Adomnan, De Iocis Sanctis I, v.1 (ed. L. Bieler, CCSL 175) 190.

98. Ibid., III, iii.

99. This suggestion is made, for example, by J. Wilkinson, JPBC91, n. 15; and 177.6. See also Barag, , ‘Glass Vessels’, 40.Google Scholar