^{page 2 note 1} Among those who have assisted me I feel bound to record my peculiar obligations to Comm. Giacomo Boni, the Director of the Excavations in the Forum, whose kindness and sympathy have in every way made my task easier, and also to the Rev. F. E. Brightman of the Pusey House and Mr. C. H. Turner of Magdalen College, Oxford, who took the trouble to look over the proofs, and to whom I am indebted for many suggestions and corrections. Their identification of the fragmentary Patristic inscriptions, and especially Mr. Brightman's discovery of the principle on which they were selected, which is important for determining the chronology of the paintings, may without exaggeration be described as brilliant. Mr. C. H. Blakiston of the British School at Rome has also given me much practical assistance which has contributed in no small degree to such accuracy and completeness as this account possesses.

^{page 4 note 1} Printed by De Rossi, in Roma Sotterranea, i. 143Google Scholar, from the MS. at Vienna (No. 795).

^{page 4 note 2} Lib. Pont. i. 385. References to the Liber Pontificalis are given by the pages of the edition of Duchesne, Mgr. (2 vols. Paris, 1886–1892Google Scholar).

^{page 4 note 3} Duchesne, Le Forum Chrétien, 42.

^{page 4 note 4} I lay no stress on the fact that S. Maria Antiqua was a diaconia, for there is no record of the date of its institution as such, and a diaconia was sometimes established in a pre-existing church. But it may be pointed out how consistent with the date assigned above to the church (or even a later one) are the conclusions of Duchesne with regard to the diaconiae, viz. (1) that they first make their appearance in the seventh century, and (2) that they are distinguished from the older presbyteral titles by their situation in the heart of the City and, frequently, in disused ancient buildings; both indications of a relatively late date: Mélanges d'Arch, et d' Hist. 1887, 239–242.

^{page 5 note 1} H. Grisar, Civiltà Cattolica, Jan. 1901, p. 232. Marucchi, O., Nuovo Bulletino di Archeologia Cristiana, vi. (1900), 313Google Scholar.

^{page 5 note 2} On the other hand, pilgrims who visited the Vatican Basilica in the eighth century passed from the shrine sanctae Mariae qttae antiqu a dicittir …. ad sanctam Maria m quae nova dicitur. De Rossi, , Inscr. Chr. ii. p. 228Google Scholar.

^{page 5 note 3} Mélanges d'Archéologie et d'Hisloire, 1897, 28 sqq.

^{page 6 note 1} Rossi, De, Inscr. Chr.1 ii. p. 71Google Scholar.

^{page 6 note 2} Duchesne, l.c. p. 30.

^{page 6 note 3} Duchesne (l.c. p. 29) makes two suggestions: (1) That the name refers to the diaconia as being, presumably, the oldest: but we know nothing about its precise date. (2) That it belongs to a picture in the Church, citing a statement of the Liber Pontificalis (i. p. 419) that Gregory III. encased in silver imaginem sancte Dei Genetricis antiqiiam. So Grisar, Civiltá Cattolica, March, 1901, p. 738. But the Liber Pontificalis gives no clue as to where the picture was.

^{page 6 note 4} Lib. Pont. i. 385.

^{page 7 note 1} Lib. Pont. i. 385.

^{page 7 note 2} Rossi, De, Inscr. Chr. ii. p. 442, l. 9Google Scholar.

^{page 7 note 3} Cf. Lanciani, Ruins and Excavations of Ancient Rome, 156, where the references are given.

^{page 8 note 1} Cf. the case of S. Maria Nova mentioned below; built before 855, not decorated before 858.

^{page 8 note 2} Lib. Pont. ii. 14.

^{page 8 note 3} Lib. Pont. ii. 145. There is no mention of the fact in the Life of Leo IV.

^{page 9 note 1} Lib. Pout. ii. 158.

^{page 9 note 2} Lib. Pout. ii. 108. The tenth indiction ended on Aug. 31st, 847.

^{page 9 note 3} Perhaps this accounts for the repetition in the outer parts of the church of paintings found also in the interior (cf. esp. p. no).

^{page 9 note 4} Lanciani, Ruins and Excavations, 245.

^{page 10 note 1} Mélanges, 1897, 15 sqq.

^{page 10 note 2} E.g. Marucchi, , in N. Bull. Arch. Cr. vii. (1900), 319Google Scholar; Grisar, Civiltà Catlolica, March, 1901, 736 (suggesting that the picture described below, p. 102, marks the site of S. Silvester in lacu).

^{page 10 note 3} Jordan, , Topographie, ii. 655Google Scholar; Urlichs, Codex Topographicus, 75.

^{page 10 note 4} Jordan, , Top. ii. 635Google Scholar.

^{page 10 note 5} Mélanges, 1897, 13, sqq.

^{page 10 note 6} MS. at Turin. Printed in Armellini, Chiese di Roma, 53.

^{page 11 note 1} Jordan, , Topographie, ii. 500Google Scholar; Duchesne, l.c. 17. If it were not for the difficulties of date, one might have been inclined to look for. S. Silvester in the building (L on the plan: cf. p. 108) between the entrance to the church and the Lacus Iuturnae.

^{page 11 note 2} Diehl, Études sur l'Administration Byzantine dans l'Exarchat de Ravenne, 241 sqq.

^{page 11 note 3} Diehl, 278.

^{page 12 note 1} First published by Didron, , Manuel d' Iconographie Chrétienne (Paris, 1845)Google Scholar.. English translation by Stokes, M., Christian Iconography (London, 1891), ii. 265Google Scholarsqq. There is also a modern Greek edition (Athens, 1885).

^{page 13 note 1} The multiplication of altars in this period is illustrated by the inscription of Gregory III. (731–741) in St. Paul's without the Walls, regulating the oblations at the six daily masses. De Rossi, , I.C.R. ii, p. 423Google Scholar, n. 41. Grisar, , Anakcla Romana, i. 169Google Scholar, and T. iii. 3.

^{page 14 note 1} Cf. p. 93.

^{page 14 note 2} But see p. 31.

^{page 15 note 1} It is unfortunate that Miss Stokes, in her translation of Didron, omitted this part of the Manual (Christian Iconography, ii. 378).

^{page 15 note 2} St. Peter appears regularly with the key at an even earlier period. The only representation of him in S. Maria is too ruined to show whether he had it here.

^{page 15 note 3} See pp. 78, 79, 98.

^{page 15 note 4} E.g. the well-known Greek Bibles in the Vatican Library (746, 747).

^{page 17 note 1} Bede, Hist. Abb. 6, ed. Plummer (vol. v. of Works, ed. Migne, p. 718).

^{page 17 note 2} Bede, l.c. 9 (p. 720).

^{page 17 note 3} pp. 63, 87, 88.

^{page 19 note 1} See pp. 62, 72.

^{page 19 note 2} The unity of the plan and the conditions of the site are conveniently brought out in Figs. 47 and 48 (published of course before the recent excavations) of Lanciani's Ruins and Excavations, pp. 121, 123.

^{page 20 note 1} The letters and numerals in brackets refer to the plan, p. 18.

^{page 20 note 2} The further one has been restored.

^{page 20 note 3} The arcade has been restored, the arches being united to the capitals by means of the splayed impost blocks (pulvini) which came into use in Italy in the course of the fifth century (G. T. Rivoira, Origini delta Architettura Lombarda, 24). The restoration is certain, as the spring of the arch had survived on the angle piers.

^{page 21 note 1} The fragments of a marble skirting in the sanctuary perhaps belonged to this original wall-lining. I am informed by Mr. W. St. Clair Baddeley that at the beginning of the excavation of the church, in March, 1900, when the observer stood close under the barrel-vault of the sanctuary, abundant traces of mosaic could be seen on the latter, though little except the bedding of the tesserae remained. This, too, must have formed part of the pre-Christian decoration of the building.

^{page 21 note 2} See, e.g., Lanciani, Pagun and Christian Rome, 28; Ruins and Excavations, 215.

^{page 22 note 1} Plinius, Panegyricus, 47.

^{page 22 note 2} Suetonius, Caligula, 22: partem Palatii ad Forum usque promovit, atque aede Castoris ei Pollucis in vestibulum transfigurata, &c.

^{page 23 note 1} The original outline of the piscina is indicated on the plan by dotted lines.

^{page 23 note 2} I have to thank Cav. G. T. Rivoira for permission to reproduce the plan of S. Sophia at Salonica from his Origini delta Architettura Lombarda, Fig. 104, p. 70.

^{page 25 note 1} The numbers in brackets indicate the position of the pictures, &c. on the plan.

^{page 25 note 2} This is the sequence, e.g. in the Byzantine Manual. Didron, ed. Stokes, ii. 268.

^{page 26 note 1} It may be mentioned here with regard to the inscriptions throughout the church, that fragmentary letters are generally printed as complete when there can be no question about their identity.

^{page 26 note 2} This is the form of the description which accompanies the scene in the twelfth-century mosaics of the Cappella Palatina at Palermo, though the representation is rather different. The same scene at Monreale has: Iacob luctavit cwn angelo. Angelas bemdixit ei dicens, etc.

^{page 27 note 1} The phraseology appears to be derived from the Vulgate. Gen. xxxvii. 27: melius est ut venundetur Ismaelitis. 28: et praetereuntibus Madianitis negotiatoribus. 36: Madianilae vendiderunt Joseph in Egypto.

^{page 28 note 1} Cf. Gen. xl. 13: dabisque ei calicem htxta officinal tuuin sicut ante facere consueveras.

^{page 28 note 2} Of this contemporary dress, regularly used here and in other monuments of the epoch, in representations of official persons and lay saints, there are familiar examples in the well-known mosaic of Justinian and his Court at S. Vitale, Ravenna. On this, and on ecclesiastical costume generally, see Grisar, , Analecta Romana, i. 521Google Scholar: Wilpert, Geivandung der Christen; and the convenient summary in Lowrie's Christian Art and Archaeology (1901), 383.

^{page 29 note 1} The sign also precedes the names and descriptions in the Old Testament scenes above.

^{page 29 note 2} Sept. 2nd in the Greek Calendar. Martyrologitim Romanum, Aug. 17th. The Byzantine Manual according to Didron (324) represents him as “jeune, imberbe.” Ruinart (Acta Sincera, ed. Ratisbon, 1859, p. 306) gives the references to him by Basil and Gregory Nazianzen, and an account of his cult in Gaul. One might have thought that a church dedicated to him was indicated by the S. Mamiatus in the list of Roman churches given by Cencius Camerarius, as published by Mabillon (cf. Armellini, Chiese, 43), the form being similar to that of Cosimatus from Cosmas. But the only MS. of authority (Riccardi, 228) has Mannatas. P. Fabre in Mélanges d'Airh. et d' Hist. 1887, 454 n; cf. 434.

^{page 30 note 1} His cult was in existence in the time of Gregory III. (731–741), whose brief regulating the services in St. Paul's without the Walls, mentions the quarta missa ad som. Grtgorium ad ianuas. Grisar, , Analecta Romana, i. 169Google Scholar, T. iii. 3. It was recognised in England soon afterwards at the Council of Clovesho (747). Haddan and Stubbs, Councils, &c., iii. 368.

^{page 30 note 2} Orient oder Rom, 124, Figs. 47, 48. The heads in the former are very close in style to the saints in S. Maria Antiqua, but, as in the Byzantine Manual (Didron, 322), both are beardless. See also Arehaeologia, lvii. (1900), 159Google Scholar, and Dalton, Catalogue of Early Christian Antiquities in the British Museum (1901), No. 398.

^{page 30 note 3} Acta SS. Oct. vol. iii. 839. In the Notitia Dignitatum (ed. Seeck, p. 31) we find sub dispositione viri illustris magislri officiorum, Schola Gentilium seniorum and iuniorum. They belonged to the Scholae Palatinae (Ammian. Marc. xiv. 7). Strzygowski, who repeats the incorrect form Κεντιλίων for Γεντιλίων, an error long ago pointed out by the editor of the Acta SS., says that the ring is the sign “ihres adeligen Ranges” (l.c. 126). But, inasmuch as it is worn by all the guardsmen in attendance on the Emperor, e.g. in the relief on the base of the obelisk of Theodosius at Constantinople (d'Agincourt, T. x. 7), on his clipeus at Madrid (Venturi, , Storia dell' Arte Italiaria, i.Google Scholar Fig. 438), and in the mosaic of Justinian at Ravenna (Garrucci, iv. T. 264, i. Hodgkin, , Invaders, ivGoogle Scholar. Frontispiece), it is clear that it is not a distinction of the officers but a badge of the corps; and I would suggest that it is the barbarian torques, alluding to the fact that, as the name shows (gentiles = barbari), these Guards were, originally at least, not Romans.

^{page 31 note 1} Hist. Franc, vii. 31. Gloria Mart. i. 96.

^{page 31 note 2} Lib. Pont. i. 420.

^{page 31 note 3} Commemorated on Nov. 28th, only in the Calendar of the City of Rome.

^{page 31 note 4} Didron, 330. With regard to the dress, see p. 35. Though the monastic saints in the eleventh-century mosaics of the Church of St. Luke of Stiris are represented in regular monastic habits, it is to be noticed that, as here, their under-garment is red (Schultz and Barnsley, Monastery of St. Luke, 51 sqq.)

^{page 31 note 5} Cf. the representation in the Church of St. Luke of Stiris, where they occur in the same group (Schultz and Parnsley, 53). “St. Saibas has a curiously trimiv.ed, short white beard, and that of Euthymius is very long.” But it should not have been stated that the crosses which they hold indicate that they were martyrs.

^{page 32 note 1} The recent excavations in S. Saba have made it clear that the church was being decorated at the same time and perhaps by the same hands as this part of S. Maria Antiqua.

^{page 32 note 2} His name was only inserted in the Roman and other Martyrologies (Sept. 16th) in the sixteenth century. The original epitaph is now in the Lateran Museum (C.I.L. xi. 4076Google Scholar). Cf. De Rossi in Bull. Arch. Chr. 1883, 134 sqq. The relics were rediscovered in SS. Cosma and Damiano in 1582 and given to the Church of the Gesu (Panciroli, Tesori Nascosti, 286), where they now lie beneath the high altar.

^{page 32 note 3} Marucchi, Cimitero e Basilica di S. Valentino, 113.

^{page 32 note 4} Duchesne, , Lib. Pont. i. XCI, 127Google Scholar.

^{page 32 note 5} He is represented in exactly the same manner (including the Greek form of benediction) in the Menologium of Basil II. in the Vatican Library (ed. Albani, ii. 78).

^{page 33 note 1} He is associated with SS. John Chrysostom and Basil in, e.g., the Greek Psalter in the British Museum (Add. 19352, f. 35 b). Cf. Nilles, Kalendaritim Manuale, 87. In the Church of St. Luke of Stiris it is Gregory ὁ Θανματουργὸς who appears in this company (l.c. 59).

^{page 33 note 2} Cf. Schultz and Barnsley, Monastery of St. Luke, Pl. 51.

^{page 33 note 3} Lib. Pont. i. 346. Armellini, Chiese, 122. Gregorovius, Rome, &c. bk. iii. ch. vi. (English translation, ii. 163).

^{page 34 note 1} Cf. Didron, ed. Stokes, ii. 388, 392.

^{page 34 note 2} See the references given on p. 28, note 2.

^{page 35 note 1} Cf. Grisar, , Analecta Romana, i. 526Google Scholar, who says that the costume is used regularly “per onorare quei santi che non erano da rappresentarsi colle vesti liturgiche o altrimenti proprie.” It might also be suggested lhat, as the pallium was the garb of philosophers, it had a special appropriateness for ascetics.

^{page 37 note 1} It recalls the figures of the sun and moon which sometimes appear in representations of the Crucifixion. E.g. Garrucci, vi. T. 459, 2, 3.

^{page 37 note 2} See the sketch, Fig. 9, p. 116.

^{page 37 note 3} The appearance must have been like that of the screens in the Church of St. Luke of Stiris (Schultz and Barnsley, Pl. 22, &c.).

^{page 38 note 1} This has a certain resemblance to the personifications of rivers with horns on iheir heads in the Greek Psalter of the British Museum (Add. 19352, f. 57 b, 125 a).

^{page 38 note 2} The doors which connect the chapel with the aisle and the sanctuary were partly built up so as to reduce their size after the building became a church. The same is the case with the corresponding chapel (G).

^{page 39 note 1} Gairucci, vi. T. 423, 9–11; Grisar, , Anal. Rom., i. 620Google Scholarsqq. Cf. also the altar of Euphrasius at Parenzo (Gairucti, vi. T. 408, 9; Rivoira, Origini, Fig. 144).

^{page 41 note 1} Some of the features appear in the more elaborate representation of the Rabulas Codex (sixth century) Garrucci, iii. T. 139.

^{page 42 note 2} Marucchi, Cimitero e Basilica di S. Valentino, 49 sqq. The discovery of the picture in S. Maria Antiqua has confirmed the truth of the old copies of the Catacomb-painting (Bosio, Roma Sotterranea, 579, and Ciacconio's copyist, Cod. Vat. 5409) against the objections raised by Wilpert (Die Katakomben-gemälde und ihre alten Copien, 40).

^{page 42 note 3} Garrucci, iv. T. 279. I, 280. 8. Part of the figure of the Virgin raising her covered hands to her face may still be seen in the Crypt of St. Peter's.

^{page 42 note 4} P. Germano, La Casa Celimontana dei SS. Martiri Giovanni e Paolo, 426, Fig. 74. The original has now practically disappeared. An ivory of about the same date in the Liverpool Museum (Mayer Collection) shows John in the same conventional manner with a book, though the Christ is of a different type (Garrucci, vi. T. 459. 3. Westwood, Fictile Ivories at S. Kensington, p. 105). Cf. also Gori, , Thes. d'et. Dipt. iii. T. xxxiiGoogle Scholar. John not infrequently retains the book in mediaeval representations, even after the right hand has assumed a different attitude.

^{page 43 note 1} E.g. the deed of gift in S. Clemente begins: infelix ego Gregorius primus presbyter, &c. Grisar, , Anal. Rom. i. 123, 172Google Scholar, T. vi. I.

^{page 43 note 2} Lib. Pont. i. under Constantine (708–715). For the office cf. Duchesne, Premiers Temps de l'Élat at Pontifical, 46: ‘personnes chargées des rapports avec les tribunaux, et notamment de l'exécution des sentences ecclésiastiques. C'est un service d'avouerie et de police.’

^{page 43 note 3} Grisar, , Analecta Romana, i. 174Google Scholar, T. iii. 5. Duchesne, , Lib. Pont. i. 514Google Scholar, n. 2.

^{page 44 note 1} Lib. Pont. i. 486. Thio, i.e. θεῖος, ‘zio,’ ‘uncle.’

^{page 44 note 2} The inscription (still in the portico) is addressed to the Virgin (1. 5): ego humillimus servulus tuns [E]ustathius inmeritiis dux quern tibi deservire et huic scae tuae diac(oniae) dispcnsatorem effici iussisli. Crescimbeni, Istona di S. Maria in Cosmedin, 62. Mai, , Script. Vet. Nov. Coll. v. 216Google Scholar. Duchesne, , Lib. Pont. i. 520Google Scholar, n. 90.

^{page 44 note 3} Acta SS. June, vol. iii. 23 (in S. Giovanni Battista, founded in 438). They are commemorated in the Western Calendars on June 16th; Martyrologium Hieronymianum, &c.

^{page 44 note 4} Mansi, , Concilia, viii. 152Google Scholar. Migne, , Patr. Lat. lix. 164Google Scholar. Jaffé, , Rigesta, ii. 693Google Scholar (Addenda), defends the authenticity of the decree of Gelasius, but cf. Grisar, , Analecta Romana, i. 46Google Scholar.

^{page 44 note 5} SS. Quirico e Giulitta behind the Forum of Augustus. The foundation of the church is much older, but the present dedication first appears in the twelfth century. Armellini, Chiese, 172.

^{page 44 note 6} Acta SS., June, vol. iii. 21. Ruinart, Acta Siiuera, 503. The dedication is even found in England.

^{page 45 note 1} M. Prou, Monnaies Carolingiennes, p. 130, No. 927: on reverse Set Cirici. On the other hand, the rediscovery of the relics at Auxerre, the translation of part of them to Nevers, and the dedication of the Cathedral there to ‘St. Cyr,’ only took place at the end of the eighth century. Pelits Bottandistes, vii. 74.

^{page 45 note 2} The same order is given in the Acta Sincera, c. xix. (Ruinart, 380).

^{page 46 note 1} The story, as is explained below, generally follows the later ‘Acta’ given in Acta SS. June 16th (vol. iii. 28). The edition of Antwerp, 1643, &c, is always cited.

^{page 47 note 1} Apuleius, , Met. ii. 21Google Scholar: porrigit dexteram, et aa instar oratorum conformat articulum; duobusque infimis conchtsis digitis, ceteros eminus porrigit. Cf. Grisar, , Analecta Romana, i. 640Google Scholar; Lowrie, Christian Art and Arch., 260.

^{page 47 note 2} The older verb catomidio (with the same meaning; see Rich, Diet. Ant. s.v.) which appears in Petronius, 132, and Spartianus, , Vita Hadriani, 18. 9Google Scholar, seems to have dropped out of use. Du Cange, Gloss, s.v. and the Bollandist editor in Acta SS. June, vol. ii. 1023 (SS. Vitus and Modestus), give examples of the later phrase. In such passage as that quoted above from the Acts of Quiricus (cf. Acts of Vitus and Modestus, June, vol. iii. 1022, §4) where catomo or catomis caedi is used, the original meaning seems to have been forgotten, and catamus = ‘the rod.’

^{page 50 note 1} Ruinart, Acta Sincera, p. 504, c. iii.

^{page 50 note 2} Ruinart, Acta Sincera, p. 502. Acta SS. June, vol. iii. 23.

^{page 50 note 3} Acta SS. June, vol. iii. 18.

^{page 52 note 1} Gori, , Thesaurus Vetentm Diptychorum, ii. T. viiGoogle Scholar. Labarte, , Histoire des Arts Inditstriels an Moyen Age, i. T. i.Google Scholar, &c. Venturi, , Storia d' Arch Italiana, i. Fig. 332Google Scholar.

^{page 52 note 2} C. Tullian in Mélanges d'Arch, et d' Hist., 1882, 28. The same idea had suggested itself to Gori (op. cit. 241), who thought that the Golden Rose, given by the Pope to individuals as a mark of honour, may have had its origin in this connexion. We should hardly be justified in supposing that the lady in S. Maria Antiqua had been a recipient of the Golden Rose, which does not appear before the eleventh or twelfth century (Cartari, , La Rosa d'Oro Pontijicia, Rome, 1681, p. 7Google Scholarsqq.).

^{page 52 note 3} On the Anglo-Saxon ivory of the Adoration of the Magi in the South Kensington Museum (142-'66), the Virgin holds a flower between the thumb and second finger of her right hand. In the Bsnedictional of Ethehvold at Chatsworth, the Virgin, in the scene of the Nativity, holds a golden lily in her left hand (Waagen, , Treasures of Art in Gt. Britain, iii. 362Google Scholar).

^{page 52 note 4} Based on Canticles, ii. 1: 'I am the rose of Sharon, and the lily of the valleys.’

^{page 52 note 5} Qy. in order that it may stand upright of itself. In the next Roman representation of vothe tapers that I can recall—the eleventh-century pictures in the lower church of S. Clemente—they have the form of coils of wax, no doubt for the same reason.

^{page 53 note 1} E.g. in the inscription of Paschal I, of the year 817, in S. Prassede (1. 18): ac s(an)i(t)orum octingentoruim quorum nomina scit om(ni)p(oten)s (Grisar, , Anal. Rom. i. 183Google Scholar).

^{page 53 note 2} So Marucchi, in N. Bull. Arch. Chr. vi. (1900) 308Google Scholar.

^{page 54 note 1} So too the twelfth century MS. of Symeon Metaphrastes in the British Museum (Add. 11870) contains many analogies.

^{page 54 note 2} E.g. VBI SCS IVLIANVS FVSTIBVS CEDITVR. Armellini, Chiese, 242. There is nothing to be seen of the pictures now.

^{page 56 note 1} The subject is described in the Byzantine Manual (Didron, ed. Stokes, ii. 320). The miniature in the twelfth-century Gospels in the British Museum (Harl. 1810, f. 261 b.)is typical. The subject appears with the same main outlines in the church of S. Luke of Stiris (Schultz and Barnsley, 49, Pl. 38) on which see Diehl in Mélanges for 1889, p. 41, who says that he knows of no earlier example.

^{page 56 note 2} Didron, ed. Stokes, ii. 321.

^{page 57 note 1} Cf. the order of the scenes in the Manual; Didron, l.c. 321.

^{page 57 note 2} Cf. Didron, ed. Stokes, ii. 356, Schultz and Barnsley, Monastery of St. Luke, 43. E.g. in S. Maria Bartholomew has a white heard; in the Byzantine representations he is a youthful personage. On the other hand, John is beardless, as usual in Western art; the Byzantine type is an old man with a long beard.

^{page 57 note 3} Lethaby and Swainson, Sancta Sophia, 290. The design in S. Maria Antiqua is almost exactly the same as one on a seventh-century door near Safa in Syria. De Vogüe, Syrie Centrale, T. 24 and p. 69; reproduced in Cattaneo, Architettura in Italia, Fig. 24.

^{page 59 note 1} As the height is great and I had not an opportunity of seeing the inscription close at hand, I give Marucchi's copy made shortly after the discovery (Nuovo Bull, di Arch. Crist, vi. (1900). 296Google Scholar. I have corrected it where possible.

^{page 59 note 2} Didron, ed. Stokes, ii. 294.

^{page 59 note 3} Salzenberg, Alt-christliche Baudenkmale von Constantinopel, T. xxx. i. So, too, in the twelfth-century Byzantine mosaics of the church of the Martorana at Palermo.

^{page 59 note 4} Ciampini, , Vet. Mon. i. 39Google Scholar, gives the words on the scroll as Hic D(ominu)s n(oste)r et n(on) impntabitur alius.

^{page 61 note 1} Not noticed by Marucchi.

^{page 61 note 2} Cf. Swete, Introduction to the Greek Old Testament, 274, for instances from Irenaeus, Tertullian, &c.

^{page 61 note 3} Didron, ed. Stokes, ii. 296.

^{page 61 note 4} Ciampini, , Vet. Mon. i. 39Google Scholar, gives the words on the scroll held by Moses: videvitis vita(m)v(est)ra(m) penden(tem).

^{page 62 note 1} Marucchi (l.c. 298) gives + SCS MARtiNVS.

^{page 62 note 2} Facsimile by Grisar, Civiltà Cattolka, Jan. 1901. p. 229.

^{page 63 note 1} Isaiah xxxviii. I.

^{page 63 note 2} Cf. the passage of Augustine which is read as a comment on I. Samuel xvii. in the Roman Breviary (Dominica iv. post Pentecosten. Lect. vi.). Venit enim virus David Christus, qui contra spiritalem Goliath, id est contra diabolum pugnaturus, crucem suam ipse portavit.

^{page 63 note 3} Cf. II. Kings xx. 5. ‘Behold I will heal thee: on the third day thou shalt go up into the house of the Lord.’

^{page 64 note 1} This appears chiefly in details such as the shape of the bed on which Hezekiah is lying, the two hills which form the background to the scene of David and Goliath, and the tufts of flowers covering the ground there and in other instances. All these may be found, e.g. in the well-known Byzantine Bibles in the Vatican Library (Gr. 746, 747).

^{page 65 note 1} Lib. Pont. i. 385.

^{page 65 note 2} Lib. Pont. l.c.: Fecit vero et imagines per diversas eccksias quas, quicunque nossedesiderat, in eis eius vnltitm depictnm repperiet.

^{page 66 note 1} Garrucci, iv. T. 279, 280.

^{page 66 note 2} The nude Christ appears in the earliest representations (e.g. the doors of S. Sabina at Rome, and the ivory box in the British Museum, both probably of the fifth century). Both types occur in the eleventh-century Byzantine Psalter in the British Museum (Add. 19352, f. 87b, 96a, 172b).

^{page 66 note 3} The form is suggested by John VII.'s inscription at St. Peter's (Garrucci, iv. T. 279, p. 97; De Rossi, , Inscr. Chr. ii. p. 418Google Scholar).

^{page 67 note 1} Cf. p. 52.

^{page 67 note 2} Portions of this can also be seen on the side-walls of the Sanctuary, but there are no indications of what came above it there.

^{page 68 note 1} The name has misled Federici (l.c. 29, 32) into thinking that they are the four Evangelists. The dress alone would make this impossible.

^{page 69 note 1} The identification is due to the Rev. F. E. Brightman.

^{page 69 note 2} Labbe-Cossart, , Concilia, iv. 1220Google Scholar. Mansi, , Concilia, v. 1377Google Scholar A. cf. vi. 959. Migne, , Patr.Lat., liv. 768Google Scholar, and see the Introduction to the Letter for illustrations of the use of the Tome in the church services at Rome.

^{page 70 note 1} Ed. Bened. i. 548. Migne, , Patr. Gr., xxxvi. 120Google Scholar. The identification is due to Mr. C. Turner and the Rev. H. A. Wilson of Magdalen College, Oxford.

^{page 71 note 1} De Spiritti Sancto, § 12., ed. Bened. (Gaume), c. viii; Migne, , Patr. Gr., xxxii. 105Google Scholar. The identification is due to Mr. C. H. Turner.

^{page 72 note 1} Ed. Bened. (Paris, 1738), viii. App. p. 15. Migne, Patr: Gr., lix. 500Google Scholar. Mr. Brightman identified the quotation.

^{page 72 note 2} The references are:

^{page 73 note 1} Mansi, x. 1083.

^{page 73 note 2} As in Ezekiel i. 6 sqq. Cf. Didron, ed. Stokes, ii. 265.

^{page 73 note 3} All that can now be seen is sanctisSIMVS ∣ PAVANs∣ ROMANus+

^{page 75 note 1} Cattaneo, Architettura, Fig. 88. For pavements of the same style in the eleventh century church of St. Luke of Stiris, see Schultz and Birnsley, op. cit., 36.

^{page 76 note 1} Federici's extraordinary idea (l.c. 21) that this belongs to the twelfth or thirteenth century and forms the signature of the mosaic worker—MATIA COMPoSE, scarcely needs refutation. For instances of these ‘libertine’ names see C.I.L. vi. 16061 sqq. and 22289 Sqq.; and cf. 22296 Matia C. L. Gnoste.

^{page 76 note 2} Cf. Salzenberg, Alt-chr. Baud. T. xxxix. 3 (St. Nicholas at Myra); and the plan of the church of St. Luke of Stiris (Schultz and Barnsley, Pl. I).

^{page 76 note 3} Lib. Pont. ii. 14: super altare maiore cyburium ex argento purissimo.

^{page 76 note 4} Reproduced in Rivoira, Origini della Architetttira Lombarda, 202, Fig. 277. Fragments of various marble shafts, also found in the church, may have belonged to the columns which supported it.

^{page 77 note 1} I.e. the backing on which we must suppose that the original marble decorations of the walls were fixed.

^{page 77 note 2} For the last part cf. e.g. Journal of Hellenic Studies, vii. (1886), 153Google Scholar, and the (qy. eighthcentury) pectoral cross in N. Bull. Arch. Crist, vi. 260, T. x. It is very common.

^{page 78 note 1} Didron, ii. 330, 400. Schultz and Barnsley, Monastery of St. Luke, 55 (cf. 44, n. 5)

^{page 78 note 2} Armellini, Chiese, 363.

^{page 79 note 1} Mai, Spicilegium Romanian, iii. (which is almost entirely taken up with an account of their miracles, and sermons in their honour) p. xi.; Armellini, Chiese, 945.

^{page 79 note 2} P. 78 (iv).

^{page 80 note 1} Just where it came, a hole has been broken through the wall in mediaeval times, in order to obtain access to a well dug in the floor of the chapel.

^{page 80 note 2} Ruinart, 386. Nilles, Kalendarium Manuale, 205.

^{page 80 note 3} Grisar, , Analecta Romana, i. 174Google Scholar.

^{page 81 note 1} M.G. Lacour-Gayet, Mélanges d'Arch. et d'Hist., 1881, 229, on the graffiti of the columns of the temple of Antoninus and Faustina. Cf. Lanciani, Ruins and Excavations, 221. For other Western examples see C.I.L. v. 3100, viii. 450.

^{page 81 note 2} In a church dedicated to the Virgin it would be natural to find a series giving the story of her life. If we restore the two names above as [Ζαχαρ]ίας and ῎Ανν[῎Ανν], it might be suggested that the scene is the Presentation of Mary in the Temple. Cf. Didron, ed. Stokes, ii. 348.

^{page 82 note 1} The identification of the subject is not obvious. The prison does not seem to be consistent with a representation of the Three Children in the furnace.

^{page 82 note 2} By the Rev. Bannister, H. M.. N. Bull. Arch. Crist, vi. (1900) 294, n. 1Google Scholar

^{page 83 note 1} Grisar, , Analecta Romana, i. 174, l. 22Google Scholar.

^{page 83 note 2} Garrucci, iv. T. 280. 1.

^{page 84 note 1} Didron, ed. Stokes, ii. 388. Considering how little the position of the pictures in the church corresponds with the directions of the Manual, it is perhaps only a coincidence that the latter puts St. Demetrius approximately in the place in which we find him here. ‘Outside, towards the singers' choir, represent the principal martyrs: St. George on the right; St. Demetrius on the left.’ In later Byzantine art he is represented as a soldier, e.g. in the church of St. Luke of Stiris (Schultz and Barnsley, 61).

^{page 85 note 1} The treatment of the subject is the familiar one in Byzantine art; e.g., Schultz and Barnsley, op. cit., Pl. 54, P. 57. n. 8. The remains are too fragmentary to show whether the angel was represented above the Children. We may well suppose that on the corresponding wall-space to the right of the Sanctuary (where all traces have vanished) the regular pendant to this scene, Daniel in the den of lions, was represented. Cf. Diehl, Convent de Saint Luc, 58.

^{page 86 note 1} Nilles, Kalendavium Manuale, 230. ᾽Αβείμ, ᾽Αντωνίου, Γουρία, ᾽Ελεαζλάρου, Εὐσεβωνᾶ, ᾽Αχείμ, Μαρκέου.

^{page 86 note 2} He is so described in the Byzantine Manual (Didron, 328). The other Eleazar is an old man with a long beard.

^{page 86 note 3} Nilles, l.c. Μνήμη τῶν ἁγίων ἑπτὰ παίδων τῶν Μακκβαίων· καὶ τῆς μητρὸς καὶ τοῦ διδασκάλου αὐτῶν ᾽Ελεαζάρου. Cf. 2 Mace. vi. 18: ᾽Ελεάζαρός τις τῶν γραμματέων; and 3 vi. I. In the Fourth Book of the Maccabees (xvi. 15) the mother is represented as present with her sons at the martyrdom of Eleazar, and encouraging them by his example. On the name Salomone see Achelis, H., Martyrologien in Abh. der K. Ges. der Wiss. zu Göttingen, NF. vol. iii. No. 3 (1900), p. 44Google Scholar.

^{page 87 note 1} Cf. the representation of the scene in the Byzantine Manual (Didron, ed. Stokes, ii. 288). ‘A lofty city. Below it many tents are visible within which soldiers are asleep. In the midst [is the tent in which Holofernes lies]. Judith, clad in sumptuous apparel, stands before him, bearing in one hand a bloody sword, while with the other she places the head of Holofernes in a wallet which her servant holds for her,’ &c. Cf. d'Agincourt, Histoire de l'Art, iii. Pl. xlii. 3, where the scene (from the eighth century ‘Bible of St. Paul's’) is accompanied by the words CAPVT OLOFERNI.

^{page 87 note 2} For examples, see Garrucci, iii. T. 150; Schultz and Barnsley, Pl. 54; etc.

^{page 88 note 1} It might be suggested that this represents the visit of the Queen of Sheba to Solomon.

^{page 88 note 2} The only youthful and beardless ascetic described in the Byzantine Manual is St. John Calybita (Didron, 335. Cf. Nilles, Kal., 70).

^{page 89 note 1} Salzenberg, Alt-chr. Baud. Anhang, p. ii. Lethaby and Swainson, Sancta Sophia, 54, and cf. Fig. 5. In the poem of Paul the Silentiary the passage occurs in the section entitled ἔκΦρασις τῦ ǎμβωνος 1. 21 sqq.

^{page 89 note 2} Lib. Pont. i. 385.

^{page 90 note 1} Facsimile in Grisar's, Anakcta Romana, i. T. ii. 7Google Scholar, and P. 167. Cf. Civiltà Cattolica, March 16th, 1901, p. 728.

^{page 91 note 2} On John's Greek origin see p. 7.

^{page 91 note 3} Garrucci, iv. T. 279.

^{page 91 note 4} The seventh-century MS. of the Acts in the Bodleian, known as the Codex Laudianus, has. at the end an entry of ownership in the form, Θεωτώκε βοήθη τούλου σου Γρμγωρίου διακόνου &c. P. Batiffol in Méianges d' Arch, et d'Hist. 1888, p. 307.

^{page 92 note 1} They have been fully dealt with by Prof.Marucchi, in the N. Bull. Arch. Crist. vii. (1901), 205Google Scholarsqq.

^{page 93 note 1} It was not that the subjects were misunderstood, for the story of Jonah was well known to Byzantine art, and its treatment was derived from the same type as the representations on the sarcophagi. E.g. in the Menologium of Basil II. (ed. Albani. Urbino, 1727) i. 61 (Sept. 22nd). Cf. Didron, ed. Stokes, ii. 286.

^{page 93 note 2} The two earliest cases are the Fieschi monument (1256) in S. Lorenzo Fuori, and the Savelli monument (1266) in the Araceli.

^{page 94 note 1} On the plaster some bilingual monk (perhaps) has scratched his name

PETRus

ΠETPOs

^{page 94 note 2} It occurs in the list in the Nobis qitoque.

^{page 95 note 1} Especially the fact that the Station for the second Mass on Christmas Day is at her church. For the history see the study of Mgr. Duchesne in Mélanges d'Arch. et d' Hist., 1887, 387 sqq.; supplemented by Grisar, , Anahcta Romana, i. 595Google Scholarsqq.

I take this from Marucchi's, account in the N. Bull, di Arch. Crist. vi. 292Google Scholar. Nothing can be made out now.

^{page 97 note 1} As the letters vary a good deal in size and spacing, there may have been more.

^{page 97 note 2} Armellini, Chiese, 525.

^{page 97 note 3} Mullooly, St. Clement (2nd ed.), 285. The inscription runs,

Quod haec prae cunctis splendet pictura decore,

Componere hanc studuit praesbyter ecce Leo.

Cf. Grisar, , Anal. Rom. i. T. viGoogle Scholar.

^{page 98 note 1} I have not come across any other example of Abbacyrus holding these objects, but they may be seen, e.g. in two Byzantine representations of another medical saint, Pantaleemon. Gori, , Thesaurus Vet. Dipt., iii. 354Google Scholar, T. IV.; iv. 7, T.I.

^{page 98 note 2} For the formula one may compare an ivory in the British Museum (described as German, 11th or 12th cent.) with ob amor(em) cs Radegid(is) fieri rogavit. See Spitzer Coll. Catalogue, I'l. IV. 58, and p. 13. Also the inscription on a well formerly at the church of S. Marco in Rome: de denis Dei et san(c)ti Marci Johannes presbyter fieri rogabit. Armellini, Chiese, 462.

^{page 99 note 1} A Greek translation of the Dialogues of St. Gregory in the Vatican Library is dated thus: ἐτελειέ ώθη δὲ βίβλος αὕτη μηνὶ ἀπριλὶῳ εἰκάδι πρώτῃ ἒτος, ςγη‛ (6308 = 800). Mélanges d' Arch. et Hist. 1888, 302. A graffito would naturally be in a less elaborate form, but cf. the inscription recording a restoration of the church at Forza d'Agrò near Messina, which ends: μνησθείη αὐτ(οῦ)ὁ:Κ(ύριο)ς· ἒτ(ει) ςχπ’ (6680 = 1171 A.D.). Not. Scav., 1885, 87. The era of Constantinople was probably used at Rome; e.g. in the dedication by Theodotus at S. Angelo in Pescheria (Duchesne, , Lib. Pont. i. 514Google Scholar. Grisar, , Anal. Rom. i. 175Google Scholar). The interpretation suggested receives some support from the identification of the Pope in the picture described below (p. 103) with Hadrian.

^{page 99 note 2} The fragmentary letters on the last line can hardly belong to a consular date, for the epitaph must be later than the end of the sixth century. Moreover the mark of contraction only belongs to the first of them.

^{page 100 note 1} Analogous expressions are common in Greek dedications. E.g. Journal of Hellenic Stuaies, xi. (1890) 236Google Scholar. Schultz and Barnsley, Monastery of St. Luke, 28( = Diehl, 10).

^{page 100 note 2} Bull. Arch. Crist. (S. iv.), 1885, 142.

^{page 104 note 1} Lib. Pont. i. 509.

^{page 104 note 2} The Byzantine Guide in Didron's version gives the description, ‘jeune, barbe arrondie’ (p. 392). According to the ‘Acts’ he τῶν ἡγουένων τῆς τάξεως (Acta SS. Sept. vol. iii. 220). I am inclined to think that, like Sergius, he had a ring round his neck; so that it may have been intended to represent him as being a member of the Imperial Guard. But the disappearance of the paint has left very uncertain traces.

^{page 104 note 3} Lib. Pont. i. 512: misericordia inotus, ob eorum mariyrum amorem.

^{page 104 note 4} The deed of gift by the dispensator Eustathius at S. Maria in Cosmedin already referred to (p. 44, n. 2). It begins thus: Haec tibi praeclara virgo caelestis regina.

^{page 104 note 5} Lib. Pont., ii. 132. Cf. 128, and i. 432. The best known instance is the inscription in S. Clemente recording the gift to the church of a Bible by the presbyter Gregorius in the time of Zacharias. Grisar, , Anal. Rom. i. 123Google Scholar, 172, T. iv. 2.

^{page 105 note 1} Generally speaking, all the graves in and about the church were found to have been rifled in later times.

^{page 105 note 2} The same method was followed in Rome till recent times. The square slabs closing the opening are a familiar feature in the floors of Roman churches.

^{page 106 note 1} There appear to be similar graves in the Basilica at Salona, which was destroyed in A.D. 639. Mr.Jackson, T. G. (Dalmatia, ii. 90Google Scholar) describes ‘a sepulchral vault with an entrance below the floor on the east side, accessible from a small square pit lined with stone. The actual entrance of the vault was closed by a stone sliding hatch running in grooves, which could be raised by an iron ring.’ Outside the Basilica (p. 92) ‘there are several sepulchral chambers like that described within the church, with a little shallow well or pit in front of the entrance, lined with slabs which are joined with lead dowels.’ In at least one case the sliding hatch is perfect with the iron ring for raising it.

^{page 106 note 2} The area of the piscina has now been completely cleared.

^{page 106 note 3} Cf. e.g. Notizie d. Scavi, 1886, 454; Bull. Comm. 1887, 50.

^{page 107 note 1} Huebner, Inscr. Hisp. Chr. 158 = Buecheler, Carmina Epigraphica, i. 724. For similar expressions (not very common in early Christian epitaphs) cf. also in Buecheler nos. 715, 756, 1435.

^{page 108 note 1} De Rossi, , Inser. Chr. i. pp. 609, 613Google Scholar.

^{page 109 note 1} C.I.L. vi. 9207, aurifex de Sacra Via; 9212, de Sacra Via auri acceptor; 9214, de Sacra Via auri vestrix.

^{page 109 note 2} See e.g. Lanciani, Pagan and Christian Rome, 325.

^{page 109 note 3} The question may be asked whether this is not the temple of Minerva mentioned in the grants of citizenship to soldiers (diplomata militaria) between A.D. 93 and the time of Diocletian. The originals are said to be fixed in muro post templum divi Aug. ad Minervam. C.I.L. iii. pt. 2, p. 859 sqq.

^{page 111 note 1} Federici (l.c. p. 46) gives some graffiti here, but I confess that I can make nothing out of the traces on the wall.

^{page 111 note 2} This agrees with the Byzantine Manual. Didron, 326 sq.

^{page 112 note 1} E.g. in the Roman Breviary, March 10th.

^{page 112 note 2} Acta SS. March, vol. ii. 12. Didron, Manuel, 326.

^{page 112 note 3} Apparently the ‘Aggée’ who appears in Didron's version of the Byzantine Guide (p. 327).

^{page 113 note 1} In their uncertain condition it is hardly worth while to attempt to identify the scenes: otherwise it might have been suggested that two of them might be incidents in the life of St. Antony, viz., the destruction, of his crops by wild animals, and the journey which he made to visit his Monasteries.

^{page 114 note 1} Tischendorf, Evangelia Apocrypha, 370.

^{page 114 note 2} Messrs. Schultz and Barnsley (Monastery of St. Luke, 48) are under a misapprehension in thinking that the name is a mistake as applied to the scene in the eleventh-century mosaic in the church of St. Luke of Stiris. They have misunderstood Diehl, to whom they refer for confirmation (Convent de St. Luc, 42.)

^{page 115 note 1} Didron, ed. Stokes, ii. 319.

^{page 115 note 2} E.g. Schultz and Barnsley, l.c. Fig. 39; d'Agincourt, T. xiii. 21 (doors of S. Paolo fuori, Rome), lvii. 6, lix. 6 (MSS. in Vatican); Gori, , Thes. Vet. Dipt. iii. T. xxxii.Google Scholar; Mélanges d'Arck. et d'Hist. 1888, 316 (eleventh-century MS. at Messina). Cf. Diehl, Convent de St. Luc, 42, for other instances.

^{page 115 note 3} Harl. 1810, f. 206 b.

^{page 116 note 1} Venturi, , Storia dell' Arte Italiana, i. 444Google Scholarsqq. Fig. 266 (p. 279). Cf. Lowrie, Christian Art, &c, 269 note. The sculptures are also reproduced in Garrucci, vi. T. 498.

^{page 117 note 1} Garrucci, iv. T. 280, 8.

^{page 117 note 2} Add. 19352. The MS. is described in Dr. Kenyon's Facsimiles of Biblical Manuscripts in the British Museum, No. vii. Cf. Waagen, , Treasures of Art in Great Britain, iv. 7Google Scholarsqq.

^{page 118 note 1} F. 31 b. The same scene appears in the Barberini Psaller (f. 44). Cf. Mélanges, De Rossi, 278.

^{page 118 note 2} E.g. the well-known reliefs from Selsey in Chichester Cathedral, of which there are casts at S. Kensington.

^{page 119 note 1} Garrucci, iv. T. 289, 2. It is not certain that the fragment belongs to the time of Paschal I.