Acta ad archaeologiam et artium historiam pertinentia: Vol. 2

Citation preview

Padded-dancer vases in the ‘Style of Timonidas’ AXEL

SEEBERG

Analysis of the Corinthian school in terms of vase-painters may be, in the words of a classic authority, ‘a somewhat unremunerative task ' but it has its uses, as the passage quoted goes on to demonstrate (1). Some vases invite this method on purely aesthetic grounds; prominent among them are certain Middle Corinthian cups, and pots (especially kraters) closely associated with them, in the ‘style of Timonidas ’. The term is Payne's — like many of his groupings, tentative, not meant to be pressed too far or taken too literally. Taken literally it is not in fact a good term, suggesting as it does a more direct connection of every work with one personality than one may feel it prudent to postulate: ‘ Circle of Timonidas’ would better describe the relationship.

Study of one subject is a poor approach to questions of style, but padded dancers in two ways have ἃ special relevance here. Elsewhere this motive leads a humble existence, on cheap vases, the backs of pots, and odd spaces left over in a frieze. The painters here in question were often led, by their superior degree of care for all parts of all vases, to emancipate it -- enrich and

elaborate the commonplace versions. And since such improvement on conventional schemes left scope for individual initiative, we may hope to find a personal element in such pictures. Another consideration which gives an interest to vases of this category, is that of their influence outside Corinth. Of late years the tendency has been to stress — rightly — the independent character

of that part of Attic black-figure which used once to be dubbed ‘ Attico-Corinthian' 6). But the influence is not to be denied altogether in the treatment of this particular subject: a sherd from Perachora (Perachora Il, no. 2488, pl. 103) offers interesting new documentation.

(1) H. Payne, Necrocorinthia (Oxford 1931), p. 183. des Komos im 6. Jahrhundert (Kénigsberg On p. 103 occur the terms ' style of Timonidas * and 1929). “Timonidas group’: for Payne's usage here, cf. Amyx' Lo Porto: F. G. Lo Porto, Ceramica arcaica dalla necropoli di Taranto, in « ASAtene+, xxxviii remarks in «Antike Kunst», v, (1962), PP. 5 544. ‘This paper is the result of work undertaken, mostly, xxxvii (1959-1960), pp. 1844. in Rome in 1962 with a grant-in-aid from the Tho- Med. Pr: D. A. Amyx, The Medallion Painter, mas Fearnley Foundation; my thanks are due to in «AJA», Ixv (1961), pp. 1599. those who made the stay possible, and to friends NC: H. Payne, Necrocorinthia (v. supra), catalogue and colleagues in Rome for all their help and hospinumbers (pp. 264549). tality. Payne: The samo, references to text and plates. Abbreviations of the titles of periodicals and stan- Perachora : Perachora. The Sanctuaries of Hera Ahdard works of reference follow the practice of « Fasti raia and Limenia. Vol. 1 (ed. Humíry Payne), Oxford 1940. Vol. II (ed. T. J. Dunbabin), OxArchaeologici +; other abbreviations are as follows: ford 1962. GKV: Jack L. Benson, Geschichte der Rorinthischen (a) Seo especially H. R. W. Smith, The Hearst Vasen (Basle 1953). k Greifenbagen: Adolf Greifenhagen, Fine affische Hyiria, în « University of California Publications în schwarsfigurige Vasengattung und die Darstellung Classical Archaeology », 1 (1944), pp. 241909.

AXEL SEEBERG

Many padded-dancer vases from the Circle of Timonidas have been published and attributed before (1); my purpose is to supplement the available material by presenting such additional relevant vases and sherds as are known to me. In doing this, I must acknowledge my great debt to Professor D. A. Amyx -- both to his published work, especially The Medallion Painter, and to his references, criticisms and suggestions, to which such value as the work may possess is due. The mistakes I claim for myself, 3

A cup to which Greifenhagen referred in 1929 — in the Pushkin Museum, Moscow -- was re-

cently published in part and is illustrated more fully here (Pls. I, Ila) (S). It is of rather a heavy model; the lip is decorated outside with a key-pattern, inside with tongues; the medallion

inside, framed by similar purple-and-black tongues and by a lotus-and-palmette chain, has the usual tondo motive of such cups — a gorgoneion. It is an odd one; more will be said later of its typological affinities. For the present, we merely note that it suggests lack of routine on the part of the decorator, which may mean that the cup was an early work of his. Above the foot, outside, there is a zone of fine double rays. There are two friezes, of which the lower houses the favourites of Middle and Late Corinthian orientalizing work -- panthers and goats, neatly drawn if rather lifeless. They pose in confronted pairs, felines to right, ruminants to left, except for one panther that was turned about, and a perched ‘eagle’ inserted behind

him (*). Above this group, two sirens confronted sit under one handle. At the other handle, instead of introducing another orientalizing motive, the painter continued the subject of the main frieze round to the reverse of the vase. ‘The better-preserved and more carefully drawn half of the tall frieze in the handle-zone bears the representation of a cavalcade of youths with spears, turned left, and flying eagles above. A conventional device to show speed, the birds seem rather out of place in this slow and solemn procession, The boys wear headbands and short chitons and have deep profiles with pointed chins and noses; their eyes look surprised, round, with sharply raised brows. One horse ambles, the rest walk. They are large, powerful creatures, top-heavy, but do not dwarf their riders as Early Archaic horses commonly do, and do not have the stilt-legs often seen even in good Middle and Late Corinthian work (!). Some have all four feet firmly planted, others — like some of the goats below -- tentatively pew the ground, foreshadowing the freer Late Archaic rendering (). On the back of the vase, the merry party of 15 padded dancers unexpectedly share something of the sport-

() V. Med. Pir, passim Lo Porto, pp. 145599. (gs. 121-124) and pp. x485qq. (fis. 125-127); CV

Louvre (6), pl. 12, 1-6 (NC 989); Perachora II, no. 2543 - all cups and kraters. Three kotylai belong to the same strain: NC 952 and 953 (GKY, p. 51), and «MonPiot », x1 (1944), pp. 23599. (pls. 3-4 and figs. 147). (2) V. D. Blavatsky, Istoriya antichnoi raspisnoi heramihi (Moscow 1953) PP. 9, 1 ; cf. Greifenhagen, p.93n. 5. I am very greatly obliged to Professor B. R. Vipper for photographs of this vase and permission to reproduce them. (3) Cf. Smith, The Hearst Hydria (seonoto2), p.244, with n. 29. 4

(4) Cf. Payne, pp. 7osqq. Long legs are a sign of high breeding in horses, not necessarily of primitive

draughtsmanship. (5) Besides Payne, I. c., see the interesting discussion of Benson (GKV, pp. rozsqq.) and LiewellynBrown (+ Schweizer Münzblätter +, iv (1953), fasc 14, PP. 4999.) on Corinthian coin-Pogasoi compared to horses in the vase-paintings. To reacha definitive result for the chronology of the coinage in this way may be difficult, even if the vases could be precisely dated, because tho change in vase-painting seems to be gradual,not sudden. The 'heraldic' Pegasol are likely to be conservative. Disproportionately largo heads like those on the Moscow cup occur in some of the earliest coin-types.

PADDED-DANCER VASES IN THE ‘STYLE OF TIMONIDAS* ing air of the cavalcade, They are slim and well-proportioned, quite unlike the Falstaffian types usually associated with this subject, and very lively; they dance in pairs with one by himself

at one end of the scene — a common form of composition. One or two hold drinking-horns. Those turned to the right (or as many of them as can be studied in the present state of the vase) have purple spots on their clothes, those of their partners are all purple (). This distinction apart, the garments are alike and of the usual kind, short tight-fitting chitons with an overfold in front (*). Some figures seem to have a misshapen or twisted foot (). Padded dancers do not often have this appearance. Those by the Late Corinthian Andromeda Painter approach it (i; but the only exact parallel is furnished by a krater in Paris (Pls. ΠΡ, III) (): the same small heads, long limbs and supple bodies, the same range of stances, the same expressive use of the arms — pointed elbows, jerked far up behind, or hands jutting out at an improbable angle to touch a neighbour. There can be no doubt that this vase, Louvre E 617, was decorated by the same artist as the cup in Moscow. Both the main faces of the krater are badly faded, but it deserves to be rescued from oblivion. It attracts attention mainly as a pot. The angle of the handles and handle-plates, the bulge of

the neck and the full, taut curve of the body depart from the usual and give it character. The

decoration is conventional. There are zigzags on the rim, griffon-birds on the handle-plates; above the rays at the base, a broad black-and-purple band supports the two tall friezes, the lower of which repeats the badly-matched pairs of goats and panthers with the same variation as on the Moscow cup.

Fowl similar to that in the frieze — perched ‘ eagles ’, regardant -- sit under each

handle in confronted pairs. The upper frieze has seven dancers on one side (two pairs and a threesome), cocks and a snake on the other (). Three dancers hold drinking-horns, all wear the normal

costume.

Resemblance to the Moscow cup is close in the animal-style also. Either this painter's

manner was remarkably constant, or the two works are close together in date. Among the manner-

isms common to both, one notes the heartless one of hitching beasts together by their tails —

not uncommon in Corinthian, but rarely carried out with such thoroughness. The goats, with their sea-horse muzzles and extremely long, elegant horns, are obviously of one herd. In build, bearing, and incised markings the animals are alike; few of the features are wholly individual, it is rather their combination that is so.

(1) Good MC vases occasionally, as here, show a on early vases, is a twisting round of one ankle so conformity of icoriographical particulars between that the toes point in opposite directions, whether dancers who are turned in the same direction —e.g., together or apart (cf. Payne, fig. 44A). The other Payne, pl. 31, 7-8 (bearded/beardless); ibid., pl. 33,0, (as here) isa deformity — usually sickle-shaped — of one and « MonPiot », xl (1944), pl. 3-4 (plain/patterned foot: cf. Payne, fig. 44G. The drawing of this latter chitons); Payne, pl. 33,10 (purple dots). Most purple version is not always very explicit, and doubt may dots on dancers’ clothes are on vases in or near the exist about the intention. * style of Timonidas'; but the exceptions (NC 774, (Ὁ V. Greifenbagen, pl. 5. 1073) seem to show that they are not merely a stylὉ) NC x177A. For a full set of photographs, istic convention. They may represent hair (Webster and permission to reproduce them here, I am greatly in «Bulletin of the John Rylands Library », xxxvi indebted to the kindness of M. Pierre Devambez. (1954), p. s81sq.): for an alternative interpretation, v. (6) A rare subject on MC kraters: NC 1160; the last section of my article in « JHS », xxxv (1965). « Hesperia », χα (1951), pl. 93; Auktion XXII, Mán(2) Ct. below, p. 7. zon und Medaillen A. G. (Basle 1961), no. 112. NC (3) Not uncommon, though it is rare to have so 1186 has cocks and snakes under one handle. The many ‘ crippled’ figures in one picture. Cf. Lo Porto jug in Boston, Fairbanks no. 553, pl. 63, with this pp. 148sqq., fig. 125, and Athens 992 (below, pp. 554). ‘subject, may be by the Cavalcade Painter (v. Med. There are two versions - one, more common Pir., p. 11 n. 35). Cf. Perachora TI, no. 2454.

AXEL SEEBERG

‘The Moscow Gorgoneion Painter (as he may be called) was a minor artist, to judge by these two early works, though his padded-dancer scenes have some distinction, His animal-style, however, links him with an interesting division of the ' style of Timonidas ’, as it were its left wing, to which belong the Cavalcade Painter and the Detroit Painter (1), The former often composed his animal-friezes in this way (see, e. g., C. Zervos, L'Art en Grèce (Paris 1946), fig. 103); the Detroit Painter did not, but the resemblance to both is strong. All differ from their master, the Medallion Painter: his figures, solid and effective by themselves, rarely achieve convincing integration in a scene without the aid of filling-ornament. The Cavalcade Painter had a lighter hand, and mastered the use of line to give unity to complicated scenes of action. The other two must have learned from him, but they were decorators at heart -- more concerned with surfaces than with subject-matter.

Of the same type as the cup in Moscow, and very similar in its stylistic affinities — though it need not be by the same hand — is a fragmentary cup from Perachora (2545) (). The obverse had a boar-hunt flanked by riders (contrast the Cavalcade Painter's in Zervos, /. 6), and there were tiders on the reverse. One of the familiar birds occupied the space under a handle, and the animal-frieze seems to have been arranged in the ‘Moscow manner’; the forms of the goats and panthers are a little looser, and the interior Gorgon was more sophisticated than that in Moscow. One point, however, it is still possible to see that the two gorgoneia had in common: the ears were set very high — and the implications of this feature may be interesting. If we compare the Moscow gorgoneion to early gorgoneia by the Cavalcade Painter (as Payne, fig. 25B) we see a difference not merely in quality, but in approach to the problem of composition, Each eye, of the Moscow version, is set carefully in the middle of its generous allotment of blank space, and aligned with ears and nose-wrinkles ~ which makes for a strong hori-ontal emphasis: the design is, in fact, ‘ bi-axial’, based on a balance of the top and bottom segments as well as on lateral symmetry. No doubt the Perachora gorgoneion was also bi-axial. Other good examples are the late Gorgoneion by the Cavalcade Painter (?) on the Vatican krater (NC 1452) (*), and Hector's shield-blazon on the Troilos krater in Paris (E 638.1) (), by another excellent artist whose animalstyle is extremely close to that of the Moscow Gorgoneion Painter. Gorgoneia constructed on this principle appear to have had a limited currency among a close group of good painters in the years of transition from Middle to Late Corinthian, They are not a logical development from the type favoured by the Cavalcade Painter in his earlier work; we cannot say how they originated. Of the examples mentioned, that in Moscow is the most primitive, In the others, the disadvantage of the design, the deadness of the middle surfaces, is countered by tinting them; on the Troilos krater, this was done by means of a brown wash, and the richer colour schemes of free painting may have suggested the way (9). (1) Med. Pir., pp. 10599. (2) Porachora ll, p. 263, pls. 106-107. Hardly by the Medallion Paintor (1 Gnomon ν, 1964, p. 405). (3) Payne fig. 27C. For the whole vase, see now P. E. Arias and M. Hirmer, Tausend Jahre griechische Vasenbunst (Munich 1960), pl. xod. For Timonidas’ own biaxial Gorgo, v. now «Antike Kunsts, vi (1963), pl. 19,2. (4) NC 1196. «MonPiot», xvi (1908), pp. 113sqq., pl. 13; Payne fig. 250; Recueil Charles Dugas (Paris 1960), pl. 15. (5) Ch. Dugas in his iconographical study, republished op. cit. (see preceding note), plausibly

suggested a model in great painting for the obverse picture of the Troilos Krater (p. 71). To arguments from iconography, add that of colour technique, The interesting and controversial article by Mogens Gjódesen in « AJA », Ixvii (1963), pp. 3338q.. reached me too late for consideration here; it contains several new illustrations of vases relevant to section 1 of the present paper. To these, and to Mr. Gjódesen's many fresh and evocative comments on them, I must be content to refer in a general way. The resemblance of the gorgoneia on the bronze krater from Vix to some which are mentioned above, seems to me less striking than he claims.

PADDED-DANCER VASES IN THE ‘STYLE OF ΤΙΜΟΝΙΡΑΒ᾽ 2.

In the material from Perachora the ‘ style of by scraps -- in a humble way, it is also present which are mentioned with them (1). Some of the are mainly concerned with fragments I-V: VI

Timonidas ' is abundantly represented, if mostly in the krater fragments 2253, 2254, and others latter are illustrated in PL. Ve, J-VII, where we and VII may or may not be related. As for

size, I, III and V might belong to one vase; the scale of IJ seems

slightly smaller, that of IV

slightly larger.

The padded dancers of III-IV are beardless, that in II bearded, in J both are combined. Such mixing of age-groups usually occurs only in scenes executed with a certain amount of care and ambition. In IIT, as in 2253, one dancer has purple dots on his chiton (*); the other costumes are normal. In I-III the headbands seem twisted in a series of loops across the forehead, like those of the sirens in 2254 (). In some of the cases this feature is combined with a special way of rendering the hair at the temples — a broad, angular ear-tress outlined in incision, the ear being omitted (1). ‘The combination recurs on two good vases: one a padded-dancer vase, the fragmentary cup

Athens 992 (Pls. IV, Vi) (), the other a fragment of a kotyle from Selinus, with a youthful horseman (6). Strictly speaking it is a parallel of iconography, not of style, and the distance in quality

is considerable. Fluctuations in care and quality do, however, happen in the work of a single

artist; the passage of time must be taken into account; and quite apart from the consideration

that there were few frustrated hairdressers among Corinthian painters, we can point to other

resemblances — more properly of style — which have the stamp of individuality. Time has been particularly unkind to the vases in question, and it is not possible to apply every criterion to (1) Perachora II, pis. 76, 78, and cf. pp. 23189. The illustrations hero are from a photograph which the late T. J. Dunbabin, with his customary generosity towards junior colleagues, placed at my disposal; it is used with the kind permission of the ‘Managing Committee of the British School at Athens. (2) Cf. above, p. 3, n. 1. (3) A similar trait is sometimes found in the Cavaleade Painter's work (e. g., Payne, pl. 32,1); ef. Med. Pir., pl. 4c (Painter of Brussels A 2182). Ob. viously related is the more elaborate rendering seen in Payne, fig. 47, and « AJA ν, xxx (1926),p. 448 fig. 3: in Attic, on the Frangois Vase and often in sculpture (GMA. Richter, Kowroi (London 1960), pp. 70, 8089.) (4) In the text to CY San Francisco, pl. 2,4, H. R.W. Smith interprets this ‘flap’ as an indication of ruffled hair — plausibly, in view of the contexts: padded dancers often have it, riders occasionally (e. g, Payne, fig. 18C), other figures hardly ever. An exception is Hippotion on the Amphiaraos Krater (NC 1471), who is perhaps characterized by hairstyle and nameas à “horseman’ — groom, squire. Two unpublished aryballoi by one hand (Toulouse, Musé St. Raymond 26.026; Hannover, KostnorMuseum K. S. 690) combine padded dancers and running figures, and all have flaps; here again Smith's

interpretation would apply, though iconographical “contagion” may give sufficient explanation. The hair would normally be brushed back behind the ears, and violent movement would cause it to slip forward so as to conceal them (‘lap’) or hang before them ('side-whiskers"): the latter version is also common, seo, o. g., Med. Pir., pl. 12a, and here, pl. VI. Side-whiskers are less readily distinguishable from such features as ‘the side-plaits of sphinxes and sirens (e. g., Délos X, no. 410; the plait of the bearded siren in Perachora II, pl. 59, no. 1526, looks like a ‘flap’). There are side-whiskerod horsemen on, e. g., the Eurytios Krater (NC 780); the groom Lapythos on the aryballos once at Breslau, Payne, fig. 45A, is a ‘ side-whisker ’ parallel for Hippotion. The two versions are not combined: painters proferred either the one or the other. Flaps are particularly often found on two series of padded-dancer aryballoi, each by one hand ~ as CV San Francisco, pl. 2,4; and as « FA» xi (1958), no. 113, fig. ro. On ‘the larger and better MC vases they are rare. (5) NC oos. Part, e’Epnus, 1885, pl. 7. The splendi new photographs, and the permission to make use of them here, I owe to the generosity of Mrs. Semni Papaspyridi-Karouzou. (6) From the Malophoros Sanctuary. e MALinc s, xxxii (1927), pl. 86, 8.

PADDED-DANCER VASES IN THE ‘STYLE OF ΤΙΜΟΝΙΡΑΒ᾽

of fashion; the style here is not strikingly different from Middle Corinthian, so that the date may be rather early. The vase seems relevant in the present context, though at present its relation to the ‘style of Timonidas’ is not easily defined (1). It invites examination of a point of iconography, for which Payne gives the cue when he cautiously speaks of “padded figures’, not dancers. What is the relationship between such pictures as this, and the more usual dancing-scenes ? P. N. Ure was aware that the distinction is not always obvious between padded dancers and the ' runners ' who also occur on Corinthian vases,as on those of other fabrics (?). The typical

runners are a subject with a different history: part and parcel of the true Orientalizing style, they are popular in the 7th century and peter out in the 6th together with animal-friezes. For a couple of generations at least, however, this history overlaps that of the padded dancers; and though the latter are more properly a ' figure-subject ', they sometimes occur in orientalizing contexts (). Your runner as well as your dancer, moreover, will occasionally have a small vase to himself (1); and runners in a very few cases are seen to overcome their native individualism and to hunt in packs, like the ‘ padded figures ” of the vase in Leningrad (5). Often the ‘runners’ are naked, but sometimes they are dressed in a short chiton. Their iconography altogether is not consistent; so probably the pictorial formula covers not one subject, but many. A plurality of runners is likely to mean simply a footrace (ἢ). Padded dancers, even when typical attributes such as drinking-horns are lacking, normally have distinguishing features which reduce the possibilities of confusion - namely (a) proportions, (Ὁ) costume, and (6) certain recurring stances. The costume differs from that of every other category of figure found in pictorial art of this period, in that a fold of material overhangs the belt in front and conceals it (?). Apparently, this feature would suffice by itself to show that a figure was intended for a padded dancer, Relying on it, a painter might vary the proportions to suit his style (*), or the movements to suit the context -- though in fact, this was not done often, Conversely, where movements andfor proportions were typical, the costume was felt to be less important. Usually this means merely that a painter working in a hurry might omit some of the incised markings (neck-opening, lower edge, even belt) — as was done on the Leningrad lekythos. More rarely, the dancers are gratuitously undressed -- the painter may even indulge in

mission to use them, I am greatly indebted to Dr. A. Peredolskaja, who also supplied this note on the condition of the vase: reassembled from sherds, the central portion of the second figure from left restored, with some repainting. (1) The resemblance to Med. Pir, pl. 13b is superficial, as Professor Amyx points out; nor does the vaso agree in type with the lekythoi associated with the Medallion Painter. Cleaning would make stylistic comparisons easier, (a) A. D. and P. N. Ure, Aryballoi and Figurines from Rhitsona în Bocotia (Cambridge 1934), p. 96. (3) E. g, Perachora I, pl. 27,4; Payne, pl. 20, 1-2. The alabastron Perachora II no. 1553, PL. 69, has one runner in the bottom frieze; the figure inthe middle frieze called a-‘ padded dancer’ is, I suspect, another, nude, seen from the back, legs lost. T. J. Dunbabin was kind enough to send me a tracing on which my opinion is based. (4) Dancer -- e. g., ΟΥ̓ Louvre (8) pl. 19,26-27,30;

CV Mains (1), pl 30,1-2. Runner - CV Heidelberg (1), pl. 114. Cf. also NC 1258, and CV Norway (1) pl. 2,11 and 3,3-4, two very odd ‘runners’ whose relation to padded dancers is difficult to determine. (5) CV Paris, Bibl. Nat. (Cab. M), pl. 10,5-6 and 16, 3-5; Payne, pl. 34,8; Lo Porto, fig. 47 b. (6) This interpretation should be safe for the aryballos-fr, Lo Porto, l c. in which he sees dancers (padded ?) and a priestess or deity: the tripod is a prize as in the parallel which he quotes (NC 552). Ct. also Perachora IL no. 1584, p. 150= and Pindar, Pyth. IX. 118? (7) Exception: horsemon fanking a dancer om the alabastron Délos X mo. 460, p. 139 fig. 4, wear the dancer-chiton — a case of ‘contagion’, as with the hairstyles on the aryballoi at Hannover and Toulouse, cited above (n. 22 ?). The garment worn, e. g., by a Boread in Payne, pl. 28,10, is, of course, different. (8) E. g, NC 1000; Lo Porto, fig. 1769.

AXEL SEEBERG

anatomical description (nipples, clavicula) -- and the costume must, as it were, be supplied from the context, or it is represented by the belt and overfold alone ('). If a Corinthian painter ever drew a figure intended for a padded dancer, without characterizing him as such by any of the above means, we shall never discover it, and so the proposition lacks practical interest. When the application of the term ‘ padded dancer ' is limited accordingly, the difficulty noted by Ure disappears. “Padded dancers, running’ are found on the following vases: Globular jug (fragmentary) Corinth C-40-87. « Hesperia », xvii (1948), pl. 78, no. D 7. Olpe Vatican 87. C. Albizzati, Vasi antichi dipinti del Vaticano (1924), p. 31, figs. 11-12. NC 1136. Ole Villa Giulia 458: ' padded men running ' according to Payne, NC 1135; I have not seen

the vase.

Plaque Berlin F. 905. Antike Denkméler, i, pl. 7, 14. Lekythos Leningrad B 4. A possible addition (cf. the Leningrad lekythos) is the fr. from Perachora, pl. Ve, IV, but similar figures occur in pair-dance scenes. Taken by itself the picture on the lekythos could represent comical lewd pursuit, for which contemporary Boeotian has parallels ( — especially since the quarry is a young boy. The pictures listed above do not corroborate this interpretation, and since Corinthian padded dancers have an old reputation for being lasziv and ausgelassen, it is a pleasure to testify that they are more clean-living than their Bocotian brethren (+). There remain at least two explanations that may be valid for all such pictures. One is that this form of composition is merely a livelier version of an extremely common scheme which shows rows of identical padded dancers, all turned

in the same direction and in one attitude (‘): that this scheme was utilized and improved by a painter in the ‘style of Timonidas’, we know (9). The other accords with the spirit of simple humour which pervades the Corinthian padded-dancer scenes — it is that fat people running are a subject calculated to provoke mirth; we may think of the pictures according to our preference, as purely pictorial inventions, or as more or less faithful representations of the antics of padded clowns. The two explanations are not, of course, mutually exclusive. (1) E. g, Med. Pir, pls. 12b, 130; CV Norway 31, 7-8; it is possible that pl 38,1 contains (1), pl. 3,1-2; NC 1000. another example. (2) E. Bielefeld, Komódienszene auf einem griech—(4) V. my article in «Acta Archaeologicas, xxxv ischen Vasenbild ? (Leipzig 1944), fig. 1. (1965). (3) In only one case does the distinction bearded(5) Med. Pir., pl. rab. beardless seem likely to refer to pederasty — Payne, pl.

A. Seeberg, Padded-dancer vases

Moscow, Prshkin Museum: Corinthian cup.

PLATE I

A. Seeberg, Padded-dancer vases

PLATE IL

4) Moscow, Pushkin Museum: Corinthian cup. b) Paris, Louvre (E 617): Corinthian rater.

A. Seeberg, Padded-dancer vases

Paris, Louvre (E 617): Corinthian krater.

PLATE III

A. Secberg, Padded-dancer vases

Athens, Natiocal Museum (092): Details of fragmentary Corinthian cup

PLATE IV

A. Seeberg, Padded-dancer vases

PLATE V

a) Cambridge, Museum of Classical Archaeology (NA 173): Corinthian krater fragment. è) Athens, National Museum (992): Detail of fragmentary Corinthian cup. c) Corinthian kra-er fragments from Perachora.

A. Sezberg,

2

Padded-dancer vases

PLATE VI

»

Len:ngraé, Hermitage Museum (B 4): Corinthian lekythos

JAMES D. BRECKENRIDGE eyes set shallow under fiat brows (1), button nose with large nostrils, thick lips fully visible under mustache; rather low forehead, almost wholly bald, hair short at sides and drawn back to reveal ears; beard broad and rather long. Little sense of surface texture, or differentiation between flesh and hair.

B) (Pl. IP). Known in some ten examples, usually identified by the copy in the Museo Nazionale delle Terme, Rome (). Shape of head quite different from Type A: face longer and narrower, forehead high and squared; cheekbones also high and prominent; much hair over top and sides of head, and hair and beard plastically differentiated from flesh. Mustache closer to corners of mouth; nose longer and more bridged, although retaining deep indentation at root. Slightly quizzical expression of Type A replaced by wholly serious, almost pessimistic mien. Identification of subject secured by inscribed herm in Farnese Collection, Museo Nazionale, Naples (). Sophocles :

4) (PL Ic). Known in 13 examples, and named for the full-length statue in the Lateran, although a more accurate impression of the original is gained from the plaster cast taken from the head of this statue before its disastrous ‘restoration ”. () Head characterized by blocky shape;

squarish face framed by hair and beard in thick locks; hair only slightly confined by the taenia. Eyes fairly deep-set, nose narrow and straight. The head broadens at the temples before disappearing under the hair; the upper part of the skull swells out over the lower. Identified by an inscribed herm in the Vatican (5). B) (PL Id). Known in over 25 examples, usually named for the not-too-well preserved head in Naples from the Farnese Collection, although the finest copy is probably the one in the British Museum (€). Head long and solemn, face rather narrow; hair on crown relatively thin, and waved downward under the taenia. Hair becomes fuller at sides and back, leaving the ears free; beard and mustache long and full. Identified by an inscribed, but severely weathered herm in the Belvedere of the Vatican, as well as by Ursinus’ engraving of ἃ now-vanished marble medallion or clipeus ('). A later development of this type seems to be represented by the group exemplified by the Arundel bronze in the British Museum (*), distinguished by a greater particularization of detail, and an expression of pathos unfamiliar in earlier portraits. (2) There is some dispute as to the priority of this type of brow treatment over that exemplified by the Berlin and Vatican copies, in which the brows are sharp, projecting, and angular: cf. Bieber, op. cit, Ῥ. 55. (2) B. M. Felletti Mai, Museo Nazionale Romano, I Ritratti, Rome 1953, No. 11, pp. tst. (3) Οἱ. R. Paribeni, II Rifratto nell'Arte Antica, Milan 1934, pl. XIX.

graphie, Munich 1901, I, p. 129, No. 1. London: A. H. Smith, Catalogue of Sculpture in the Department of Greek & Roman Antiquities of the British Museum (hereafter ‘ B.M.C.’), London 1904, III, No. 1831, Bp. 1328. ὀ(7) Vatican: W. Amelung, Vat. Cat., Berlin 1908, II, 1, Belvedere, No. 69b, p. 176 and pl. 11, Ursinus: Bernoulli, op. cit, I, p. 124, fig. 23. (8) H. B. Walters, Catalogue of the Bronzes, Greek,

(4) Cî. P. Amat, H. Brunn & F. Bruckmann, Griechische und rómische Porträts, Munich 1801 (hereafter ‘ A-B'), 113-115, On the problems raised by this restoration, cf. note I, p. 14 below. (5) G. Lippold, Die Stutpturen des Vaticanischen Museums (hereafter ‘ Vat. Cat."), Berlin 1936, ΠῚ, 1, Sala delle Muse, No. 492, pp. rof. and pl. 10. (6) Naples: J. J. Bernoulli, Griechische Ihomo-

Roman & Etruscan, im the Department of Greek & Roman Antiquities, British Museum (hereafter ‘ B. M. C. Bronzes ἢ, London 1899, No. 847, p. 153. A head in Copenhagen, NCG 411, is in a sense transitional between “he two types, although closer to the Arundel group: V. Poulsen, Les portraits grecs, « Publications de la Giyptothèque Ny Carlsberg », 5, Copenhagen 1954, No. το, pp. 37f.

10

MULTIPLE PORTRAIT TYPES

Euripides : A) (Pl. ILa). Known in six copies, generally named for the Rieti herm in the Ny Carlsberg Glyptotek which has inscribed on its breast a quotation from the poet’s ‘ Alexander’ (1); but probably the most accurate copy is the one in the British Museum (PI. III d) (?). In the latter, the head is long, face fairly narrow, the crown round and thinly covered with flowing locks of hair which grow fuller at the sides, concealing the ears. Beard full and long, mustache leaves mouth fully exposed. Nose large, long, and rather bulbous; eyes fairly deep-set, under straight brows; expression serious but mild. In other examples, culminating in the Rieti herm, the brows furrow, the mouth turns down, and the expression gives more and more a feeling of grimness and even bitterness, B) (PI. ΤΙ δ). Known in over 25 examples, once again named for the copy in the Farnese Collection in Naples, inscribed with the subject’s name (). The copy in Mantua, on the other hand, is in the best state of preservation (*). Head also long, but with wide, domed forehead, gently curved brows, straight thin nose, and mild expression. Hair on top of skull quite thin and combed forward; at sides and back growing thick and very long, again covering ears. Beard by

contrast

full but

rather

short.

Having described our types, the next step — an obvious if perilous one — is to examine surviving documentary evidence which might help in assigning any of these portraits to specific artists, places or dates of origin. Unfortunately, the texts are, in concrete information, remarkably meager; what we do have, however, are a few references which would place at least one example of the portraiture of each of these individuals at more or less the same time and place.

Whether

this is significant or merely fortuitous must remain to be seen. On the one hand, there is the record that Lycurgus, during his twelve years as ruler of Athens (338-326 B.C.), not only completed construction of the Theater of Dionysus on the slope of the Acro-

polis, but placed there bronze statues of the greatest dramatists, Aeschylus, Sophocles, and Euripides (ἡ. With regard to Socrates, we have the information that Lysippus was supposed to

have created a portrait which the Athenians placed in the Pompeion (*); and while our authority states that this was done immediately after Socrates’ death, all that we know of Lysippus' chronology, as well as of Socrates’ public reputation in Athens immediately following his execution,

compels us to refer this work to the same period, perhaps the ’30 's of the fourth century, as the group of the three great dramatists. Our first problem might then appear to be that of deciding which of the extant portraits, if any, represent these known and more or less dated dedications. Such procedures, however, have occasioned much of the confusion already existing in the study of this subject; given a handy set of round holes, it is all too tempting to see if the square peg cannot, with a bit of effort,

(1) NCG 4r4b: V. Poulsen, op. cit, 386 (2) Smith, B.M.C., III, No. 1833, pp. (3) AB, 1212. (4) A. Levi, Sculture greche e romane Ducale di Mantova, Rome 1931, No. 56, pl. XLI. (5) Plutarch, Moralia, 841 C-D and F D

No. 11, pp. 133f. del Palazzo pp. 384. and (Vit. X orat.,

Lycurgus); also Athenaeus, I. 19e, and Diogenes Laertius, II, 5, 43. Cf. Pausanias' description, I. 21, 18, where Aeschylus is included only by implication. Kleiner, Über Lysipp, «Neue Beitrige zur klassischen Altertumswissenschaft. Festschrift Bernhard Schweitzer », Stuttgart 1954, pp. 227%, esp. p. 238, would date the statues to the years 334-331. (6) Diog. Laert, II, 43. I

JAMES D. BRECKENRIDGE

be whittled down to fit. We propose instead to see if the dating of the known portrait-types cannot be better established by reference to stylistic comparisons with more securely dated works, before trying to assign any of them to particular occasions. Although not identified by any inscribed example, our portrait-type A of Socrates has been universally accepted as such, first, because of its general resemblance to the subsequent, identified type, and second, because of its still closer correspondence with the caricaturistic descriptions given of Socrates by his friends — descriptions which are, in fact, among the earliest approaches to physical delineation of the human individual in Greek literature 2). Because of these affinities, this type of Socrates, with its stylistic links to late fifth-century art, has been placed at the head of the line of his portraiture. Yet to find a date at which its prototype could have been executed, we must accept as a terminus a quo the foundation of Plato's Academy, in 387 B. C., since that was quite literally the only place in Athens where a statue in honor of Socrates could have been placed prior to the middle of the fourth century. We have no documentation for the existence of such an image of Socrates in the Academy, but the likelihood that there was one there is enhanced by the fact that a portrait of Plato by the sculptor Silanion was commissioned for the Academy by a Persian student, almost certainly during the subject’s lifetime (*). We seem to have just one portrait-type of Plato (PI. II c), known in 18 copies; it is identified by an inscribed herm in Berlin (*), The work bears some comparison with the Socrates of Type A (!), especially considering that we are dealing with portraits whose originals were almost certainly by different, although roughly contemporary, artists. Division of the hair into large locks of regular cyma-and S-curve, integrated into the larger masses, is similar in both images, as is the way in which the hair emerges from the borders of the face, and the system of delineating the eye sockets. Above all, the similarity between the works is evident in the subordination of lesser to larger shapes, features to head, hair to beard, and other elements, always controlled by the allover form. The lerger silhouettes determine the inner details, without allowing the abstract shape to be broken by projecting elements, as in the comparatively ‘ baroque’ Socrates of Type B. Finally, insofar as copyists' products are to be trusted, there appears to be a significant lack of interest in differentiating material textures,

such as hair and flesh.

᾿

The portrait of Plato does indeed correspond to such meager descriptions as we possess of

the philosopher's appearance, and especiallyof his grave and unsmiling mien — although to say that Plato never smiled would scarcely seem to imply the permanent sneer which characterizes most

copies of his portrait. As in the case of Socrates, the image fails to give us any sense of the true man behind the mask; as F. Poulsen put it, ‘The portrait is extrao-dinarily abstract and typical, there is no trace of intellectual comprehension, and likeness to the individual is limited to a few such traits as the breadth of the forehead and the cut of the beard’ (*). Yet this portrait, by all (x) Xenophon, Symposium, iv. 19-v., 10; Plato, ἃ Swiss private collection:H. Bloesch, Antike Kunstin Menon, 808, and Symposium, 215 A-B, with compa- der Schweiz, Zurich 1943, No. 21, pp. Soff., and pl. 45. rison to Silenus and Marsyas statues. On Socrates’ (4) Cf. esp. E. Pfuhl, Die Anfiinge der griechireputation and its relevance to his portraiture, cf. F. sehen Bildnishwnst, Munich 1927, pp. 28f. Poulsen, «From the Collections of the Ny Carlsberg (3) Loc. cit, p. 42. T. B. L. Webster, Art and Glyptothek », I, 1931, pp. 23ff. Literature in Fourth: Century Athens, London 1956, pp. (@) Diog. Laert, ΠῚ, 25. On the donor, Mithri- — roff, advances the -hesis that a deliberate use of such dates, of. ¢ Ath. Mitt. », XXVIII, 1903, pp. 348. contrast between axterior appearance and interior (3) CER. Boehringer, Platon. Bildnisse und Nach- content (or character) was a fundamental modus opeweise, Breslau 1935. The finest copy is probably onein randi of early fourch-century intellectual expression 12

MULTIPLE PORTRAIT TYPES

the evidence, could have been done directly from the model! If we compare other literary

references to works by Silanion, such as his portrait of the sculptor Apollodorus, who was shown

as ‘not a human being but anger personified’ (1), we arrive at the conclusion (consonant with the character of the likeness of Plato) that it was not this sculptor’s practice to concentrate on

the individualizing characteristics of his subjects, but rather to use these merely to modify the features of a more generalized type, with which the individual could then be identified: in short, ἃ method still very close to the idealizing procedures of the fifth century. Despite the fact that as much as two decades could separate the first portrait of Socrates from that of Plato in original execution, both show sufficient similarity in style and concept to suggest that they belong to the same general phase in the development of fourth-century Attic style, and particularly fo the same stage of evolution of the approach to the problem of the portrait, the stage in which the subject is still depicted in a strongly generalized form. Around these two works it may be possible to group ἃ wider circle of stylistically related portraits, all of which may be dated (in their prototypes) to the general period c. 380-360 B.C. Within this group, it might be possible for an acute eye to discriminate the work of different artists, perhaps three; but it is safer to bear in mind

the intervention

of copyists’

hands,

and

avoid over-elaborate formulations. In any case, it should be remembered that differences in style within the group are at least as apt to be the result of variations in personal’ manner as of fine chronological distinctions, within the general grouping imposed both by broader stylistic comparisons and by similarity in the conceptual approach to the portrait itself.

Closely similar in concept as in style to the portraits of Plato and Socrates is that of the famous rhetorician Lysias (PI. Il d), who died in 380 (*); treatment of locks of hair, overall conformation to silhouette, differentiation of textures, all are closely related in the three works, while the handling of the mouth, and its relationship to the framing mustache and beard, seem closest to that of the Socrates.

The portrait of Thucydides, who died in the first years of the

century, belongs to this group as well, for its general stylistic characteristics are similar (). The same broad skull, squared forehead, and generalized features, but with somewhat different stylistic mannerisms in the treatment of detail, are to be seen in another group of portraits, one of which V. Poulsen has recently suggested may be of Democritus (PI. ΠῚ a) (t). This ‘ Democritus ' portrait is of particular interest because it was for a long time regarded as an addi-

tional, and earliest, portrait of Sophocles (5). While this identification is clearly untenable, the strong similarity between this portrait and that of Sophocles of the Lateran type (PI. Le), which

was responsible for its initiation, might better be considered as owing to chronological proximity of origin. On the basis of these stylistic similarities of structure, stereometry, relation of phy-

(1) Pliny, Natural History, XXXIV. 81-2. Οἱ, E. Schmidt, Silanion der Meister des Platonbildes «Jhb. d. Inst. », 47, 1932, pp. 230ff., and ibid, 49, 1934. pp. 180ff. (2) H. Stuart Jones, A Catalogue of the Ancient Sculptures Preserved in the Municipal Collections of Rome : The Sculptures of the Museo Capitolino, Oxford 1012, Stanza de' filosofi,No. 96, p. 257 and pl. 60. 3) E. &, the bust at Holkham Hall published by F. Poulsen, Greek & Roman Portraits in English Country Houses, Oxford 1923, No. x, pp. 27ff. Anattempt to distinguish two Thucydides types was made G. Dickins, J. H. S., XXXIX, 1914, pp. 2934 For an ‘up-to-date list of copies of this portrait, cf. A. Hekler,

Bildmisse berühmior Griechen, Dritte Auflage (ed. H. v. Heintze), Mainz 1962, pp. 556. (notes to p. 22); the new edition is helpfully brought up to dato on all personalities covered. — (4) Et beromt Filosofpar, «Meddelelser fra Ny Carlsberg Glyptotek», VIII, 1951, pp. 33ff. This likenessis found linked in a double herm in the Louvre (A-B,771-3) with a stylistically comparable head which Poulsen suggests might be identified as Heraclitus. ὀ (s) Ev g., K. Schefold, Die Bildnisse der antiken Dichter, Redner und Denker, Basel 1943, pp. 72 Cf. F. Poulsen « From the Colls. ν, I, 1931, pp. Goff, with note, p. 62, on resemblances to idealized heads from Attic tomb reliefs of c. 400 B.C. 13

JAMES D. BRECKENRIDGE

sical details, and surface treatment, we should prefer to date the prototype of the Lateran portrait, our Type A, to the same period as the group of portraits under examination. However much violence this may do to the wide acceptance of this likeness as a copy of the Lycurgan image of c. 330, the fact remains that its stylistic character, and above all the type and degree of idealization, conform much more closely to other heads in poriraits and grave reliefs created before the middle of the fourth century than to those made afier it (1). ‘There is evidence enough in our sources to indicate a considerable increase in the production of portraits in Athens as well as elsewhere in Greece after the beginning of the fourth century. The victors of the Peloponnesian Wars naturally made votive dedications at Delphi; and whereas such dedications following the Persian Wars a century earlier hed included only a rare image purporting to represent any historical participant, the new monuments were largely composed of portrait statues of the actual generals and admirals responsible Zor the victory (ἢ). In Athens, which had been chary of granting distinctions to its leaders, a new attitude began to be felt once Spartan power waned and the Second League gave at least a brief illusion of resurgence. A few scattered dedications of a portrait-like nature had of course been placed in public sites for the better part of a century, but they had usually been private offerings like the images of Anacreon and Xanthippus put up on the Acropolis by the latter's son Pericles, or the portrait of Pericles himself, by Cresilas ‘ who made noble men still nobler” () ; surviving copies of these works serve to demonstrate how the concept of the portrait remained strongly idealized in the second half of the fifth century. Now such dedications came to be made on public rather than private initiative. It is well known that Demosthenes, who, with his accustomed admiration for what he considered the superior virtues of earlier days, rather decried the whole business, stated that the first man to be so honored publicly since the Tyrannicides (who were something of a special case) (x) Cp. Reinach as cited in n. 1, p. 9 above; also fig. 25) may give some useful clues to the appearance A. Hekler, ¢ Jhb. d. Inst. 2, XLII, 1927, p. 68. To of the original, but itis dangerousto rely on Hekler's comparison with a head on a grave relief of curacy of such reduced versions. Under the circumthe mid-fourth century, Athens, National Museum, stances, it seems sefestto base stylistic judgement on No. 832, we may add similar heads in Copenhagen, the surviving cast of the unrestored head alone. NCG 213 and 219a; the list could be enlarged without (2) Cp. Pausanias, X.roí, on the Marathon difficulty. Given the conservatism of funerary sculp- monument, with idem, X.o.7ff., X.15.4, etc. ture, so often remarked upon, we may assume a lag (8) Anacreon was identified by R. Kebulé von of about one artistic generation from progressive | Stradonitz, Amakreon, « Jhb. d. Inst. », VII, 1892, pp. modes of expression, such as actual portraiture. 1198., with the statue in Copenhagen, NCG 409, In judging the style ot the Lateran Sophocles, whose original he attributod to Cresilas. V. Poulsen precision is especially difficult to achieve because, has followed Fiürtwángler and others in advancing although the statue is often used as a cornerstone of an attribution to Phidias: Les portraits grecs, No. 1, fourth-century chronology, on the assumption of pp. 13 and 25fí, as well as in Fidias og Portracthunstits faithful reproduction of the Lycurgan memorial, em, + Meddelelser fra N.C.G. », 9, 1952, pp. 3f. and its closest true parallel is the full-length portrait of Phidiasische Bildaisse, Festscrift Schweitzer, pp. Aeschines (A-B, 116-8), usually dated to the early 202ff. He also suggests that the head in Copenhagen, years of the third century (cf. p. 17, below) NCG 438, may be copied from the accompanying Part of the difficulty is undoubtedly occasioned statue of Xanthippus mentioned by Pausanias, 1.25. by the fact that the restorer Tenerani based his ‘Les portraits grecs, No. 2, pp. 271. Cf. G. Haine drastic reconstruction of the Lateran statue upon Anakreon und Xanthippos, « Jhb. d. Inst. ν, LXX] the Naples Aeschines itself! (« J.H.S. », XLII, 1922, 1956, pp. 1ff. The best copy of the Pericles portrait P. 52). The statuette illustrated by Miss Richter is that in the British Museum: Smith, B.M.C., Lon(Greek Portraits, IV. Iconagraphical Studies : A Few — don 1892, I, No. 540, pp. 288f. Suggestions (+ Coll. Latomus ν, LIV, Brussels 1962,

τ

MULTIPLE PORTRAIT TYPES

was Conon, commander of the fleet which crushed Sparta's navy at Cnidos in 394 (1); from this time forward the practice of making official dedications of statues to public heroes (usually accompanied by other honors, privileges and immunities) became a common means of recognition of outstanding public service. It wasin this atmosphere of increasing regard for the abilities of the individual citizen - and the manifestation of such regard in the specific form of a portrait dedication ~ that we may place the group of portraits we have described, comprising a sequence roughly from the Socrates to the Plato likeness. The sort of return to the virtue of an heroic past for which Demosthenes had been calling was actually put into effect when, on the morrow of Chaeroneia, in 338, Lycurgus was appointed

Tamias, financial administrator of Athens. For twelve years he was in effect dictator of the vanquished city, and in that period he set about systematically to restore, if not the military power, at least the confidence and sense of inherited greatness of his fellow citizens. His actiVities found expression in a great number of public works, of which the Theater of Dionysus was only one example; he reformed public education along lines which recall that he himself had studied at the Platonic Academy; above all, he held up the great men of the Athenian past as examples for the emulation of his contemporaries (8).

These are the circumstances in which the famous group of statues of the three great dramatists were made for the Theater; it would explain too the ordering of a portrait of Socrates from Lysippus, which is generally assumed to have been done at this time, and which was the first image of the philosopher to be placed in a public site just as we have no evidence of any earlier public dedication to the playwrights. The Socrates of Type B (PI. I 2) has generally been identified with this work of Lysippus, on the basis of what can be reconstructed

of that artist's

style (); what, then, of the three poets? The assumption that the Lateran statue of Sophocles reproduces the portrait from the Theater has seldom been challenged. On the other hand, we have already suggested reasons why it seems closer stylistically to works created before the middle of the fourth century. The Sophocles Farnese, our Type B, conforms on the other hand much more closely to Athenian style of the third quarter of the century (4), as does the Rieti portrait af Euripides, our Type A (9). While not so similar as to suggest the work of the same hand, these two portraits evidently belong to the same stage of stylistic development (cp. PI. ΠῚ c, d): The shape of head and face is in both cases oval, rather than the globe of Socrates A or the squared sphere of Sophocles A; the hair is thicker and more confused in detail, and much more sharply differentiated in texture from the soft, smooth flesh surfaces; above all, the likenesses convey a strong sense of inner personality not entirely dissimilar between the one and the other, stern and unsmiling as would befit the greatest of tragedians. All these factors suggest a common

point of origin, in circumstances not unlike those which have been described in the Athens of Lycurgus (ἢ. (1) In the oration against Leptines, 70, delivered kischen Steinsarkophage, Berlin 1952, No. 148, PP. in 355 BC; presumably this is the statue mentioned — 67i, and pl. 95. by Pausanias, 1.3.2. (5) CL V. Poulsen, Les portraits grecs, pp. (2) Cf. his biography, in the Lives of the Ten 38%. Orators (Plutarch, Moralia, 841-844);cp. the remarks ὀ(6) More information is supplied in Miss Richof F. Poulson, From the Colls., I, 1931, p. 44ter's chapter, The statues of the three tragedians in (3) Cf. F. P. Johnson, Lysippos, Durham 1927, the theatre of Athens, Grech Portraits, IV, pp. 24fí, and esp. Kleiner, op.ci although of course her identification of the three (4) A copy found in Etruria dates from the beginn- surviving types does not correspond with that given ing of the third century BC: R. Herbig, Die jüngeretrus- — here.

15

JAMES D. BRECKENRIDGE

Furthermore, the portrait which is generally accepted as representing the third member of the supreme triad, Aeschylus (PI. IIT 2) (1), is also similar in general configuration, if less immediate as a depiction of character. This type has, interestingly enough, been found in a double herm coupled with the Farnese type of Sophocles (5). The Theater of Dionysus, of course, contained many other statues, among them those of the fifth-century military heroes Themistocles and Miltiades (). While attempts to identify or reconstruct these images are not especially profitable (*), the evidence on these statues contributes to the general picture we have that these last days of Athenian independence were a time of great interest in portraying the great men of the more glorious past, and hence a period when a num-

ber of portrait-types of such heroes were either invented or modified, just as the likenesses of Sophocles and Socrates appear to have been ‘revised’ (). In the case of such modernized portraits, the style of the prototype, if any, as well as what this historically-conscious city knew of the art of the past — and Athenians were then as now living in a museum of classical art — conditioned the appearance of the revised likenesses. The new poztraits have in common, however, an added quality of a deliberately invoked sense of the character and the significance of their subjects, in strong contrast to the detachment and impersonality of such earlier portraits as have survived without such revision. Some writers on the Farnese Euripides, our Type B (Pl. II 8), were so struck by the lifelike impression given by extant copies that they were impelled to consider it a contemporary or only slightly posthumous work (ἢ) - a matter unthinkable in terms of what we know of the artistic style, not in portraiture alone, of the close of the fifth century. Instead, Studniczka was certainly correct in asserting its close stylistic affinity to the portrait which he identified, by means of an Ursinus drawing, as that of Aristotle (Pl. Va) ("). In both these portraits we encounter, in a new way, a sense of vivid and immediate contact with a powerful and exciting

(1) Best copy probably the one in Naples, A-B, 4or-2; the identification was made by K. A. Mac Dowall « J.H.S. 3, XXIV, 1904, pp. 81ff ( In the Palazzo Colonna: οἵ. W. Amelung, 7! ritratto di Sofocle, « Atti della Pontificia Accademia Romana di Archeologias, III (Memorie, I. i), 1024, PP. 11off, and fig. 6. This juxtaposition seems to us more valid than the usual comparison of this AeschyJus portrait with the Lateran Sophocles as restored. V. Poulsen, op. cit, pp. 4f, makes a further dis tinction between existing copies of the Aeschylus portrait. On the other hand, the portrait in the Capitol

XLVI.x6r.13 f,, ed Dindorf, II pp. 515 f, and scholion on this passage, ed. Dindorf, ΠῚ, pp. 535f. (4) Οἵ. e. g Richter, op. cit, IV, pp. off (5) For some suggestions as to such revisions, cf. V. Poulsen, op. cit, pp. X6f. (6) E. gv H. Schrader, «Die Antike v, II ‘1926, pp. 124f.; A. Heller, Bildnisse Derühmlar Griechen, — Berlin 1940,p. 25; S. Lippold, Die griechische Plastik («Handbuch der Archaeologie », II, 1), Munich 1950, pp. 2734. (7) Das Bildnis des Aristoteles, Leipzig 1908. Schefold, op. cit, 2. 36, suggests an attribution of both portraits to Lysippus; cf. Kleiner, op. cit, p.

Ῥ. 252 and pl. 60) recently advanced by Hafner (+ Jhb.

234 £. that the Aristotle ' could very well bea crea-

(Stanza de’ filosofi, No. 82, Stuart Jones, op. cit,

238, and the curious remark of Johnson, op. cit., pp.

d. Inst.» LXX, 1055, pp. 1058.) as a more authentic likeness (following a challenge to the MacDowall identification presented by M. Floriani Squarciapino in « Archeologia Classica », V, 1953, PP. 5541) seems to have litte besides its baldness to recommend it = as remarked by Miss Richter, op. cit, IV, p. 25, n. 4. Stylisticaly this head conforms to the archaizing phase of later Hellenistic art, with its puttylike flesh surfaces and sulky brow. (3) As mentioned by Aelius Aristides, Oratio

tion of Lysippos, though on grounds of style alone it could not confidently be assigned to him ', The identification of this portrait-type with Aristotle has been challenged by J.H. Jongkees, Fulvio Orsini's Imagines and the Portrait of svistole Archaeologica Traje na», IV), Groningen 1960. ‘The reply included by Miss Richter in Greek Portraits, IV, pp. sof, covers most of Jongkoes objections competently ; while Orsini's pictures arecertainly tobe handled with great caution, their evidence need not be rejected out of hand. 16

MULTIPLE PORTRAIT TYPES

personality — something still missing in the brooding power of the Lycurgan dramatists, for example. Comments about the Aristotle such as Pfuhl's, ' The first portrait of an individual in the strongly realistic sense of the word’, () show the impact of these portraits in contrast with what had gone before, and demonstrate that they represent ἃ new stage in the development of the Greek portrait. It is at least in part this sense of immediacy which has led to the general assumption that

the Aristotle portrait must be based on a contemporary prototype, a likeness done directly from life. An inscription on a herm of the late imperial period records an epigram stating that Alexander the Great had a portrait of his teacher set up in Athens (*); this of course would have had to have been within the lifetime of its subject, since the pupil predeceased’ the master. There is another mention of an Aristotle portrait in the Testament of Aristotle's successor, Theophrastus (who diedin 287), implying that a portrait of Aristotle was in existence before that date (). On the other hand, we cannot afford to be misled by the impression given by a portrait itself, of liveliness or of the opposite, or even of convincing fidelity to subject; as Gombrich has

observed, it is only in accordance with the aesthetic criteria of a given epoch that ‘ the correct portrait... can be constructed to any required degree of accuracy’ (‘). After all, the same sense of life is felt in the Farnese Euripides, which cannot have been even remotely contemporaneous with its subject. The fact is that these two portraits can be related to a larger circle of works which, by

the identity of one among them, must be dated to a period a full generation after the death of

Aristotle (). The same features which distinguish the aforementioned heads: the broad, domed forehead, with sparse locks combed forward over the high brow; the straight eyebrows, level gaze, and serious but far from grim expression of the sensitive mouth; the bushy side hair and rather short beard; and above all the sense of fidelity to life and personality, are shared by other portraits of men of the same generation, such as the orator Hyperides (PI. V c), who

died like Aristotle in 322 B. C. (*), and also Aeschines (PI. Vd) (}). But they are also to be found in portraits of members of the next generation such as Theophrastus (Pl. V2) (). The key to the dating of the entire group is the inscribed herm of Olympiodorus (PL IV) (*), since this represents the leader of the revolt of 288 which freed Athens from Macedonian rule. As it is unlikely that he would have been portrayed before that event, we have a chronological ‘fix’ for the

group as a whole: for, while we need not assume that all these portraits are exactly contemporary, it can hardly be expected that they would have been executed over a period of much more than, say, two decades (9), Given our fixed points, it is unlikely that any one of them

NOG 437, and the full-length statue in Naples, A-B, 116-8. (8) Villa Albani, No. 1034; cp. Schefold, op. cit, p. 99, Nos. 3 and 4. (6) Nasjonalgalleriet, Katalog over Shulptur og Kunstindustri, Oslo 1952, No. 4, pp. 156 Originally published by F. Poulsen, La collection Ustinow : La sculpture, «Norske Videnskapsselskapets Skrifter » IL. Hist.-filos. Klasse, 1920, pp. 21ff. Found in Caesarea, Palestino, together with an apparently matching herm of Sophocles, Type B: ibid. pp. 181. (1o) Cf. V. Poulsen, lc. cit., and Les portraits grecs, pp. 5t and 188. Of course the same period saw the production of the much-copied Demosthenes by Polyeuctus (Plutarch, Demosthenes, 3oí; idem, Vit. X

(1) Op. cit., p. 8; cf. W. Jaeger, Aristotle, Oxford 1934, p. 322. (2) CIG, No. 136. (3) Diog. Laert., s. v. Theophrastus, V. st. V. Poulsen implies, op. cit., pp. 19 and 51, that the Testament ordered a new portrait of Aristotle to be made; this is not borne out by the text as preserved, (4) E. H. Gombrich, Ar! and Illusion, New York 1959, p. 90. (5) V. Poulsen, Ein neues Homerbildnis, «Acta Archaeologica», XIII, 1942, pp. 15b, and esp. pp. 156f. (6) Ct. Schefold, op. cit, pp. τοϑέ. (7) Lippold, Vat, Cat., III, τ, Sala delle Muse, No. 502, pp. 26£, and pl. 22. Cp. the head in Copenhagen, 7

JAMES D. BRECKENRIDGE

could have been made much before the year 300 B. C.; on the contrary, the probability is that, just as had been the case a century earlier, it was the new autonomy gained in 288 that supplied the stimulus for the creation of this new set of images of the last generation of giants of classical Athens. While it is not the intention of this essay to offer a complete survey of the development of Greek portraiture in the fourth century, neither should we wish to give the impression that Athens was the only locality where the art of portraiture was developing at this time. In particular, the Athenian material ignores completely the question of the development of the ruler portrait, so important to the succeeding Hellenistic era. On the other hand, the evidence from outside Athens, at least until the time of Alexander the Great, is too occasional to provide secure foundation for ἃ comprehensive sequential analysis. Nor does it serve to illuminate the problem we are attacking, which is just why, in the cases az least of the portraits of Socrates and of the two great dramatists, one portrait type should have been replaced by another in the course of less than a century's development. In the case of these portraits, on the other hand, we do have material of a particularly pertinent sort for attacking what Hanfmann has called the key to the history of portraiture, the study of the history of the concept of the personality (1). The relation of artistic developments, whether in style or subject matter, with contemporary intellectual concepts in other spheres, is a field fraught with peril. It will not do to present,

‘however unconsciously, the picture of Lysippus going off from his studio to attend the lectures of Aristotle in order to learn how to make his portrait of ‘ Alexander with the Lance’. Nevertheless, there are times when the wishes and intentions of intellectuals have controlled not only

the content but the manner of the artist-executants: such were some periods in the middle ages, for example, or the time of the Italian Neo-Platonists. And such, it would seem, could have been the case in the period we are examining here. For we have seen that the portraits of Socrates and of Plato were put up within the Academy itself, at the order of the teachers or students there, and at a time when the practice was not such a common one that we can assume it to have been a matter of course. It is difficult to doubt that these portraits must have been made to conform, albeit not so much explicitly as by coincidence of attitude, to the wishes of these patrons; and that these men had a definite idea of what a portrait of an individual

represent.

should

Similarly, Lycurgus was closely allied with the philosophers of his day, and undoubtedly was aware of their teachings, psychological as well as pedagogical, when ordering portraits for his Theater, In the third phase of development, we have found Theophrastus directly concerned with ordering portraits: another firm link with the central tradition of Athenian philosophy. The art of portraiture does in fact show adirect relation to intellectual concepts; that is why,

whenever discussing the subject historically, we are sooner or later compelled to use the phrase, ‘ portraiture in our sense of the word’. For our sense of the word does not hold good for all

times, any more than portraiture itself can be found practiced at all times and places in the

history of art. The importance of the period here under consideration lies in the demonstrable fact that it was the time when the portrait ‘in our sense of the word’ was first created.

We

can

do no better than to adopt Schweitzer’s analysis of the requiremerts of this kind of ‘true’ por-

orat., Dem., 8448. (Moralia, 844-8); Pausanias, 1.8.2; Amelung, Vat. Cat, I. 1, 1903, Braccio Nuovo, No. 62, pp. Soff. and pl. XI; NCG 436a, V. Poulsen, Les portraits grecs, No. 27, pp. 55), which may be un-

derstood as a deliberate chines (Richter, op. cit., (1) G.M.A.Hanfmann, fraifure, «Coll. Latomuso, 18

foil to the portrait of AesIV, pp. 354). Observations on Roman PorXI, Brussels 1953, pp. 33f.

JAMES D. BRECKENRIDGE

likeness of the human, and mortal, and unintellectual, and multiform, and dissoluble, and changeable’ (). Plato's Socrates does not contradict Xenophon's, Sut rather suggests that any depiction of physical appearances is without ‘real’ meaning, This is far from being the case with Aristotle. ' There seems to be no case in which the soul can act or be acted upon without involving the body; e. g. anger, courage, appetite, and sensation generally. Thinking seems the most probable exception; but if this too proves to be a form of imagination or to be impossible without imagination, it too requires a body as a condition of its existence... It therefore seems that all the affections of soul involve a body -passion, gentleness, fear, pity, courage, joy, loving, and hating; in all these there is a concurrent affection of the body’ (). Aristotle, who for the first time introduces a discussion of the integrated components of what we should call the personality, rather than merely a catalogue of emotions and sensations, does not wholly abandon the idea of an unchangeable aspect of this personality; this he calls mind (nous) in contrast to his body-linked soul (psyche): ‘Mind... is separable, impassible, unmixed, since in its essential nature it is activity’ (). It is worth noting that, matters of terminology aside (and Aristotle's own set of special usages was at variance with those of his predecessors), modern psychologists still find useful the concept of an inner, impersonal core of the intellect,

though not a rational one, Thus Aristotle, the first of the Greek philosophers to introduce something like empirical method into the study of phenomena, produced a theory of the integration of the human personality with the physical body which has its counterpart in his own portrait, the first ‘ biographical " one in which the life and thought of its subject, almost existentially speaking, are presented in a kind of visible summation (4).

This concept of the portrait was not reached, obviously, at cne step from the Pericles to the Aristotle images; just as the philosophers debated the matter of soul-versus-body, so the ar-

tists arrived by more or less empirical processes at the mastery of physiognomic and psychological realism reflected in the portrait of Aristotle. The steps in this process are marked by the revisions in the original likenesses, which are what we have been studying. In the three significant periods of portrait production, whose phases seem largely to coincide with changes in Athens’ fortune, we may trace distinct steps from the heroicized effigy of the Periclean age to the psychological portrait of Hellenistic times. The artist of the initial phase, c. 380-360 B. C., was for the first time oriented toward creating a true ‘likeness’, not just a hand-

some image conditioned by certain of the specific features of his subject; but he was not quite sure of the way to reach such a likeness. The lack of any fecling of relation between external and internal character, we might say, accounts at the intellectuel level for the mask-like impersonality of the portraits created at this period, while the as yet only partial liberation from the need to generalize, reflected in Socrates’ own remarks, explains the often very slight differentiation between these portraits and the figures in other sculptures of the same period

The primary consideration which we must keep in mind is the almost antithetical contrast

between exterior and interior character; the lack of intellectual qualities in the portrait of Plato, (1) Phaedo, So A-B (tr. Jowett, Dialogues of Plato, New York 1937, 1, p. 465). (2) De anima, I. 1 (4032) (tr. Smith, Works of Aristotle, ed. Ross, Oxford 1931, III). Aristotls concern with the importance of environment in determining personality is perhaps overstressed by C.

Shute, The Psychology of Aristotle, « Columbia Studies in Philosophy», I, New York 1941, but the fact that ho had such concern is obviously pertinent. (3) Ibid., IILS (4308). (4) Cp. Pfuhl, op. cit, p. 8; Schefold, op. cit, p. 37; Webster, op. cit, pp. gol., etc. 20

MULTIPLE PORTRAIT TYPES

for example, produced a sense almost of shock in the first generation of scholars to recognize its identification (). In this connection, the word ‘mask’ is of interest to us. Alfüldi has suggested orally to Miss Richter that in such a comedy as Aristophanes’ ‘ Clouds ', masks resembling the subjects might have been used (?). The idea of a resemblance between masks for the Old Comedy and the personages represented, as implied by Pollux (?), may reinforce the cogency of this as a general observation, but in the specific case of the ‘Clouds’, we happen to have the evidence, for what it is worth, of a Hellenistic scholion which states that Socrates was made up to look like a Silenus, exactly the figure to which his friends compared his appearance (ἢ. Thus his mask was most probably as much a caricature as his character as presented in the lines of the drama, Even so, it is interesting to conjecture that such a satyr-mask, identified with this play, might have served as one guide for the sculptor who created the earliest actual portrait of Socrates, In the second phase, then, that occupying the period c. 340-320 B. C., the increasing skill which, under the influence of this trend of thought, artists had achieved in the accurate representation of individual appearance, was employed in a new way. The portraits of Socrates and of the dramatists made during the Lycurgan period do have an intellectual and emotional content which is quite evident when they are compared with their predecessors. . This is achieved, however, in a particular way: the emotion expressed is a single one, and more or less immediate; each indivi dual is in the grip of one particularly characteristic thought, one mood. In this, rather than in any specific advance in surface naturalism as such, lies the difference between these portraits and those of the succeeding period, c. 300-280 B. C. In the latter, this transitory quality of mood is replaced by a less time-bound sense of enduring character. Instead of a momentary aspect

of temperament, we receive an impression of the full personality of the man, Euripides is sad, but no longer angry. This progression, which is an advance in depth of characterization, is then the reason why,

in the case of the most significant subjects of Athenian portraiture, the images

had to be revised

before the process could be considered complete. For a person so important to Greek thought as

Socrates, the sort of impersonal caricature represented by the Munich bronze headwas inadequate for later generations, accustomed to reading so much more from the portrait of a great man (5). Just why certain of these portraits should have been created, or revised, at certain specific

moments and not at others, we cannot hope to know except on the rare occasions when some document

has survived.

We

are too much

at the mercy of the random

survival of evidence,

and the equally random possibilities of identification, to be able to assert with certainty that a

portrait of Euripides was not made before 350, or that a new likeness of Sophocles was not created in the 280's. On the other hand, we can point to the relative popularity of the two poets in these respective periods as an indication that, if our evidence can be relied upon, there is at least some reason to believe that the probabilities bear out the surviving incidence of portraittypes (9). (1) Cf. Richter, Greek Portraits, II, pp. aof. (2) Ibid IL, pp. 39f Aeschylus and Euripides appear in the Frogs.

(5) The Romans, with their eclectic tastes and their interest in veritas for its own sake, were most responsible for the preservation of the superseded

Ὁ) Onomasticon, IV.143. Cf. G. Ken, e Usterr. carly types, as they were for reducing the full Jhfte. s, XLII, 1955, pp. Sf, and T. D. L. Webster length statues invariable in Greek portraiture to ‘Bulletin of the John Rylands Library», XXXILI, herms and busts: ef. Schefold, op. cit, pp. 30%, 1049, pp. 38. and 196, (4) Schol. V (ext) to line 223: W. J. M. Starkie, — (6) CL. P. T. Stevens, Euripides and the AtheThe Clouds of Aristophanes, London 1911, p. 59; mians, «J. H. S.», LXXVI, 1956, pp. 876. cf, also pp. xxxdi and 152. 2r

JAMES D. BRECKENRIDGE

It does seem possible to suggest, moreover, that the reason why the artist of later periods may have felt free to revise these accepted likenesses as their own ideas of the nature of portraiture developed was at least in part because they knew that none of the existing likenesses possessed the authority of absolute contemporaneity with their subjects. Sketches there may have been (1), or caricaturistic masks, but there was nothing in the way of a received likeness directly from life to confine later artists to portraying their subject in one way, and one way only. Evidence for this seems to lie in the negative fact that, as far as we know (*), the portrait of Plato, although suffering from all the disadvantages of first-period viewpoint, was never revised. For if the conclusions of modern scholars as to its date are correct, it was known to have been created during the subject's lifetime, and hence to possess the simple virtue of presumptive accuracy. With the generation following Aristotle and Lysippus, however, the creation of the portrait “in our sense of the word’ was completed, and it is significant that no further major changes in iconography seem to have been made, In a sense, the ‘ Aristotelian’ portrait has remained the definitive conception for Western tradition ever since. (1) Cf. Richter, op. cit, II, pp. 4of. On the other hand, it cannot be expected (as Hekler observed, Greek and Roman Portraits, London 1012, p. x) that even such works would have violated to any si ficant extent the aesthetic standards of their age — a point not contraverted by Miss Richter in Greeh Portraits, III. How were likenesses transmitted in ancient times ? Small portraits & near-portraits in terracotta, « Coll. Latomus », XLVIII, Brussels 1960. (2) It is understood that a new portrait-type of Plato is to be published soon. Whether this is to enter the canon as either a supplement or asubstitute

for the accepted type must await publication but certainly the generally accepted type discussed here would be hard put to find a new name, given its evident stylistic date of origin. This article has not attempted to deal with other interesting examples of multiple portrait-types such as that of Homer, since the latter were so clearly based upon imaginative conceptions; the types we discuss, on tho other hand, portray individuals who could have been seen by artists of the classical or post-classical period, within the time that the portrait in the sense discussed here was created.

22

J. D. Ereckenridge, Multiple porlrait types

PLATE I

a) Soera’es. Munich, Glyptothek. δ) Socraws. Rome, Museo Nazionale. (Photo: Anderson). e) Sophocles. Rome, Vill. Medici. (Photo: Daatsches Archiologisches Institut Rom). 4) Sophocles. London, British Museum. (Courtesy 2f the Trustees of the British Museum)

J. D. Breckenridge, Multiple portrait types

PLATE II

b a) Euripides, Ccoenhagen, Ny Carlsberg Glyptotek, Ὁ) Euripiles Copenbagen, Ny Carlsberg Glyptotek. ἢ) Plato, Vatican, Museo, Sala delle Muse. (Photo: Andersva). d) Lysias. Rome, Museo Capitolino.

J. D. Breckenridge, Multiple porirais types

PLATE ΠῚ

a) ᾿ Democritus". Copenhagen, Ny Carlsberg Glyptotek. b) Aeschylus, Copenhagen, Ny Carlsberg Glyptotek. d Sophocles. Copenhagen, Ny Carlsberg Glyptotek. d) Euripides. London, British Museum. (Courtesy of ‘the Trustees of the Britisk Museum).

J. D. Breckenridge, Multiple portrait types

Olympiodorus. Oslo, Nasionalgallerie:.

PLATE IV

J. D. Breckenridge, Multiple portrait types

PLATE V

m

a) Aristotle. Vienna, Kunsthistorisches Museum. Ὁ) Theophrasi4s. Rome, Villa Albani (Photo: Alinari) ¢) Hyperides. Copenhagen, National Museum, Department of Oriental and Classical Art. d) Aeschines. Copenhagen, Ny Carlsberg Glyptotek.

Hellenistic doorways and thresholds from Morgantina ROYNE KYLLINGSTAD and ERIK SJOQVIST

Up in the central Sicilian mountains, close to the village of Aidone in the province of Enna, lies the ancient city of Morgantina where an archaeological expedition from Princeton University has carried out excavations since 1955 *. Their progress can be followed in the American Journal of Archaeology in the form of yearly, preliminary reports (!). In a few cases, specific problems are broken out of their contexts, and dealt with in separate short papers. That is done whenever the topic is sufficiently circumscribed and the documentation clear enough, and when a separate publication can be reasonably considered as of some use to fellow archaeologists working with similar material (). This is the case with the present paper and may serve as an excuse for its limited scope

and its pedestrian subject. ‘A considerable part of the central quarter of the city, located around its agora, has been uncovered in the excavations. Public buildings surround the open space of the market place and private residences rise on the hillocks on each side. The material here discussed comes from both these kinds of architecture. It should be pointed out that the main structural principles applied in both cases, are essentially the same: foundations and lower courses of external walls,

corner

constructions, and doorways are generally in ashlar masonry without mortar, while interior and dividing walls are of rubble masonry. The superstructures were mainly of mudbrick and timber, sometimes of rubble masonry.

The preferred building material was the local shelly limestone, and all our threshold blocks are of that material, when not otherwise stated.

A few words should also be said about the general chronology and the criteria on which it

is based. The earliest urban settlement of Morgantina goes back to the second quarter of the sixth century B. C. This archaic period came to an abrupt end in the year 459 B. C. when the (+) This study was undertaken during the seventh campaign of the Princeton Archaeological Expedition to Sicily, in the spring of 1061. The Expedition's architect of that year, Mr. Reyne Kyllingstad, who joined us through the good offices of the Norwegian Institute in Rome, made all the drawings here published. The selection of the material is also his work. He should further be credited with the attempted reconstruction of the doors, their form and function. Based on his notes and observations, the article is written by Sjéqvist whose responsibility also was that of classification and dating, and of adding some of the comparative material.

(1) Such reports have appeared in e AJA» 1957, pp. 151-159; 1958, PP. 155-164; 1959, PP. 167-173; 1960, pp. 125-135; τοῦτ, pp. 278-281; 1962, pp. 135143; 1963, pp. 163-171, and 1964, pp. 137-147. Other articles of general interest ere found in «AJA 1958, pp. 79-90 (by K. Erim), in « Rend. Acad. Lincei Classe delle Scienze morali, 1959, pp. 39-48 and 1960, pp. 291-300; in « Kokalos », 1958, pp. 107-118; 1962, pp. 52-68, and in «Ill. London News », 1957, 9 Nov., pp. 788-791. (+) Ci. ¢ AJA», 1059, pp. 275-277 (on inkstonds), ibid. 1960, pp. 78-83 (on medicine bottles), and in « Amer. Numism. Soc. Museum Notes», Vol. 9 (1960), pp. 65-73 (on halfeoins, by R. R. Holloway). 23

ROYNE KYLLINGSTAD - ERIK SJÒQVIST

city was captured and destroyed by the Sikel leader Ducetius. It seemingly took almost a hundred years for the city to recover from the blow, even if the last quarter of the fifth century and the first half of the fourth are not entirely empty pages in its history as revealed in the excavations. The revival comes with Timoleon in the third quarter of the fourth century and with Agathocles of Syracuse in the last decades of the same century. None of the material he-e discussed belongs to these relatively early periods of the city’s history: Public and sacred architecture is not miss-

Po. 1. ing from any of these, but doorways and thresholds are of a simple type that reveals nothing of their function or construction.

One

is lead to believe that thresholds, posts and lintels were

entirely of wood, because the doorways are nothing but simple openings in the walls. The third,

second and first centuries B. C. introduced more sophisticated structural methods, and these are the object of this study. The finer dating of our material is based on a happy combina-

tion of different kinds of evidence. Stratigraphical and structural criteria are of course of primary importance. To these should be added typological considerations of the overly abundant

ceramic material and stylistic evidence gathered from the great amount of terracotta figurines found in archaeologically controllable layers. Last but not least a crop of some 8,000 legible coins, a number of which was found in sealed deposits, add many steps to our chronological ladder. This external evidence combined with the historical information on the shifting destinies of Morgantina found in the ancient authors from Thucydides to Livy and Strabo, gives reassuring strength to the datings here proposed. None the less, some questionmazks remain in the mar-

24

HELLENISTIC DOORWAYS AND THRESHOLDS FROM MORGANTINA gin of our chronological table. In doubtful cases, as in the thresholds nos. 3, 4, 5, and 12, the tolerance has to remain

rather wide.

The first threshold block to be discussed (no. 1) is typical of a group with very specific characteristics, here called Type x. It consists of one block corresponding to the width of the door (dimensions: 1,44 x 0,84 x 0,45 m). It is cut down on its interior side so as to fit the wooden framework of the door, as illustrated in fig. 1. Two angular cuttings flank the ends of the block and leave the central part on a higher level. But even this central inner part is somewhat lower than the external side of the block. This proves that the doors opened inwards and abutted against the raised exterior edge of the threshold. In the corners of the angular cuttings

are traces of bronze sockets obviously meant to catch the

pivots of the swinging doors. In the centre of the inner threshold is a bolt hole, some 0,03 m from the raised external part of the block. The distance corresponds with the thickness of the door. The bolt hole is carefully cut into an oblong rectangular shape, and strengthened on its inner side by an iron mounting held in place by fused lead. This represents an unusually careful method of construction. The exterior side of the threshold is slightly more worn than the interior, but on the whole the block is well preserved. The door leads Fio. 2. from one room to another in the large residential house, called the House of the Official. It belongs to its initial building period and can be dated to the decade between

260 and 250 B. C.

Our threshold no. 2 is a slight variation of this same type (fig. 2). The principle for fixing the door — this principle will be discussed later on — is the same. The two angular cuttings, the holes for receiving the pivots, the high external edge, and the lower inner level are all there. The only difference is that each half of the double door was held to the threshold by separate

bolts. Of further interest is that the door jambs show cuttings for holding a cross bar, serving

asan extra safeguard when closing the door. The block is much worn and the bolts have cut curved ruts in the central part of the threshold. These ruts are sectors of two circles the centres of which lie in the pivot holes.

This door leads from a small vestibule into

the

kitchen and latrine

of the well equipped private residence, known as the House of the Doric Capital (1). It is approximately contemporary to the House of the Official, perhaps a decade later, thus dating between

the years 250-240 B. C.

Quite similar to this threshold are those numbered 3-5 (figs. 3-5). The two first ones are worn and it should be noted that no. 3 shows a particular type of wearing. The right half of the

door was obviously for part of its life out of plumb, and was scraping against the threshold when

swung open or shut. Nos. 3 and 4 cannot be dated with precision. They are located in Area I,

Trench 69, and belong to houses not completely excavated. Further they were covered by very little soil, consisting mostly of the ploughed surface layer, so that reliable ceramic and strati(1) The latrine is situated immediately to the right of who enters. The independent bolting of the two halves of the rather narrow door may be motivated by a desire to keep the latrine out

οἱ sight for those passing the vestibule, and still to leave the kitchen door half-open. The cross beam may have been in function when the latrine was used. 25

HELLENISTIC DOORWAYS AND THRESHOLDS FROM MORGANTINA

owest doors and one of the few single doors, so far traced in Morgantina. Consequently it has only one angular cutting and the bolt hole is placed at the opposite end of the block. The total height of the door can be reconstructed to 1,88 m. thanks to the good preservation of the jambs which stand up to the height where the lintel block was placed. As the height of the threshold is 0;19 m. the actual opening

was only 1.69 m. high. The fitting of the lintel is, as

the placing of the threshold, a modification of an originally different scheme. This is revealed, by the masonry of the jamb to the right of who looks out

through the door, as illustrated in fig. x. The pry-

laneion itself dates from the reign of Hieron II of Syracuse (275-215 'B. C.), was considerably damaged

in 211 B. C., and later repaired. It served its purpose

as long as urban life continued.in Morgantina, i. e. until about 30 B. C. Our door received its present shape during the restoration work after 2rr B. and probably after the lapse of several decades. Its approximate date is the second quarter of the

second century B. C. The thresholds described above have one main characteristic in common: the angular cutting is perpendicular to the door. They can all be grouped.

together as Type x, if one keeps in mind the variations represented by no. 6 (with the cuttings

describing obtuse angles) and no. 7, which is the only single door in the group. A different system of fitting the woodwork into the threshold block is revealed by a single threshold (no. 8) which alone may account for our Type 2 (fig. 8). The door is one of the main entrances to the House of the Official. This may account for its unusual width, a circumstance that made it necessary to build the threshold in four blocks. The upper step, leading down from the street. pavement, is secondary and conditioned by the rising street level, but the threshold itself dates from the original building period of the house and can be dated to 260-250 B. C. The cuttings for the door frame are

Fis. 8.

straight and parallel to the threshold, not

angular and perpendicular. In one corner

the bronze socket of the pivot is preserved; pierced by a central hole. There is a central bolt hole, and the left half of the door was further secured by a second bolt. The unusual width of the door, and the fact that it lead out to the street may account for this special safety measure. 27

ROYNE KYLLINGSTAD - ERIK SJOQVIST A smaller door belonging to the same house and opening to the same street is represented by threshold no. 9 (fig. 0). As the drawing shows, this was not originally a door sill, but one of the continuous foundation blocks of the exterior wall of the house, modified and recut into a threshold when the new door was opened. This becomes particularly clear when studying the right hand comer of the door where the superimposed block was recut to fit the door frame. It can be dated to the rebuilding period of the House of the Official, i. e. to the middle of the second century B. C. It represents really a new type in as much as it is provided with straight perpendicular cuttings, not angular as Type 1, nor stzaight parallel ones, as Type 2. For the sake of clarity, we would suggest to call it Type 3. The pivot holes are well preserved. There are two bolt holes close to the centre. The right one is cut close to the edge of the raised, exterior par: of the threshold, which scems Fic. 9. to indicate that the bolt was built into the thickness of the door itself, The left one is placed at the usual distance, and indicates that the door was about 0,03 m. thick. Similar to this one, but without holes for pivots or bolts, is no. ro (fig. 10). The reason for their absence is hard to explain, Perhaps a wooden beam with the necessary mountings was fitted into the deep longitudinal cutzing in the block. It is a narrow little door leading into the central pavillion of the macellum and dates from about 150 B. C. ‘Very different is ournext type (Type 4), represented by threshold no. 1r (fig. rr). It serves the open front of the summer triclinium. of the so-called House of the Arched Cistern, and its total width is 5,00 m. It was a multiple folding screen door, probably not Fic. 10. reaching the complete height of the opening. The threshold consisted

of three blocks of finely grained white limestone of choice quality. What is preserved is only the right side corner block, but marks in the foundations show that it was matched by a block of equal length to the left, and that the central block was somewhat longer. Their respective lengths were 1,64, 1,64, and 1,72 m. The preserved cuttings for the door post are of the same design as those of Type τ, i. e. of angular, perpendicular shape. The bronze mounting for the pivot is square and pierced by a central hole. It was found slightly dislocated cs the drawing shows. The door opened toward the peristyle, not inwards toward

Fio. m.

the room,

Morgantina.

as is the rule in all other cases in

This must be motivated by its special

function as a multiple screen door which could be completely folded up against the docr posts, or opened in

sections only, if so desired, The only bolt hole preserved lies at a distance of 0,90 m. from the

right post. That must be the width of the first wing of the door. This figure is no equal multiple of the total width of the opening, which is 5,00 m. Any reconstruction must therefore cal28

HELLENISTIC DOORWAYS AND THRESHOLDS FROM MORGANTINA

culate with an uneven width of the separate members of the folding screen. One could reasonably think of two wings on each side, 0,90 m. wide. This would cover 3,60 m. of the total opening. The remaining distance (1,40 m.) may have been covered by a double door, closing the central part of the opening. Such an arrangement would

have facilitated the opening and closing of the

door.

‘The threshold belongs to the original building period of the house and can be dated to the third quarter of the third century B. C. The fifth ‘and last type of our threshold to be discussed here is represented by nos. 12 and 13 (figs 12-13). No. 12 has a total length of 2,37 m. and is composed of five blocks of different sizes, while No. 13 is monolithic and only 1,24 m. wide. What they have in common is that thé

rectangular perpendicular cutting with the pivot

Fis, 12, of the in no. in the agora. re 12)

hole exists only at one end of the door — the left end, seen from the interior - while along the right side of the block runs a longitudinal groove up

to the right hand post. In no. 12 the interior side

threshold is cut down to a lower level only to the point where the groove begins, while 13 the interior level as a whole is lower. Both these doors are shop-fronts, no. 12 located so-called Northwest Shops, and no. 13 in the ‘ South Market’, both adjacent to the The two square holes cut in the door post of no. 12 (see figudo not correspond to the cuttings in the threshold and

may have served for fixing a wooden counter when the door was

open. Their dates cannot be given with satisfactory certainty. No. 13 belongs to the first half of the second century B. C. Most probably no. 12 dates from the same period, but it is not excluded that it is somewhat earlier. ‘The typological considerations given above can be thus summarized. Our Type 1 with the angular perpendicular cuttings (nos. 1-4) is well established in 260 B. C. and quite common down to the last quarter of the century. It is not excluded that its origin is even earlier (no. 5). Other constructions of the same date are represented by Type 2 (no. 8) and Type 4 (no. τι). The former is a broad external door of an important residential building, and the latter a folding screen door in another private house. Toward the middle of the second century B. C. Type 1 was modified (no. 6) and Type 3 (nos. 9-10) was introduced. The single door (no. 7) -- a unique specimen among our material - belongs to the second quarter of the second century, and the shop front door (Type 5, nos. 12-13) was introduced at about the same time, or possibly somewhat earlier. We have next to consider the form and function of the wooden doors themselves,

or — more

correctly — whether and to what extent the elaborate cuttings observed in the threshold blocks in Morgantina can throw any light on this rather obscure problem (1). (x) Historians of ancient architecture and excavators only seldom pay any attention to these matters,

and that for obvious reasons. Vitruvius’ classification of the various types of doors does not help us 29

ROYNE KYLLINGSTAD - ERIK SJOGVIST

There can be no doubt that our doors swung on vertical pivots, following the classical system used throughout antiquity. The pivot holes close to the posts, the bronze sockets preserved in nos. τ, 7 and 8, and the curved ruts in nos. 2-4 clearly prove this fact. It is probable that the

Ill LI Fio. 14.

l

pivots were strengthened by some metal mounting, in

order to make them more endurable. Our excavations have not produced any object that readily would lend itself to such a purpose. The question what they looked like is therefore left to our speculation. A certain guidance is offered by the few pivot sockets found in the threshold blocks, and they tell vs, at least, that all were not alike, The ones found in nos, 8 and 11 are of square shape and provided with a cylindrical central hole, To fit into such a socket, the pivot itself must have been correspondingly narrow. It mast thus have been of

metal.

A construction like that demonstrated in fig. 14

much (Vit. IV. 6). H. Philipp's article Thyra in «RE. (IL Reihe, Halbbd. rr, coll. 737-742) gathers much useful material and is the only comprehensive study of the matter that has been available to.us. A dissertation in Giessen by H. Klenk, Die antike Τῶν, 1924, is a theoretical and philological treatise, adding nothing to our specific problem, For Roman comparative material wereferto J.Durm, Die Baukunst der Romer, and ed. 1905, pp. 349-351. Three excavation publications have proved particularly helpful: Délos, Vol. 8: 2 by J. Chamonard dedicates pp. 261303 to doors and windows; Priene, where Th. Wiegand deals with them in the main publication, pp. 304306; and Olynthus (Excavations at Olynthus, Vol. 8, 1938, pp. 249-263) where we owe our information to J. Waiter Graham. Thè ancient town of Glanum in Provence is very rich in cut threshold blocks (H. Rolland, Fouilles de Glanum (Saint-Rémy deProvence), Paris 1946, pp. 72, 79, 93, 123 and figs. 55, 61, 74, 94), some of them dating from 3rd-and cent. B. C. None of these places has however produced thresholds. cut in the same way o: similarly to our Morgantina types 1-4, with the possible exception of a late specimen in Glanum located in the sanctuary of Hercules and dated to the first cont. I. C. (H. Rolland, Fouilles de Glanum 1947-1956, Paris 1058, p. 107, plan IX and pl. 37,4). A rarely quoted but fundamental study is that of S. Ivanof, Soglie Pompeiane in « Ann. dell'Inst. Corr. Arch. », Vol. 31 (1859), pp. 82-108, with illustrations in Mon, inediti, Vol. 6, pl. 28. It shows how close the relationship is between the material from Delos, Gianum and Pompeii.

The excavations at Karanis in Egypt, undertaken by the Kelsey Museum of the University of Michigan in the years 1924-1928 will prove of great importance for our subject when fully published. A preliminary report appeared 1931 by the hands of Arthur E. R. Boak and Enoca E. Peterson (Karanis, topographical and architectural report of excavations during the sea sons 1924-28. Univ. of Mich. Studies, Humanistic Series. Vol. s) The remarkably good preservation of the woodwork in doors and window.shutters of various kinds makes Karanis a mine of information on our subject. Dr. Peterson who is preparing the final publication has most generously provided us with additional answers to many questions and given us much detailed and useful information. For this we are deeply greteful. We cre also much indebted to Dr. George H. Forsyth, director of the museum, for allowing this information to be passed on to us. One group of material from Karanis belongs to the late first cont. A. D., another to the third and early fourth cent. As comparative material to our Hellenistic doors it is thus of a relative value only. Lintels, posts, pivots, and —in most cases — also the thresholds were of wood. Bolts and locking devices were very elaborate, and in some cases seemingly similar to ours. In general, bowever, all the material quoted above provides us with very few convincing parallels, and it is hero enumerated mainly as general references. Wo are painfully aware of the probability that much excavated but not yet published material may have escaped our attention.

should be postulated. The sockets in the corners o f threshold no. 1 are not as well preserved, but enough is left of them to show that they were of a different kind. There is no central hole in them, but the whole socket is concave and cup-shaped. . No narrow cylindrical metal peg could ever

30

HELLENISTIC DOORWAYS AND THRESHOLDS FROM MORGANTINA

ii

have turned in them. We have to reckon with a wider pivot which could well have consisted of a sturdy wooden peg, strengthened by a convex metal mounting as sug: gested in fig. rs. Variations on this type are naturally possible (1). The whole lower angle of the door could have been strengthened by a mounting of the type published by Durm (fig. 16) (). In such ἃ case the socket was probably cylindrical to receive the pivot, as demonstrated in fig. 17. We have no reason to believe that this system was used in Hellenistic Morgantina, as traces

of such rather substantial mountings are lacking. Fis. 15. We miss original lintel blocks from the excavations and are therefore without any tangible lead regarding the construction of the top pivots and their fixtures. Fig. 18 represents a possible pivot type corresponding to the bottom pivot in fig. 14 (threshold n. 11), and fig. 19 is based on the same system as in fig. 15 (threshold no. 1). In this case a metal mounting would have been necessary only on the

inside of the pivot, as the weight and pressure of the swinging door would follow the direction of the arrow (see figure). One should also calculate with the possibility -- even probability that the lintel in many cases was of wood. An inserted pivot peg of hardwood would under such conditions hardly have needel Fio. 16. any strengthening at all (?). How were the doors brought into position and how were they kept in place? Here our angular cuttings give us the answer. Fig. 20 demonstrates the

method better, we believe, than many words. The door was held obliquely; the top pivot was fitted into the

socket of the lintel; the bottom pivot was lead in and along the angular cutting, and the door was heaved and swung into place so that the bottom pivot caught its socket. There was no need of blocking the angular cutting when the door was in position, as pressure from outside was taken up by the angle itself. The weight

of the door, when opened or shut, pressed against the raised outside of the threshold

and the post.

It is a

practical and rather ingenious device which, together

with the bolts and the lock, makes the door burglarproof. An additional advantage lies in the fact that it was equally easy to unhook the door from inside when repairs were needed. Fie. 17.

‘These operations were still easier when working with

the later variation of Type 1 where the cuttings form

(1) Delos has produced sockets similar to oufs and corner. The piece comes from Egypt and is of Roman pivot mountings of basically the same type as the date. one here suggested. Delos, Vol. 8: 2, p. 285, fig. 157. (3) Such is the rule in Karanis, as Dr. Enoch E. (2) Durm, op. cit, p. 352, fig. 388, lower right Peterson kindly informs us. Cf. above noto 1, p. 29. 3r

ROYNE KYLLINGSTAD - ERIK SJOQVIST

an obtuse angle, allowing the operator to hold the door in an even more oblique position, when

putting it into place. A similar handling was possible in the case of threshold no. 8 (Type 2), only that the direction of the heaving and shoving was lateral, not perpendicular. A blocking of

the cutting would also have been necessary when the door was in position. That became mandatory also in the late Type 3 (thresholds nos. 9 and το), otherwise the door could have been pressed open from the outside. How the hinges, necessary for the satisfactory function of the multiple screen door (threshold no. 7) worked eA is obscure. One should however keep in mind that it must have been of a light construction as its only purpose was that of screening off the summer triclinium. from the shaded peristyle of the house. Under such conditions sturdy leather straps nailed to the door frames Fic, 18. may have given sufficient strength and flexibility (1). There remains to discuss our Type 5, the wide shopfronts with the longitudinal grooves in the threshold blocks. Their function was explained already

by Ivanoff in 1859 and definitively demonstrated by

Mau (?). F. H. Bacon showed that a similar system was in use in

the

windows

of the first floor of the

Hellenistic stoa in the agora at Assos (*), and lately

Gorham P. Stevens analyzed a shop-front in the Library of Pantainos in Athens applying the same solution (€). The long groove in the threshold corresponded to a similar one in the lintel, so that upright boards could be inéerted in the slots and be kept safely in place. Our fig. 21 shows the method applied to a threshold similar to our no. 13, narrow enough to require only one upright inserted board. The swinging door valve could be locked to the board and the door safely closed. When the shop front was wider, as in to block the opening and lock the boards, as suggested by Stevens, but the principle remained the same. It was a simple and cheap way of solving the problem of how to have a long counter in the open front of the

\

ll

our threshold no. 14, one had to insert more boards

taberna for the display of goods in the daytime, and

at the same time safeguard the house during the night. ‘There are innumerable examples of this method used in Roman markets and townhouses from late Repu-

Fie. 19.

(2) Real hinges such as those from Pompeii publish- fig. 184-185, presents and discusses an actual plaster ed by Durm, op. cit, p. 350, fig. 386 are not to be cast of such a door. excluded, but may seem less likely. (3) Investigations at Assos, Vol. 1, p. 93, fig. Y (2) Ivanof, in «Anm. Inst. Corr Arch. a, Vol. gr and pp. rosi. (1859), pp. roaf and illustration in Mon. inediti, (4) «Hesperia», Vol. 18 (1949), pp. 269-274. Vol. 6, pl. 28.10. Overbeck-Mau, Pompeji 4, p. 378. 32

HELLENISTIC DOORWAYS AND THRESHOLDS FROM MORGANTINA

blicän and Imperial times. Gorham Stevens asks the question where the method was used for the first time, and although leaving the question open, suggests Asia Minor as a possible place of origin (). He does so on the strength of the appearance at Assos of shop windows, closed in an analogous way. The date of the big four-storied stoa at Assos may be debatable, but we

Fio, 20.

concur with R. Martin, who considers it under Pergamene influence, and datable to the first half of the second century B. C. (’). That would make the windows at Assos approximately contemporary with the doors at Morgantina, and make them antedate any other known specimen of the type by about a century. Stevens may be right when considering Asia Minor as the place (a) Ibid, p. 274.

(2) R. Martin, Recherches sur l'agora grecque. Paris. 1951, pp. 486, 487, 507.

33

ROYNE KYLLINGSTAD - ERIK SJOQVIST where the ‘invention’ was made, Pergamon was an active architectural centre at this time, and Assos, as a neighboring town, in many ways reveals Pergamene influence. But it should also be remembered that Sicily of the third and early second century was a technically very advanced country, and that Syracuse was one of the biggest cities of its time. We have reason to believe that some technical innovations originated here in the hometown of Archimedes, In the modest

Fis. ar. realm of domestic architecture this can be exemplified by our excavations at Morgantina. The examples include the levigated floor mosaic (1), the ‘horizontal arch’ (*) and possibly also the rotating flour mill (9). It is therefore not excluded that our thresholds and the various methods of mounting and closing the doors revealed by them, are another addition to this list.

17

() K. M. Philips Jr, Subject and technique im (2) «AJA» 66 (1962), po. 139, 140. Heltenistic-Roman mesaics, «Art Bulls Vol 42 (3) D. White, A survey of milltones from Mor(1960), pp. 243-262. gantina, « AJA », 67 (1963), pp. 199-206. 34

Les couloirs annulaires dans l’architecture

funéraire antique HEMMING

WINDFELD-HANSEN

Les mausolées antiques de forme circulaire ou carrée ont parfois une particularité curieuse dans leur plan: ils comportent un couloir annulaire voüté en berceau. Ce couloir qui a une forme circulaire entoure une ou plusieurs chambres sépulcrales, placées dans le noyau massif de l'édifice. Dans la plupart des cas, un dromos croise le couloir annulaire pour mener à la chambre centrale, mais dans quelques-uns, le couloir annulaire est le seul accés à l'hypogée ou à la salle funéraire. On connait maintenant tout un groupe de mausolées antiques dont les chambres sépulcrales sont ou ont été entourées de couloirs annulaires, Un certain nombre de ces monuments funéraires n'existent plus et ils ne sont connus que par les relevés d'architectes italiens de la Renaissance et de l'époque baroque, alors que d'autres ont été découverts au cours de fouilles de l'époque moderne. C'est principalement en Italie que l'on rencontre les mausolées à couloirs annulaires, et ils datent presque tous de l'époque impériale, En dehors de l'Italie, on ne connait que trois édifices funéraires dont le plan présente cette dispositiori. Il s'agit, comme nous le verrons, de monuments numides qui se trouvent ou se trouvaient dans l'Algérie actuelle, deux de la période hellénistique, le dernier remontant à la fin de l'Antiquité. On se rend compte de la grande importance que présentaient ces couloirs annulaires dans l'architecture funéraire antique si l'on considère qu'ils existent dans les deux grands mausolées impériaux de Rome: celui d'Auguste et celui d'Hadrien. ΤΙ convient aussi de mentionner brièvement que les couloirs annulaires des mausolées romains ont marqué de leur empreinte non seulement l'architecture funéraire de l'Antiquité tardive, mais aussi l'évolution des rotondes paléochrétiennes à division intérieure concentrique, c'est-à-dire à déambulatoire annulaire placé autour d'une piéce centrale ronde. Koethe (1) est un des premiers qui aient supposé que les couloirs annulaires des mausolées romains ont eu une grande importance pour l'évolution de l'architecture du moyen-âge. Il pensait que les cryptes paléochrétiennes de plan semi-annulaire et les déambulatoires de plan semi-circulaire des choeurs des églises médiévales avaient comme modèles ces couloirs annulaires des mausolées antiques. Dans son « Martyrium », Grabar (*) a démontré que les déambulatoires intérieurs des martyria et des églises à plan circulaire consacrées aux martyrs ont leur prototype dans les

couloirs annulaires des mausolées antiques.

(1) H. Koethe, Das Konsantinsmáwsoleum und verwandie Denkmäler, «Jahrbuch des deutschen archäologischen Instituts », XLVIII, 1933, p. 200, n. 1.

(2) A. Grabar, Martyrium, Recherches sur le culto des reliques et l'art chrétien antique, 1, Paris 1946, pp. 146-147, 172, 191, 192, 258-260, 343-344, 366-367. 35

HEMMING WINDFELD-HANSEN + eta

Il n'existe pas d'exemples de couloirs annulaires dans les édifices funéraires antiques de la période archaïque et de la période classique, et ces couloirs n'ont pas de prototypes dans Parchitecture funéraire préhistorique des pays méditerranéens. D'autre part, certains tombeaux préhistoriques ἃ tumulus et quelques mausolées circulaires antiques sont entourés de passages ou de

plates-formes qui doivent être mis en relation avec les couloirs annulaires des mausolées romains et numides. En Europe occidentale et septentrionale, les dolmens des périodes néolithique et chalcolithique sont fréquemment entourés d'un ou de plusieurs cercles de pierres. Le plus souvent, ces cercles sont composés de blocs de pierre ou de dalles juxtaposés formant une enceinte autour du tumulus qui couvre ou qui couvrait à l'origine le dolmen (!). Mais dans d'autres cas, la base du tumulus est entourée d'un cercle extérieur se composant de grands blocs verticaux, séparés

les uns des autres (7). Parmi ces cercles extérieurs de pierres qui formaient une délimitation du monument funéraire, on doit mentionner celui qui entourait le tumulus du grand tombeau

mégalithique de New Grange en Irlande (9). Les cercles de pierres, dont il existe des traces autour de certains des tumuli qui couvrent des tombeaux à coupole dans la nécropole de Los Millares en Sspagne, sont aussi des délimitations de l'aire funéraire. Parmi ces monuments, on peut citer le tombe no. 16 dont le tumulus a un diamètre de 11 m. À 13 m. du centre de la chambre sépulcrale se trouvent les vestiges de deux cercles concentriques de dalles de pierre (ἢ). En Grande-Bretagne, des fouilles exécutées dans des tombeaux à tumulus appartenant à la civilisation campaniforme et datant de la période chalcolithique et de l'âge du Bronze ancien (1700-1500 av. J.-C.) ont démontré que la tombe placée au centre de ces monuments était souvent à l'origine entourée d'un fossé circulaire qui probablement peu de temps après les funérailles avait été comblé de terre ct couvert d'un tumulus (9. A Coity, dans le comté de Glamorgan, on a mis à jour un tombeau circulaire de l’âge du Bronze moyer. (1100 av. J-C.), le Pond Cairn, où l'on a déblayé un étroit passage annulaire, qui était placé entre la tombe centrale couverte d'une pile cylindrique de tourbe et un remblai en forme d'un large anneau construit de pierres sèches (9. Dans le méme ordre d'idées, il faut mentionner une série de tombeaux à tumulus de l'âge

du Bronze, qui étaient à l'origine de larges constructions circalaires en bois. Ces constructions, (3) J. Déchelette, Manuel d'archéologie préhistorique, celtique et gallo-romaine, I, Paris 1908, pp. 444447: O. Menghin, Europa und cinige angrensende Gebiete ausser dom dgdischen und italischen Kreis, dans «Handbuch der Archiologie », II, 1, Munich 1954, PP. 95-96, figg. 42-46. (2) TI no faut pas confondre ces cercles funéraires de pierres avec les cromlechs de Bretagne et les grands monuments circulaires d'Angleterre appelés «henge monuments», comme par ox. le Stonehenge pris de Salisbury. Ces monuments, qui se composentde plusieurs cercles de pierres levées, souvent réunies par des linteaux, ne sont pas des tombeaux, mais des édifices de caractère sacré. Voir: J. Déchelette, los cit; R. J. C. Atkinson, Stonehenge, Londres 1960, passim.

(3) E. O. James, Prehistoric Religion, Londres 1957, p. 88, fig. 7; Glyn Daniel, The Megalith Builders of Western Europe, Londres 1962, pp. 113-115, fig. 20. (4) G. et V. Leisner, Dis Mogalithgräber der it rischen Halbinsel, orster Teil: Der Süden, Berlin 1943, PP. 31-22, pl. τά, 1 (5) De tels fossés ont été trouvés sous un tertre près de Ysceifiog dans le comté de Flintshire et sous un tombeau è tumults «Sutton 268» à Llandow dans le somté de Glamorgan. Voir: C. Fox, Life and Death in the Bronze Age, Londres 1959, pp. 1-11, figg. 1-6, pls. 1-3 et pp. 62-70, figg. 40-42 a, pls. 1421. (6) C. Fox, op. cit, pp. 105-127, figg. 59-63, pls. 29-38. 36

LES COULOIRS ANNULAIRES

qui ont été déblayées au cours de fouilles exéoutées en Grande-Bretagne, aux Pays Bas et en Allemagne occidentale, étaient composées de plusieurs cercles concentriques de poteaux, qui formaient des palissades autour d’une cabane centrale. Celle-ci était aussi une construction en bois et abritait la tombe placée au centre de l'aire funéraire (1). Dans les régions centrales et orientales de la Méditerranée, en particulier en Italie, les-tumuli de certains tombeaux préhistoriques sont aussi entourés d'un cercle de pierres. En plusieurs endroits, ce cercle est le seul vestige qui indique maintenant l'emplacement du tumulus, depuis longtemps disparu (?). Dans d'autres cas, les tumuli étaient élevés sur des soubassements Tonds formés de pierres amoncelées, mais qui à l'extérieur avaient l'apparence d'un mur (). En Grèce, un grand nombre de tombeaux à coupole de la période mycénienne étaient couverts de tumuli qui sont maintenant plus ou moins détruits (+). On ne peut donc plus constater, si ces tombeaux à tumulus avaient été entourés d'un mur de clôture. Mais les célèbres tombes à fosse de la période mycénienne ancienne, qui ont été retrouvées dans le grand cercle funéraire de la citadelle de Mycènes, sont entourées d'une clôture circulaire de 26 m. 50 de diamètre (). Cette clôture qui se compose d'une double rangée de dalles dressées supportant un couvercle de plaques horizontales ne remonte pas plus haut que l'époque de la reconstruction de la citadelle (1350-1300 av. J.-C. Des cercles funéraires plus anciens datant de l'Helladique moyen (1900-1600 av. J.-C.) ont été trouvés au cours de fouilles plus récentes prés de Mycènes et è Malthy en Messénie (1). Dans le méme ordre d'idées, il faut aussi mentionner que le sanctuaire de Pélops ou Pé dans l'Altis è Olympie était à l'origine entouré d'un téménos formé d'un cercle de pierres d'environ 33 m. de diamètre (?). Ce téménos faisait le tour du monticule où était enterré Pélops. Le tumulus conique fut toujours considéré par les Grecs comme la forme primitive de la tombe (8). Cette forme de tombe était généralement utilisée pendant la période géométrique et plus tard pendant la période archaïque (*), mais ces tumuli grecs étant presque toujours de très petite dimension, il est rare que l'on ait trouvé des traces du monticule lui-méme au cours des fouilles. Les tombeaux à tumulus de caractère monumental en Grèce, comme par exemple le Séros, élevé dans la plaine de Marathon pour les morts de la bataille contre les Perses en 490 av. J-C., sont tellement détruits qu'il n'est plus possible de se prononcer sur leur forme primitive (9). buch der Archdologie », TL, 1, p. 198, fig. 74. Les tombes de Gravina en Apulie ont une construction analogue. Voir: F. von Duhn, Italische Graberkunde, II, Heidelberg 1939, pp. 288, 318, pl. 35a-e. (4) N. Valmin, Tholos Tombs and Tumuli, «Acta instituti Romani regni Sueciae», IT, 1932, pp-216-227. (5) W. B. Dinsmoor, The Architecture of Ancient Greece, Londres 1950, pp. 25-26, fig. 11; A. W. Lawrence, Grech Architecture, Londres 1957, p. 75, fig. 42. (6) A. W. Lawrence, loc. cit; F. Matz, Kreta, Myhene, Troja, Stuttgart 1957, D. 119. (n) P. Goessler, Das Pelopsgrab in Olympia und seine hultische Bedeutung, «Bericht über den VI. internationalen Kongress für Archäologie », Berlin 1940, pp. 184-186. (8) H. V. Herrmann, Omphalos, (Orbis antiquus, no. 13), Münster 1959, p. 40. (9) loc. cit. (ro) V. Stais, dans «Athenische Mitteilungen», XVIII, 1893, pp. 46 ss.

(1) Parmi ces tombeaux circulaires, on peut mentionner le tumulus + Sheeplays 203» près de Llantwit dans le comté de Glamorgan. Voir: C. Fox, op. cit, pp. 128-143, figg. 65-69, pls. 39-41. Pour les tombeaux de cette forme en Allemagne occidentale et aux Pays Bas voir: O. Menghin, op. cit., pp. 96, 100, 107, fig. 57; W. Glasbergen, Barrow Excavations in the Eight Beatitudes, The Bronze Age Cemetory between Toterfout and Halve Mijl, North Brabani, « Paleohistoria », II, 1954, pp. 1-134, ΠῚ, 1954, pp. 1-204; S. J. de Laet, The Law Countries (Ancient Peoples and Places), Londres 1958, pp. 120-122, fig. 31. (à) F. von Duha, Italische Graberkunde, 1, Heidelberg 1924, PP. 124, 134, 187, 197-108, 199, 220-223, 225-226, 287-288, pl. 16, fig. 50, PI. 17, fig. 57, pl. 18, fig. 58; id., art. Steinkreisgrab, dans «Reallexikon der Vorgeschichte », vol. XII, Berlin 1928, pp. 406-408. (3) Par ex. les tombeaux circulaires dans l'ile de Leucade. Voir: F. Matz, Die Agdis, dans « Hand37

HEMMING WINDFELD-HANSEN

En Italie par contre, il existe un grand nombre de tombeatx étrusques à tumulus bien conservés, notamment de la période archaïque. Ces tumuli se trouvent en particulier dans les nécropoles de Caere (Cerveteri) et autour d'autres villes de l'Etrurie méridionale, comme par exemple Biedà, ainsi qu'aux alentours des villes de Vetulonia et de Populonia dans la région côtière de YEtrurie septentrionale, Dans les nécropoles de Podere di S, Cerbone et de Poggio della Porcareccia près de Populonia, on a dégagé de grands tombeaux à tumulus dont la orépis est entouré d'un dallage circulaire formé de plaques de pierre calcaire (1). Le Mausolée d'Alexandre le Grand que Ptolémée II Philacelphe (284-246) fit édifier à Alexandrie a été remplacé par Ptolémée IV Philopator (221-205) per un monument funéraire encore plus grand appelé le S4ma. Celui-ci devait servir de mausolée commun aux membres de la dynastie des Ptolémées et fut pour cela aussi appelé Prolémaion (*). Il paraît ressortir des sources littéraires qu'aussi bien ce nouveau mausolée que son modèle primitif avaient une forme de tumulus (). L'archéologue allemand Thiersch a soutenu que le Mausolée d'Alexandre était identique à la colline artificielle conique appelée Panéion (ἢ) dont parle Strabon dans sa description d'Alexandrie (), Strabon indique que le Panéion était parcouru de chemins en spirale qui montaient jusqu'au sommet de la colline. Thiersch pensait que ces chemins étaient des couloirs en spirale placés à l'intérieur d'un tumulus, mais il ressort clairement de la description de Strabon qu'il s'agissait uniquement de sentiers dans un parc. Dans un ouvrage un peu antérieur (5, Thiersch avait d'ailleurs reconnu que le Pandion ne pouvait pas être identifié au Séma des Ptolémées. Sa théorie, suivant laquelle le Mausolée d'Alexandre aurait été le modèle des édifices funéraires romains à couloir annulaire, ne s'est donc pas confirmée. L'énorme tumulas conique qui forme le sommet artificiel du Nemruc Dagh, de 2 150 m. d'altitude, dans le massif de l'Anti-Taurus, en Asie Mineure, est un des seuls exemples encore existant de mausolées princiers de l'époque hellénistique ('). Le cumulus proprement dit, de 49 m. de hauteur et d'environ 150 m. de diamètre, est sur trois côtés bordé de larges terrasses. Deux de celles-ci portent les vestiges d'un grand nombre d'autels et de statues colossales alignées qui servaient au culte funéraire, Durant les fouilles effectuées dans les années 1953-1956, le mur d'enceinte du tumulus a été retrouvé, et sur ce mur on a déconvert une voie dallée de plaques de pierre. Cette voie entourait le cône du tumulus et unissait les terrasses entre elles. Plusieurs inscriptions en grec indiquent que le monument a été édifié comme tombeau pour le

roi Antiochus I° (60-34 av. J.-C.) de Commagène. On ne sait pas si le couloir annulaire a été utilisé dans l'architecture funéraire des grands ‘empires hellénistiques il existe toutefois en Cyrénaique (*), soumise pendant la plus grande partie

(6) H. Thiersch, Pharos, Antike, Islam und Occidont, Leiprig 1909, p. 144. (7) K. Human et O. Puchstein, Reisen in Kleinasien und Novdsyrien, Berlin 1800, pp. 234-353, pls. XX-XL; T. Goell, The Excavation of the « Hierothesion » of Antiochus I of Commagene on Nemrud Dagh (1953-56), + Bulletin of the American School of Oriental Research », no. 147, oct. 1957, PP. 4-22. (Δ) 1. R. Pacho, Relation d'un voyage dans la Marmarique, le Cyrénaïque ot les oasis d'Andjelah et de Mavadis, Taxte, Paris 1827, pp. 371-372, Atlas, Paris 1829, pl. XXIV-XXV; R. Murdock Smith et E. A. Porcher, History of the recent Discoveries at Cyrene made during the Expedition to the Cyrenaica in 1860-61, Londres 1864, p. $8, pl. 45.

(1) À. Akerstrom, Studion über die etruskischen Graber, « Acta Instituti Romani regni Sueciae », ΠῚ, 1934, PP. 149, 151, 152, 153; A. Minto, Populonia, Florence 1943, pp. 76-101, figg. 20 a-b, 26-20, pls. XVII, 1; XIX, 1. (2) H. Thiersch, Die alevandrinische KénigsneHropole, « Jabrbuch des deutschen archiologischen Instituts » XXV, 1910, pp. $597; E. Breccia, Alexandre ad Aîgyptum, Bergamo 1922, pp. 9698. (3) Lucain, Pharsale, VIII, 694 ss; X, 19 ss. (4) H. Thiersch, op. cit., pp. 85-87; R. Pagenstecher, Nekropolis, Leipzig 1919, pp. 6-7. (5) Strabon, XVII, 1, το; E. Breccia, op. cit, p. τοῦ 38

LES COULOIRS ANNULAIRES de la période hellénistique ἃ l'Égypte des Ptolémées, des exemples de petites tombes circulaires

qui forment la transition entre les tombeaux à tumulus dont le mur d'enceinte circulaire ou le socle est entouré d'un passage dallé de plaques de pierre et les mausolées ronds où ce passage extérieur est remplacé par un couloir annulaire intérieur. On trouve un exemple d'un tombeau circulaire de ce genre à Ghermés ou Mgarnes, à 13 km. à l'Ouest de Cyrène (PL. XII 5). Le tombeau circulaire lui-même, qui a un diamètre d'environ 3 m. 50 est entouré d'un mur concentrique. Ce mur, dont le diamètre est d'environ 7 m. ne forme pas une clôture extérieure, mais appartient au monument Iui-méme, car il est placé sur son socle carré. Entre le mur concentrique et le tombeau, il y a une sorte de couloir annulaire qui est un passage découvert. I. - LES

MAUSOLÉES

ROMAINS

À COULOIRS

ANNULAIRES.

xa) Mausolées circulaires à deux couloirs annulaires concentriques.

Mausolée circulaire avec un caveau annulaire placé autour d’un pilier centralet entouré de deux couloirs annulaires: le Mausolée d’Auguste à Rome. Le seul mausolée circulaire conservé avec un caveau entouré de deux couloirs annulaires est le Mausolée d'Auguste à Rome (PI. I a) (1). Ce gigantesque monument dont l'Empereur Auguste a fait entreprendre la construction en l'an 28 av. J.-C. est unique en son genre parmi les mausolées circulaires romains, non seulement par l'importance de ses dimensions (diamétre: 89 m. 30), mais aussi parce qu'il est une combinaison du simple tombeau à tumulus et du mausolée circulaire en

forme de tour. Le centre du grand tumulus, dont l'existence est connue grâce à la description que Strabon (ἢ

a fait du mausolée, était surmonté par une construction cylindrique couronnée d'une statue de

l'Empereur, élevée sur un piédestal: C'est grâce à une disposition intérieure propre à ce mauso-

lée, la présence de plusieurs murs concentriques, qu'il a été possible de construire cette super-

structure, Il s'y trouve au total cinq murs circulaires placés autour d'un pilier central qui se dressait jusqu'au sommet du monument et qui supportait la dite statue de l'Empereur. Ces murs

divisent l'intérieur du mausolée en cinq zones annulaires concentriques dont les trois intérieures constituent l'hypogée méme, alors que les deux zones extérieures forment une armature, avec des murs établis suivant des axes rayonnants ou en arcs concaves. Dans la zone extérieure, les murs curvilignes et le mur d'enceinte du mausolée délimitent 12

cavités de forme pratiquement semi-circulaire, alors que dans la zone suivante les murs rayonnants (1) A. M. Coliniet G. Q. Giglioli, Scavi del Mausoeo d'Augusto, « Bullettino Comunale », LIV, 1926, Pp. 191234; A. Bartoli, L'architttura del Mausoleo di Augusto, « Bollettino d'Arte », VII (série II), 1927, PP. 30-45; R. A. Cordingley et I. A. Richmond, The Mausoleum of Augustus, « Papers of tho British School at Rome », X, 1927, pp. 23 ss.; S. B. Platner et Th. Ashby, Topographical Dictionary of Ancient Rome, Londres 1920, pp. 332-335; G. Ὁ. Giglioli, ΤΙ sepolcrelo imperiale, «Capitolium», VI, 1930, pp. 532567; G. Gatti, I1 Mausoleo di Augusto, « Capitolium », X, 1934, pp. 457-464; id., « Bullettino Comunale », LXVI, 1938, pp. 273-275; A. Muñoz, La sistemazione

del Mausoleo di Augusto, « Capitolium ν, XIII, 1938, pp. 491-508; G. Lugli, Monumenti antichi di Roma, III, Rome 1938, pp. 194-211; M. E. Blake, Ancient Roman Construction in Italy from the prehistoric Period to Augustus, Washington 1947, pp. 171-172, 264; G. Lugli, La tecnica edilizia romana, Rome, 1957, BP. 180, 187, 315, 320, 329, 430, 506, 508; L. Crema, L'architettura romana (Enciclopedia Classica, section III, vol. XII, t. I, Turin 1950, pp. 243-244, figg. 263265; E. Nash, Bildlexiton sur Topographie des antihen Rom, IL, Tübingen 1962, pp. 38 ss. (2) Strabon, V, 3, 8.

39

HEMMING WINDFELD-HANSEN

séparent 12 chambres fermées de plan trapézoidal. Toutes ces cavités et chambres avaient pour fonction de maintenir en place les masses de terre du tumulus qui recouvrait une partie du mauso1ée. Simultanément, les murs rayonnants reportaient le poids des murs de l'hypogée et de la superstructure cylindrique sur l'enceinte circulaire, tandis que les murs en arc concave allégeaient la pression exercée par les masses de terre du tumulus sur cette enceinte. Les trois zones intérieures de l'hypogée du Mausolée d’Auguste sont limitées à l'extérieur par le 3° mur circulaire uniquement percé par un dromos voüté en berceau. Le pilier central est entouré d'un caveau sépuleral annulaire qui est limité par le 1" mur circulaire et recouvert d'une voîte en berceau. Les deux couloirs annulaires qui entourent cette chambre sépulcrale sont placés respectivement entre les 1° ct 2° et 2° et 3° murs. Le dromos aboutit devant le 2° mur, où il débouche dans le couloir annulaire extérieur. Ce mur était probablement percé de deux portes qui donnaient accès au couloir annulaire intérieur. Ces portes sont comblées par des masses de magonnerie éboulées, et la porte qui menait du couloir intérieur à Ja chambre sépulcrale a également été bouchée par l'effondrement des voûtes de ce caveau. Mausolée circulaire avecun escalier central touré de deux couloirs annulaires.

en

colimagon

en-

On considère généralement que le Mausolée d’Auguste est le seul monument funéraire τὸ main formé de plusieurs murs concentriques, mais une reconstitution de plan dans les dessins et les gravures de Gio. Batt. Montano semble indiquer qu'il & existé un autre mausolée circulaire ayant cette structure intérieure (PI. I0) (1. Du plan, il ressort que ce mausolée était formé

de trois murs concentriques placés autour d'un escalier en colimagon central. Entre les murs, il

y avait deux couloirs annulaires concentriques. Dans le mur extérieur qui formait l'enceinte du Soubassement cylindrique, une porte donnait accès à un dromes qui traversait les couloirs annulaires et menait

à l'escalier central.

15) Mausolées circulaires à un seul couloir annulaire.

A l'époque d'Auguste, les constructions ἃ murs concentriques ont aussi été utilisées dans quelques grands mausolées circulaires ayant un corps cylindrique élevé. On trouve dans ces édifices un noyau de forme cylindrique construit en opus caementicium, dans lequel sont aménagés un ou plusieurs caveaux sépulcraux. Un mur circulaire qui forme l'enceinte du cylindre et qui est revêtu de pierres de taille à l'extérieur entoure le noyau comme ane enveloppe, séparée de celui-ci

par un étroit couloir annulaire. Dans la plupart des monuments zonservés, ce couloir est aussi haut que le mur d'enceinte du cylindre et il est recouvert d'une voûte en berceau. Mausolées circulaires avec quatre caveaux rectangulaires placésdanslesaxes principaux del'édificeetentourésd'uncouloir annulaire.

Le Mansolée de Munatius Plancus prés de Gate (PL. Ic) (+) est un édifice circulaire bien conservé en forme de tour. Cet édifice repose sur un soubassement circulaire bas. Un corps cyli drique, de 29 m. so de diamètre, muni d'une base et se terminant par un entablement, se dresse (1) Paris, Bibliothèque Nationale, Cab. des Est, Hb. 22, pl. 7o; Gio. Batt. Montano, Libro ΠῚ, Rome 1684, pl. 41.

(4) G. Q. Giglioli, La Tomba di L. Munazio Planco, «Architetturaed Arti decorative», I, 1921-1922, pp. 507. 525; M. E. Blake, op. cit., p. 170; S. Aurigemma et 4o

LES COULOIRS ANNULAIRES

sur ce soubassement. L'intérieur du mausolée se compose d'un noyau cylindrique entouré du mur d'enceinte du cylindre. Sur le cóté Nord du corps cylindrique, se trouve la porte d'entrée donnant accès à un court dromos qui traverse le mur d'enceinte du cylindre. Le dromos débouche dans un couloir annulaire séparant le mur d'enceinte du noyau cylindrique. Ce couloir, dont les parois sont revétues de masonnerie réticulée, est recouvert d'une voûte en berceau, très élevée. Dans le noyau cylindrique du mausolée étaient aménagés quatre caveaux rectangulaires identiques, placés dans les axes principaux de l'édifice correspondant aux quatre points cardinaux. Dans les mêmes axes, on trouve des corridors qui mènent du couloir annulaire aux caveaux, et on ne peut arriver à ceuxci qu'en parcourant une partie du couloir. L'inscription inaugurale du mausolée indique qu'il a été édifié comme tombeau pour le général et homme d'état romain Lucius Munatius Plancus. Toutefois, l'année de sa mort étant inconnue, il n'a pas été possible de dater le monument d’une manière précise. Munatius Plancus est né à Tibur vers l'an 9o av. J.-C. et on pense qu'il est mort dans la 2* décennie avant notre ère, Fellmann (1) place donc le mausolée vers l'an 20 av. J. au plus tôt.

Un autre mausolée circulaire qui se trouve aussi près de Gaète est un peu plus petit, mais

d'une forme correspondant à celui de Munatius Plancus. Ce mausolée (PI. I d) (2), d'environ 24 m.

de diamètre, se composé aussi d'un noyau cylindrique et d'un mur d'enceinte concentrique. Entre Je noyau et cette enceinte, il y a un couloir annulaire voûté en berceau, dont les parois sont recouvertes d’opus reticulatum. Dans le noyau se trouvent quatre caveaux, placés suivant les axes principaux du monument, autour d'un pilier central carré. Les trois caveaux qui donnent au Nord, à l'Est et au Sud sont de plan rectangulaire identique et voütés en berceau. Le caveau donnant à l'Ouest est par contre

de forme elliptique, mais il est probable qu'il avait à l'origine la méme forme que les autres. Entre les caveaux se trouvent des cavités triangulaires inaccessibles. Aurigemma () indique que le mausolée a. été le tombeau de Lucius Sempronius Atratinus, préfet de la flotte d'Antoine de 38 à 34 av. J.-C., et consul pendant cette dernière année. Selon les sources littéraires, il est mort en l'an 7 ap. J.-C. Ce monument est donc postérieur au Mausolée de Munatius Plancus (*).

Mausolée circulaire avec deux caveaux rectangulaires placés dans le méme axe et entourés d'un couloir annulaire. Un plan et une élévation dus à Pirro Ligorio nous font connaitre un mausolée circulaire maintenant disparu (PI. IIa) (5). Son plan rappelait celui des deux mausolées prés de Gaète. Dans ce tombeau qui, d'après les indications de Ligorio, était situé sur la Via Labicana dans les environs de Rome, il y avait dans le noyau cylindrique deux caveaux rectangulaires placés dans le méme axe que la porte d'entrée. Les caveaux étaient entourés d'un couloir annulaire muni sur sa paroi intérieure de six niches semi-circulaires. A. De Santis, Gaeta, Formia, Minturno («Itinerari dei musei», n. 92), Rome 1955, pp. 7-8, fig. 1, pl. I; R. Fellmann, Das Grab des Lucius Munatius Plancus bei Gaeta, « Schriften des Institutes für Ur. und Frühgeschichte der Schweiz», 11, Bâle 1957, passim; L. Crema, op. cit, pp. 244-247, figg. 267, 269. (ἡ R. Fellmann, op. cit, p. 31. (2) S. Aurigemma et A. De Santis, op. cit, PP. 8:9; L. Crema, op. cit, pp. 244-247, fig. 270.

(3) S. Aurigemma et A. De Santis, op. cit., p. 9. — (4) Pauly-Wissowa, « Realencyclopädie der classischen Altertumswissenschaft », II A, 2, Stuttgart 1923, cols. 1366-1368. (5) Oxford, Bodleian Library, Cod. Canonici ital. no. 138, f. 142 v. Voir: Th. Ashby, The Bodleian MS. of Pirro Ligorio, «Journal of Roman Studies », IX, 1919, p. 200. ar

HEMMING WINDFELD-HANSEN

Mausolée circulaire avec une chambre sépulerale centrale à plan cruciforme, entourée d'un couloir annulaire. Le Mausolée des Plautiens prés du Ponte Lucano sür la Via Tiburtina (1) est un tombeau circulaire en forme de tour dont le cylindre élevé, d'énviron 18 m. de diamètre, a conservé le mur circulaire. Celui-ci forme maintenant l'enceinte, revétue de pierres de taille, autour d'un cylindre creux. Dans son aspect actuel, le mausolée rappelle Je grand tombeau circulaire appelé Torre di Micara à Frascati (2) où l'enceinte extérieure entoure aussi un zylindre creux. Un plan de Giuliano da Sangallo (Pl. II 2) () montre toutefois qu'à l'intérieur du cylindre du Mausolée des Plautiens, il y avait jusqu'à la fin du 15° s. une chambre sépulcrale centrale, à plan cruciforme. Elle était placée dans un noyau cylindrique, séparé de l'enceinte encore conservée par un couloir annulaire, Dans sa forme primitive, ce mausolée peut être daté dc la fin de l'époque d’Auguste, La partie supérieure du cylindre dans lequél se trouvait cette chambre sépulcrale a probablement été ajoutée au milieu ou dans la deuxième moitié du 1° s. ap. J.-C. (4). Mausolée circulaire avec une galerie hélicoidale menant a une chambre sépulcrale centrale ἃ plan cruciforme: le Mausolée d'Hadrien à Rome. ᾿

Le Mausolée de l'Empereur Hadrien (Château Saint-Ange) à Rome (PI. II c) (€) est étroitement apparenté aux mausolées circulaires qui ont un noyau cylindrique entouré d'un mur d’enceinte séparé du noyau par un couloir annulaire. II est vraisemblable que ce mausolée avait été projeté à l'origine comme un édifice circulaire ayant le caracsère d'une simple tour. Il semble que les plans aient été modifiés au cours de la construction et le monument fut achevé en l'an. 139 ap. J.-C. par Antonin le Pieux avec un soubassement carré entourant la partie inférieure du

cylindre, La maçonnerie à l'intérieur du corps cylindrique est généralement considérée comme étant homogéne. Il ressort cependant des relevés du mausolée qu'il y avait à l'intérieur du cylindre un noyau central ayant la forme d'une tour carrée. Ce noyau se dresse jusqu'au sommet de l'édifice et portait une sculpture ~ probablement un quadrige - couronrant le monument. Dans ce noyau (1) A. Nibby, Amalisi storico - topografico - antiquaria della carta dei dintorni di Roma, III, Rome 1837, pp. 640-642; L. Canina, Edifisi di Roma antica è sua Campagna, Rome 1848-1856, V, pp. 109-111,

(3) Rome, Bibliothèque Vaticano, Cod. Barb. Lat. no. 4424, 1. 41 v. Une reproduction en fac-similé de ce dessin se trouve dans: Ch. Hillsen, Il libro’ di Giuliano da Sangello, Leipzig roro, f. 4t v (description,

VI, pls. CXXLCXXIL figg. 1-6; Th. Ashby, The Classical Topography of the Román Campagna, II, «Papers of the British School at Rome», IIT, 1906, p.127; G. T. Rivoira, Architettura romana, Milan 1921, PP. 17, fig. 14, 275, fig. 266; B. Goto, Ein Rundgrab in Falerii, Baugeschichto des rómischen Adels- und Kaisergrabes, Stuttgart 1939, p. 11, pl. IV E-F; M. E. Blake, op. cit, p. 171; R. Fellmann, op. cit, PP. 74-75; L. Crema, op. cit, p. 250, fgg; 276, 279. (2) G. Me. Cracken, The Villa and Tomb of Lucullus at Tusculum, « American Journal of Avchaeology», XLVI, 1925, pp. 325-340; M. E. Blake, op. sit, pp. 169-170; R. Péllmann, op. cit, pp. 68.69;

ibid., p.58). (4) R. Fellinann, op. cit, p. 75. (5) H. Jordan et Ch. Hiüsen, Die Topographie der Stadt Rom im Altertum, 1, 3, Berlin 1907, pp. 663667; E. Rodocanachi, Le Chateau Saint-Ange, Paris 1909, passim; S. B. Platner et Th. Ashby, op. cit, PP. 336-338; M. Borgatti, Il Mausoleo di Adriano, Rome 1931, passim; S. Rowland Pierce, The Mausoleum of Hadrian and the Pons Aelius, « Journal of Roman Studies, XV, 1925, pp. 75-103; G. Lugli, Monumenti antichi di Roma, TIT, Rome 1938, pp. 693. 708; id., La feonica edilizia romana, pp. 209, 215, 306, 333, 439, 605-606, 673, 681; R. Fellmann, op. cit, PP. 77-78; L. Crema, op. cit, p. 484, figg. 622-623.

42

LES COULOIRS ANNULAIRES

en formé de tour sont placées non seulement la grande chambre sépulcrale à plan ciuciforme, voûtée en berceai, mais aussi les deux chambres qui se trouvent au-dessus de cette cella, la chambre supérieure étant ronde et couverte d'une coupole.

Autour du noyau central, un mur circulaire de grandes dimensions forme l'enceinte du cylindre de 64 m. de diamètre. Ce mur était à l'extérieur revêtu de deux couches différentes de pierres de taille, Le noyau central et le côté intérieur de l'enceinte sont construits en opus caementicium et l'espace qui les sépare parait étre rempli de terre et de pierres sèches. La construction avec un mur d'enceinte a donc aussi été utilisée dans ce mausolée, mais l'enceinte n'y est pas

une mince enveloppe autour d'un noyau massif comme dans le Mausolée de Munatius Plancus.

Elle est devenue au contraire l'élément dominant de la structure du cylindre, alors que le noyau lui-méme se réduit à une construction en forme de tour relativement étroite au centre de l'édifice.

Dans le Mausolée d'Hadrien, on trouve une galerie annulaire hélicoïdale, placée sur le côté intérieur de l'enceinte du cylindre. Cette galerie — à laquelle on accède du vestibule voîité en berceau qui est situé au bout du dromos — forme une rampe circulaire de 3 m. 20 de largeur, voütée en berceau. Celle-ci entoure le noyau central et mène à un point situé verticalement au-dessus du vestibule, De ce point, un couloir conduit au sépulcre méme. Mausolées circulaires avec une salle ronde centrale entourée d'un couloir annulaire concentrique. Le seul exemple partiellement conservé d'édifice circulaire de l'époque impériale qui ait une salle ronde à coupole, entourée d'un étroit couloir concentrique, est un petit mausolée situé près

de Casale Portonaccio au Sud de la Via Tiburtina dans les environs de Rome (PI. III a) (1). Il

s'agit ici d'une rotonde isolée, dont la salle à coupole a un diamètre de 6 m. 20. Le couloir annulaire qui est très bas est coupé par un mur sur un de ses côtés.

Dans la Vigna Pepoli à Rome, on trouve les ruines d'un grand édifice circulaire (PI. III 2) (ἢ,

qui était probablement un mausolée en forme de tour ronde, mais dont il n'est conservé qu'une

partie du cylindre. D'un plan de Lanciani, qui s'appuie sur Je rapport des fouilles du milieu du siècle dernier, il ressort que dans la partie inférieure du cylindre, il y avait une salle circulaire centrale, munie de cinq niches rectangulaires, Cette salle était entourée d'un couloir annulaire auquel on pouvait accéder des deux côtés du dromos. La maçonnerie en opus reticulatum qui revêt le noyau intérieur doit dater du début de l'Empire. Parmi les relevés et les reconstitutions des architectes italiens de la Renaissance et de l'époque

baroque, on trouve trois exemples de mausolées circulaires en forme de rotonde, maintenant disparus, dont la salle circulaire à coupole était entourée d'une sorte de couloir logé dans l'épaisseur du mur. L'un de ces édifices est reproduit sur un plan de Bramantino (Pl. IIT c) (*) et aurait

été situé prés de Tre Fontane au Sud de Rome. Les deux autres mausolées circulaires de ce type ont l'aspect de tours et ne sont connus que par des reconstitutions de Gio. Batt. Monta-

no. Dans l'un des cas, il s'agit d'un mausolée qui aurait été situé dans la Campagne romaine

() Th. Ashby, The Classical Topography of the Roman Campagna, IL, «Papers of the British School at Rome», IIl, 1906, p. 95 (@) A. Fossati, «Ballettino dell'Institutos, 1838, Pp. 49-50; Note perla pianta di Gio. Batt. Noli, Rome 1884, p. 40 (n. 1611 et 1525); R. Lanciani, Forma Urbis Romae, Milan 1893-1901, pl. 45; H. Jordan et

Ch. Hilsen, op. cit, p. 186; G. Lugli, Monumenti antichi di Roma, IV (Supplemento), Rome 1940, p. 163. (3) Milan, Biblioteca Ambrosiana, Taccuino del Bramantino, pl. XXIII. Un fac-similé de ce dessin so trouve dans: G. Mongeri, Le rovine di Roma al principio del secolo XVI, Milan 1875, pl. 23. 43

HEMMING WINDFELD-HANSEN

(Pl. III d) () et dans l'autre d'un mausolée inconnu (PI. III e) (?). Dans tous ces tois monuments disparus, le couloir annulaire avait la forme d'un escalier hélicoidal logé dans l'épaisseur du mur. Mausolée circulaire avec un caveau d’un couloir annulaire.

octogonal

central

entouré

D'après un plan de Francesco da Sangallo, il apparaît qu'il y avait près de St-Sébastien sur la Via Appia un mausolée circulaire (Pl. IV a) @), dans le noyeu duquel se trouvait un caveau octogonal. Celui-ci était creusé de niches de forme alternativement semi-circulaire et rectangulaire et entouré d'un étroit couloir annulaire. L’enceinte du mausolée était également creusée de niches, séparées par des pilastres, Il parait ressortir du plan que l'enceinte formzit un mur circulaire séparé du noyau par le couloir annulaire. Mausolées couloir

circulaires

que

l'on

croyait

autrefois

munis

d'un

annulaire.

Le grand mausolée circulaire appelé Carceri Vecchie qui se trouve sur le côté gauche de la Via Appia entre 8, Maria Capua Vetere, l'antique Capoue, et Caserta, n'avait jusqu'à ces dernières années jamais fait l'objet d'une enquête archéologique. Il n'existait aucun relevé exact de ce monument, dont la chambre sépulcrale était inaccessible. C'est pourquoi les connaissances que l'on avait de la forme intérieure du monument s'appuyaient pour une grande partie sur les plans des architectes italiens de la Renaissance. Un dessin de Giuliano da Sangallo (PI. IV 8) (€) montre les Carceri Vecchie en plan et en reconstitution perspective. Sur le plan, on voit au milieu du mausolée une chambre en croix grecque. Cette chambre est entourée d'un étroit couloir annulzire. Dans les axes secondaires du mausolée, des portes mènent du couloir ἃ de petites pièces rondes placées entre les bras de la croix. Sur un autre plan des Carceri Vecchie dû è Pirro Ligorio (PI. IV e) (€), la disposition du monument est sensiblement la méme que sur le dessin de Sangallo. Dans un ouvrage de De Franciscis et de Pane sur les mausolées romains de la Campanie, paru en 1957, cn trouve pour la pre-

mière fois une description satisfaisante et un relevé exact des Carceri Vecchie (PI. IV d) (*). Il en ressort que le mansolée est reproduit d'une maniére entièrement erronée tant sur le plan de Sangallo que sur celui de Ligorio. Il est vrai qu'il y a bien dans l'édifice une chambre sépulcrale centrale en forme de croix grecque, mais elle est petite et placée dans un noyau cylindrique d'opus caementicium au milieu du monument. A l'origine, un dromos long et étroit, voûté en berceau, menait à la chambre,

(x) Londres, Sir John Soane's Museum, dessins de Montano, vol. IIT, pl. 47; Paris, Bibliothèque Nationale, Cab. des Est., Hb. 22, pl. 62; Gio. Batt. Montano, Libro III, Rome 1684, pl. 37. (2) Gio. Batt. Montano, Libro III, pl. 40. (3) Rome, Bibliothèque Vaticane, Cod. Barb. Lat. no. 4424, f. 39 r. Une reproduction en fac-similé du dessin se trouve dans: Ch. Hülsen, It libro di Giuliano da Sangallo, f. 39 (description, ibid p. xxxvi et p. 56). (4) Rome, Bibliothèque Vaticane, Cod. Barb. Lat.

no. 4424, £. 8 x, Une reproduction en facssimilé se trouve dans: Ch. Hülsen, op. cit, f. 8 (description, ibid., p. 15-16). Il existe une réplique de ce plan ἃ la Biblioteca Comunale, Sienne, Cod. S. IV. 8, f. 16 v. (5) Naples, Biblioteca Nazionale, Cod. XIII. B. 10, Libro XLIX Je numéro du feuillet manque). Une copie de ce dessin se trouve à Bibliothèque Vaticane, Cod. Vat. Lat: no. 3439, f. 33 r. (6) A. De Franciscis et R. Pane, Mausolei romani in Campania, Naples 1957, pp. 36-38, 87-104, figg. 74-84; L. Crema, op. cit., p. 326, figg. 371-373.

44

LES COULOIRS ANNULAIRES

Le relevé montre en outre que la chambre n'est pas entourée d'un couloir annulaire et que les quatre petites pièces rondes reproduites sur les plans de Sangallo et de Ligorio n’existent pas. En revanche, on a trouvé sur le cóté postérieur du mausolée, un couloir long d'environ 20 m., longeant concentriquement le mur extérieur de l'édifice. Le mur d'enceinte du mausolée avait un décor extérieur de colonnes engagées et de niches, construit en opus mixtum composé d'assises de magonnerie réticulée et de briques alternées. Cela indique que les Carceri Vecchie ont été construites au 2° s. ap. J.-C.

Deux gravures de Gio. Batt. Piranesi (Pl. Va et 5) (1) reproduisent un plan et une coupe du grand mausolée circulaire appelé Monte del Grano qui était vraisemblablement situé sur l'antique Via Tusculana et qui se trouve maintenant dans le faubourg de Quadraro, ἃ 3 milles romains de la Porta S. Giovanni. Ce mausolée, en forme de tumulus, est muni d’un long dromos, voûté en berceau, qui mène ἃ l'hypogée placé au centre du monument. Cet hypogée se composait ἃ l'origine de deux chambres rondes ayant approximativement le méme diamétre, placées exactement l'une au-dessus de l'autre, Sur le plan de Piranesi, la chambre inférieure est entourée d'un couloir, bas et étroit, recouvert d'une voüte en berceau, qui est aussi représenté sur la coupe. Il est cependant douteux que

ce couloir ait existé, Sur des gravures du 17) s., l'une dans Raffaele Fabretti, De aquis εἰ aguaductibus veteris Romae (3) et l'autre dans Pietro Santi Bartoli, Gli antichi sepolcri (*) où le Monte del Grano est aussi reproduit en coupe, le couloir annulaire manque. Dans un article d’Ashby et de Lugli sur la villa impériale appelée ad duas Lauros sur la Via Labicana, ce grand mausolée est minutieusement décrit (+). Mais le couloir annulaire montré par Piranesi n'est ni mentionné dans la description, ni reproduit sur le relevé du monument qui se trouve aussi dans cet article (PL Vo). 2) Mausolées carrés à couloir annulaire. Mausolée en forme de tour dont le podium caveau cruciforme central entouré d'un

carré contient un couloir annulaire.

Le mausolée appelé Sepolcro dei Servili près du Tombeau de Romulus sur la Via Appia (Pl. VI a) €) est un édifice en forme de tour qui rappelle les grands mausolées circulaires parce qu'il ἃ un corps cylindrique. Celui-ci est toutefois surélevé sur un haut soubassement ou podium de plan carré, A l'intérieur de ce podium, il existe un hypogée bien conservé se composant d'un caveau

cruciforme

central

entouré

d'un

couloir annulaire

étroit

trés élevé,

dont

les parois

et la

voûte en berceau, recouvertes de stuc, étaient ornées d'un décor peint. Le Sepolcro dei Servili est daté de la fin de l'époque d'Auguste. Ce monument remonte donc à la méme période que la plupart des grands mausolées circulaires à couloir annulaire. (1) Gio. Batt, Piranesi, Le antichità romane, Rome 1756, vol. II, pls. 3r (plan) et 32 (coupe).

cia Romana di Archeologia +, série III, vol. II, 1928, PP. 179-182.

(2) Raffaele Fabretti, De aguis et aquaductibus veteris Romae, Rome 1680, pl. XVI (p. 50). (3) Pietro Santi Bartoli, Gli antichi sepoler, Rome 1699, pl. 90. (4) Th. Ashby et G. Lugli, La villa dei Flavi Cristiani cad duas Lauros e il suburbio imperiale ad oriente di Roma, «Memorie della Accademia Ponti-

(5) L. Canina, op. cit, VI, pl XVII; J. Rispotelli et H. Marucchi, La Via Appia, Rome. 1908, pp. 126-128; G. T. Rivoira, op. cit, pp. 11-12; G. Giovannoni, La tecnica della costruzione presso i ros mani, Rome s, a. p. 86, pl. VII; M. E. Blake, op. cit, pp. 294, 339; R. Fellmann, op. cit, p. 74, fig. 17, 3. 45

HEMMING WINDFELD-HANSEN Mausolée carré d'un couloir

avec une salle annulaire.

circulaire

centrale

entourée

Deux dessins de Sallustio Peruzzi () reproduisent le plan d'un édifice carré dont les

côtés sont ornés de pilastres (Pl. VID). Au milieu de l'un des côtés, une porte donne accès à

l'intérieur de l'édifice. On voit au centre une salle circulaire omée de huit pilastres. Cette salle est entourée d'un couloir concentrique séparé de la piéce centrale par un mur circulaire dans lequel étaient ménagées huit portes étroites placées entre-les pilastres, Les huit pans de inur entre les portes ont l'aspect de piliers cruciformes qui paraissent former ur cercle de piliers autour de la salle centrale, Le couloir annulaire, sûrement voüté er: berceau, était creusé à l'extérieur de grandes niches semi-circulaires placées dans les axes principaux et secondaires. Il n'y avait cependant que sept de ces niches, la huitième, dans laquelle se trouvait la porte, étant de forme

rectangulaire. Chacune de ces grandes niches semi-circulaires était ornée de trois petites niches et

dans la niche d'entrée il y avait également de petites niches dans chacune des parois latérales.

Etant donné que les fenêtres font défaut dans les murs de l'édifice et qu'il n'y a qu'une seule porte d'entrée, il est vraisemblable que la salle circulaire était une chambre sépulcrale et que

le monument était un mausolée, D'après les textes explicatis qui accompagnent les plans de Peruzzi, ce mausolée aurait été situé à proximité des aqueducs de la Via Latina dans les envi-

rons de Rome. L'architecte français Jacques Androuet Du Cerceau a probablement utilisé les deux plans du mausolée de la Via Latina de Peruzai pour effectuer une reconstitution en plan, élévation et coupe du méme monument (Pl. VI c) (?). Sur le plan de Du Cerceau, la salle centrale est de plus grandes dimensions que sur les dessins de Peruzzi. Tant l'élévation que la coupe paraissent être des reconstitutions fantaisistes sans valeur archéologique. Gio. Batt. Montano a également reconstitué ce mausolée sur plusieurs de ses dessins, en plan et en perspective avec une coupe partielle, et une reproduction de cette reconstitution se trouve aussi parmi les gravures de Montano (PL VI d) Q). | 3) Rolondes funéraires avec un soubassement circulaire ἃ couloir annulaire. Quelques mausolées romains de l'époque impériale ont la forme de rotondes à coupoles,

élevées sur un soubassement circulaire formant podium. Dans quelques-unes de ces rotondes, le

soubassement renferme un hypogée ayant le caractère d'une crypte, entouré d'un couloir annuaire. Comme dans la plupart des mausolées circulaires que nous avons examinés, il se trouve dans ces soubassements un noyau central de forme ronde entouré d'un mur d'enceinte qui est séparé du noyau par un étroit couloir annulaire. Sur cette enceinte reposait à l'origine soit le mur extérieur de la rotonde, lorsque celle-ci était de forme simple, soit les colonnes

tique extérieur concentrique, lorsque la rotonde avait une forme périptère. (2) Florence, Musée des Offices, no. 665 Av. Uno reproduction en fac-similé se trouve dans: A. Bartoli, Monumenti antichi 4 Roma nei disegni degli Uffisi di Firenze, Rome-Florence 1914, pl. COCLXXXVIII, fig. 681 (description, ibid.; p. 120); Florence, Musée des Offices, no. 689 A v. Un fac-similé se trouve dans: A. Bartoli, op. cit., pl. CCCLXXV, fig. 655.

d'un por-

(a) Jacques Androuet Du Cerceau, Les Grands Temples, deux gravures désignées « Les Petits Temples, deux gravures désignées « B». (3) Londres, Sir John Soane's Museum, dessins de Montano, vol. II, pls. 7et 11; Paris, Bibliothèque Nationale, Cab. des Est, Hb. 22, pl. 33; Gio. Batt. Montano, Libro IL, Rome 1684, pl. 26.

46

LES COULOIRS ANNULAIRES Rotonde funéraire (Ὁ) avec un soubassement circulaire muni d'un caveau cruciforme central entouré d'un couloir annulaire.

Au Nord de la Via Tiburtina, près du chemin de traverse entre Settecamini et Guidonia se trouvent les restes d'un grand édifice romain à plan circulaire appelé Tor Inviolata ou Torraccio dell'Inviolata (Pl. VII a) (). De l'édifice primitif, il ne reste qu'une partie du soubassement rond dont la face extérieure comportait sept niches semi-circulaires. Sur un des cótés de l'édifice, il y avait un élargissement rectangulaire dont seule la fondation est visible. Dans le noyau du soubassement circulaire, on trouve un caveau central de plan cruciforme, bien conservé. Ce caveau est relié à deux pièces dotées de deux absides, placées dans l'axe principal de l'édifice. Un couloir annulaire, voûté en berceau, entoure le caveau central, reliant en méme temps les pièces bi-apsidales, Il s'agit sans aucun doute d'une crypte funéraire dont la pièce centrale.a été la chambre sépulcrale. Au-dessus du soubassement partiellement conservé, s'élevait probablement un temple funéraire en forme de rotonde. La facade de ce temple était formée par un portique à colonnes dont le podium reposait sur l'élargissement rectangulaire du soubassement. Les murs du mausolée de Tor Inviolata sont construits avec un noyau de maçonnerie en blocage, revêtu d'un mélange d'opus reticulatum et de maçonnerie en briques. Le monument est daté par Giovannoni du 1° ou du 2° s. ap. J.C. Rotonde funéraire avec un soubassement circulaire muni de deux caveaux allongés entourés d'un couloir annulaire.

Le mausolée appelé Sepolcro di Gallieno prés du neuvième milliaire de la Via Appia (*) est une rotonde funéraire à soubassement circulaire, mieux conservée que la Tor Inviolata.

Cette

rotonde, d'environ 13 m. de diamètre, a depuis longtemps perdu sa coupole et la moitié du mur circulaire qui la supportait a aussi disparu. Dans le soubassement circulaire qui est maintenant partiellement. enterré sous des masses de terre et de pierres, on trouve deux caveaux allongés parallèles, voîtés en berceau, qui sont coupés par un couloir transversal également voüté en berceau.

Sur une gravure.de Pietro Santi Bartoli (Pl. VILc) (5), on voit ce monument reproduit en

élévation et en coupe partielle. Bartoli l'appelle Temple de Domitien-Hercule (Templum Domitiani

Aug. sub Herculis vuliu), en raison de l'interprétation erronée d'une estampille de brique trouvée dans la maconnerie de la rotonde. Sur la gravure de Bartoli la rotonde est entourée d'un portique extérieur à colonnes se composant, semble-t-il,

de 18 colonnes corinthiennes. Ces colonnes qui sont

de grandes dimensions sont placées sur un podium élevé qui entoure le soubassement. De la coupe, il ressort qu'il y avait dans ce podium deux couloirs concentriques voîtés en berceau, placés autour des caveaux. Les colonnes reposaient sur le mur circulaire séparant les deux couloirs. Il est

(1) Th. Ashby, The Classical Topography of the Roman Campagna, ΤΙ, « Papers of the British School at Rome», III, 1906, p. 105; G. Giovannoni, op. cit, p. 55, pl. XII, 4; L. Crema, op. cit, p. 325, fig. 477. a (2) A. Nibby, op. cit, ΠῚ, pp. 555-556; L. Canina, op. cit., V, pp. 40-41, VI, pl. XLVIII; G. Tomassetti, La Campagna romana, I, Rome 1910,

pp. 111-112 (le monument est ici confondu avec le Torraccio del Palombaro aussi appelé la Barretta del Prote); G. T. Rivoira, op. cit., p. 229; U. Leoni et G. Staderini, Sull' 4ppia antica, Rome 1930, p. 237: G. Lugli, La tecnica edilizia romana, pp. 616, 674, 692, pl. CLXXXIV, 1. (3) Domenico de” Rubeis, Romanae magnitudinis monumenta. etc., Rome 1699, pl. 50. 47

HEMMING WINDFELD-HANSEN

cependant probable que le couloir extérieur est une addition arbitraire de Bartoli et que le dia mètre du podium n'a pas beaucoup dépassé celui du portique à colonnes, Un dessin de Sallustio Peruzzi (PI. VII δ) (1) qui représente la même rotonde en plan et en élévation montre que le couloir annulaire intérieur a existé. Les proportions de l'élévation sont’ erronées, mais le dessin présente quand méme un grand intérét, car il montre qu'il y avait sur la face extérieure de la rotonde des restes d'une voûte en berceau annulaire placée au niveau du sol de la salle circulaire. A l'endroit où cette voûte était placée, il y a maintenant une large et profonde entaille dans le mur qui prouve que les briques dont se composait la voîite ont été enlevées. La voûte en berceau recouvrait probablement le couloir annulaire intérieur qu'on voit sur la coupe de Bartoli. 2 La rotonde appelée Sepolcro di Gallieno doit avoir été un mausolée: Cela ressort du plan de l'étage inférieur, qui se présente comme une crypte dont les deux pièces allongées devaient être les caveaux sépulcraux. A l'intérieur et à l'extérieur, les murs de la rotonde ont un revétement de très fine maçonnerie de briques qui semble dater de 3* s. ap. J.-C. Rotondes funéraires avec un soubassement circulaire muni de deux couloirs qui se croisent à angle droit et sont entourés dun

couloir

annulaire.

Près de la Via Portuensis, à l'Est du Portus Traiani à Porto (Fiumicino), on voit les ruines d’une

grande rotonde qui, depuis a Renaissance, a porté le nom de Tempio di Portuno (Pl. VIII a) (:). Cette rotonde se dresse sur un soubassement circulaire élevé qui forme un podium ayant un

diamètre beaucoup plus grand que celui de la rotonde elle-même. A l'intérieur du soubassement,

on trouve un noyau central de forme ronde, entouré d'une enceinte circulaire percée de portes,

suivant l'axe transversal. Entre le noyau et l'enceinte se trouve un coulcir voüté en berceau, Le noyau est divisé en quatre quadrants par deux couloirs également voütés en berceau qui se croisent à angle droit selon les axes principaux de l'édifice. Dans chacun de ces couloirs quatre niches rectangulaires sont ménagées vis à vis les unes des autres dans les parois. Environ un quart du mur de la rotonde est conservé. Ce mur comportait primitivement dans

sa face intérieure quatre niches rectangulaires aux axes principaux et quazre niches semi-circulaires aux axes secondaires. Entre ces niches, il y avait des colonnes supportant un entablement. La salle circulaire, de 14 m. 50 de diamètre, était recouverte d'une coupole dont il ne reste maintenant qu'un fragment au-dessus de la partie de mur conservée. A l'origine, la rotonde était entourée d'un portique extérieur dont les colonnes ont disparu depuis longtemps. Mais sur le podium, on voit encore les marques de quelques-unes de leurs bases. Le portique était probablement recouvert de petites voûtes d'arétes de forme trapézoidale, En

se basant sur la magonnerie de briques, Lugli date la rotonde de Porto de l'époque des Sévére (193-235 ap. J.-C).

(1) Florence, Musée des Offices, no. 664 A v. Une reproduction en fac-similé se trouve dans: A. Barto, op. cit, pl. CCCLXXXVII, fig. 679 (description, ibid, p. 120). (2) À. Nibby, op. cit, IL, pp. 648-650; L. Canina, op.cit., VI, pl. CLXXXVI; W. Altmann, Die italischen Rundbasiten, Belin 1906, pp. 69-71; G. T. Rivolra,

ὁ, cit, pp. 231-232, fig. 223; G. Lugli et G, Filibeck, I porto di Roma imperiale e l'agro portuense, Rome 1935, pp. 93 ss.. fig. 61; G. Lugli, Edifici rotondi, « Studies presentedto David M. Robinson », Saint Louis 1953, II, p. 1212; id. La fecwica edilizia romana, pp. 668, 674, 676, 686, 687, 691; L. Crema, op. cit, p. 563, fig. 745.

48

LES COULOIRS ANNULAIRES

Autrefois, on considérait ce monument comme un temple consacré à Portunus, mais il y a tout lieu de croire que les huit niches rectangulaires qui se trouvent dans les couloirs du soubassement étaient destinées à des sarcophages. S'il y a une crypte sépulcrale dans le soubassement, la rotonde elle-méme était forcément un temple funéraire. Au cours des recherches effectuées dans les années 1860 près de Via Portuensis, à environ 5 milles romains de Rome, dans la région où les frères Arvales avaient leur sanctuaire, on a trouvé dans les caves sous le casino de la Vigna Ceccarelli un soubassement circulaire, d'environ 20 m. de diamètre, qui appartenait è une rotonde maintenant disparue (PI. VII) (1). A l'intérieur de ce soubassement, on peut voir - comme dans le soubassement du Tempio di Portuno ἃ Porto = un noyau central rond entouré d'une enceinte. Entre le noyau et le mur il y a un couloir annulaire, muni sur l'extérieur de sept niches rectangulaires placées aux axes principaux et secondaires de l'édifice, Une huitième niche est remplacée par la porte d'entrée, Sur le côté intérieur du couloir, il n'y a par contre que quatre niches placées dans les axes secondaires. Le noyau est divisé en quatre quadrants par deux couloirs qui se croisent à angle droit au milieu de l'édifice, De la rotonde elle-même, il ne reste que des fragments de l’ornementation architectonique, mais sur un dessin de Pirro Ligorio () l'étage supérieur, maintenant disparu, est reproduit en plan. Sur celui-ci, le mur de la salle circulaire est doté à l'intérieur de 15 niches de forme alternativement semi-circulaire et rectangulaire, tandis que la 16° niche est remplacée par la porte. L'intérieur de la rotonde est divisé en une partie centrale circulaire supporte par τό colonnes doubles et un déambulatoire annulaire, ΤΙ n'est pas possible de déterminer la fonction de cette rotonde dans le sanctuaire des frères Arvales, mais comme celui-ci ne peut étre strictement délimité, il est possible que notre édifice se trouve situé en dehors de la zone sacrée proprement dite et soit simplement un mausolée ana-

logue, par son plan, au Tempio di Portuno de Porto. En se basant sur les détails architectoni ques, Lanciani pense que la rotonde a été édifiée soit à la fin du 2° s., soit au début du 3° s. ap. J-C. 4) Hypogée avec un caveau circulaire entouré d'un couloir annulaire.

IL n'est pas conservé un seul exemple d’hypogée où le caveau soit entouré d'un couloir annulaire, mais il ressort d’une gravure de Pietro Santi Bartoli (PI. VIII) () que cette forme de tombeau a aussi été utilisée dans l'architecture funéraire romaine. La gravure reproduit en plan et en coupe un hypogée à caveau circulaire qui se trouvait au pied de l'Aventin, à Rome. Le texte indique que le tombeau était placé à une profondeur exceptionnelle au-dessous de la surface du sol. Un escalier étroit et très raide menait à un petit

vestibule,

d'où une

porte

don-

nait accès au caveau, recouvert d'une coupole hémisphérique. A la hauteur de la naissance de la voûte étaient placées trois petites niches semi-circulaires destinées à des urnes cinéraires. Le caveau était entouré d'un couloir annulaire étroit et bas, couvert d'un plafond. Ce couloir ne faisait pas tout le tour du caveau, mais aboutissait à une petite pièce carrée dont le sol était creusé d'un puits, Un conduit de lumière vertical montait d'une ouverture dans le sommet de la (1) G. Henzen, Scavi nel Bosco Sacro dei Fratelli — (s) Turin, Archivio di Stato, Cod. a. IL. s, vol. Arvali ete., Rome 1868, passim; R. Lanciani, Appen- V.X.Z., Libro XX, f. 65 v. dice sull'edificio rotondo della Vigna Ceccarelli, ibid., (3) Pietro Santi Bartoli, Gli antichi sepoleri, Rome pp. 105-106, pls. IV-V; W. Altmann, op. cit., pp. 63- 1699, pl. 45. Dans l'édition de 1727, cette gravure 69; R. Lanciani, Storia degli scavi di Roma, Rome est reproduite sur la pl. so. Voir aussi: H. Jordan 1902-1912, vol. LII, pp. 170-172, IV, p. 57. et Ch, Hilsen, op. cit, 1, 3, p. 186, n. 13. 49

HEMMING WINDFELD-HANSEN

coupole à la surface du sol. Cet hypogée devait se trouver à l'extérieur du pomérium et était probablement situé au pied du versant Sud ou Sud-Ouest de l'Aventin, car en ces endroits de Servius suit justement le haut de la colline."

IL - Les

MAUSOLÉES

NUMIDES

À COULOIR

le mur

ANNULAIRE

Le seul pays en dehors de l'Italie dans lequel il existe des mansolées anticues à couloir annulaire est l'Algérie dont la partie orientale appartenait dans l'Antiquité à la Numidie, et la partie occidentale ἃ la Maurétanie, Ces pays faisaient partie à l'origine de la zone d'influence libyco-numide qui a durant toute l'Antiquité conservé beaucoup de son originalité malgré les fortes influences de la civilisation punico-hellénistique et plus tard de la civilisation romaine (1) Parmi les monuments funéraires antiques de l'Algérie, trois seulement sont munis de couloirs annulaires, Ces mausolées ne sont pas du même type et datent de périodes différentes. Pourtant

ces trois monuments ont des traits communs qui témoignent qu'ils ont été construits par des

Berbères numides, influencés sans doute par les civilisations importées, mais aussi héritiers d'une tradition d'architecture funéraire autochtone. 1) Tombeau à tumulus avec un caveau central entouré d'un couloir annulaire.

Le plus simple et aussi le plus ancien de ces trois mausolées aumides couloir annulaire est un tombeau à tumulus de pierres, qui se trouve dans le voisinage du mausolée circulaire monumental appelé le Médracen, au Nord-Est de Batna dans le Corstantinois (Pl. IX a) (). Dans ce tombeau à tumulus, d'environ 25 m. de diamètre, un couloir annulaire entoure un caveau rectangulaire central aaquel mène un court dromos. Le caveau, le couloir δὲ le dromos étaient construits en pierres de taille et étaient couverts d'un toit en dos é’4ne formé de plaques de pierre placées obliquement. Le tumulus lui-même était un amas de blocs de pierre irréguliers. Ce tombeau de caractere primitif fait partie de la méme zone funéraire que le Médracen et date de la même époque que ce grand mausolée que son ornementation architecturale punico-kellénistique permet de dater du 3° ou peut-être du 2° s. av. J.-C. 6). 2) Mausolée circulaire avec une galerie annulaire menant ἃ un hypogée central. Le grand mausolée circulaire appelé Tombeau de la Chrétienne, d'environ 6r m. de diamètre, près de Tipasa à l'Ouest d'Alger (PI. IX c) (ἢ) rappelle beaucoup dans sa forme extérieure le Médracen près de Batna. Les deux mausolées ont un corps cylindrique décoré de 60 colonnes (1) S. Gsell, Histoire ancienne de l'Afrique du Nord, VI, pp. 261-265; Ὁ. Welter, Zwei vorrômische Nord, 8 vol, Paris 1913-1929, voir surtout: vols. Grabbauten im Nondofrika, « Romische Mitteilungen », V et VI; G. Charles-Picard, La civilisation de l'Afri- XLII, 1927, pp. 84 ss. que romaine, Paris 1959, 9p. 324 ss. (4) S. Gsell, Les monuments antiques de V Algérie, (2) Brunon, Mémoire sur les fouilles exdoutées aw I, pp. 69-74, fig. 21; id., Histoire ancienne de l'Afrique Madras'en, «Recueil. dos notices et mémoires de du Nord, VI, pp. 265-273; id., Promenades archéolola Société archéologique de la province de Constan- — giques aux environs d'Alger, Paris, 1926, pp. 143-160; tino», XVI, 1873-1874, PP. 325, 336-338; S. Gsell, G. Welter, op. ct, pp. 84-116; M. Christofie, Le op. cit, VI, p. 199, n. 2. Tombeau de la Chrétienne, Paris 1951, passim; P. (3) Branon, op. cit, pp. 303-350; S. Gsell Les Romanelli, La Tomba della Cristianae il suo mistero, monuments antiques de l'Algérie, Paris 1901, I, pp. «Archeologia Classica», IV, fasc. 2, 1952, pp. 27465-69, fig. 20; id, Histoirs ancienne de l'Afrique dw 83; G. Charles-Picard, op. cit., p. 324. 50

LES COULOIRS ANNULAIRES

engagées à mi-épaisseur et de quatre fausses portes, placées aux axes principaux qui correspondent exactement aux quatre points cardinaux. Ce corps cylindrique porte dans les deux cas un cône tronqué formé de gradins concentriques et aussi bien dans le Tombeau de la Chrétienne que dans le Médracen, le matériau utilisé est la pierre de taille, Toutefois, le corps cylindrique du Médracen est plus bas, La structure intérieure du Tombeau de la Chrétienne est aussi entièrement différente de celle du Médracen ot il n'y a qu'un dromos qui méne à un caveau central rectangulaire, L'entrée de l'hypogée du Tombeau de la Chrétienne se trouve sous la fausse porte de l'Est. Cette entrée, une porte basse, donne accès à un couloir, orienté Est-Ouest, qui mène à un caveau rectangulaire de méme orientation. Dans la paroi Nord du caveau se trouve une porte dont le linteau est décoré d'un relief qui représente un lion et une lionne. Cette porte donne sur un autre couloir et au bout de celui-ci, un escalier de sept marches conduit à une galerie annulaire qui est recouverte d'une voûte en berceau. La galerie, qui a 140 m. de longueur, entoure l'hypogée placé au centre de l'édifice. Sur une distance d'environ 120 m., la galerie se maintient à égale distance de la périphérie, mais ensuite elle s'infléchit subitement à gauche, c'est-à-dire vers l'hypogée. Celui-ci se compose d'un petit vestibule transversal et d'un caveau rectangulaire, tous deux voûtés en berceau. Dans trois des parois du caveau, il y a de petites niches rectangulaires placées au milieu. Il est probable qu'il s'agit Ιὰ du caveau funéraire proprement dit dans lequel étaient placées les urnes cinéraires. Les détails architectoniques et les proportions prouvent que le Tombeau de la Chrétienne doit étre moins ancien que le Médracen où la tradition de l'architecture punico-hellénistique est plus marquée (1). Le Tombeau de la Chrétienne a été construit par des artistes et des artisans

numides pour lesquels le langage de l'architecture punico-hellénistique n'était pas très familier, mais qui étaient influencés par les mêmes courants artistiques que ceux qui ont empreint l'architecture dans le Sud de l'Italie au 3° et 2° s. av. J.-C. (1). 3) Mausolée octogonal avec un caveau circulaire entouré d'un couloir annulaire.

Le troisième et dernier monument funéraire numide avec couloir annulaire était un petit mausolée, dont il existait encore des ruines à Blad Guitoun près de Ménerville dans la Kabylie

occidentale à la fin du siècle dernier (PI. IX 8) (). L'édifice, qui était octogonal, se dressait sur une plate-forme également octogonale. Du côté Est, cette plate-forme s'agrandissait d'une partie rectangulaire. Chacun des côtés de l'octogone était orné de quatre colonnes ioniques engagées ἃ mi-épaisseur, dont deux placées aux angles. La partie supérieure du mausolée manquait, mais les fragments d'architecture retrouvés permettaient d'affirmer qu’au-dessus de l'étage inférieur il y avait un étage supérieur également décoré de colonnes engagées.

ΤΙ n'y avait plus aucune trace de la couverture de l'édifice, mais Gsell pense qu'elle avait la

forme d'une pyramide octogonale à gradins de pierres de taille. Au milieu de chacun des cótés

de l'octogone était taillée une fausse porte avec un encadrement richement décoré. La véritable porte d'entrée du monument était à moitié souterraine et se trouvait sur la partie antérieure de

l'élargissement rectangulaire de la plate-forme. La porte donnait accès à un couloir qui conduisait

(1) G. Welter, op. cit, pp. 97-98 ap. J.-C. a été réfutéo par G. Charles-Picard, op. (2) S. Gsell, Promenades archéologiques, pp. 156- cit, p. 324, n. 44160; G. Welter, op. cit, p. 108 se; G. Charles(8) S. Gsell dans ¢ Comptes rendus de l'Académie Picard, op. cit, p. 324. La théorie de Romanelli des Inscriptions et Belles Lettres», 1898, pp. 481(loc. cit) selon laquelle le Tombeau de la Chré- 409; id., Les monuments antiques de l'Algérie, TI, pp. tienne serait un monument du s* où du δ᾽ siècle 412417, figg. 170-171. 5r

HEMMING WINDFELD-HANSEN

à l'hypogée. Celui-ci se composait d'un caveau circulaire central, de 3 m. 75 de diamètre, couvert d'une coupole et entouré d'un couloir octogonal vers l'intérieur et circulaire vers l'extérieur. Ce couloir ne faisait cependant pas le tour du caveau central, il était arrété du cóté droit par une maçonnerie massive, peut-être ajoutée plus tard. Le cavean était décoré de huit pilastres munis de chapiteaux-impostes portant des arcs surbaissés, Le noyau du mausolée était de maçonnerie en blocage. Non seulement la chambre sépulcrale, mais aussi le mur qui entourait le couloir, étaient revétus à l'exzérieur et à l'intérieur de pierres de taille parmi lesquelles on trouvait aussi des moellons encastrés. Les pierres de taille étaient en calcaire poreux et la maçonnerie manquait de régularité. L'ornementation du mausolée et le caractère de la maçonnerie semblent indiquer que ce monument ne peut avoir été construit avant le 4° s. ap. J.-C. et Gsell le date d'ailleurs du début du 5° s. Il s'agit probablement du tombeau de quelque souverain local. II. - LES MAUSOLÉES CIRCULAIRES AVEC UNE ARMATURE RIEURE ET L’ORIGINE DES COULOIRS ANNULAIRES LES ÉDIFICES FUNÉRAIRES ANTIQUES.

INTÉDANS

x) Tombeau grec à tumulus en Asie Mineure divisé par des murs intérieurs.

Dans les tombeaux à tumulus préhistoriques de la région méditerranéenne, les amas de terre et de pierres qui forment les soubassements circulaires et les tumuli coniques sont généralement indivisés. Il en est de méme de la plupart des tombeaux à tumulus lydiens et phrygiens en Asie Mineure qui datent des 8°, 7° et 6* s. av. J.-C. (). La première fois qu'on trouve une division de la masse intéieure d'un mausolée circulaire, c'est dans le Tombeau de Tantale, la plus grande des tombes à tumulus dont il y ait des restes sur le Jamanlar Dagh, appelé aussi Mont Sipyle, près de Smyrne (?). A l'intérieur du soubassement citculaire de ce grand édifice se trouvait un cavean rectangulaire recouvert d'une voûte en encorbellement. Ce caveau était entouré d'une série de murs rayonnants qui transmettaient

la poussée du noyau central entourant le caveau à l'enceinte du soubassement. Le cavean était aussi entouré d'un mur concentrique intérieur, Tous ces murs divisaient la masse intérieure de moellons du mausolée en espaces en forme de secteurs ou de trapèzes dans lesquels les tas de pierres étaient maintenus en place.

Le Tombeau de Tantale, de 29 m. 60 de diamètre, est maintenant — commela plupart des autres tombes à tumulus de la nécropole du Jamanlar Dagh-- si endommagé qu'il n'est plus possible de se faire une 1465 certaine de sa structure intérieure. Mais la construction originale du mausolée est connue par des descriptions et des plans du siècle dernier, élaborés alors que le monument n'avait pas encore été bouleversé par des fouilles ()). Les tombeaux circulaires du Jamanlar Dagh étaient considérés autrefois comme d'origine lydienne, mais il ressort des recherches plus récentes qu'ils ont probablement été construits par des colons grecs au 7° ou δ᾽ s, av. TC. (9. (1) G. Perrot et C. Chipiez, Histoire de l'art dams F. Schachermeyr, op. cit, p. 382; F. et H. Miltner, l'Antiquité, V. Paris 1890, pp. 82, 265-285; F. Scha- Bericht über cine Voruntersuckung in All-Smyrna, chermeyr, Materialien zur Geschichte der dgdischen +Jabreshefte des Gsterreichischen archiologischen Wanderung, « Athenische Mitteilungen+, XLI, ro6, Instituts», XXVII, 1931, « Befblatt », cols. 149-158. PP. 376-385. (3) G. Perrot et C. Chipiez, op. cit., V, fig. 15. (2) G. Perrot et C. Chipiez, op. cit, V, pp. 48-53: — (4) F. et H. Mltner, op. cit, p. 158, n. 42. 52

LES COULOIRS ANNULATRES

2) Tombeaux numides à tumulus divisés par des murs intérieurs. Dans le cercle d'influence de la civilisation libyco-numide en Afrique du Nord, on trouve un grand nombre de monuments funéraires. Le plus souvent, ce sont des constructions mégalithiques qui ne sont pas d'origine préhistorique mais datent soit des derniers siècles avant notre ère, soit de l'Empire romain, Beaucoup des tombeaux circulaires de pierres que l'on trouve dans le Sud Tunisien, dans l'intérieur de l'Algérie ou au Sahara ne sont pas plus anciens que l'époque précédant la conquéte islamique (1). Ces monuments funéraires ont ou bien la forme d'un cercle de pierres entourant un dolmen maintenant dégagé ou bien celle d'un tumulus de pierres placé autour d'un caveau. Ils peuvent aussi apparaître sous la forme d'une tour cylindrique basse, appelée bazina, ayant au centre un creux en forme de puits.

Ces formes primitives de tombes présentent un très grand intérét, car elles montrent qu'une tradition funéraire préhistorique s'est maintenue chez les Numides de l'Afrique ‘du Nord, aussi bien pendant la période hellénistique que sous la domination romaine. Par conséquent plusieurs des tombeaux des princes numides ne sont que la transformation monumentale des tombes primitives circulaires qui ont été revétues à l'extérieur d'un décor architectonique d'origine punico-

hellénistique (). Comme dans la grande tombe grecque à tumulus appelée le Tombeau de Tantale, prés de Smyrne, on trouve souvent dans les soubassements des tombeaux numides à tumulus des murs rayonnants qui divisent en secteur la masse intérieure de pierres brutes. Dans d'autres cas, un caveau en forme de cercueil ou un dolmen, placé au centre du tumulus de pierres, est entouré d'un mur intérieur placé concentriquement à l'enceinte. Les murs circulaires peuvent

étre les restes de l'enceinte d'une tombe à tumulus plus an-

cienne qui, à une époque ultérieure, a été agrandie, ou la fonction de ces murs peut avoir été

de maintenir en place les masses de pierres du tumulus et du soubassement circulaire (). Les nombreuses tomibes numides à tumulus qui n'ont pas de murs circulaires intérieurs montrent cependant que ce genre de construction n'est pas indispensable pour édifier un tumulus de pierres dans ces régions.

3) Mausolées et édifices circulaires romains avec murs intérieurs de contrefort et murs annulaires concentriques. ' Structure intérieure des mausolées circulaires romains de forme simple. La forme la plus simple du mausolée romain circulaire a un tumulus conique élevé sur un soubassement ou socle rond aussi appelé crépis. Ce soubassement est muni à l'extérieur d'une enceinte circulaire revétue de pierres de taille, relativement basse par rapport au diamètre du mausolée. C'est principalement ce type de tombeau romain qui poursuit la tradition des tombes

étrusques à tumulus (1).

(1) S. Gsell, Histoire ancienne de l'Afrique du Nord, VI, p. 181-189; M. Reygasso, Monuments funéraires préislamiques de l'Afrique du Nord, Paris 1950, passim; G. Charles-Picard, op. cit, p. 153, ΕΗ (Ὁ G et C. Charles-Picard, La vie quotidien-

ne à Carthage au temps d'Hamwibal, Paris 1958, p. 54. (3) S. Gsell, Histoire ancienne de l'Afrique du Nord, VI, pp. 183-184. (4) G. T. Rivoira, op. cit, p. το; R. Fellmann, op. cit, pp. 90-93. 53

HEMMING WINDFELD-HANSEN

Certains mausolées circulaires de la période hellénistique peuvent avcir servi de modèles

aux architectes romains (). Peu de tombeaux romains en forme de tumulcs sont plus anciens

que l'époque d' Auguste et aucun ne semble avoir été construit avant l'époque de Sylla (*). Dans de nombreux tombeaux circulaires à tumulus des nécropoles des environs de Rome, non seulement le tumulus lui-même, mais aussi la partie intérieure du soubassement sont formés d'une masse homogène de terre, contenue par Je mur d'enceinte (ἢ), Dans quelques monuments de ce type, une petite cavité au centre du soubassement est destinée à -ecevoir des urnes cinéraires (9) Dans d'autres cas, il y a dans ces mausolées circulaires une chambre sépulcrale voütée placée au centre du soubassement, au niveau de la surface du sol. Ce caveau et le dromos qui y con-

duit, sont soit construits en pierres de taille, soit revêtus d'ofus reticulatum ou de briques. Le caveau lui-même est bétonné dans un bloc massif d'ofus caementicium qui se dresse le plus souvent comme une tour à l'intérieur du tumulus et qui supportait une sculpture couronnant le monument; au contraire, l'espace entre le noyau central et l'enceinte est rempli de terre et de magonnerie de pierres sèches (). Vers la fin de la période républicaine, il apparait dans les nécropoles des environs de Rome une nouvelle forme de mausolée circulaire avec un cylindre élevé qui donne au monument le caractère d'une tour et sur lequel repose un tumulus de forme conique relativement bas (). Pour renforcer encore l'aspect monumental du mausolée et son caractére de tour, le cylindre est souvent élevé sur un socle ou un podium de plan carré (7). Ces mausolées romains en forme de tour ne continuent pas une tradition de l'architecture funéraire étrusque. Il faut plutót chercher leurs modèles parmi les mausolées circulaires des empires hellénistiques de la Méditerranée orientale. Dans les nécropoles de Rome et de nombreuses cités italiennes, on trouve le long des routes romaines les ruines d'un grand nombre de ces mausolées circulaires qui remontent essentiellement au début de l'Empire. Dans ces monuments, l'intérieur du cylindre — comme le podium — est en règle générale construit en magonnerie bétonnée: c'est une masse compacte d'opus caementicium, qui est entourée d'un mur circulaire revêtu de pierres de taille qui forme l'enceinte du cylindre. Dans ce cas-1à, le tumulus qui repose sur le cylindre a le plus souvent la forme d'un monticule conique de terre (+). A l'intérieur de ces mausolées, on trouve habituellement un caveau central et un dromos qui y conduit (). Tombeaux circulaires me armature.

avec un systéme de murs

intérieurs

com-

Dans certains tombeaux circulaires à tumulus, la masse de terre à l'intérieur du soubassement circulaire n'est pas cohérente, mais elle est divisée par un système de murs qui rayonnent e plus souvent à partir du centre ou qui forment une rangée d'arcs concaves couchés, placés (1) F. Matz, «Gnomon», XVII, 1941, p. 215; R.

Fellmann, loc. cit. (a) R. Fellmann, op. cit, p. 93; L. Crema, op. cit, pp. 131, 243. (3) B. Gotze, op. cit, p. 9. sur la Via Appia. (4) Par ex.: la Tombe des Curiates Voir: L. Canina, op. cit., V, p. 33, VI, pl XXXIII; G. Pinza, Ricerche intorno ai monumenti ritrovati al V miglio dell'Appia, « Jahresheîte des dsterreichischen archdologischen Instituts», X, 1907, p. 205. (5) Par ex: le Mausolée de Lucilius Petus sur la Via Salaria et lo Torrione sur la Via Praenestina.

Voir: C. Pietrangeli, II somursento dei Lucii, Rome

1941 (a L'Urbe v, V, no. x1), passim; id., ΠῚ Torrione » della Via Prenestina, Rome 1941 («L'Urber, VI, no. 5), passim. (6) L. Crema, op. cit, pp. 244-246. (7) B. Gótze, op. cit, pp. 9-12; L. Crema, op. cit, pp. 248-251. (5) B. Gütze, op. cit, p. 9. (9) Par ex: le Mausolée de Caecilia Metella sur la Via Appia. Voir: J. Rispotelli et H. Marucchi, op. cit, pp. 139-148; U. Leoni et G. Staderini, op. cit., p. 106-114; R. Fellmann, op. cit., pp. 66-68, fig. 16, r.

54

LES COULOIRS ANNULAIRES

derrière l'enceinte du soubassement. Ces murs intérieurs divisent les masses de terre du tumulus et les maintiennent en place à l'intérieur du tombeau; de plus les murs rayonnants reportent le poids du noyau central du tumulus sur l'enceinte et les murs concaves allègent la pression des masses de terre contre cette enceinte.

Dans les tombes étrusques à tumulus, qui sont les modèles les plus proches des tombeaux circulaires romains, il n'y a pas une telle armature dans le monticule. Il n'existe pas non plus une forme quelconque de division des masses de terre ou de pierres dans le soubassement des mausolées circulaires hellénistiques et il en est de méme des rares exemples de tombeaux à tumulus de la période grecque classique. L’armature au moyen de murs intérieurs dans les mausolées circulaires romains est un des éléments de la technique romaine de la construction. Elle est basée sur le méme principe que celui utilisé par les Romains pour la construction de fondations et de terrasses, afin d'empêcher l'éboulement de grandes masses de terre. Ce principe de construction qui est décrit par Vitruve (1) consiste à diviser la masse de terre de part et d'autre par des murs qui maintiennent cette masse en place et qui transmettent simultanément la pesée

sur les piliers de contrefort du côté extérieur du mur de la fondation ou de la terrasse (). La première fois que l'on trouve un rudiment d'armature intérieure dans un mausolée romain, c'est dans l’une des deux tombes à tumulus dites Tombes des Horaces prés du cinquième milliaire

de la Via Appia (Pl. X a) (*). Ce monument qui est la seule tombe romaine à tumulus qui puisse étre sürement datée de la fin de la période républicaine (80-44 av. J.-C.) a un soubassement dont le mur circulaire est revêtu de blocs de péperin. Derriére.ce mur, on trouve 12 pans de murs concaves qui forment une rangée circulaire d’ares couchés. Dans l'autre de ces deux tombes (), on trouve par contre à l'intérieur du tumulus une armature formée de six murs rayonnant

du

centre du monticule et divisant la masse de terre en six secteurs, A proximité de Todi, on a déblayé un tombeau à tumulus avec le méme système de murs intérieurs (Pl. X δ) (9) et à Marcigliana prés de la Via Salaria au Nord de Rome, on trouve un mausolée circulaire (PI. X c) (*) qui a un système intérieur de murs de contrefort en forme d'étoile.

Il y a ici, en plus des six murs radiaires, whe rangée circulaire de six pans de murs.en arc concave placés derrière l'enceinte du socle. Dans aucun des tombeaux de ce genre, il n'y a de caveau central, mais dans le tombeau de Marcigliana et dans un mausolée similaire déblayé à West Mersea dans le Sud de l'Angleterre (PL. X d) (7) les murs rayonnent à partir d'une petite cavité inaccessible au centre du monticule où étaient vraisemblablement placées les urnes cinéraires. Il a cependant aussi existé des mausolées circulaires romains en Italie dans lesquels le mur entourant un caveau central était relié à l'enceinte circulaire de l'édifice par une armature de murs, Entre ces murs, il y avait des chambres ou des cavités inaccessibles qui formaient une rangée autour de la pièce centrale. Etant donné que les mausolées circulaires ayant cette structure intérieure ont probablement toujours été recouverts d'un tumulus de terre, il est possible que ces cavités aient été remplies de terre.

Dans certains mausolées circulaires, touré de cavités curvilignes ayant plutót de murs concaves, placés entre le caveau en 1958, avait cette disposition intérieure.

on trouve un caveau central de plan cruciforme, enune forme semi-circulaire. Ces cavités étaient formées et l'enceinte circulaire. Un mausolée déblayé à Canosa Le monument qui peut étre daté du début de l'Empire

(1) Vitruvo, VI, 8. (2) G. Giovannoni, op. cit, pp. 53-54; L. Crema, op. cit, pp. 130-131, 243. (3) G. T. Rivoira, op. cit, pp. 14-15; L. Crema, op. cit, p. 131, fig. 260.

(4) G. T. Rivoira, loc. cit; L. Crema, op. cit. p.243. (5) L. Crema, op. cit, fig. 259. (6) G. Giovannoni, op. cit, p. 54, pl XII, L. Crema, op. cit, fig. 261. (1) Τὸ Crema, op. cit, p. 485, fig. 262. 55

HEMMING WINDFELD-HANSEN

avait un corps cylindrique au-dessus d'un podium carré peu élevé (1). Dans d'autres mausolées circulaires maintenant disparus, avec armature intérieure, il y cvait au contraire une chambre ronde centrale.

On connait un mausolée de ce type gráce aux fouilles effectaées dans les années 1856-1857 près de la Via Latina dans les environs de Rome (Pl. X e) (). Du plan du monument, il ressort que la pièce ronde était inaccessible et que le mur qui l'entourait avait une-forme octogonale à l'extérieur, Entre ce mur et l'enceinte circulaire du mausolée, on a trouvé huit murs rayonnants qui séparaient huit chambres, dont sept étaient de forme trapézoidale, alors que la huitième était un petit caveau rectangulaire auquel on accédait par une porte dans le mur d'enceinte. Un mausolée circulaire dont on voit les ruines sur la Via Flaminia, au Nord de Rome (Pl. X f) €) avait une forme similaire. Mais il n'en reste plus maintenant que le noyau central dans lequel se trouve un caveau rond dont les murs, creusés de quatre niches trapézoidales, sup-

portaient primitivement une coupole. Ce caveau est entouré d'un mur octogonal d’où partent

un grand nombre de petits pans de murs qui forment apparem-ment un système de contrefort autour du noyau.

La forme primitive du mausolée est reproduite sur un plan de Pirro Ligorio (PL X g) (1) d'après lequel le noyau octogonal était entouré d'un mur circulaire qui formait l'enceinte du mau-

solée et qui était relié au noyau au moyen de 17 murs rayonnants. Ces murs qui sont partielle-

ment conservés et rappellent des piliers de contrefort, séparaient un nombre correspondant de très étroites chambres trapézoidalés. Dans ce mausolée les murs étaient revétus d'un opus reticu-

latum. qui date le monument de l'époque d'Auguste. Un autre dessin de Pirro Ligorio montre le plan d'un mausolée circulaire inconnu (Pl. XI a) ©) qui avait probablement le caractère d'une tour, car le corps cylindrique reposait sur un socle ou un podium de plan carré. Ce cylindre comporte une pièce circulaire centrale entourée d'un mur concentrique; Entre ce mur et l'enceinte du cylindre, il y avait quatre murs curvilignes concaves qui, avec le mur de l'enceinte, limitaient quatre chambres curvilignes, probablement inaccessibles. Dans trois des mausolées décrits ici, on trouve ainsi une sorte de division concentrique avec une pièce centrale entourée d'une zone annulaire, divisée en petites chambres entièrement

fermées.

Un édifice conique qui n'est connu que par un déssin de Pirro Ligorio (Pl ΧΙ δ) (9) est d'un caractère différent. Ce monument qui est reproduit en plan et en perspective était, d’après la description, un mausolée de la Via Labicana aux environs de Rome. De la reproduction

en

perspective, il ressort que le cône a reposé sur un podium carré, mais le plan montre aussi qu’entre

le cóne et le podiumil devait y avoir un socle circulaire, omis sur la reproduction en perspective. A l'intérieur de l'édifice, il y avait une pièce circulaire centrale à laquelle conduisait un dromos. Huit murs rayonnants divisaient l'espace compris entre le mur circulaire de la pièce et le mur entourant le socle en sept chambres de forme trapézoïdale. Celles-ci communiquaient entre elles par des portes et formaient ainsi une sorte de couloir annulaire autour de la pièce centrale, On (1) F. Bertocci, Un soo mawsoleo a Canosa, Une copie de ce dessin se trouve è la Bibliothèque 4 Palladio », XI, 1961, pp. 86-91. Vaticane, Cod. Vat. Lat. no. 5439, £ 33 r. (a) L. Fortunati, Relazione generale degli scavi e (s) Oxford, Eodieian Library, Cod. Canonici scoperte fatte lungo la Via Latina, Rome 1859, no. ital no. 138, f. 78 r. Voir: Th. Ashby, The Bodleian

34 sur le plan d'ensemble. (3) Th. Ashby et Fell, The Via Flaminia, «Journal of Roman Studies », XI, 1921, p. 147, fig. 9. (4) Naples, Bibliothèque Nationale, Cod, XIIL -B. 10, Libro XLIX (le numéro du feuillet manque)

MS. of Pirro IX, 1919, p. (6) Oxford, ital, no. 138, p. 195.

56

Ligorio, « Journal of Roman Studies », 188, fig. 22. Bodleian Library, Cod. Canonici £ 114 v. Voir: Th. Ashby, op. cit,

LES COULOIRS ANNULAIRES

peut donc dire que cet édifice forme la transition entre les mausolées circulaires dont le cylindre est muni d'une rangée de chambres ou de cavités inaccessibles placées autour de la pièce centrale et les mausolées circulaires qui ont une division concentrique formée de couloirs annulaires séparés par des murs concentriques. Edifices circulaires construits avec des murs datant de la fin de la période républicaine.

concentriques,

De la fin de la période républicaine, on a quelques exemples d'édifices circulaires romains formés de deux murs concentriques. Il ne s'agit cependant pas de mausolées mais dans le premier cas, d'un édifice cultuel, et dans l'autre, d'une construction utilitaire profane. La plus ancienne de ces constructions est la rotonde qui couronne la terrasse supérieure du sanctuaire de la Fortuna Primigenia à Préneste (Palestrina) (Pl. XI «) (1). Cette rotonde se trouve derrière le grand hémicycle qui a la forme d'un double portique à colonnes dans lequel le Palais Colonna-Barberini est partiellement engagé. La rotonde généralement corisidérée comme

le sanctuaire méme de la Fortuna est formée de deux murs concentriques autour

d'une salle cir-

culaire, Ces deux murs sont construits en opus incerfum avec un noyau d’opus caementicium et il y ἃ entre eux un couloir de o m. 87 de largeur. Gullini et Fasolo () pensent que la partie antérieure de la rotonde était munie d'une large porte, flanquée de colonnes engagées et que cette ouverture donnait accès à la salle circulaire de culte du temple. Kabler 6) pense par contre que les deux murs annulaires de la construction appartenaient à un étage inférieur et qu'ils

étaient surmontés par une légère rotonde à colonnes qui s'élevait au-dessus du toit de l'hémicycle et formait la partie centrale terminale de cet énorme édifice. Sur la reconstitution de Kühler du sanctuaire de Fortuna, le temple rond est présenté comme un monoptére circulaire. Mais les dimensions relativement importantes de la rotonde et les deux

murs concentriques semblent indiquer plutôt qu'il était périptére et que le mur de la cella était supporté par le mur circulaire intérieur plus épais, tandis que la rangée de colonnes reposait

sur le mur extérieur plus étroit. La maçonnerie de ces murs correspond à la maçonnerie des autres parties du sanctuaire et cela parait indiquer que la rotonde remonte à la méme époque

que le reste de l'édifice qui, d'après Gullini et Fasolo (‘) date du milieu du 2° s. av. JC, mais que Lugli () et Kühler (5) placent seulement à l'époque de Sylla, c'est-à-dire vers 80 av. J.-C.

L'autre exemple d'édifice circulaire de la fin de la République présentant deux murs con-

centriques est une grande citerne que l'on a trouvée ati cours de fouilles dans une villa romaine

située prés de la source Aqua Claudia, à proximité d'Anguillara Sabazia, au bord du lac Bracciano (?). De cette construction, il ne reste que les deux murs circulaires et un pilier rond central. A l'origine, cette citerne devait donc se composer de deux espaces annulaires concentriques placés autour du pilier central et recouverts de votites en berceau. La citerne est sûrement contemporaine de la villa qui, d'après Vighi ('), doit remonter au milieu du 1% s. av. J.-C.

(1) F. Fasolo et G. Gullini, 1 santuario della (4) F. Fasoloet G. Gulli, op. cit,p. 301-323. Fortuna Primigenia a Palestrina, Rome 1953, pp. (6) G. Lugli, dans «Archeologia Classica», 1954, 187 ss; H. Kahler, Das Forlunaheilighum von Pale: pp. 305 s . sirina-Pracnese, «Annales Universitatis Saravien- — (6) H. Kabler, op. cit, pp. 208-209. sis, Philosophie» 1958, p. 207. Ὁ) R. Vighi, dans «Notizie degli scavi di anti(a) F. Fasolo ct G. Cullini lc. cit. chità », 1940, p. 406, figg. 18-19. (5) H. Kähler, loc. cit. (5) R. Vighi, loc. cit 57

HEMMING WINDFELD-HANSEN IV.- LES RITES DE CIRCUMAMBULATION ET LA SIGNIFICATION COULOIRS ANNULAIRES FUNÉRAIRES.

DES

A l'époque moderne, quelques savants qui ont étudié l'architecture funé-aire antique et ses monuments ont mentionné les couloirs annulaires des mausolées romains et numides. Certains ont essayé d'en expliquer l'usage primitif. Dans l'article de 1927 qui traite des deux mausolées nord-africains, le Médracen et le Tombeau de la Chrétienne, Welter (*) avance la théorie qu’aussi bien la galerie annulaire qui se trouve dans le dernier de ces tombeanx que les couloirs annulaires qui existent dans d'autres édifices funéraires ont eu une fonction rituelle. Il pense que ces couloirs ont servi aux rites de circumambulation de caractére lustral utilisés dans le culte des morts. Depuis lors, plusieurs autres archéologues qui se sont occupés de ces couloirs et galeries annulaires ont été du méme avis. Dans son ouvrage sur les mausolées circulaires romains, Gó-ze () parle des couloirs annulaires qui se trouvent dans quelques-uns de ces monuments. Comme Welter, il pense qu'ils ont servi à des parcours de lustration. Une description des couloirs annulaires dans les mausolées antiques est donnée dans le livre de Fellmann sur le Tombeau de Munatius Plancus à Gaète (). Lui aussi est d'avis que ces couloirs ont eu une fonction rituelle et, comme Gütze, il cite à ce propos des textes grecs et latins qui mentionnent des parcours c-rculaires rituels, effectués dans des cérémonies funéraires. Dans un article de Bohm (4), il est fait mention des cérémonies romaines de circumambulation dans le culte des morts, mais la description la plus détaillée de toutes les différentes formes

de ces rites funéraires pendant l'Antiquité se trouve dans le premier chapitre de l'ouvrage d'Eitrem « Opferritus und Voropfer der Griechen und Rômer » (); l'auteur y traite en particulier des cérémonies de cireumambulation qui accompagnaient les sacrifices aux dieux, mais aussi

toutes les autres formes de ces parcours rituels et magiques (7).

Des anciennes civilisations de l'Orient et en particulier de la Mésopotamie, les Grecs avaient hérité la croyance en la force magique du cercle et du mouvement circulaire (7), une croyance que l'on trouvait aussi aux Indes et chez la plupart des peuples indo-européens (f). Cette

force pouvait avoir un simple caractére magique, ainsi lorsqu'un magicien liait un homme en

tragant autour de lui un cercle qu'il ne pouvait pas franchir ()) ou lorsqu'on s'entourait d'un cercle pour se protéger des démons (9). Dans le monde des idées religieuses, le cercle et le mouvement circulaire avaient un effet cathartique et ils étaient utilisés dans les rites de purification 2). L'objet qui avait un effet magique de lustration était le plus souvent porté en cercle autour de ce qui devait étre purifié. (1) G. Welter, op. cit, pp. 113-115. kunde, 1938, pp. 219-287; F. Robert, Thymél, Re(2) B. Gótze, op. cit., p. 1r. cherches sur la signification et Ta destination des mo(3) R. Fellmann, op. cit, pp. 87-89. numents circulaires dans l'architecture religieuse de la (4) Bohm, art, «Lustratio», dans Pauly-Wissówa, Grèce, Paris 1939, pp. 319-322; L. Hautecoeur, Mysti«Realeneyelopidie » XIII, 2, Stuttgart 1927, cols. que et architecture, Symbolisme du cercle et de la cow 2036-2037. pole, Paris, 1954, PP. 34 85. (5) S. Eitrem, Opferritus und Voropfer der Grie(7) L. Hantecosur, Op. cit, pp. 31-32. chen und Romer, Kristiania 1915, pp. 6 ss. (8) S. Eitrem, op. cit, pp. 8, 12, 18, 20-21, 30,

(6) Pour les circumambulations voir aussi: W. 43, 51. B. Kristensen, Kringloop en totaitet, « Mededeslingen (o) ibid, pp. 21, 44. der Koninklijke Nederlandsche Akademie van We(to) ib d, pp. 9, 16-17, 20. tenschappen », Nieuwe Reeks, 1, Afdeeling Letter. — (rr) ibid, pp. 14, 15, 28. 58

LES COULOIRS ANNULAIRES

Il pouvait s'agir d'un foyer, d'un temple, d'une assemblée du peuple (ecclesia) ou de la thymélè au centre de l'orchestra des théâtres, qui étaient tous à l'origine des lieux de caractère cultuel (1). On pouvait d'ailleurs procéder de façon inverse: l'objet qu idégageait la force purifi-

catrice était alors placé au centre et ce qui devait étre purifié porté en cercle autour de l'objet

lustra (). Chez les Romains, le mouvement circulaire avait aussi un effet cathartique et était utilisé entre autres dans les fêtes annuelles des ambarvalia, des ambilustra ct des amburbia Q). Aussi bien chez les Grecs que chez les Romains, il était de la plus haute importance de procéder à une purification de l'autel avant un sacrifice. Le ou les officiants pratiquaient ce rite purificatoire appelé peristia chez les Grecs (*) et Jusératio chez les Romains (5) en faisant un tour ou le plus souvent trois tours autour de l'autel où le rite divin avait lieu (). Dans les littératures grecque et latine, on trouve de nombreuses descriptions indiquant que les prêtres procédaient è une purification de l'autel en tournant trois fois autour de celui-ci (7) avant d'offrir les sacrifices sur les autels consacrós aux divinités célestes. Dans d'antres cas, c'étaient les animaux destinés au sacrifice à qui on faisait faire le tour de l'autel où ils allaient être immolés (5); dans d'autres cas encore, on exécutait une danse rituelle en cercle autour

de l'autel (). Les parcours circulaires autour des autels des divinités célestes avaient toujours lieu en tournant vers la droite, car on considérait que le mouvement circulaire dans le méme sens que le soleil portait bonheur, alors que le mouvement en sens inverse était considéré comme néfaste (9). Les cérémonies cathartiques de circumambulation pratiquées à l'occasion des sacrifices chtoniens ne sont pas décrites d'une maniére précise dans les sources littéraires antiques, ces sacrifices ayant un caractère occulte. Il est vraisemblable qu'au cours de ces cérémonies, on tournait tant ἃ gauche qu'à droite autour des autels consacrés aux divinités infernales. Le sous-sol circulaire de la grande rotonde appelée Tholos ou Thymélè du Hiéron d'Asclé-

pios à Epidaure (#) que l'on peut dater des années 360-330 av. J.-C. (12) contient en effet un

curieux labyrinthe (Pl. XII a). Celui-ci est formé de trois murs concentriques séparés par des couloirs annulaires et se trouve sous le sol de la cella de la rotonde. Dans chacun de ces murs,

il y avait une porte, alors que chaque couloir était fermé par une cloison placée prés de la porte

ouvrant sur le couloir suivant. Ces cloisons étaient placées de telle manière que, pour passer d'un couloir à l'autre, on était obligé de faire un tour complet autour de la chambre circulaire placée au centre du soubassement. Les parcours s'effectuaient de manière telle que dans le couloir extérieur il fallait tourner à gauche, dans le couloir suivant - celui du milieu — tourner à droite, et enfin dans le couloir intérieur, tourner de nouveau

à gauche

avant

d'atteindre la chambre.

Au cours des fouilles, Ro-

bert (5) a trouvé qu'il avait existé dans cette chambre un petit énagistérion, c'est-à-dire un autel chtonien creux destiné aux sacrifices sanglants. Ce labyrinthe, avec ses trois couloirs annulaires, qui a une ressemblance marquée avec quelques-unes des plus anciennes représentations sur les (1) L. Hautecoeur, op. cit., p. 34-35.

(7) ibid., pp. 7-8, 24-25, 26.

(2) S. Eitrem, op. cit, pp. 14, 15. (8) ibid, pp. 25-26. (9) ibid, pp. 25, 28-20. K. Latte, Rümische Religionsgeschichte, dans (3) «Handbuch der Altertumswissenschaft », Munich 1960, (ro) ibid, pp. 43-45, 55; F. Robert, op. cit, pp. 41-42. Voir aussi: S, Eitrem, op. cit. pp. 17,18. pp. 319-320. (4) 1. Hautecoour, op. cit, p. 34. (11) F. Robert, op. cit, pp. 296-305. (5) Bohm, loc. cit. Voir aussi: S. Eitrem, op. cit, 14) W. B. Dinsmoor, Architecture of Ancient PP. 16, 17-18. Greece, pp. 235-236. (6) S. Eitrem, op. cit, pp. 6, 7-8, 9, 24-25. (13) F. Robert, op. cit, pp. 338-341. 59

HEMMING WINDFELD-HANSEN

monnaies du Labyrinthe primitif de Cnossos en Crète (), a vraisemblablement servi à des cérémonies cathartiques, en relation avec les sacrifices chtoniens de caractère occulte qui avaient lieu dans cette rotonde. La Tholos d'Epidaure est le seul édifice circulaire grec dans lequel nous trouvions un labyrinthe souterrain formé de couloirs concentriques qui avaient une fonction cultuelle. Mais comme Robert () l'a démontré, il s'agit ici d'un sanctuaire chtonien rattaché au culte d'Asclépios et non d'un édifice funéraire, Aussi bien chez les Grecs que chez les Romains, en plus des effets cathartiques décrits cidessus, le cercle et le mouvement circulaire avaient aussi un carastère apotropaique de conjuration des influences malignes et il est souvent difficile de distinguer entre les propriétés de lustration et celles d'apotropée du cercle. Si le cercle et les cérémonies de circumambulation sont si souvent utilisées dans le culte des morts, c'est en particulier par suite du caractère de préservation contre les influences malignes, attribué au mouvement circulaire. Déjà dans les cérémonies funèbres, le défunt était lié à un cercle sépulcra! imaginaire du fait que l'on faisait un parcours circulaire rituel, soit autour du bücher d'incinération, soit autour de la tombe elle-même, immédiatement après l'inhumation, Dans la description que fait Homère des funérailles de Patrocle dans l’Iliade (?), il dit qu’Achille et ses Myrmidons firent irois tours à cheva autour du corps de Patrocle, A propos des funérailles d'Archémoros, Stace (‘) mentionne que lesl Argiens armés, avec leurs rois en tête, tournérent trois fois à gauche — δὰ orbe sinistro — autour du bücher en flammes, après quoi la procession fit volte-face pour faire trois tours dans le sens inverse. Les Ephèbes chevanchèrent aussi trois fois autour du tombean de Néoptoléme à Delphes (ἢ. La description que fait Plutarque (9) de la visite d'Alexandre le Grand ἃ la tombe d'Achille è proximité de Troie

montre

que

les cérémonies

de circumambulation

étaient

aussi

utilisées

par les Macédoniens hellénisés, D'aprés cette description, Alexandre offrit des sacrifices tant à Athéna qu'à Achille dont la stèle funéraire fut ointe d'huile. Après cela Alexandre et sa suite coururent nus autour du tumulus pour déposer finalement une couronne sur la stèle funéraire, Dans la decursio romaine autour du rogus ou du bücher d'iacinération d'un chef militaire, utilisée plus tard aussi à l'occasion de l'incinération des empereurs (7), les cavaliers chevauchèrent dans un sens puis dans l'autre autour du bücher. Cette decursio doit avoir rappelé la danse des cavaliers appelée Trojae ludus pratiquée aussi bien par les Etrasques que par les Romains (). Un tel Trojae ludus est décrit par Virgil dans l'Enéide, à propos des concours funèbres en l'hon-

neur d'Anchise (°). Lorsque le défunt était inhumé, il pouvait être éternellement lié à sa dernière demeure et Yon tragait autour de sa tombe un cercle, soit sous la forme d'un tumulus, soit sous celle d'un cercle de pierres ou d'un mur d'enceinte circulaire formant le socle du monticule conique (*). L'urne du défunt, son cercueil ou son caveau étaient le plus souvent placés au centre de ce monument circulaire. Le défunt était ainsi enfermé dans le cercle sépulcral comme les animaux sauva-

ges qui, au cours de la chasse, étaient capturés dans un enclos circulaire servant de piège (2). (1) Humborg-Karo, art. ¢Labyrinthos», dans Pauly-Wissowa, «Realencyclopádies vol. XII, 1,

(7) S. Eitrem, op. cit, p. 11; F. Robert, op. cit, p. 321.

Stuttgart 1924, cols. 312-323; F. Robert, op. cit, Ῥ 306, (2) F. Robert, op. cit., pp. 325-338. (3) Iliade, XXIII, 12-14. (4) Stace, Thébaïde, VI, 213-219. (s) Héliodore, Ethiopigues, 3, 5, 2. (6) Plutarque, Vie d'Alexandre, 15.

(8) F. Robert, op. cit, pp. 316-318; E. Mehl, art. ¢Troiaspicl=, dans Pauly-Wissowa, «Realencyclopädie » Suppl. VITI, Stuttgart 1956, cols. 888905. (0) Enéide, V, 545 ss. et principalement 580-587. (ro) L. Hautecoeur, op. cit, pp. 26-28. (r1) ibid, p. 3». 60

LES COULOIRS ANNULAIRES

Simultanément, le cercle visible autour de la tombe attirait l'attention des vivants surla présence d'un lieu saint (1). Etant donné que la tombe était un locus religiosus consacré aux divinités infernales, les vivants ne devaient pas la fouler avant d'avoir procédé au rite funéraire; il était nécessaire en particulier de faire des sacrifices aux mAnes du défunt (); en plus, il fallait effectuer un ou plusieurs parcours circulaires à gauche autour de la tombe (!) avant d'y pénétrer, Un parcours de ce genre peut être interprété comme un «rite de passage » (ἢ) nécessaire pour passer du monde des vivants à la sphère réservée aux morts, placée sous ja protection des puissances chtoniennes et infernales.

Lorsqu'on quittait ensuite la tombe pour revenir au monde réel, on faisait ἃ nouveau le même parcours circulaire autour de la tombe, mais cette fois dans le sens inverse, en se libérant par là des morts et des effets magiques dangereux de la tombe (9. Chez les Romains, ce rite de circumambulation fut peu à peu remplacé dans les cérémonies de sacrifice aux divinités célestes et dans le culte funéraire par une volte purement symbolique devant l'autel ou la tombe (9) Un témoignage de parcours circulaire rituel ou de danse en rond autour des tombes à tumulus chez les Grecs est fourni dans la scène reproduite sur un lécythe à figures noires qui se trouve au Musée National de Naples (7). On y voit une petite tombe à tumulus autour de laquelle deux guerriers et une femme dansent, tandis qu'un vieillard appuyé sur un baton contemple l'acte rituel (Pl. XII c). La présence de circumambulations rituelles dans le culte des morts est ainsi attestée par de nombreux exemples, tirés des littératures grecque et latine. Mais il existe aussi des témoignages archéologiques qui montrent que les cérémonies de circumambulation ont eu lieu autour des tombeaux à tumulus et des mausolées circulaires. Les murs d'enceinte des tombeaux préhistoriques à tumulus ont peut-être formé des barrières rituelles entre le monde des vivants et celui des morts. Les cercles de pierres qui entourent tant de ces tombeaux à tumulus ont aussi probablement été édifiés pour marquer la. délimitation rituelle de l'aire funéraire, et les rites de circumambulation ont vraisemblablement eu lieu autour de ces enceintes et de ces cercles de pierres. Les cercles concentriques de pierres qui formaient des clótures autour de certains des tombeaux mégalithiques à tumulus dans la nécropole de Los Millares (cf. ci-dessus p. 36) peuvent avoir servi à délimiter la zone annulaire où ces parcours rituels devaient s'accomplir (?). En Grande-Bretagne, les déblaiements d'une série de tombeaux à tumulus de l'àge du Bronze moyen et récent (1300 à goo av. J.-C.) ont donné des preuves certaines que des rondes rituelles ont été dansées autour de ces tombeaux (*). En effet, on a trouvé une couche de terre battue trés dure, formant une large zone autour de la tombe elle-méme, et dans cette couche,

(1) loc. cit (2) Fustel de Coulanges, La cité antique, Paris 1876, pp. 13-14; H. J. Rose, Ancient Greek Religion, Londres 1946, pp. 30, 41; H. V. Herrmann, op. cit, PP. 53-59. (3) S. Eitrem, op. cit, pp. 29, 43, 45. (4) ibid, p. zr. (5) ibid., p. 43. (6) ibid., pp. 11, 24, 45, 46. (7) J. Harrison, Delphica, « Journal of Hellenic Studies », XIX, 1899, p. 229, figg. 9-10; S. Eitrem,

op. cit, p. 29; H. V. Herrmann, op. cit, p. 41, fig. 3. (8) Dans le méme ordre d'idées, il faut aussi mentionner que le tumulus du grand dolmen à golerie de Kercado près de Carnac en Bretagne est entouré d'une voie dallée. Un cercle concentrique de pierres formait une clôture autour de cette voi qui avait probablement une fonction rituelle. Voir: E. O. James, Prehistoric, Religion, pp. 79-80, fig. 5. (Ὁ) C. Fox, Life and Death im the Bronze Age, pp. XXVI, 98, 185. 6x

HEMMING WINDFELD-HANSEN

on voit encore des traces de pas (i). Les danses rituelles auîour des tombes n'ont eu lieu que pendant un assez court espace de temps: depuis l'époque où les cendres du mort furent ensevelies dans une fosse au centre de l'aire funéraire jusqu'à celle où la tombe et le rond de danse environnant

furent couverts d'un tumulus de terre et de tourbe peu élevé.

Les fouilles n'ont pas révélé que des circumambulations rituelles ont eu leu autour des mausolées grecs à tumulus, pas non plus pour les tombeaux étrusques à tumulus, car les dallages ou plates-formes qui entourent les crépis des grands tombeaux circulaires dans les nécropoles de Populonia n'ont probablement été aménagés que pour protéger les fondations des tombeaux contre l'écoulement de l'eau qui tombait du grundarium (cf. ci-dessus, p. 38)

La voie dallée dont on a retrouvé les traces sur la terrasse portée par le mur d'enceinte du

mausolée d'AntiochusI" sur le Mont Nemrud Dagh (cf. ci-dessus, p. 38) doit avoir servi aux

processions de circumambulation qui ont sürement eu lieu autour de cet héroon. Le curieux

passage annulaire qui se trouve dans le petit mausolée circulaire de Ghermes (cf. ci-dessus, p. 39)

est une autre preuve que les circumambulations rituelles ont &6 pratiquées autour de monuments funéraires de la période hellénistique dans les pays de la Méditerranée orientale. Des céré-

monies de parcours circulaire pouvaient se dérouler à l'abri des regards, le passage étant découvert

mais entouré

d'un mur

circulaire.

Pour expliquer la présence de murs circulaires intérieurs dans certains des tombeaux nu-

mides à tumulus, Gsell (?) a formulé l'hypothèse que ces murs pouvaient avoir une fonction

rituelle puisqu'ils formaient une ceinture autour du tombeau méme, tout en admettant que ces murs circulaires pouvaient avoir un rôle fonctionnel ou qu'ils étaient les restes de tombeaux plus anciens. Gsell () émet d'ailleurs l'hypothèse que les couloirs annulaires des mausolées numides pouvaient tirer leur origine des murs circulaires qui existaient dens ces tombes à tumulus nordafricaines primitives. Le couloir annulaire qui se trouve dans le tombeau situé près du Médracen (cf. ci-dessus, p. so) pent étre considéré comme le prototype de la galerie annulaire du Tombeau de la Chrétienne (cf. ci-dessus, p. 50-51). Les parois de cette galerie sont creusées de nombreuses petites niches destinées à recevoir des lampes à hüile (+). Ces lampes servaient à éclairer la ga lere, ce qui montre que celle-ci a été construite pour des processions rituelles. La raison pour laquelle on a construit deux couloirs annulaires dans le Mausolée d'Auguste (cf. ci-dessus, p. 39-40) ne peut pas étre d'ordre technique. S'il s'agissait seulement de renforcer la

stabilité du noyau de l'édifice, on aurait préféré diviser aussi les espaces entre les trois murs concentriques intérieurs en de petites cavités séparées par des murs rayonnants. Les deux couloirs

concentriques de ce mausolée ont vraisemblablement été utilisés pour des parcours rituels, car

ici le dromos ne méne pas directement à la chambre sépulcrale, comme dans la. plupart des autres mausolées circulaires romains, mais aboutit devant le 2° mur circulaire. Celui qui veut visiter l'hypogée du Mausolée d'Auguste est donc obligé d'entrer dans le couloir extérieur au lieu de passer par une des deux portes étroites percées dans le 2° mur culaire à gauche et à droite de l'endroit où le dromos débouche dans ce couloir. Le visiteur est donc contraint de suivre tout le couloir extérieur et de faire ainsi un tour complet autour de la chambre sépulerale, La méme chose se produit une fois qu'il est arrivé dans le couloir intérieur par l'une des deux portes mentionnées. Et c'est donc seulement après avoir fait deux fois le tour de la chambre sépulcrale qu'il peut franchir la porte cuverte dans le 1* (2) Par ex. les tombeaux à tumulus «Sutton 268 et «Pond Cairns, tous les deux dans le Comté de Glamorgan. Voir: C. Fox, op. cit, pp. 97-105, fig. 41 et pp. 105-127, fig. 59.

(ἢ) Nord, (3) (4)

S. Gsell, Histoire ancienne de l'Afrique du VI, pp. 183, 188. ibid, p. 199, n. 2. M. Christofle, Le Tombeau de la Chrétienne, p. 29.

LES COULOIRS ANNULAIRES

mur et arriver à la dernière demeure du Divus Augustus et de tous les membres défunts de sa famille. Le couloir annulaire du Sepolcro dei Servili (cf. ci-dessus, p. 45) sur la Via Appia doit aussi avoir servi de cadre à la circumambulation rituelle qui précédait l'entrée dans la chambre sépulcrale. Si le couloir de ce petit monument avait été un intercapedo destiné à protéger le sépulcre de l'humidité extérieure, il aurait été inaccessible. La même explication vaut sûrement aussi pour les deux grands mausolées situés près de Gatte, celui de Munatius Plancus et celui de Sempronius Atratinus (cf. ci-dessus, p. 40-41), où le couloir anmulaire est le seul accès par lequel on puisse parvenir aux chambres sépulcrales.

La galerie annulaire du Mausolée d'Hadrien (cf. ci-dessus, p. 42-43) est une transformation monumentale des étroits couloirs annulaires qui se trouvent ou se trouvaient dans plusieurs des mausolées circulaires du début de l'Empire. Du fait que la chambre sépulcrale de ce mausolée était placée à un niveau beaucoup plus élevé que le dromos et le vestibule, il a été nécessaire de construire une rampe par laquelle le convoi funèbre ou la procession cultuelle pouvaient passer lorsque le sépulcre sacré des empereurs Antonins était ouvert. Ce qu'il y a de curieux, c'est que cette rampe est établie de manière telle que ceux qui voulaient se rendre à la chambre sépulcrale et qui suivaient la galerie devaient faire le tour complet du sépulere dans le sens contraire à celui

du soleil, c'est-à-dire vers la gauche. Lorsque les visiteurs quittaient la chambre sépulcrale, ils

devaient à nouveau suivre la rampe et faire encore une fois le tour de la cella, mais cette fois vers la droite. Lorsque les processions effectuaient ces parcours pour entrer dans la chambre sépulcrale

du Mausolée d'Hadrien et ensuite pour en sortir, elles accomplissaient justement le méme rite de circumambulation que les dévots pratiquaient en visitant le caveau central du Tombeau de la Chrétienne. Cette circumambulation rituelle qui appartient exclusivement au culte des morts

(cf. ci dessus, p. 61) précédait probablement aussi l'entrée dans les chambres sépulcrales des autres mausolées romains et numides à couloir annulaire et se répétait en sens inverse lorsqu'on voulait quitter le lieu sacré de la tombe qui était consacré aux divinités chtoniennes (1). (1) Je me propose de consacrer un travail ultérieur à l'analyse typologique des monuments cireu-

laires dans l'architecture funéraire antique et à une recherche sur leur origine et leur signification.

65

H. Windfeld-Hansen, Les couloirs annulaires

PLANCHE I

a) Le Mausolée d'Auguste à Rome, plan (d'après Gatti). b) Maxsolée antique, plan (d'après une gravure de Gio. Batt, Montano). c) Le Mausolée de Munatius Plancus près de Gadte, plan (d'après Fellmann) 4) Le Mausolée de Sempronius Atratinus près de Gate, plan (d'après Crema).

H. Windfeld-Hansen, Les couloirs annulaires

PLANCHE Il

H. Windfeld-Hansen, Les couloirs annulaires

PLANCHE ΠῚ

a) Mausolée prés de Casale Portonaccio au Sud de la Via Tiburtina, plan approximatif (d'après l'auteut). b) Mausolée dans la Vigna Pepoli à Rome, plan. R. Lanciani, Forma Urbis Romae, pl. 45. c) Bramantino, Mausolée près de Tre Fontane au Sud de Rome, plan. Milan, Biblioteca Ambrosiana, Taccuino di Bramantino, pl. XXIII. d) Gio. Batt. Montano, Mausolée dans la Campagne romaine, plan. Libro III, pl. 37. €) Gio. Batt, Montano, Mausolée antique, plan. Libro III, pl. 40,

H. Windfeld-Hansen, Les couloirs annulaires

PLANCHE IV

B

E

a) Francesco da Sangallo, Mausolée près de St-Sébastien sur la Via Appia, plan. Bibliothèque Vaticane, Cod. Barb. Lat. no. 4424, f. 39r. 6) Giuliano da Sangallo, Le mausolée appeïé Carceri Vecchie près de S. Maria Capua Vetere (Capoue), plan. Bibliothèque Vaticane, Cod. Barb. Lat, no. 4424, f. 87. v) Pirro Ligorio, Le mausolée appelé Carceri Vecchie près de S. Maria Capua Vetere, plan. Naples, Biblioteca Nazionale, Cod. XIII. B. το, Libro XLIX. d) Le mausolée appelé Carceri Vecchie, plan (d'après Pane)

H. Windfeld-Hansen, Les couloirs annulaires

PLANCHE V

2»

a) Gio. Batt, Piranesi, Le mausolée appelé Monte del Grano, plan. Le antichità romane, vol. II, pl. XXXI. Ὁ) Gio. Batt. Piranesi, Le mausolée appelé Monte del Grano, coupe. Le antichità romane, vol. II, pl. XXXII. è) Le mausolée appelé Monte del Grano à Rome, plan et coupe (d'après Ashby et Lugli).

H. Windfeld-Hansen, Les couloirs annulaires

PLANCHE VI

4) Le mausolée appelé Sepolcro dei Servili sur la Via Appia, plan (d'après Canina). 8) Sallustio Peruzzi, Mansolée près de la Via Latina, plan. Florence, Musée des Offices, no. 665 A v. c) Jacques Androuet Du Cerceau, Mausolée près de la Via Latina, plan, «Grands Temples», gravure désignée « Β » d) Gio, Batt. Montano, Mausolée près de la Via Latina (la légende sur la gravure est erronée), plan, perspective et coupe partielle, Libro II, pl. 26.

H. Windfeld-Hansen, Les couloirs annulaires

Tian [errat e

PLANCHE VII

Bor

a) Mausolée appelé Tor Inviolata au Nord de la Via Tiburtina, plan (d'après Giovannoni). δ) Sallustio Peruzzi, Mausolée appelé Sepolcro di Gallieno sur la Via Appia, plan et élévation avec coupe partielle, Florence, Musée des Offices, no. 664 A v. e) Pietro Santi Bartoli, Mausolée appelé Sepolcro di Gallieno sur la Via Appia (la légende sur la gravure est erronée), élévation et coupe partielle, Domenico de’ Rubeis, Romanae magnitudinis monumenta, pl. 50.

H. Windfeld-Hansen, Les couloirs annulaires

PLANCHE VIII

D

Brod

2

1

a) Mausolée appelé Tempio di Portuno à Porto (Fiumicino) près de Rome, plan de |étage inférieur (d'après Lugli et Filibeck). 2) Rctonde prés de la Via Portuensis, plan de l'étage inférieur (d'après Lanciani). v) Pietro Santi Bartoli, Hypogée au pied de l'Aventin à Rome, plan et coupe. Gli antichi sepolcri, pl. 45.

HE IX

H. Windfeld-Hansen, Les couloirs annulaires

PLANCHE X

a) L'une des deux tombes à tumulus dites Tombes des Horaces, plan (d'après Crema). 2) Tombe à tumulus près de Todi, plan (d'après Crema). c) Tombe à tumulus à Marcigliana pris de la Via Salaria, plan (d'après Crema). d) Tombe à tumulus à West Mersea (Angleterre), plan (d'après Crema). e) Mausolée près de la Via Latina, plan (d'après Fortunati). f) Mausolée sur la Via Flaminia, au Nord de Rome, plan (d'après Ashby et Fell). g) Pirro Ligorio, Mausolée sur la Via Flaminia, au Nord de Roms, plan. Naples, Biblioteca Nazionale, Cod, XIII. B. 10, Libro XLIX,

H. Windfeld-Hansen, Les couloirs annulaires

a) Pirro Ligorio, Mausolée inconnu, plan. Oxford, 5) Pirro Ligorio, Mausolée sur la Via Labicana, plan. £ 114 v. e) Le sanctuaire de la For-una Primigenia Gullini

PLANCHE XI

Bodleian Library, Cod. Canonici Ital. no. 138, f. 787 Oxford, Bodleian Library, Cod. Canonici Ital. no. 138, à Préneste, plan de l'hémicycle et de la rotonde (d'après et Fasolo)

H. Windfeld-Hansen, Les couloirs annulaires

PLANCHE ΧΙ

a) La Tholos dans l'Hiéron d'Asclepios à Epidaure, plan du soubassement (d'après Robert). b) Mausolée à Ghermèe (Cyrénaique), plan (d'après Pacho). c) Naples, Museo Nazionale: Lécythe è figures noires. (d'après Harrison).

Nuovo

contributo

allo studio

del Palazzo Erculio di Piazza Armerina HANS

PETER

L’ORANGE

« Fra quanti scavi archeologici vengono, nel nostro tempo, ad arricchire di nuove scoperte il patrimonio dell'antica storia e dell’arte, l'esplorazione della villa romana di età imperiale di Piazza Armerina è certamente destinata, come si dice, a fare epoca. In una località, che non era per nulla famosa di tradizioni dell'antichità, è apparso uno dei monumenti più solenni dell’arte romana; non soltanto il più notevole della Sicilia, ma certamente — senza tema di esagerare per amore di argomento -- fra i più cospicui di tutto il mondo romano ». Facendo mie queste parole del compianto prof. Biagio Pace, mi permetto di esprimere la gratitudine che i cultori dell'archeologia classica d'ogni parte del mondo devono all'insigne storico siciliano per la grandiosa iniziativa da lui presa per gli scavi di Piazza Armerina e per i primi contributi di chiarificazione che egli ha portato allo studio del monumento (1). La nostra riconosceriza va anche al Sopraintendente degli Scavi, prof. Bernabò Brea e αἱ dott. Gino Vinicio Gentili che ha diretto gli scavi e ne ha dato i rendiconti (). Nella vivace discussione sul singolare monumento () intendo qui, ancora una volta, recare il mio contributo. Sulla base delle notizie preliminari sugli scavi di Piazza Armerina, pubblicate dal Pace e dal Gentili fino all'anno 1952, ebbi a proporre un'altra data da quella supposta: proposi cioè di datare tutto il complesso musivo di Piazza Armerina agli anni intorno al 300 d. C. Ove si prescinda da alcuni restauri e ritocchi d'epoca più tarda, tutto l'anzidetto complesso musivo mi sembrò che si accordasse con un unico grandioso concetto, il quale, originariamente, doveva essere

in rapporto organico col palazzo. Sulla base di elementi di deciso carattere erculio-tetrarchico nelle rappresentazioni musive, oltre che per la disposizione imperiale degli ambienti, che esse ador-

nano, proposi di ravvisare nel comprensorio del palazzo una reggia dell'imperatore Massimiano Erculio. Le mie argomentazioni e deduzioni furono esposte in un articolo pubblicato in « Symbolae Osloenses » nel 1952, il cui titolo era «È un palazzo di Massimiano Erculio che gli scavi di Piazza Armerina portano alla luce ? ». Il miglior conforto alla nostra datazione ed interpretazione del palazzo mi pare sia stato

dato dallo stesso autore degli scavi del palazzo, il Gentili, il quale già nella seconda edizione della

(1) B. Paco, Note sulla villa romana di Piassa Armerina, « Rendiconti dell'Accad. dei Lincei», serie VIII, fase. rraz, 1051; in seguito, dello stesso, gli articoli sulla rivista «Sicilia» di Palermo, I, n. 4 © sulla « Revue Française» VI, n. 6r. L'ultimo lavoro del Pace su questo argomento è il volume: I mosaici di Piazza Armerina, Roma 1955. (2) G. V. Gentil, «Not. Scavi », 75, 1950, pp. 201 spi; idem, I mosaici della villa det Casale di Piassa Armerina, «Boll. d'Arte» 37, 1952; idem, La villa

10

romana di Piassa Armerina, e, La villa imperiale di Piazza Armerina (entrambi in «Itinerari dei Musei e Monumenti d'Italia», 1951 e 1054); Iconografe tefravehiche nei mosaici della villa Erculia di Piatra Armerina, «La Giara», IV, n. 4, 1955; Le gare del _ circo nel mosaico di Piazsa Armerina, « Boll. d'Arte», 42, 1957, PP. 7 968. La Villa Evculia di Piazza Armerina, I mosaici figurati, Roma 1959. (3) Una descrizione degli scavi è stata data da B. Neutsch, « Archäolog. Anzeiger» 1954, coll. 553 588.

65

P. ORANGE

sua guida «La Villa Imperiale di Piazza Armerina» (1954), aderiva alla nostra concezione del palazzo, nei suoi punti essenziali, riconoscendo nello stesso «la grande villa della famiglia imperiale di Massimiano Erculio » (1) ed abbandonando il concetto, esposto nella prima edizione (1951), che si trattasse di una 4 grande villa di età evidentemente costantiniana ». Nel suo ultimo lavoro sul palazzo, « La villa Erculia di Piazza Armerina, i mosaici figurati » (1959), il Gentili espone il suo punto di vista che, sostanzialmente, corrisponde a quello da me sostenuto nell'articolo citato, nel 1952 in «Symbolae Osloenses» á La migliore conferma dei miei risultati di studio si è avuta quando il Gentili, dopo scavi ed esami continuati sul luogo per sette anni, affermò (?): «Un fatto è certo, che alcune delle maggiori composizioni musive cono strettamente legate... alle concezioni politico-religiose della prima Tetrachia, mentre la quasi totalità dei mosaici riflette la forma e lo stile dell'arte di quella età. . Indubbiamente un piano organico ed unitario ha presieduto alla costruzione dai suoi corpi di fabbrica in quello scorcio di tempo che sta a cavallo tra il m ed il 1v secolo». Che poi si tratti di un «ritiro» o d'una villa per la caccia (), questo è soltanto un particolare nel concetto totalmente nuovo del palazzo, ora accettato anche dal Gentili. Altre voci, però, si sono levate contro la nostra datazione ed interpretazione del palazzo. Così il Pace (ἢ si oppose energicamente al nostro concetto, assumendo che il palazzo e la sua decorazione musiva non costituiscono un'unità, bensì sono venuti sorgendo man ‘mano attraverso più generazioni durante l’ultimo rv secolo e il primo v secolo, come una specie di otium philosophicum per una delle grandi famiglie latifondiste dell’aristocrazia tardo-romana, i Nicomachi (p. 103). Anche Giuseppe Lugli, basandosi sulla costruzione e sui muri del complesso, pensa che il palazzo e la sua decorazione siano sorte in più riprese (*). Recentemente Michelangelo Cagiano de Azevedo ha ripreso la tesi del Pace cercando di rafforzarla con nuovi argomenti (p. 89). Riassumiamo qui le nostre ragioni per la datazione ed interpretazione del complesso musivo sviluppate nel suddetto articolo e aggiungiamo nuovi argomenti in favore della tesi, apportati sia dal Gentili, sia da noi stessi. IN CHE TEMPO HA AVUTO ORIGINE DI PIAZZA ARMERINA.

L'OPERA

Datazione fondata sui dati di scavo.

MUSIVA

DEL PALAZZO -

Nel suo ultimo libro sul palazzo di Piazza Armerina dell'anno 1959 il Gentili riassume i risultati degli scavi che hanno importanza fondamentale per la datazione (). «Che i mosaici della Villa Erculia siano opere del periodo tetrarchico, fatta naturalmente eccezione per le più tarde rifazioni, già denunciate, è provato dai dati di scavi. Sotto le superfici musive, che per ragioni di restauro e di consolidamento sono state distaccate in numerosi ambienti della villa, gli elementi suggellati sono apparsi costantemente uniformi: ceramiche sigillate (1) La villa imperiale di Piazza Armerina, 1954, passim : si vedano anche i lavori successivi del Gen-

fili citati sopra alla p. 65, nota 2. (Ὁ 1 mosaici figurati, op. cit, pp. ra sg. (3) Gentili, La ville imperiale, op. cit, p. 9 © passim. Su questa discussione si veda ἢ mio contributo negli «Studi in onore di Calderini © Paribent», III, 1936-57, PP. 596 sg

(4) 1 mosaici di Piazza Armerina, op. cit., passim. (5) G. Lugli, Contributo alla Storia edilizia della

villa romana di Passa Armerina, « Rivista dellIstituto Nazionale di Archeologia e Storia dell'Arte 2027, NS. 11-12, 1963, pp. 28 mE. ἠ ἠ(6) Gentil, T mosaici gurati, op. cit, D. 74.

66

PIAZZA ARMERINA

chiare, che nella classificazione preliminare tentata dal Lamboglia si inquadrano, per le forme generali del vaso e per l'ispessimento degli orli, nel tipo D, ascritto tra la seconda metà del sec. πὶ ed il periodo precostantiniano; monete di bronzo del corso del secolo terzo, prevalentemente del tipo degli antoniniani, che da Gallieno, Quieto, Claudio il Gotico giungono sino a Probo, e tra cui non mancano tipi con caratteri barbarici. Sotto il mosaico del Circo, assieme ad un antoniniano purtroppo non leggibile, si è raccolto un medio bronzo piuttosto sciupato di Treboniano Gallo (251-254) » «Nello scavo superiore alle pavimentazioni musive e sopra i livelli antichi dei cortili si sono invece incontrati bronzi di Massenzio e dei secondi Flavi. Ma elemento importantissimo, e, mi pare, di chiaro significato cronologico, essendo una moneta evidentemente caduta inavvertitamente all'artigiano durante il corso stesso del lavoro, è l’antoniniano di Massimiano Erculio, recuperato sulla malta che cementava la lastra marmorea della soglia nella esedra sud-orientale del frigidarium. 11 tipo dell'imperatore Maximianus e nel verso Augusto in tenuta militare rivolto a destra con farazonium a ricevere da Giove nudo col manto sulla spalla sinistra, un globo sormontato dalla Vittoria, è certamente degli ultimi anni del secolo ur d.C.» « Questo documento irrefutabile puntualizza, in maniera assoluta, la cronologia dei mosaici

di Piazza Armerina...»

Datazione fondata sullo stile.

Nonostante l'enorme superficie occupata dai mosaici e nonostante la varietà dei temi e dei motivi che vi compaiono, il lavoro, come è stato anche accertato da J. Lavin e dal Gentili, dà un'impressione relativamente unitaria. Le differenze stilistiche non mi paiono più grandi di quelle, per esempio, che si trovano nel quasi contemporaneo Arco di Costantino tra i fregi storici ed i rilievi dei basamenti, tutti e due lavorati negli stessi anni, fra il 312 ed il 315 (). Secondo (1) H.P. L'Orange -- A. von Gerkan, Der Spatantike trambe le opere. Lo stile ¢ volumetrico » sviluppanBildschmuch des Konstantinsbogens, Berlin. 1939, pp. dosi prima di tutto nel nudo, caratterizza la forma 192 5gB., 29 588.,34 568.Il Pace vede profondi contrasti degli animali, dei giganti, di Ercole, ecc, mentre stilistici tra i differenti gruppi dei mosaici e conclude si fa meno sentire nei cacciatori vestiti della « Grande che questi devono essere stati creati nel corso di al- Caccia +. Le differenze dipendono molto dal soggetto e cune generazioni (Pace, I mosaici, op. cit, pp. 93 dalle tradizioni tematiche (si confrontino lo differenze sg). Com'è naturale, nell'enorme lavoro si distin- stilistiche nei rilievi costantiniani sull’Arco di Coguono artisti e correnti ben differenti. Ma, nonostante stantino: L'Orange - von Gerkan, loc. cit). Nonoqueste differenze, mi pare che anche gli stessi mosaici stante tali differenze, come le vediamo al massimo che al Pace sembrano differire il più tra di loro, cioè grado p. es. tra la «Grande Caccia» e la giganto«la Strage» o la « Grande Caccia», siano congiunti machia (che secondo il Pace appartiene al gruppo da una così profonda affinità che l'appartenenza allo della «Strage 1), lo stretto nesso tra le duo opere si stesso tempo, e forse alla stessa bottega, non possa rivela in certi tratti del disegno del paesaggio: cio’ venir messa in dubbio. Se si confrontano il toro della in certi rami e piante isolati e di forma specialissima, « Strage» (Pace, I mosaici,op. cit., Fig. 11) col rino- con i quali sono riempiti gli intervalli tra le figure ceronte della «Grande Caccia» (op. cit. Fig. 24), di (cfr. Pace, op. cit., Fig. II e Figg. 10 e 27). I mosaici più i cavalli della «Strage» (op.cit., Fig. VII) con delle duo ripartizioni cerimoniali del complesso del quelli della «Grande Caccia» (op. cit, Fig. X), il modo palazzo appartengono, così per il tema come per la di contornare, ombrare e modellare la forma pare forma, ad una unità ὁ si sviluppano, come vedremo poco diferente, ed i «frazionamenti geometrici», in seguito, secondo un unico grande piano logico: è la composizione «a settori», «l'espansione volume. dunque poco probabile che si tratti di un lavoro trica in avanti » delle forme— per usare le parole con continuato nel corso di alcune generazioni. La tesi Je quali il Pace descrive lo stile della « Strage » - mi del Pace (op: cit., pp. 95 sg.) che la vastità dell'opera. sembra si manifestino in modo molto simile in en- deve aver richiesto un lungo periodo di attuazione 67

H. P. ORANGE

me, tali differenze stilistiche, così nell'arco di Costantino come ἃ Piazza Armerina, nascono da differenze nelle correnti stilistiche e dai temi iconografici e non da differenze d'età (1). Certe peculiarità specialissime delle forme e dello stile si ripetono nelle differenti parti dell'estesissima decorazione musiva e mostrano chiaramente che, se il palazzo veramente, come lo vede Giuseppe Lugli, è stato costruito in più riprese, queste, almeno nei due corpi cerimoniali qui considerati, devono essere state vicinissime e subordinate allo stesso piano decorativo. Così, per esempio, le stesse ghirlande a corone di lauro con orli «a dente di sega » (pp. 72 sgg) o «a millepiede » (pp. 73 sg.), come anche il singolarissimo elefante con c pelle areticolo » (pp. 75 sgg.) si ritrovano nelle differenti parti della decorazione musiva del palazzo, così nella galleria della Grande Caccia come nei portici del grande atrio, che, secondo Lugli, sarebbero stati costruiti in due diversi tempi. La totalità dei mosaici, se soltanto se ne eccettuano delle innovazioni relativamente limitate, come i mosaici della « sala delle dieci ragazze », trova a mio modo di vedere, il suo posto ~ tanto per lo stile quanto per i motivi e per i temi — nella grande arte dell’epoca intorno al 300 d. C. (ἢ. La stessa datazione viene mantenuta da E. Dyggve sulla base di considerazioni di stile relativamente all'architettura e al suo ornamento (). B. Neutsch ha confrontato la decorazione architettonica di Piazza Armerina con quella del palazzo di Diocleziano a Spalato e delle Terme di Diocleziano a Roma (1). E. H. Kahler ha osservato — come anche il Lugli (pp. 69 sg.) -- che nel grande atrio si trovano capitelli appartenenti ad una produzione di serie, la quale è anche rappresentata tra i capitelli del palazzo a Spalato (). La datazione del palazzo supposta dal Lugli si basa sostanzialmente sulla muratura, che è, nel suo complesso, unitaria, ma ha un accentuato carattere locale e provinciale e non consente perciò alcun confronto con le forme di data certa, caratteristiche della ecnica e del materiale dei centri maggiori. Su tale base non è possibile pensare ad altro che ad una datazione relativa. Ciò che si può accertare in base al detto materiale è soltanto l'alterna sequenza delle successive riprese edilizie, e non la loro collocazione in un determinato periodo. mon regge, poiché noi non sappiamo quanti fossero gli artisti o gli artigiari al lavoro (cfr. Gentili, Mosaici figurati, op. cit. p. 1). (1) Essenzialmente della stessa opinione è I. Lavin,

The hunting mosaics of Antioch and their sour ces, «Dumbarton Oaks Papers» 17, 1964, PP. 244 568., specialmente Nota 284. Ὁ) Secondo il Pice, però, i mosaici di Piazza Armerina devono essere datati al «corso inoltrato ~ © non mai al principio ~ del sso. rv e agli inizi del v», mentro $ mosaici delle « Ragazzo © delle terme sono ancora posteriori (op. cit, p. 105). I confronti fatti dal Paco con altre opere d'arte del mondo tardo-antico (con i mosaici del palazzo imperiale a Costantino. poli, con i mosaici di Antiochia, di Treviri, di Leptis Magna, di Aquileia, con i mosaici parietali cristiani, con riliovi, dittci, miniature © monete (op. cit, pp. 98 sgg) per fissare questa cronologia si limitano alla considerazione di generalità stilistiche di scarsa rilevanza per una datazione esatta. Nei mosaici dello terme, che secondo il Gentili sono tra i più tardi mosaici di Piazza Armerina, appaiono figure femminili con l'acconciatura tipicamente tetrarchica (si veda oltre p. 80 sg.). Le considerazioni del Pace sullo stile

dell'architettura e dell'ornamento (op. cit, pp. 32 gg.) non comportano più che una datazione generica in epoca tardo-antica,

(3) E. Dyggve, «Symb. Osl.» 29, 1952, pp. 122 sgg. (4) B. Neutsch, op. cit, col. 567. (5) H. Kahler, Zur Datierung des Sarkophags von Manastirine im archéologischen Museum von. Split, «Mullus, Festschrift Th. Klausers, München Westfalen 1964, S. 175: « Wie es Gamals auch anderen Orts üblich war, wurden Dinge, dio serienmássig hergestellt werden Zonnten, wie etwa Stulenschafte oder Kapitelle, bereits in unmittelbarer Nahe der Stein riche ausgeführt. So stammen etwa die τοῦ Kapidelle der Süulenstrassen (nel palazzo di Diocleziano a Spalato), die die beiden Nebentore im Osten und ‘Westen miteinander verbindan und die von der Porta Aurea zum grossen Peristyl im eigentlichen Wohntralet führen (val. den Plan bei Niemann 2.0.4 Abb. 4.), nach dem, was von ihnen erhalten blieb (H. Kahler, Die Gebälke des Konstantinsbogens II 1 (1953) 34). aus den prokonnesischen Brichen, die gleichzeitig auch die Kaiservilla von Piazza Armerina auf Sizilien beliefert haben: (vgl z.B. das Kapitell: NotScav. 1950, 314 Abb. 15).

PIAZZA ARMERINA

Il Lugli fonda la cronologia assoluta in parte sulla tecnica e sullo stile architettonico ed in parte su di un materiale che propriamente non appartiene alla storia edilizia, cioè sui mosaici. Così, in base allo stile architettonico, Lugli data l'atrio ovoidale e la sala triloba al tardo secolo v (). Ma questa datazione non è in alcun modo vincolante, tanto che Dyggve ha potuto richiamarsi a costruzioni parallele del 300 circa (+). Le copertine a volta, alleggerite da tubi fittili innestati tra loro, ricordate da Lugli, non impegnano del pari ad una datazione posteriore, ma costituiscono, come osserva lo stesso Lugli, un sistema particolarmente in uso nell'età di Massenzio e Costantino ()). In rapporto alla datazione effettuata in base ai mosaici, Lugli segue Pace nel distinguere vari tipi stilistici nell'arte del mosaico e li ripartisce tra diverse generazioni. Così Lugli (1) -dai mosaici e dalla cronologia relativa della muratura -- data il complesso delle terme alla fine del m secolo; il palazzo di abitazione (cioè il grande atrio con le stanze adiacenti, eccettuata la stanza 30) e l'ingresso monumentale al tempo tra il 350 ed il 370. Ci sorprende il fatto che Lugli, il quale non riscontra alcuna differenza tra il tardo mosaico di Bonifatius, verosimilmente medioevale (*), ed il tappeto-mosaico nel portico del grande atrio nel quale il mosaico di Bonifatius è inserito (e che lo stesso Lugli data al periodo 300-330) (ἢ, possa poi far tanto conto delle sfumature che distinguono per es. la «Strage» dalla «Grande Caccia» così da distaccarle di un mezzo secolo.

Se si considerano — come fanno Gentili, Lavin ed io stesso con loro - i mosaici di Piazza

Armerina come un’opera unitaria tanto sotto l'aspetto stilistico quanto sotto l'aspetto iconogra-

fico, le varie riprese edilizie supposte dal Lugli dovrebbero essere strettamente congiunte. Ciò va detto in ispecie per i due complessi da cerimonia, cioè il vestibolo col mosaico dello adventus, il grande atrio con le protomi ferine, il corridoio con la « Grande Caccia», l'atrio ovoidale con le protomi ferine emergenti da tralci e Vaula tricora con i mosaici della « Strage», della gigantomachia, dell'apoteosi di Ercole e della scena di Licurgo ed Ambrosia ('). Il Lugli osserva (*) che le tre scale che dal corridoio della «Grande Caccia» portano al «Grande Atrio» demoliscono un tratto del mosaico dell'atrio stesso e si domanda: «sono contemporanei oppure esiste una successione di tempi? ». Lo stesso rapporto si ripete esattamente per la scala che congiunge l’aula tricora con l'atrio ovoidale. Qui, dunque, si ripresenta la stessa questione (?). Secondo me, soltanto i mosaici

possono dare una risposta a questa domanda, e, poiché essi sono di stile identico in entrambi i

casi, dobbiamo necessariamente concludere che sono contemporanei. Nell'opera di costruzione si riscontrano dunque, nelle singole sezioni, delle impreviste modificazioni ed improvvisazioni le quali attestano che si lavorò affrettatamente per la sollecitazione di una pronta conclusione del lavoro: una circostanza, questa, che appunto corrisponde alla condotta imposta da un ordine imperiale. Niente ci vieta di credere che delle vecchie costruzioni siano state inserite nel nuovo palazzo, con modificazioni più o meno radicali di muri e di decorazioni. Un caso parallelo si ritrova nella Villa Adriana.

Un buon fondamento per una datazione assoluta è dato dai capitelli e ci sorprende che il Lugli nella sua storia edilizia del palazzo non si valga del detto materiale, come già avevano fatto il Dyggve, il Kahler ed il Neutsch. Il tipo di capitello che predomina nel palazzo di Piazza Armerina è di una ben nota forma tetrarchica standardizzata, come lo stesso Lugli ha notato

(@) (@) (3) (4) (5)

Lugli, op. cit; pp. 59 see Dyegve, op. cit, pp. 125 sg. Lug op. cit, p. 30. Op. cit, pp. 78 sex. Op. cit, Fig. 24; Gentili La villa imperiale, op. cit, p. 29.

(6) (7) (8) (0)

Lugli, op. cit, p. 47. V. sopra, p. 67, nota 1. Op. cit, p. 49. Op. cit, p. ór.

H. P. ORANGE

«La maggior parte dei capitelli », dice Lugli (), «rientra in quel tipo in serie che è stato riscontrato anche a Spalato». Noi troviamo capitelli di questo tipo (Tav. XLIII a-c) sparsi nelle di verse parti del palazzo di Piazza Armerina, nella galleria della « Grande Caccia», nel grande atrio, nelle terme, Oltre che a Piazza Armerina e nei dintorni del palazzo di Diocleziano a Spalato (Tav. XLIII A) (ἢ, ci è conservata una quantità di capitelli della stessa serie di produzione nel Museo Nazionale delle Terme di Roma: capitelli che senza dubbio provengono dalla costruzione delle Terme di Diocleziano (Tav. XLIII f, e). Capitelli del medesime tipo, che possono provenire anche da Piazza Armerina, si trovano altresì nel Museo di Siracusa (Tav. XLIII d, e). In tale tipo di capitelli, che presenta la forma piatta delle foglie di acanto, caratteristica della tarda antichità e con stilizzazione ornamentale dei contorni terminali, è notevole una figura che formalmente contrassegna i prodotti di questa serie come un marchio di fabbrica: i dentelli di foglie contigue compongono un disegno astratto, quasi geometrico, di rombi, rettangoli, triangoli, e segmenti d'arco, che si susseguono dall'alto in basso sempre in ordine costante. Caratteristica della serie è anche la lavorazione leggera, più o meno abbozzata delle helices. Non può essere puro caso il fatto che noi troviamo questo zipo di capitello nel palazzo di Diocleziano a Spalato, nelle sue terme di Roma e nel palazzo di Piazza Armerina. In tutti e tre i casi si tratta di fornimento in grande del tipo dominante di capitelli tetrarchici, che si deve supporre che venissero prodotti in massa per le grandi costruzioni edilizie del tempo ed esportati poi dal centro di produzione nelle province, I soli elementi della decorazione architettonica del palazzo di Piazza Armerina che consentono una cronologia assoluta si richiamano così univocamente alla tetrarchia. È anche importante il fatto che i capitelli si trovino sparsi nelle diverse parti del palazzo. Ciò attesta chiaramente l'esistenza di un piano unitario in tutto il complesso (). A lato dei capitelli di questo tipo tetrarchico predominante si trovano anche dei capitelli di forma differente, tra cui dei capitelli ionici, ma ron un solo di dichiarato tipo post-tetrarchico.

Mi riesce difficile seguire il nostro illustre collega quando afferma che la villa ebbe un carattere del tutto intimo e familiare (). Il piano di disposizione dei due grandi complessi edilizi principali è conforme alla sequenza cerimoniale dei palazzi imperiali ed i mosaici degli stessi ‘hanno un carattere eminentemente rappresentativo con un ricco e differenziatissimo sviluppo della

ideologia erculea (v. sotto pp. 96sgg.). Quando in una sala aggitinta al grande atrio vengono rap-

presentate delle persone individualizzate in scene di sacrificio e di banchetto, esse compaiono secondo lo schema solenne stabilito per gli imperatori sacrificanti e 2er i banchettanti eroici (p. 97, nota s). Nello stesso stile imperiale sono l'ingresso in forma di arco trionfale - come lo stesso Lugli lo caratterizza - ed il mosaico dell’adventus nel vestibolo che dal cortile dell'ingresso porta al grande atrio. Il Lugli riconosce che la grande aula basilicale con absidecoperta a volta, la quale rappresenta il culmine del più grande reparto cerimoniale, « richiama alla mente le basiliche dei palazzi imperiali: di Domiziano sul Palatino, di Adriano a Tivoli, dei Varii sul Laterano (palatium Sessorianum), di Diocleziano a Spalato, di Costantino a Treviri ed a Costantinopoli » (5). Se tali basiliche esistono anche in palazzi privati, non hanno, certo, una

espressione che si accordi con un carattere intimo e familiare. Limitiamoci ai mosaici. Le singole figure, i gruppi di figure e le intere composizioni — per

(1) Op. cit., p. 46. (2) R. Kautzsch, Kapillistudien, Berlin 1936, Tav. I nr. 2 e 7; II nr. 10.

tutto il complesso edilizio e tutti i singoli elementi della costruzione, per esempio i capitelli, le basi, i fusti delle colonne, la trabeazione, ecc.

(3) Ci manca, come anche il Lugli osserva (op. — (4 Op. cit, pp. 77 58. © 52. it, p.39), una pubblicazione completa che illustri (6) Op. cit, PP. 39 986. 7o

PIAZZA ARMERINA

esempio nei due mosaici della caccia — si sono affrancati dalle norme classiche del ritmo e delle proporzioni. Troviamo qui un modo di esprimersi semplice, rozzo e duro, una forma direttadi nar-

razione, contraria a tutta la fraseologia tradizionale. La forza e l'espressività di ogni movimento

hanno un carattere spontaneo, genuino, spesso esagerato, Il linguaggio immediato, senza ricerca del bello, lo sviluppo della forma dall'azione e dall'oggetto senza l'osservanza di tutte le regole consuete corrisponde in quest'arte ad una violenta accentuazione dell'espressione, fino a contorsioni convulse del corpo e delle membra e a contrazioni e smorfie del volto (1). I giganti feriti della gigantomachia, il cavallo e il cavaliere caduti della « Strage di Ercole », certe teste contorte di cavallo nella « Strage », nelle « cacce» ecc. sono, a tale riguardo, tipici (Tav. Ig, 8). È un'arte

spontanea ed espressiva, drammatica e movimentata, che reagisce violentemente a tutti i canoni

della bellezza classica. Sovrabbondano i tratti aneddolici e-di genere, che tradiscono il diretto interesse reale dell'artista per il suo soggetto più che per la resa formale e insieme la sua gioia di raccontatore e la sua fantasia inventiva.

ΤΙ carattere artistico qui definito è peculiare della cosiddetta corrente popolare, che alla fine del ur e al principio del rv secolo, si impone come stile ufficiale in Roma e s'incontra in un gran numero di sarcofagi dell'epoca come anche nei fregi storici dell'Arco di Costantino (} (312-315). Da Roma questa corrente penetra nelle province confinanti e invade grandi aree del mondo oc-

cidentale, mentre si fa valere meno nella parte ellenistica dell'impero. Lo stile ha una ricca fioritura nei territori africani più vicini alla Sicilia, nelle province dell’Africa proconsularis, della Numidia e della Mauritania, cio? nella regione oggi corrispondente alla Tunisia e all’Algeria. Del tempo intorno al 300 d. C. in queste province si conserva un ricco tesoro di mosaici che, tanto nello stile quanto nel temà, sono molto affini ai mosaici di Piazza Armerina. Quadri di caccia

riccamente differenziati — anche con scene di sacrifici e di cattura di bestie mediante trappole —;

protomi di animali entro corone di alloro e protomi di animali che sporgono dalle volute di girali; quadri marini con Nereidi, Tritoni e ogni specie di mostri del mare, con Arione sul suo delfino; Orfeo musico tra gli animali; quadri di circo con amorini o veri aurighi; amorini e putti in scene d'acqua. e di caccia, di vendemmia, in girali d'acanto, tra uccelli e altri animali: tutto questo, espresso nelle forme tipiche dello stile popolare, si trova sia nei mosaici africani sia in quelli di Piazza Armerina. Il Gentili e numerosi studiosi di archeologia africana hanno notato questa stretta relazione tra i mosaici africani e quelli siciliani (*), Alla comunanza stilistica e tematica si aggiunge quella del materiale: i mosaici di Piazza Armerina, secondo il Gentili (ἢ), sono per una gran parte lavorati in un materiale di pietra africana. « La varia gamma dei colori, veramente sorprendente, era ricavata da nuclei di pietre calcaree, di provenienza prevalentemente africana, come può desumersi dal confronto con i materiali usati nei mosaici della Tripolitania, della Tunisia e dell’Algeria ed in qualche caso di non improbabile origine siciliana-- forse pietra di Trapani ». Nelle pagine seguenti prenderemo in esame certi tratti specialissimi di motivo e di stile che, (1) L'Orange, Studien sur Geschichte des spütamthen Ports, Oslo 1933, PP. 31, 39, 44. (2) G. Rodenwaldt, Eine spatantike Kunststrd‘mung in Rom, « Rom. Mitt. + 36-37, 1931-32; L'Orange - von Gerkan, op. cit, Tavv. 6-21, Textband Figg. 58, 61; E. Gerke, Die christ, Sarhophage der vor — Ronstantin. Zeit, Tavv. 39, 3; 9, τ; R. Bianchi — Bandinelli, Continuità Ellenistica nella Pittura di Età medio © tardo-romana, « Rivista dell'Istituto Naziomale d'Archeología e Storia dell'Arte», Nuova Serie, 2, 1953 (1954), pp. 46 seg.

(3) Così il Gentili (Mosaici figurati, op. cit., pp. rx; 34 568) rileva la somiglianza tra le cacce di Piazza Armerina ola caccia alla lepre di Oudna e di El-Djem (fg. 23) © sottolinea (o. cit, pp. 38 sg) le speciali relazioni esistenti tra i mosaici con soggetti marini nei due gruppi, per l'identica resa formale della fauna che affolla il mare e «per quella resa convenzionale, eppure così efficace, delle onde mediante tratti oriz. zontali, di tonalità grigio-verde, talora con flettatura nera, continui o spezzati a zig zag. (4) Gentili, op. cit, p. rr. n

H. P. ORANGE secondo noi, ancorano stabilmente i mosaici di Piazza Armerina sul fondo dello stile popolare romano degli anni intorno al 300 d. C. e che legano tra loro in maniera indissoltbile Piazza Armerina e il nostro gruppo di mosaici nordafricani. Ghirlande d'alloro alloro con orli

con orli «a dente «a millepiedis.

di sega»

e ghirlande

di

Una singolarissima forma dell'antica ghirlanda d'alloro che, indubbiamente, si ricollega a ‘una particolare scuola d’artisti, s'incontra di confinuo a Piazza Armerina e nel nostro gruppo di mosaici nordafricani degli anni attorno al 300 d.C. Cerchiamo di definire le particolarità di questa forma di ghirlanda collocando, ad esempio, il mosaico pavimentale con medaglioni a protomi nel grande atrio di Piazza Armerina (Tav. II) a lato d'un affine mosaico pavimentale del Museo del Bardo, anche questo con medaglioni a protomi (Tav. III). Il mosaico proviene dal tepidario delle terme di una villa riccamente decorata con mosaici a Thuburdo Maius (Tunisia), che mostra un gran numero di punti di contatto con Piazza Armerina e sulla quale, perciò, ritorneremo in seguito (1). Insieme con l’altra decorazione musiva della villa il suddetto mosaico con medaglioni a protomi viene, nel catalogo del Museo del Bardo, datato al mr secolo e, ultimamente, da L, Poinssot e P. Quoniam agli anni intorno al 300 d. C. (). Qui, come nel mosaico dell'atrio a Piazza Armerina, il tema consiste in protomi delle bestie dell’arena incorniciate da corone d'alloro. Mentre, perd, le protomi a Piazza Armerina sono limitate alla testa, a Thuburbo abbracciano tutto l'avancorpo. Una lampante somiglianza che si estende fino ai particolari più minuti della forma si ritrova invece tra le corone d'alloro: Già lo specialissimo cambiamento cromatico da una corona all'altra è identico: a Piazza Armerina si alternano corone giallo-brune e corone verdi così come a Thuburbo corone tonde giallo-brune e corone esagonali verdi (*). Ogni singola foglia di alloro è modellata nello stesso modo a Thuburbo e a Piazza Armerina: la parte inferiore, verso il picciolo, è fatta di tessere oscure così da sembrare ombrate, mentre l'estremità superiore che sporge nella luce è intessuta di tessere chiare. Per la intera lunghezza della corona è ripetuto in modo del tutto regolare sempre 10 stesso modello costituito da foglie del tipo descritto: le foglie sono disposte a tre a tre per quasi tutta la larghezza

della corona quasi a formare la figura di un fiore di giglio e tra fiore e fiore spunta da ambo i lati una piccola foglia dello stesso tipo. Questo modello di foglie emerge su di un fondo scuro omogeneo, il quale è formato dagli strati più profondi delle foglie e si estende fino agli orli esterno e interno della corona disegnandovi un assai caratteristico contorno a denti di sega. Una stranissima variazione di questa ghirlanda di alloro a denti di sega s'incontra a Piazza Armerina come a Thuburbo ed inoltre in tutto il gruppo dei mosaici nordafricani, che stiamo considerando, e lega questi e quelli di Piazza Armerina ancor più strettamenze tra loro. Riproduciamo come esempio un mosaico pavimentale del Museo del Bardo (Tav. IV o) il quale, per tutta la sua impronta stilistica, è strettamente connesso ai mosaici considerati di Piazza Armerina e (1) Senza approfondire i confronti, già il Gentili Thuburbo Majus, « Karthago s, IV, 1953, PP. 155 sgg: ha richiamato l'attenzione sulla parentela tra i mo- — Bétes d'amphitféatre sur trois mosaïques du Bardo, saici delle due ville (op. cit, pp. 40 sg, p.20 notai). Karthago », III, 1952, pp. 127 SE. (2) A. Merlin -L. Poinssot, Guide du Musée Alaowi, ὀ (3) Tavole a colori: Gentili, op. cit, Tav. XV; Musée du Bardo, Tunisi, 1950, 4. ed. a cura di P. Pace, op. cit, Figg. VIII, IX. Cir. la descrizione dei Quoniam, Tav. IVe pp. 71 sgg.;Inv.2757. L. Poinssot— colori in « Kartago 1, IV, 1653, P. 157; senza colori, P. Quoniam, Mosaiques des Bains des Protomés à Pace, Figg. 12-17. 5

PIAZZA ARMERINA

di Thuburbo e, come questi, presenta un riempimento regolare di tutto il campo mediante medaglioni di alloro: questi medaglioni, però, non racchiudono protomi, bensì figure intere di animali. I medaglioni sono tutti inscritti nella nostra ghirlanda d'alloro con orli a dente di sega e la ghir-

landa, tanto qui quanto a Piazza Armerina, è fasciata da una banda. Tutto il campo occupato dai medaglioni è incorniciato da una ghirlanda di alloro, che l'inserimento di fiori, frutti e maschere rende più larga e più fastosa delle ghirlande o dei medaglioni, pur mantenendo sostanzialmente lo stesso modello di foglie. Una particolarissima variante è, però, introdotta lungo gli orli della ghirlanda: invece di denti a sega corre qui una fitta successione di sottilissime strisce, simili a zampe di millepiedi. Come esempio di questa ghirlanda di alloro con orli a millepiedi di Piazza Armerina, riproduciamo due dettagli del mosaico del cosidetto « Cubicolo con scena erotica» (Tav. IV a, D). Anche qui la variazione del modello a millepiedi si accompagna al modello più generale a denti di sega. Si trovano anche forme intermedie tra quella a denti di sega e quella ἃ millepiedi e le due forme possono essere anche combinate in una medesima ghirlanda, come si vede nel cornicione esterno che circonda il mosaico con amorini vendemmianti nella nostra casa a Thuburbo (1). Nel cornicione, che all'interno circonda il campo centrale, l'orlo a millepiedi domina da solo. Queste due forme della ghirlanda di alloro si incontrano a Piazza Armerina — oltre che nei medaglioni del grande atrio () e nel « Cubicolo con scena erotica» () — anche, per esempio, in una corona nella cosiddetta « Edicola del peristilio » (1), e ancora in una ghirlanda che corre a modo di cornicione attorno a tutto il quadro, non diversamente che nel mosaico del Museo del Bardo (Tav. IV c), nella cosiddetta «Diaetadi Orfeo» (*), così pure in una ghirlanda che serve come decorazione festosa delle navi che portano le bestie a Roma nella «Grande Caccia» (Tav. Ve) (* Diamo ora in appresso qualche altro esempio delle due varianti, in aggiunta a quelli già dati, della fine del 11 o del primo 1v secolo nei mosaici di Tunisia e di Algeria. 1, Dalla casa di Thuburbo proviene, oltre al mosaico con medaglioni a protomi, trasferito come questo al Museo del Bardo, anche un mosaico con quadri di pesci, granchi, calamari, uccelli, fiori, cesti con frutta, collocati in campi esagonali e incorniciati dalla nostra ghirlanda di alloro con orli a dente di sega (Tav. V δ). Come quello a medaglioni anche questo mosaico viene datato intorno all'anno

300 (ἢ. 2. Un mosaico di Susa (Hadrumetum), corteo di Nettuno, attualmente nel Museo del Bardo (Tav. VI), mostra una leggera variazione della nostra ghirlanda, sempre però con orli a denti di sega e, secondo la descrizione, con corone alternativamente verdi e giallo-rosse; per lo stile delle figure il mosaico difficilmente può essere datato a epoca posteriore all'anno 300 d.C. circa (9. 3. Nello « Apollo-Melkarth », mosaico a Hippo Regius (Tav. V a), le figure sono inserite in corone di alloro con orli a denti di sega; una delle figure, una donna che si fascia il petto e la cui testa è ornata da una corona, ha l'acconciatura di forma tetrarchica simile al paranuca di un elmo 6) (p. 80). 4. 11 mosaico del pavone, ora nel Museo del Bardo, ha la stessa ghirlanda di (x) L. Poinssot -- P. Quoniam, e Karthago +, IV,

me-tavole II n. 125; R. de la Blanchère, Collections

(6) Ibid, Fig. 21, p. 73e Mosaici figurati, Tav. XXV

(ὁ) Gauckler, Inventaire des Mosaiques, volume-

1953 fig. 8. (2) Gentili, La 12, p. 64. (3) Ibid., Fig. (4) Ibid., Fig. (5) Ibid., Fig.

du Musée Alaowi, pp. 15 sgg. ὁ Tav. x con ricostruVilla imperiale, op. it, Figg. Xr e zione a colori; A. Merlin - L. Poinssot, Guide du Musée Alaoui, p. 34 € Tav. X; L. Foucher, Inventaire 23, p. 74. des Mosaiques, Sousse, «Inst. National d'Archéol. 4, p. 60. et Arts», Tunisi 1960, Tavv. XXVIII; XXIX 57, 119 29, p. 79. e pp. 56 seg.

(7) Poinssot-Quoniam, op. cit, «Karthagos, IV, tavole ΠῚ n. 41; Ist. Archeol Germ., Roma, Inst. fig. 3. Neg. 61,539. In seguito le negative del dotto istituto (8) P. Gauckler, Inventaive des Mosaigues, volu- | saranno citati: Inst. Neg. n

73

H. P. L'ORANGE

alloro; viene datato al principio del rv secolo (!). 5. In un mosaico da ΕἸ Djem, forse nel Museo del Bardo, la nostra ghirlanda di alloro riempie tutto il cornicione che circonda ventiquattro campi quadrati con quadri di uccelli, pesci ed altri animali e in uno di essi un uomo che giuoca a dadi; il mosaico viene datato al 11 secolo dopo Cristo (?). 6. Nel mosaico delle Muse a Hippo Regius la nostra ghirlanda riempie un sistema labirintico di cornicioni geometrici che inquadrano i busti delle Muse; il mosaico fa parte di uno strato che sta sotto il mosaico della caccia (Tav. XIV) e dunque è anteriore a quest'ultimo, però, probabilmente, soltanto di pochi anni (). 7. Un mosaico nel Museo del Bardo ( (Tav. VIT a) con un quadro marino ha un cornicione ἃ forma di ferro di cavallo riempito dalla nostra ghirlanda di alloro, questa volta con orli a millepiedi (*); stilisticamente il mosaico non può essere staccato dal gruppo di opere qui trattate. Particolari forme burbo Maius.

di

acanto

a

Piazza

Armerina

e

a

Thu-

Mentre le protomi dei medaglioni a Piazza Armerina sono limitate alla sola testa dell’animale, le protomi nei girali di acanto, sempre a Piazza Armerina, presentano tutto Yavancorpo dell'animale proprio come si incontra a Thuburbo Maius. Come queste protomi africane, le protomi di acanto di Piazza Armerina anzidette, hanno in comune anche la parte terminale a forma di corolla di fiore, dalla quale l’animale sporge come da una cornucopia. Confrontiamo una protome di cavallo in girali di acanto a Piazza Armerina (Tav. VIIZ 5) con le protomi di cavallo e di leone nel mosaico a medaglioni a Thuburbo (Tav. VIII d). Le particolarissime corolle con le foglie ἃ forma di lingua devono essere confrontate nei due casi, in modo dettagliato; a Piazza Armerina la corolla, la quale, meglio che nella nostra Tav. ὙΠῚ ὁ, si vede nella riproduzione colorata, Gentili, Mosaici figurati, Tav. XIV b, sporge dal girale d'acanto, mentre a Thuburbo la corolla è senza nesso organico con un qualche vegetale e si chiude come un collo o una cintura attorno al corpo dell'animale. Ma anche questo strano fiore-collo staccaio si ritrova. tra i girali d’acanto ἃ Piazza Armerina, per esempio nel protome di uno struzzo (Tav. VIII c); staccato anche dallo animale, lo si vede sotto gli uccelli nella Tav. VIII a. Che nel caso del nostro fiore-collo si tratti veramente della corolla tradizionale che sporge dalle volute d’acanto, si ricava dalle protomi d'acanto in mosaici a Oudna (Uthina) e Kourba (Curubis) (9), cke dal Poinssot e dal Quoniam (7) sono stati confrontati con le protomi ferine del mosaico a medag'ioni di Thuburbo (). Entrambi i mosaici di Oudna e di Kourba vengono datati all'ultimo nr o al primo rv secolo. Il Gentili estende il confronto anche ai mosaici di Piazza Armerina (1). (6) Poinssot — Quoniam, op. cit, « Karthago », () A. Merlin - L. Poinssot, Deux mosaïques de Tunisie, «Mon, Piots, XXXIV, 1934, PP. 145, 154 IV, 1953, Figg. 6 e 7: la figuraὃ stata capovolta. Pro(2) P. Ch. Picard, « Fasti Arch.» 5, 1050, 1.3577, ‘tomi animali che sporgono dai fiori di girali d'acanto sono, com'è noto, un motivo tradizionale nell'arte elleFigg. 77 e 78. () E. Mareo, Une nouvelle mosaïque des Muses nistico-romana: J. M. C. Toynbee — J. B. Ward Pera Rippone, Libyca (Bull. du Serv. des Antiq. arch. Kins, People Scrolls, « Papers of the British School épigr.) 6, 1958, pp. 123 sgg. Figg. 2-10. at Rome, 18, 1950, pp. I Figg., Tavv. 1-26. (Ἱ Ibid., «Karthagos, IV, 1953, p. 160. (4) Inst. Neg. 61,539. (8) P. Gauckler, « Mon. Pot, ILI, 1896, pp. 177 (5) Altri esempi del Museo del Bardo di mosaici con ghirlanda di alloro a denti di sega: Inst. Neg. sgg.; idem Inventaire IL, pp. 126 sg. n. 374, 165, 497. 61,537: 61,565; L. Foucher, Inven- Du Coudray la Blanchère — P. Gauckler, Cat. Musée 61,539;op.61,511; taire, cit, Tav. XLVI 57, 204, Sousse, La ghir- Alaoui, 1897, Tavv. 23 A. 101, 28A. 140; M. Bernard, landa arricchita di frutti e maschere con orli a mille. «Bull. archéol. du Comité» 1906, Tavv. 1, 2. (0) Mosaici figurati, op. cit, p. 40© p. 24, nota 2. piedi: Inst. Neg. 61, 528. 74

PIAZZA ARMERINA

Nelle terme della summenzionata villa a Thuburbo Maius, a lato del tepidario con i mosaici di medaglioni a protomi ferine, si trova il frigidario decorato con un grande mosaico pavimentale contemporaneo a quello del tepidario. Come quest’ultimo, anche il mosaico del frigidario mostra, così nello stile come nel tema, affinità con i mosaici di Piazza Armerina e viene datato all’epoca tetrarchica (1). Attorno al campo centrale di forma quasi quadrata con scena di amorini vendemmianti si vedono lungo ogni lato strette sezioni con animali caccianti. Attorno a questo campo centrale si trova poi una larga zona occupata da un ricco quadro di acanto (). La zona, sia verso l'interno che verso l'esterno, è chiusa entro un cornicione quasi quadrato, riempito da una ghirlanda d'alloro con orli che presentano ora il modello a denti di sega, ora il modello a millepiedi, Come questa ghirlanda, così anche l'acanto (Tav. VII c) costituisce un elemento in comune con i mosaici di Piazza Armerina (Tav. VII 5). Tratti particolari, che s'incontrano nello acanto in entrambi i mosaici, sono gli steli coperti da una successione di foglie ad orli frastagliati, L'elefante

con

pelle

a reticolo.

Una grande somiglianza generica esiste tra le immagini ferine di Piazza Armerina e quelle

del nostro gruppo di mosaici nordafricani. È tipico di queste immagini, in entrambi i casi, un vivo naturalismo espressivo con un forte senso del carattere e della ferocia propri delle singole razze ferine, Spesso l'individualizzazione delle belve produce nei due gruppi tipi assai affini, per

esempio nei leopardi e nei leoni della « Grande Caccia » di Piazza Armerina e di una caccia di Hippo Regius (Tav. XIV a, δ) ο nei cani della « Piccola Caccia » di Piazza Armerina (Tav. XII a) e di una caccia in un mosaico di El Djem (Tav. XIITa). Una somiglianza del tutto particolare esiste tra gli elefanti stranissimi dei mosaici di Piazza

Armerina e la rappresentazione dell'elefante nell'arte romana e nordafricana della fine del x se-

colo e del principio del v (*). Il tratto speciale nell'elefante di Piazza Armerina (Tav. IX a-d) è la stilizzazione della pelle rugosa in forma di una rete o di un reticolo a rombi. Questa rete si può adattare in modo più libero così da aderire alle vive e vere forme naturali, o può essere eseguita in maniera più regolare cosicché un reticolo geometricamente perfetto si astrae dalle forme naturali e si estende a tutta la superficie del corpo dell'animale. Nei mosaici del lungo nartece (Pianta n. 4) si trovano due esemplari di questo tipo di elefante con pelle ἃ reticolo (Tav. IX a, δὴ -- una volta a forma naturalistica, un'altra volta a forma astratta — nel quadriportico del grande atrio (Pianta n. 3) si trova almeno un esemplare (Tav. IX c) e nella «Diaeta» di Orfeo (Pianta n. 6) che si apre sul detto quadriportico, ancora un esemplare (Tav. IX d), tutti e due ἃ forma naturalistica. Numerosi esempi dell'elefante con pelle a reticolo si trovano nell'ambiente artistico della capitale dall'età di Gallieno fino al tempo di Costantino. Qui possiamo soltanto dare degli esempi 1. Exuviaé di elefante che coronano la testa di una personificazione dell’Africa sul «Sarcofago di Annona» nel Museo Nazionale delle Terme (Tav. XI a, 8). I ritratti consentono di datare il sarco(1) L.Poinssot- P. Quoniam, op. cit, 4 Karthago » IV, 1053, pp. 164 sgg.e Figg. 8-10: «Sans nul doute contemporaines, les deux mosaïques des Bains des Protomès peuvent être attribuées, pensons nous, à l'époque de la Tetrarchie». AncheG. Ch. Picard data il mosaico alla fine del mr secolo, Mosaiques Afri-

caines du III* siècle ap. J. C., «Rev. Archeol.» 1960, IL pp. 17 888. P. 35. (a) Poinssot- Quoniam, op. cit, Fig. 8. (8) L'Orange, « Symb. Odl. » op. cit, pp. 116 sg: Β. Neutsch, op. cit, col. 578, 171; Gentili, Mosaici figurati, op. cit, Fig. 59. 75

H. P. ORANGE

fago con la maggiore precisione agli anni 260-270 (1). 2. Elefanti ai lati di un sarcofago di tipo ovale nel.chiostro di San Paolo fuori le Mura (Tav. X a, 6). Per lo stile il sarcofago difficilmente può essere posteriore all'anno 300 d. C. circa. 3, Quadriga di elefanti su un coperchio di sarcofago nel chiostro di San Lorenzo fuori le Mura (Tav. X c). Il sarcofago, stilisticamente, fa gruppo con i fregi storici dell'Arco di Costantino (312-315 d.C.) (#). 4. Due elefanti circondanti un medaglione con ritratto femminile su un frammento di un coperchio di sarcofago a Ostia, Castello (Tav. Xd). L’acconciatura del ritratto consente una datazione all'epoca di Costantino. 5. Elefante in una scena di venatio su un rilievo di sarcofago incastrato nel muro di cinta di Villa Borghese, angolo Giardino Zoologico, Via Pietro Raimondi (*). Secondo quanto ho potuto vedere, lo stile suggerisce una datazione nei decenni intorno al 300 d. C. 6. Elefante in un frammento di sarcofago del III secolo, incastrato nel muro del chiostro di San Paolo fuori le Mura. 7. Elefante in un mosaico con rappresentazioni di animali trovato sull’Aventino e attualmente nel Vaticano, «inserted in the walls of the corner rooms opening out from the octagonal court », di elefanti M. E. Blake data il mosaico agli anni tra Gordiano I e Costantino (). 8. Quadriga (). Torlonia Museo del sarcofago d'un porto di su un arco trionfale sul celebre rilievo con scena Lo stile del rilievo non permette una datazione posteriore al 300 d. C. Al di fuori di quest'ambito strettamente romano, si trova un esempio dell'elefante con pelle a reticolo e cioè in un frammento di mosaico della villa dei Laberii a Oudna in Tunisia che si data intorno al 300 (Tav. XI c) (9. Qui si riscontra la forma più libera del reticolo, adattato alle forme naturali dell'animale. Che io sappia, non si trovano nell'arte contemporanea ellenisticoorientale nessun esempio di elefanti con questo particolarissimo tratto della pelle a reticolo. Così gli elefanti sull’arco di Galerio a Salonicco sono sempre rappresentati con pelle naturale (ἢ. Così l'elefante con pelle a reticolo è in special modo legata all'arte romana intorno al 300, anche se appare in altre epoche (*). Ci è conservato, come giustamente ha rilevato il Pace ()), un esempio di elefante con pelle a reticolo di epoca posteriore, cioè quello della quadriga di elefanti sul dittico dei Lampadi del v secolo (Tav. X e). Però il reticolo qui ha un carattere differente da quello degli elefanti del 300 circa: la rete è a maglie più piccole e dà l'impressione di una superficie squamata. Se si confrontano i due tipi con gli elefanti di Piazza Armerina, non vi è più dubbio che questi ultimi appartengono al gruppo del 300 circa. La

caccia

con

la trappola.

I quadri di caccia di Piazzà Armerina e nel nostro gruppo di mosaici nordafricani costituiscono una base che si presta particolarmente ai confronti. Mettiamo a paragone la « Grande » ela « Piccola Caccia » (Tav. XII a, 5) con mosaici simili a quelli della caccia alla lepre di Oudna (1) G. Rodenwalds, + Jahrb, des Deutschen Archiiolog. Instituts», 51, 1936, pp. τοῦ sgg. L'Africa con le exuviae d'un elefante sulla testa sulle monete: M. Bernhart, Handbuch zur Münshunde der rom. Kaiserseit, Tav. 74, 3-5, 8 © p. 104. (2) L'Orangovon Gerkan, op. cit., p. 224 e Fig.63. (3) Neutsch, op. cit, col. 578, n. 171. (4) B. Nogara, I mosaici antichi del Vaticano & del Laterano, Milano 1910, Tav. IX, Figg. 1-4; M. E. Blake, Mosaics of the Late Empire in Rome and Vicinities, Mem, American Acad. in Rome, 17, 1940, pp. 115 sg, Tav. 31, 1-4.

(5) E. Q. Visconti, Catalogue of the Torlonia Museum, Roma, 1883, n. 430. (6) F. Bernerd, Le cheva! dans les mosaïques de l'Afrique du Nord, «Bull. Archeol du Comité des Travaux Historiques » 1906, Tav. 1, 2. (1) K. F. Kinch, L'Are de triomphe de Salonigue, Paris 1890, Tav. VIII, cfr. Inst. Neg. 55. 740. (8) R. Bianchi Bandinelli, E. V. Caffarelli, G. Caputo, F. Clerici, Leptis Magna, Roma 1963, Tav. 56: elefanti con pelle a reticolo del rrr o del rv secolo. (9) Pace, op. cit., p. 104 76

PIAZZA ARMERINA

(Uthina) e di ΕἸ Djem (Thysdrus) (Tav. XIII a) e quello della caccia al cinghiale di Cartagine (Tav., XIII δ), che sono datati all'ultimo rrt o al primo 1v secolo (1). Tutte le particolarità sopra definite dello stile popolare (p. 71) caratterizzano del pari le cacce siciliane e quelle nordafricane. La descrizione del terreno e del paesaggio è -- dalla disposizione generale fino ai più minuti dettagli — la stessa. Uomini, animali e piante sono rappresentati secondo lo schema generale di « Hochstaffelung », cioè le singole figure, anziché essere collocate in prospettiva, sono sovrapposte Yuna sull'altra. In un modo del tutto particolare le singole piante, pietre e rocce sono presentate non come parti di un tutto continuo, ma staccate da questo complesso e collocate ciascuna a sé stante e spesso appoggiate su di una propria linea di base. È specialmente caratteristico che

dei piccoli arbusti e cespugli siano sparsi isolati per tutto il campo, come se spuntassero da una nuda arena. Allo stesso modo anche le forme delle pietre e delle rocce vengono sempre staccate dalla continuità del paesaggio e isolate su di una propria base. Similmente uomini e animali vengono collocati su una loro striscia o piano isolato d'ombra, che in casi estremi viene a formare come ‘una piattaforma (cfr. Tav. XII a e XIII a). Alle forme del paesaggio, del terreno e dell'orizzonte viene dato risalto con una larga striscia scura, come una forte pennellata pastosa. Alberi e cespugli leggeri stanno sul fondo, spuntando dalla larga striscia dell'orizzonte (*). Un rapporto di somiglianza del tutto speciale si riscontra — come hanno visto.il Picard e il Gentili — tra la « Grande Caccia» a Piazza Armerina e la caccia a Hippo Regius (Tav. XIV a, 2), che dobbiamo collocare qui per poter stabilire un confronto dettagliato (ἢ). Il mosaico a Hippo dal de Pachtère e dal Marec viene datato alla fine del m secolo, mentre il Picard ritiene che non sia anteriore alla fine del governo di Costantino. All'ala estrema a sinistra della grande caccia a Piazza Armerina si trova un drammatico quadro di trappola per le fiere (Tav. XV a-c). La trappola è posta in un avvallamento del terreno: attorno a questo avvallamento sta una fila di cacciatori, spalla a spalla, con grandi scudi i quali in parte si accavallano per costituire un unico schermo di difesa; con la mano destra i cacciatori impugnano due aste. Davanti ad essi sta una fitta successione di cespugli artificial. mente collocati a mo’ di siepe, che si deve pensare tanto alta da nascondere i cacciatori stessi. Sulla testa i cacciatori portano delle corone di foglie, verosimilmente per mascherarsi. A un certo punto la fila dei cacciatori e della siepe è interrotta per dare spazio a una gabbia aperta, di un tipo che si incontra continuamente nella « Grande Caccia » e che ora è collocata nell'area della caccia, ora viene trasportata, con dentro la bestia catturata, su un carro o su una nave. La gabbia ha un meccanismo di chiusura manovrato per mezzo di una manopola di forma trapezoidale disposta sul tetto della gabbia stessa: una capra scorticata, con il ventre squarciato, è appesa nell’interno della gabbia. Un cacciatore, che si deve pensare inginocchiato sul tetto della gabbia, (x) Inst. Neg. 61.514; L. Poinssot, « Revue Tunisienne», 1940, pp. 226 sg.; C. Ch. Picard, «Rev.

Arch», 1949, II, pp. 147 sgg.; P. Quoniam,+ Karthago», II, 1952, pp. 121 egg; Merlin — Poinssot, op. Sit, Tav. 16, p. 48; e «Mon. Piot », XXXIV, 1034, PP. 154 sg, Tav. IX 2; Poinssot - Quoniam, « Karhago s, 1958, op. cit, p. 167. (2) Le figure di belve in caccia all'aperto nel mosaico dol frigidario nella casa di Thuburbo Maius (v- p.75, nota 1), che abbiamo visto tanto legata a Piazza Armerina, purtroppo sono tanto distrutto e restaurate che ci sono scarse possibilità di confronto. 1 frammienti del quadro del paesaggio sono particolarmente importanti. Come nei mosaici di caccia ἃ Piaz-

za Armerina, anche qui cespugli isolati stanno a indicare la vegetazione e larghe strisce ombrose a de-

scrivere le forme del paesaggio, con alberi leggeri spuntanti dallo strisce dell'orizzonte, cfr. Tav. XIII a, bi Gentili, Mosaici figurati, Tavv. XVI, XVII, XXIV e XXXIL. (3) F.G, de Pachtère, Inventaire des mosaïques de la Gaule ei de VAfriqué, Paris 1909 — 15, n. 45 (437); E. Marec, Hippone la Royale, Alger 1954, pp. 109 © 48, Fig. 55 a pp. 104-105; Libyca6, 1, 1958; Ch. Picard, «Rev. Arch.», 1960, II, p . 38 sgg: I. Aymard, Essai sur les chasses romaines, Paris 1951, Tav. III, pp. 228 sg; 287 sg; 406; 450 ogg. Sulla trappola: Oppianus, Cynegetica, IV, 354 986.

H. P. LORANGE

afferra con le mani la manopola trapezoidale, pronto a manovrarla subito, poiché due leopardi ruggenti sono proprio sul punto di gettarsi nella gabbia. Questa trappola, che nella « Grande Caccia » di Piazza Armerina è soltanto un piccolo particolare dell'insieme, a Hippo Regius costituisce il tema principale (Tav. XIV a, δ). Ci troviamo

in un paesaggio con alberi sparsi e con formazioni rocciose nello sfondo. Un avvallamento nel terreno è circondato da una siepe artificiale appoggiata ad alberi e sostenuta da una rete di funi.

Cacciatori dentro o dietro la siepe tengorio una fila di grandi. scudi davanti alla siepe così da costituire uno schermo continuo di difesa; tra gli scudi essi stendono in avanti delle torce. In tre punti la siepe forma un’ansa sporgente all'indietro cosicché dietro allo schermo dei cacciatori risultano dei piccoli bacini chiusi. Questi bacini sono pieni di animali domestici che con i loro versi richiamano da lontano le belve. Nel bacino più a destra viene spiegato come funziona la trappola, L'apertura, che conduce al bacino, è chiusa da una gabbia con manopola trapezoidale, cioè dello stesso tipo che abbiamo visto nelle scene di trappola della « Grande Caccia » di Piazza Armerina e come là, due belve ruggenti ~ a Hippo un leopardo e una leonessa — stanno saltando verso la gabbia che qui, come a Piazza Armerina, deve essere immaginata aperta:e mostrante ‘un'esca all'interno. Un carro con una gabbia, pure della stessa forma, avanza verso il campo cintato per l'ulteriore rifornimento. A destra, dietro il campo chiuso, altre quattro gabbie di questo tipo stanno allineate e pronte. Entro la zona cintata vengono cacciati non soltanto animali feroci, ma anche struzzi e gazzelle. Come a Piazza Armerina i cacciatori tengono in mano due aste. A destra, all’aria aperta, un gruppo di uomini sta banchettando: una scena che si trova anche nella «Piccola Caccia» a Piazza Armerina (1). La corrispondenza stilistica tra il mosaico della grande cacciz a Piazza Armerina e il mosaico di Hippo è altrettanto convincente di quella tematica. Le forme tipiche dello stile popolare tanto nel mosaico di Piazza Armerina quanto in quello di Hippo sono vive e in entrambi abbondano i tratti aneddotici e di genere, Troviamo le stesso piante isolate, 16 stesse pietre e forme rocciose staccate dall’insieme continuo del paesaggio; uomini e animali ccllocati nello stesso modo su un

piano d'ombra isolato come su di una piattaforma; lo stesso modo di contornare linee di terreno e di orizzonte; gli stessi alberi leggeri sul fondo, spuntanti dalla striscia pastosa dell'orizzonte. In particolar modo si deve confrontare la figurazione degli alberi nel mosaico di Hippo e nella «Piccola Caccia » a Piazza Armerina. Le foglie anteriori, lavorate in modo da conservare la loro struttura formale sporgono chiare sul fondo degli strati più profondi del fogliame, il quale appare come una continuità oscura omogenea: lo stesso contrasto che abbiamo visto nella caratteristica lavorazione delle ghirlande d'alloro di Piazza Armerina e nei mosaici nordafricani. Il motivo delle trappole è popolarissimo nelle rappresentazioni di scene di caccia nell'arte romana del m e del rv secolo. Una caccia che spinge gli animali. specialmente cervi, verso una rete stesa tra alberi nel bosco, come la vediamo nella «Piccola Caccia» a Piazza Armerina, si ritrova continuamente sui sarcofagi dell’epoca. Di un carattere più speciale è il motivo della trappola con siepe, gabbia ed esca che abbiamo trovato a Piazza Armerira e a Hippo e che

serviva alla cattura delle belve per l'anfiteatro. Un grandissimo mosaico romano di caccia, verosimilmente d'epoca tetrarchico-costantiniana che ho potuto studiare col gentile aiuto di Dinu Adamesteanu, mostra una certa generica affinità di stile con Piazza Armerina e con il nostro gruppo nordafricano. Proveniente da una casa sull’Esquilino, questo mosaico è conservato în grandi frammenti nel Palazzo dello Sport all'E.U.R. (2) Gentili, Mosaici figurati, op. cit, Fig. 4 e Tav. XXL Questa scena di banchetto come anche una scena di sacrificio dello stesso mosaico, sono trattati

da me in un articolo nello scrittoin onore di E. Arslan Un sacrificio imporiale di caccia nei mosaici di Piazza Armerina (in stampa). 78

PIAZZA ARMERINA

(vicino a Roma) e si trovava prima all'Antiquarium Comunale di Roma. Come motivo principale i frammenti mostrano una trappola con gabbia ed esca (Tav. XVI a) che deve essere confrontata con Piazza Armerina e con Hippo ('). Una grande ansa del paesaggio è stata chiusa dai cacciatori: sul davanti lo sbarramento consiste in una rete di funi fissata ad alberi; Ἰὰ dove lo sbarramento s'incurva all'indietro è indicato con tre o quattro strisce, fiancheggiate da grandi rami frondosi. In mezzo a questa curva si trova una gabbid aperta che si deve immaginare — non altrimenti che a Piazza Armerina e a Hippo — inclusa nel recinto dello sbarramento con l'apertura rivolta verso il campo chiuso; perché si potesse vedere il meccanismo della trappola, la gabbia è stata raffigurata fuori dello sbarramento. Dall'apertura, dove sta appesa l’esca, una passerella infissa all'orlo inferiore dell'entrata è messa lì come un invito alle bestie. Per mezzo di due funi la passerella può essere rialzata così da chiudere l'apertura. Un cacciatore inginocchiato sul tetto della gabbia afferra, come a Piazza Armerina, la manopola: per mezzo di questa, che si deve intendere collegata con le funi, la passerella può essere sollevata. L'uomo è li pronto a manovrare all'istante poiché due orsi stanno per balzare nella gabbia.

Datazione fondata sul ritratto e sul tipo fisico del tempo, sulle vesti e sulle insegne.

Ritratto e tipo fisico del tempo. I mosaici di Piazza Armerina fanno gruppo, come abbiamo visto, così dal lato stilistico come da quello tematico, con l’arte musiva intorno al 300 d. C. delle vicine province dell’Africa nord-occidentale e chiare linee di rapporto conducono dai mosaici di Piazza Armerina e delle vicine località africane a Roma come al proprio centro di irradiazione. Ai decenni attorno al 300 d. C. -- ed esclusivamente a quest'epoca — appartengono inoltre il ritratto individuale, il tipo generale dell'uomo, l'acconciatura femminile e le forme maschili dei

capelli e della barba che s'incontrano a Piazza Armerina. Un autentico ritratto maschile a Piazza Armerina è il « vecchio » con il berretto tondo nell’ala destra della « Grande Caccia » (Tav. XVIII a). È di tipo puramente tetrarchico e invano gli si cercherebbero paralleli nella seconda metà del rv secolo. Qui il tipo maschile dell’epoca degli imperatori-soldati in tutta la sua rozza robustezza è tuttora vivo. Caratteristica è la struttura ossuta e angolosa del viso, la barba corta, i capelli corti che ancora non formano un arco ma conservano il taglio corto e gli angoli alle tempie. Collochiamo, per il confronto, a lato del « vec-

chio» del mosaico di Piazza Armerina, tre delle teste dei tetrarchi in porfido a Venezia (Tav. XVIII e, d) ed una testa sull’Arco di Costantino (Tav. XVIII 5). La forma del viso dura ed ossuta e il taglio angoloso dei capelli congiungono il nostro « vecchio » particolarmente con le teste dei tetrarchi. Nella testa sull'Arco di Costantino il taglio ad angolo è sparito, i riccioli formano un arco, Nei mosaici di Piazza Armerina gli uomini sono ora barbuti ora senza barba, però in prevalenza sbarbati. Questo alternarsi di forme è caratteristico del ritratto dell’epoca progredita tetrarchica e si osserva per esempio nei gruppi di porfido a Venezia, dove due imperatori appaiono con barba e due senza (v. p. 84). Sull'Arco di Costantino i veterani, che rappresentano la generaὉ W. Helbig, Führer, 3. ed. n. 1073. A. Héron de Villefosse, « Bull. Arch. du Comité 198, pp. 257 sg; M. Blake, «Mem. American Acad. in Rome, 17, 1940, pp. 116 sg. e Tavv. 31, 5; 27, 2, 3. La Blake

lo data al principio del rv secolo. J. Aymard, «Mélanges ’Archeologie et d'Histoire de l'École Franqaise de Rome», 1937 (54), pp. 52 sg. Tav. IL e Les chasses romaines, op. cit, Tav. XXXII. 79

H. P. L'ORANGE

zione più vecchia, portano in genere la barba (Tav. XVIII 5), mentre le giovani truppe da combattimento sono quasi sempre sbarbate (1). Nei rilievi oratio e liberalitas sullo stesso arco, le guardie dietro ἃ Costantino sono giovani e in generale senza barba, mentre i senatori, esponenti della. tradizione conservatrice, sono quasi sempre barbuti. Tra i cittadini dell’oratio al Foro romano si alternano il tipo con barba e quello senza e si vede che i più vecchi — e per esempio, i calvi — portano la barba. Anche a Piazza Armerina si riscontra lo stesso mutamento delle generazioni. Nella scena dell’adventus il lampadoforo, che è un uomo di età avanzata, porta la barba a differenza dei giovani che lo accompagnano (Tav. XXIX) e il nostro « vecchio » della «Grande Caccia» è circondato da giovani senza barba. Sull’Arco di Galerio a Salonicco le teste con e senza barba stanno vicine l'una all'altra (ἢ (Tav. XXVIII cd). Una differenza tra la prima epoca tetrarchica e quella successiva consiste anche nella foggia dei capelli. Caratteristico per quella più antica è, come detto, il taglio corte e ad angolo, tramandato dal secolo degli imperatori-soldati. La tetrarchia più tarda, come appare ad esempio nei ritratti riprodotti sulle monete di Massenzio, si lascia crescere più abbondantemente i riccioli, non più nella forma ad angolo (). A Piazza Armerina le forme si mescolano, però, con prevalenza dell'ultima. Già il Gentili () ha osservato che le acconciature femminili a Piazza Armerina datano i mosaici all’epoca attorno al 300 d.C. Dappertutto dove le donne appaiono con l'acconciatura di un determinato tempo, esse presentano sempre le forme particolari dell’acconciatura della tetrarchia e dell'epoca di Costantino (Tav. XVII a-i, XVI 5-4). È questa acconciatura che vediamo in nove teste di Nereidi del mosaico di Arione (Pianta n. 9, Tav. XVII a, in un ritratto di signora ed in una testa di Nereide in mosaici delle terme (Tav. XVI ὃ, c) e in una testa di stagione nel cosiddetto cubicolo con scena erotica (Pianta n. 8, Tav. XVI d). Questa foggia di acconciatura la conosciamo in tutti i minuti particolari e da ritratti in scultura e da quelli sulle monete dei decenni intorno al 300 d.C. (Tav. XVI e). L'acconciatura è caratterizzata dai capelli che vengono tirati sulla nuca dove sono raccolti in una treccia larga e piatta, che da qui viene voltata verso la sommità del capo e fino alla fronte, dove la treccia si alza sul viso come un pettine. È inoltre caratteristico che la massa dei

capelli là dove è piegata ed intrecciata sotto alla nuca formi degli sbuffi e ciocche su ambo i lati del collo, mentre sopra queste sporgenze i capelli vengono intrecciati e compressi verso la nuca, formando così come il paranuca di un elmo. Parliamo di acconciatura a forma d’elmo (5). Questa forma s'incontra prima sulle monete di Severina, la consorte di Aureliano, di Magnia Urbica (Tav. XVI ¢), la consorte di Carino, e di Galeria Valeria (Tav. XVI fi), figlia di Diocleziano e consorte di Galerio, e si mantiene fino all'epoca di Costantino, durante la quale è portata da Fausta, la consorte di Costantino morta nel 326, e, fino a quest'anno, anche da Elena, madre di Costantino. Di questa forma si era sviluppata, già verso la svolta del secolo, una variante: l’acconciatura cosiddetta a corona o turbante, nella quale la larga treccia è divisa e si posa come una corona ὁ un turbante attorno alla sommità del capo. Evidentemente si tratta di quest’ultima variante nella testa della Nereide, Tav. XVII. La forma appare prima sulle monete di Galeria VaJeria (!), ma non si impone generalmente prima dell'anno 326, quando la treccia chiusa sparisce

(Ὁ L'Orangevon Gerkan, op. cit, pp. 44, 97. (9) L'Orange, Studien, Figg. 62, 63.

46 sgg K. Wessel, Rémiscke Frauonfrisuren von der Severischen bis sur honstantinicchen Zeit, «Archiolog.

(3) L'Orange, Studien, p. 105. (4) Mosaici figurati, op. cit, p. 70.

Anzeiger», 1946, 47, pp. 62 sgg.; L'Orange, Der subtile Stil, « Antike Kunst» 4, 1961, pp. 71 588.

(6) B. M. Felletti Maj, « Critica d'Arte», 6, 1941, (5) R. Delbrück, Antike Porphyrwarke, Berlin roga, p. 124, «Helmirisure; Spitantike Kaiserporräts pp. p. 76; Wessel op. cit, pp. 62 568. 8o

PIAZZA ARMERINA

del tutto. Quest'ultima forma si rinnova, però, nell'età di Teodosio; ma allora senza carattere di acconciatura ad elmo, cioè senza la pressione sulla nuca e senza le ciocche sporgenti ai lati di questa (1). L'acconciatura a turbante vicino all’acconciatura ad elmo costituisce una combinazione possibile soltanto negli anni attorno al 300. Il

berretto

tondo.

Nel mosaico della «Grande Caccia» s'incontrano due insegne d'autorità che sono in modo speciale legate allo stato tetrarchico: cioè quelle che noi con espressioni semplificative chiameremo il «berretto tondo » ed il « bastone a fungo », Studiamo prima il berretto tondo. Nella « Grande Caccia » tutti gli uomini — i cacciatori, i marinai ecc. - sono a capo scoperto, con tre sole eccezioni: in questi tre casi eccezionali i personaggi portano un particolarissimo copricapo cilindrico basso. Nella «Grande Caccia » gli uomini con questo berretto sono persone altolocate (*) e portano come tali il bastone a fungo (pp. 87 sg). Uno di questi uomini, cioè il «vecchio » nell’ala destra del mosaico (Tav. ΧΧΤΙ δ), viene coperto -- come le persone del più alto rango (*) -- da due guardie con scudi grandissimi. Gli altri due uomini col berretto tondo sono significativi già per la loro stessa posizione nel quadro: in luogo alto verso l'orizzonte, nel centro del mosaico (p. 95. Tav. XXIIa). Come vedremo, appartenevano ad un gruppo a quattro

in parte distrutto. Erano tutti e quattro, indubbiamente, dello stesso tipo con berretto tondo e bastone a fungo e formavano un gruppo centrale dominante tutto l'enorme quadro di mosaico. Se ne deduce, dunque, che il berretto di pelo, al modo stesso del bastone a fungo, è un privilegio, un'insegna di autorità. Come tale insegna pubblica il berretto tondo è stato portato -- al modo stesso del bastone a fungo — in un solo periodo dell'impero, cioè durante la tetrachia. In questo periodo l'imperatore stesso porta questo berretto: così Diocleziano nel medaglione-ritratto nel suo Mausoleo di Spalato (Tav. XIX e), inoltre tutti e quattro gli imperatori, senza dubbio della prima tetrarchia,

nei gruppi di porfido di Venezia (Tavv. XVIII c, d), ancora un imperatore tetrarchico in un frammento di porfido nel Museo Nazionale di Nis in Jugoslavia (*) (Tav. XVIII e-g); infine; due imperatori tetrarchici accoppiati a dei fluviali in due erme di pietra locale a Salona (Tav. XIX a-d). La testa di Nis, vicinissima per stile e tipo alle teste del gruppo di Venezia, si stacca, come il gruppo veneziano, ma in modo più libero, da un fondo (di una colonna ?) e forse faceva parte di un simile gruppo.

Si sono però alzate voci sostenenti che il berretto tondo è stato portato nella vita ufficiale non soltanto nei tempi tetrarchici e che così non costituisce un criterio per la datazione della

«Grande Caccia » a quella epoca. H. Stern ha voluto dimostrare che il berretto tondo si nella vita pubblica anche dopo il periodo tetrarchico, riferendosi alla rappresentazione di muarius in una copia rinascimentale a Vienna di una figura nel Calendario del 354 (Tav. XXT Secondo Stern si tratta di un vicomagister che porta il nostro berretto tondo ad un sacrificio (1) L'Orange, Der subtile Stil, op. cit, pp. 62 sgg. (2) I mosaici di Piazza Armerina hanno portato nella discussione un nuovo momento decisivo che comporta un mutamento della nostra precedente opinione circa il berretto cilindrico degli imperatori, L'Orange-- von Gerkan, op. cit, p. 44, nota 5. I fori nei berretti dei tetrarchi di Venezia dovono aver ser-

usò Jaa). uf-

vito a incastrare gioielli, coccard o cose del genere, L'Orange, Studien, op. cit, p. 17, nota 4. (3) «Symb. Osl.», op. cit, p. 118. — (i) M. Grbié, Rómische Kunstschatse aus dem serbischen Donaugebiet, in: E. Swoboda ¢Rémische Forschungen- in Niederisterreich », TI, p. 84 e Tav. XII 4. Br

H. P. L'ORANGE

ficiale. Ma questa interpretazione evidentemente è impossibile. Come tutti sappiamo, secondo il rito romano non si sacrificava con un copricapo in testa, ma colla testa coperta dalla toga o dal mantello. Il berretto di Januarius, ben descritto da Stern come «un bonnet de fourrure d'où tombe un voile» (1), non ha nulla da fare con il copricapo antico da noi trattato (+), ma è un tipico copricapo rinascimentale degli stessi anni della Vindobonensis, cfr. per esempio il famoso ritratto d'ignoto di Antonello da Messina nella Galleria Borghese (Tav. XXI δ). La figura originaria del Calendario deve essere stata un togato che capite velato compiva un sacrificio: la parte della toga soprastante la testa e pendente dietro la nuca e verso le spalle si trasformò nel tipico copricapo rinascimentale. Lo stesso adattamento alle forme attuali del proprio tempo si vede nella rappresentazione nel Vindobonensis del: vestimento femminile di Februarius () (Tav. XXI ὁ). Mentre le copie Luxemburgensis (1). e Bruxellensis 6) (Tav. XXI d) rispettano le forme originarie antiche del peplos e dello himation, la Vindobonensis le ha trasformate radicalmente in un vestimento rinascimentale. Considerando che il Januarius dello Stern è il solo documento datato esistente per sostenere la tesi che il berretto -tondo è stato portato nella vita ufficiale in epoca post-tetrarchica, la base di questa tesi pare piuttosto fiacca. Nel periodo tetrarchico lo stesso berretto tondo appare anche come copricapo militare. Così lo vediamo per la prima volta nei rilievi dell'Arco di Galerio a Salonicco dell'anno 300 circa, Soldati col berretto tondo si vedono tra i cammelli che portano le provvigioni della armata (Tav. XXc). Nonostante il cattivo stato nel quale si trovano i rilievi, il contorno del nostro berretto spicca chiaramente sul fondo (Tav. XX a). Su un rilievo calcareo funerario nel Museo di Treviri ‘colla rappresentazione di una coppia di coniugi, l'uomo, senza dubbio un militare, porta il nostro berretto tondo (Tav. XX δ). Anche qui si tratta di un monumento dell'anno 300 circa, come si può dedurre dall’acconciatura della donna (f) (cfr. la nostra Tav. XVI e-R). Sull'Arco di Costantino troviamo il berretto tondo come copricapo militare e qui — come nel rilievo di Treviri —

esso è chiaramente caratterizzato come fatto di pelliccia. In un fregio del ciclo storico, del 3r

315 d. C., che rappresenta la partenza per la battaglia dell'armata di Costartino, un reparto di

milites porta il berretto e anche qui, come nell’Arco di Galerio, sorveglia i carri di vettovaglia-

mento (Tav. XX d); tutti hanno la clamide sopra la tunica, corrispondente ad un grado più alto

di quello di semplici soldati (*. Già lo Styger (*) ha notato la barba come un tratto che distingue questi uomini con berretto tondo dagli altri soldati. La barba, essendo, come abbiamo visto, una caratteristica del nr secolo e dei primi tempi tetrarchici, indica una certa età in questi soldati — essi sono dei veterani (pp. 79 sg.). Sull’Arco di Costantino, dunque, i milites con berretto tondo rap-

presentano la vecchia generazione, Nei mosaici di Piazza Armerina il berre:to è portato tanto da vecchi quanto da giovani, il che corrisponde ad una datazione di età anteriore a Costantino.

Sull’Arco di Costantino è l’ultima volta che vediamo il berretto tondo in una scena di vita reale. Come devono essere interpretati gli uomini con berretto tondo addetti ai carri di vettovagliamento dell’armata ? Il servizio dei carri riguarda i rifornimenti del vitto e del foraggio e, pertanto, mediatamente, concerne il servizio di polizia connesso con le requisizioni, sempre più pe-

() H. Stem, Le calendrier de 354, Paris 1053, (4) Stor, op. cit, Tav. VIII, 1. pp: 266 sgg. ὁ Tav. XVII, s. (3) Stem, op. cit, Tav. XVI, x fe) Anche W. C. Volga mette in dubbio le (0) E. Vorrenbagen, Jabrestericht des Provinzialautenticità della figura di Januarius per quanto ri- museums zw Trier « Trierer Zeitschrift, 5, 1950, Guarda il copricapo, De figura mensis Januariî « pp. 166 ag. © Tav, 5,1. codice deperdito exscripla, «Mnemosyne s, 59, 1931, (7) L'Orange-von Gerkan, op. cit, pp. 44 98 p. 304. Tevv. 6a, b; 19, c. G) Stern, op. cit, Tav. XVIII, 4. (5) «Rümisehe Quartalschrifty27, 1915, p.39 notaτ 82

PIAZZA ARMERINA

santi sotto gli imperatori-soldati, con le necessarie scorte dei trasporti e con la sorveglianza dei depositi dell'esercito, È tipico che nella tarda antichità — e verosimilmente proprio sotto Diocleziano — i frumentarii che avevano à che fare con l'approvvigionamento dell'esercito siano sostituiti dagli agentes in rebus, che, oltre ad essere ufficiali addetti alle provviste, sono anche agenti di polizia. Sotto un regime sempre'più rigidamente autoritario tali truppe di guardia e di polizia si vedono attribuire un campo d'azione sempre più vasto e un potere sempre maggiore e si possono paragonare alle truppe di polizia delle dittature moderne. Il praefectus practorio si rivolge agli agentes in rebus per ordinare gli incarceramenti e le esecuzioni: essi sono i guardiani delle carceri e gli ispettori dei carcerati e hanno una parte importante nelle persecuzioni dei Cristiani (). In coda ai carri sul rilievo dell'Arco di Costantino, due ufficiali con berretto di pelo avanzano su un carro fastoso, verosimilmente una carruca, trainata da quattro cevalli, il che corrisponde ad una alta espressione di potere e di prestigio (?): si tratta di alti ufficiali del servizio d'approvvigionamento e polizia (Tav. XX d). Tanto gli imperatori col berretto tondo, come rappresentati nei loro ritratti, quanto gli uomini in berretto tondo come

li vediamo

nei rilievi sull’Arco di Galerio, sull’Arco di Costantino

e sul

rilievo sepolcrale di Treviri appartengono tutti alla vita del tempo di queste opere, cioè ai decenni intorno all'anno 300. Soltanto in quell’epoca il berretto tondo s'incontra come copricapo tipico della vita statale e imperiale delle organizzazioni amministrative e militari. Dietro l'uomo col berretto tondo in tutte queste immagini sta la potente realtà storica dell’epoca. In origine si trattava

evidentemente

di un copricapo

locale dell'Illyricum,

che verosimilmente

fu lanciato

da Diocleziano e pian piano dalla classe dominante passò nelle varie attività e corporazioni organizzate dallo stato tetrarchico e favorite dall'imperatore. Dopo l'Arco di Costantino, però, l'uomo col berretto tondo scompare per sempre dalla scena della storia e non s'incontra mai più in rappresentazioni della vita contemporanea. Cosi lo si cercherebbe invano su tutti i monumenti storici di quelle epoche tarde alle quali il Pace ed il Cagiano de Azevedo datano la « Grande Caccia ». Non più nelle scene storiche, ma soltanto nella iconografia sacra cristiana l'uomo col berretto tondo si conserva ed è ripetuto: persiste nella memoria come lo sbirro ed il boia della grande persecuzione dioclezianea contro i Cristiani e in appresso diventa il tipo stesso del persecutore degli uomini di Dio e si ripete nella immagine della guardia che afferra Pietro, delle guardie che sorvegliano Pietro in carcere, dei soldati che compaiono nella Passione di Cristo e che vigilano presso il suo sepolcro ()). Dalla vita l’uomo col berretto tondo è passato nelle immagini della Passione e in quelle degli atti dei martiri. Nel mosaico di Piazza Armerina vediamo l’uomo dal berretto tondo al culmine della potenza, come incarnazione dello stato e della legge al centro

del suo mondo

reale; nell’arte paleocristiana, invece, noi

(1) L'Orangevon Gerkan, op. cit, p. 45, note t. 7; CCXVIIL τ, 2; CCKXVI, 2; CCXXIX, 1; e2. CCXXXMM, s; COKXXXIX, 11; CCLIT, 1. Tavv. (3) Op. cit, pp. 53 sg. vol. ΠῚ (supplemento), Tavv. CCLXXXVI, 1, 8; (3) J. Wilpert, 7 sarcofagi cristiani antichi, Città CCLXXXXVII 1. M. Dalton, Cat. of the Tvory del Vaticano 1929-36. Tavv. vol. I, Tavv. XXXVIII, Carvings of the Brit, Mus, Tav. 4. K. Wessel, Fine 3; XL; LXI, 3; LXXXVI, 3; LAXXXI; LXXXXIL, Gruppe oberitalienischer Elfenbeinsarbeiton, « Jahrb. d. 1; LKXXXVI; LXXXXIX, 1, 5; C, 3, 6,9; CI, 4, 5,8; Deutschen Archäolog. Inst. » 63-64, 1948-49, pp. 125 CHIL, 2, 4, 7, 8; CIX, t, 2, 4, 6; CV, 2, 3; CVI, 2; sgg, Figg. 6-8. Rinvii in L'Orange - von Gerkan, CVIL τ, 2, 3, 6, 9; CX, 1, 3; CXL 1, 4; CXIL 2; 3, op. cit, p. 44 nota 6. Gli esempi dati dal Pace di CRU, x, 3; CXIV, 1, 2; CXV, τ, 2; CXXII x, 2, 3: uomini con berretto di pelo del tardo rv © del v seCXXIII, 2, 3; CKXVI, 2; CXXVII, τὶ colo (Pace, I mosaici, p. 109 nota 22, cfr. p. 117) appar1, 2,5 CKXIX, 2; CXXXXITI, x; CLIT, tengono tutti ad una fissa tradizione iconografica CLV, 3, 4, 5; CLVI; CLVII, 1; CLVIII, τ, 2, 3. Tavv. cristiana e non riflettono soldati o guardie della vita vol. If, Tavv. CLXXX, 2; CCVI, 7; CCKII, 2; CCXV, contemporanea. 83

H. P. ORANGE

lo vediamo in un'immagine-ricordo come il servitore del tramontato stato pagano. Il berretto tondo diventa nell’iconografia cristiana un segno o attributo fuori del tempo, non altrimenti, ad esempio, del copricapo frigio divenuto attributo fisso nella immagine dei re magi (1). L'uomo dal berretto tondo è dunque diventato tipo rigorosamente limitato alle scene di Passione e di martirio dell'iconografia cristiana e dopo l’Arco di Costantino lo si cercherebbe invano in rappresentazioni tratte dalla contemporanea vita reale.

È scomparso dal palazzo e

dall’amministrazione civile. Dio ha annientato i persecutori, «cosi che di loro non resta né ceppo né radice », ut eorum nec stirps nec radix ulla remaneret (2), dichiara Lattanzio negli anni successivi all’Arco di Costantino. « Dove si trovano adesso i nomi degli Jovii e degli Herculii, orgogliosi e famosi tra i popoli... ? Dio li ha annientati e rasi via dalla terra ». Ubi sunt modo magnifica illa et clara per gentes Joviorum et Herculiorum cognomina... delevit ca dominus et erasit de terra (). Nella « Grande Caccia » di Piazza Armerina non si tratta — lo ripeto -- di una scena tradizionale dell'agiografia cristiana. Al contrario, assistiamo ad una cattura in massa di belve per l'anfiteatro, cioè ad una scena della vita reale e contemporanea, non altrimenti che nei rilievi storici sull’Arco di Galerio e sull’Arco di Costantino. Dobbiamo, dunque, contro l'opinione del Pace () e del Cagiano de Azevedo (5) mantenere fermamente la tesi che l'uomo dal berretto tondo di Piazza Armerina è un argomento incontrovertibile per la datazione del mosaico all'età tetrarchica. Il Cagiano de Azevedo ha voluto dimostrare che i quattro tetrarchi di porfido a Venezia non hanno diritto a questo nome, essendo essi immagini imperiali di epoca posteriore (ἢ). L'illustre studioso parte dall’osservazione di alcuni dettagli del lavoro che, secondo lui, implicherebbero necessariamente che tutte le figure all'origine fossero imberbi. La conclusione che se ne trae che, se imberbi, allora non possono rappresentare i tetrarchi, non mi pare lecita. Diocleziano appare senza barba nel ritratto sul fregio del suo mausoleo di Spalato ( (Tav. XIX e), e così pure nei frammenti della sua testa in un rilievo con sacrificio sull'Arco di Galerio a Salonicco (*). Nel gruppo imperiale a quattro dello stesso arco sono conservati notevoli resti della testa di Massimiano, presentanti anch'essi una testa imberbe (?) (Tav. XXVIII 8). Galerio sul detto arco in una scena di supplicatio (19) appare senza barba (Tav. XXVIII a). Tra le figure anonime le teste imberbi si alternano con teste barbate (Tav. XXVIII c, d). Secondo le testimonianze incontestabili dei monumenti scultorei dell’epoca, già durante la prima tetrarchia gli imperatori cominciarono a levarsi la barba seguendo o creando quella moda che doveva costituire la nuova forma del rv secolo (11). Più rilevante mi pare il tipo fisionomico dei gruppi di Venezia, che è

(1) =von (2) (3) (4) mota

«Symb. Osl. », op. cit, p. 117 e L'Orange Gerkan, op. cit, pp. 43 588. Lactantius, De mort. perseculorum. 50. Op. cit, 52. I mosaici, op. cit, p. 104 nota 23; p. 117 18.

(Β) K-F. Kinch, L'Arc de Triomphe de Salonique, Tav. V e p. 36. L'Orange, Studien, p. 100, nota 2. (9) Studien, p. 100 e Fig. 248. Kinch, op. cit, Tav. VI. (1o) Kinch, op. cit, Tav. IV. (11) Anche sulle monete si rifletta questo cambia-

(5) M. Cagiano de Azevedo, I proprietari della mento della moda nel corso dell'epoca tetrarchica. villa di Piassa Armerina, «Scritti di storia dell'arto — Pero, come ha dimostrato J. Maurice (Numismatique in onore di Mario Salmi, Roma τοῦτ, pp. 18 sg. — Constantinienne, Parigi, 1911, I, pp. 4 s), il valore (6) Cagiano do Azevedo, I cosiddetti terarchi di iconografico delle monete di questa epoca che non Venezia, « Commentari ν, 1962, pp. 160 sgg. si trattenova dal sostituire tipi imperiali già coniati (7) R. Delbrück, Ame Porphyrwerke, Tav. 61, ἃ nome di un altro imperatore, è più che dubbio. Se Fig. 49, p. 124. L'Orange, Die Bildnisse der Telvar- una tale sostituzione del ritratto coniato per un altro chen, «Acta Archaeologicas, II, 1931, Tav. III; imperatore fosse tollerabile, allora è chiaro che il ri Studien, Fig. 39 © p. 100. tratto coniato per lo stesso imperatore in un tempo 84

PIAZZA ARMERINA

evidentemente precostantiniano, ancora col carattere brutale dell'imperatore-soldato dell’ultimo ut secolo, viso massiccio e rettangolare, tratti duri e rudi, taglio angolare dei capelli contrastante col taglio arcuato dei ritratti costantiniani. Secondo me è impossibile inserire le teste dei gruppi veneziani nella ben conosciuta sequenza tipologica dei ritratti imperiali da Costantino a Teodosio.

.

ΤΙ punto di partenza per l'identificazione dei quattro imperatori di Venezia deve essere il singolare copricapo portato da costoro, cioè il berretto tondo. Come abbiamo visto, questo copricapo appare sui monumenti storici soltanto nei decenni attorno al 300 cioè sull’Arco di Galerio e sull’Arco di Costantino, in epoca tetrarchica, dunque. Ma c'è qualcosa di più determinante. Nel palazzo di Diocleziano a Spalato s'incontra Diocleziano stesso con questo copricapo nel medaglione-ritratto sul suo mausoleo. Un altro importantissimo argomento per l'identificazione dei nostri quattro imperatori con i tetrarchi ci è dato da due ritratti unici a Solin (Salona) nella vicinanza immediata di Spalato e provenienti indubbiamente dal palazzo di Diocleziano. Penso αἱ due ritratti abbinati a dei fluviali in doppie erme e portanti tutti e due il nostro berretto tondo (1) (Tav. XIX a-d). Mi pare che il carattere essenziale di questi ritratti ad erma sia sfuggito al Cagiano de Azevedo. Non si tratta in questo caso, come pensa l'illustre collega (, di erme di piccole dimensioni «adatte per pilastrini di transenne » e non rassomigliano affatto «ad analoghe sculture del tardo restauro del Nymphaeum Traiani in Efeso » (*), ma, al contrario, di ritratti in grandezza

naturale (come risulta dalle dimensioni da noi fornite negli Acta Archaeologica 2, 1931, p. 32 nota 2 e p. 36 nota 4). In ambedue le erme un ritratto chiaramente individualizzato, con clamide e berretto cilindrico tetrarchico, è abbinato ad un dio fluviale. È soltanto questa deità fluviale che è un Mischgestalt, con «caratteri semiferini » (‘), mezzo uomo e mezzo toro, ma i ritratti non sono affatto dei Mischgestalten, come pensa il Cagiano de Azevedo, evidentemente per un malinteso. Al contrario, sono persone chiaramente caratterizzate, sia di viso che di abbigliamento e di copricapo. Stilisticamente le erme appartengono al Palazzo di Diocleziano (9), dal quale, come detto, devono provenire.

L'aggiunta del dio all’erma mette in rilievo l'insigne importanza della persona rappresentata ("). Si tratta di un tipo di dio fluviale che si trova in adattamento architettonico nel Palazzo di Diocleziano a Spalato ed in gran numero in Dalmazia e nelle province vicine e deve rappresentare un esteso culto locale (?). Nell’accoppiamento del ritratto al dio, la potenza fertilizzante, κουροτρόφος, del dio fluviale viene incorporata nell'uomo. La persona viene glorificata

come fertilizzatore ed apportatore di felicità al mondo, cosa che conviene solamente all’ imperatore, Le erme acquistano dunque un significato che richiama le parole che un panegirista del tempo rivolge a Diocleziano e Massimiano ( glorificando la loro potenza fertilizzante: omnes terras omniaque maria plena esse vestri... Nullus ager fallit agricolam. .. Rumpunt horrea conditac messes etc. etc. I due ritratti, seppure di lavoro rozzo e provinciale come anche il ritratto di precedente poteva, posteriormente, essere ripetuto senza fine, non badando ai particolari come il cambiamento della forma della barba. (1) L’Orange, op. cit. «Acta Archaeologica, Tavv. LIL, Figg. 2-6; Studion, Pigg. 36-3888, 17588, 1005g. (2) Cagiano de Azevedo, op. cit,p. 169. (3) F. Miltner, XXIII Vorläufiger Bericht über die Ausgrabungen in Ephesos, «Oesterreich. Jahresheftes 45, 1960, p. 326 e Figg. 184-186. (4) Cagiano de Azevedo, op. cit, p. 169.

(5) L'Orange, op. cit, «Acta Archaeologica», pp. 37 sgg. (6) Non sono più sicuro, come trent'anni fa, che i tratti delle divinità flaviali siano assimilati a quelli degli imperatori, op. cit. «Acta Archaeologica»,p. 35, (7) L'Orange, op. cit, «Acta Archaeologica », pp. 40 sgg. (8) Mamertini paneg. genethl. Max. Aug. x4 SE. (E. Galletier, Paneg. Latins, I, pp. 63 sg). L'Orange = von Gerkan, op. cit, pp. 158 stg. 85

H. P, L'ORANGE

Diocleziano nel suo mausoleo, non possono non rappresentare imperatori, e allora imperatori dell'epoca del palazzo di Diocleziano, cioè della prima tetrarchia (1). Ancora una volta incontriamo l'imperatore col berretto tondo nell'Ilyricum. A Ni, l'antica Naissus, centro della popolazione illirica, si è rinvenuto il frammento di una testa con berretto tondo (Tav. XVIII eg), forse facente parte di un gruppo simile a quello veneziano, e — da notarsi bene -- lavorato nello stesso porfido. Ancora non ho potuto vedere l'originale, ma dalle fotografie mi pare possibile che si tratti di una replica dei gruppi di Venezia. Così anche questi verrebbero messi in una certa relazione all’Illyricum, come tutte le altre immagini conservateci

degli imperatori dal berretto tondo (*), È ben noto che tutti gli imperatori della prima tetrarchia provengono da questa provincia: Diocleziano nacque nella regione di Spalato, probabilmente a Salona, dove anche eresse il suo ofium; Galerio e, verosimilmente, anche Massimino Daza,

figlio della sorella di Galerio, provengono dalla regione di Serdica, non lontano da Naissus; L'Illyricum è la madrepatria di Massimiano Erculeo; a Naissus nacque il figlio di Costanzo Cloro, Co-

stantino. (v. il supplemento alla fine della p. 104). Non è sostenibile neppure la tarda datazione dei gruppi veneziani sulla base del motivo che

unisce tutti e quattro gli imperatori: l'abbraccio che, secondo il Cagiano de Azevedo, non sarebbe entrato nel repertorio figurativo aulico prima del gruppo costantinopolitano detto « Phi

ladelphion », eretto in memoria dell'incontro e dell'abbraccio che si scembiarono i figli di Costantino incontrandosi a Costantinopoli dopola morte del padre (*), L'abbraccio, la stretta di mano, l'unione intima in gruppi di figure molto-accostate sorgono, secondo il Cagiano de Azevedo, soltanto come espressione della fratellanza di sangue degli imperatori e non come mani festazione di una loro fraternità, derivante dal concetto politico della Concordia. Questa tesi viene contraddetta dalle nostre fonti letterarie che mostrano incontestabilmente che tali forme di stretta unione degli imperatori erano caratteristiche proprio della Concordia Tetrarchica. Mamertinus nel Panegiricus Genethliacus a Massimiano Augusto descrive con enfasi come gli imperatori Diocleziano e Massimiano, quando parlano insieme, si tengano sempre per mano o, quando si mostrano in pubblico, siano sempre strettamente uniti l'uno a fianco dell'altro: Mihi ante oculos pono cotidiana vestra colloquia, coniunctas in omni sermone dexteras. .. (t). Clamare omnes prac gaudio, iam sine metu vestri et palam manu demonstrare: « Vides Diocletianum ὃ Maximianum vides ? Ambo sunt, pariter sunt! Quam iunctim sedent! Quam concorditer colloquuntur! » (8). Si

(1) Da avvicinare alle doppie erme di Solin è cosa di simile; se ne deduce che in tutti c tre i casi si un'altra erma doppia del 300 d. C. circa, nella quale tratta dello stesso copricapo che negli imperatori a il ritratto d'un imperatore è accoppiato alla testa Venezia aveva uno. speciale ornamento. di Crono-Saturno: anche in questa combinazione si (3) Cagiano de Azevedo, op. cit, pp. 169 sgg. Mi vuole esprimere l'idea dell'Abundentia o Fecunditas è impossibile accettare la tesi di P. Verzone (I due temporum, della Felicitas temporum lodata dai pa- gruppi in porfido di S. Marco in Venezia ed il Philanegiristi e rappresentata sulle monete del tempo. delphion di Costantinopoli, «Palladio», Rivista di H. Fubrmann, Zum Bildnis des Kaisers Diocletian, Storia dell'Architettura, 1958, pp. 8 sgg.) che i gruppi veneziani provengano dal Philadelphion di Costanti« Rôm. Mitt. » 53, 1938, PP. 35 seg. (2) Secondo me non vi può essere dubbio che si nopoli e rappresentino il binomio, ripetuto due volte, tratti di un copricapo chiuso e non di un diadema di Costantino e di suo padre Costanzo Cloro. IL ritratto aperto come il Cagiano de Azevedo sostiene nel caso di Costantino ci è ben conosciuto ὁ differisce in tutto e non quello dei quattro imperatori a Venezia (op. cit, pp. 175 dalle testo veneziane. 1:11 tipo tetrarchico 5). Come in questi, così anche nelle erme di Salona costantiniano che caratterizza dette teste. e nella testa di porfido a Nis, il berretto tondo è senza (4) Mamertini Paneg. Genethl. Max. Aug. x2. (Ed. indicazioni di pelo ed è somigliante ad un feltro liscio, Galletier, Paneg. Latins, I, p. 62) senza, però, i buchi che nel berretto degli imperatori (5) Op. cit. 21. (Ed. Galletier, Paneg. Latins, I, a Venezia servivano ad incastonare gioielli o qualche p. 62) 86

PIAZZA ARMERINA

può avvicinare a questi testi il celebre gruppo dei tetrarchi, strettamente uniti, sull'arco di Galerio a Salonicco (1) La fraterna unione delle figure, la stretta di mano o l'abbraccio appaiono essere così delle espressioni spontanee di quella Concordia tetrarchica e di quella collegialità imperiali che furono alla base del sistema statale di Diocleziano. Quando quattro statue imperiali sono unite tra di loro nel modo dei gruppi di Venezia e del Vaticano, l'associazione ai tetrarchi si presenta subito alla mente. E quando tali imperatori, come nei detti gruppi, sono di un tipo esclusivamente tetrarchico, non possono, secondo me, esistere dubbi sulla legittimità di questa identificazione. Esiste anche un altro importante punto di contatto tra il gruppo veneziano ed i tetrarchi illirici e cioè la particolarissima spada dalla elsa a testa d’aquila che, come dagli imperatori del gruppo veneziano, viene portata da una figura loricata in un frammento di rilievo nel Museo di Spalato () (Tav. XXVIII e). Stilisticamente il frammento appartiene al Palazzo di Diocleziano. La tarda datazione di questo tipo di spada mantenuta

è dunque insostenibile.

dal Cagiano

de Azevedo

(*)

Nonostante tutte le differenze esistenti tra i gruppi di porfido di Venezia e del Vaticano non

posso in alcun modo

accettare l'asserzione del Cagiano

de Azevedo che lo stile e l'iconografia

rendono assurdo un confronto tra i due gruppi (‘. Al contrario, i due gruppi sono uniti tra di loro da un complesso di tratti specialissimi, costituente il loro carattere di opere uniche ed inseparabili in tutta l’arte imperiale romana. In quest'arte una sola volta — e precisamente nci gruppi di Venezia e del Vaticano — abbiamo davanti ai nostri occhi, lavorati nello stesso porfido,

nello stesso alto rilievo e sul fondo di una colonna, quattro imperatori che si abbracciano esattamente nello stesso modo, In considerazione di questa somiglianza essenziale e generale, l'esatta

corrispondenza dei dettagli della peculiarissima corazza acquista un significato decisivo come legame infrangibile tra le nostre sculture. Il bastone

a fungo.

Più breve può essere la nostra trattazione dell'altra insegna tetrachica che troviamo nella «Grande Caccia » e che abbiamo chiamato il bastone a fungo. Tutti gli uomini col berretto di pelo nella «Grande Caccia» portano, oltre a questo berretto,

un particolarissimo bastone quale insegna della loro autorità (Tav. XXII a-c). Si tratta di un

alto bastone che va restringendosi verso il basso poter essere infisso én terra per meglio servire da traversa convessa, simile al cappello di un fungo chiamato bastone con manico a fungo. Vediamo tutta la composizione

della « Grande

ala destra della « Grande Caccia proprio sotto il cappello ἃ fungo, versa (Tav. XXII ὃ). Il bastone è sempre una persona altolocata ciatori (Tav. XXII ὁ). Questo

bastone

e verosimilmente terminava a punta così da appoggio. Di sopra, il bastone termina in una e per questa ragione il bastone viene da noi i personaggi del gruppo centrale, dominante

Caccia » con questo bastone in mano.

Il « vecchio » nella

» si appoggia allo stesso bastone impugnandolo con la destra mentre la sinistra si appoggia sulla superficie convessa della traa fungo è portato anche da una persona a capo scoperto, che e mostra la sua autorità punendo con la frusta uno dei cac-

s'incontra soltanto

su monumenti

(2) Kinch, op. cit, Tav. VI. (2) L'Orange, op. cit., «Acta Archaeologica», Fig. 16 e p. 52.

tetrarchici come

distinzione di persone

(3) Cagiano de Azevedo, op. cit., p. 177. (4) Op. cit, p. 168.

87

H. P. L'ORANGE

insignite d'autorità. Nel rilievo dei Suovetaurilia sulla Base dei Decennali nel Foro Romano, del l’anno 303 d. C. (Tav. XXIII δ), un uomo togato, il quale sta a capo della processione sacrificale, tiene questo bastone nella mano sinistra. Egli si volge indietro verso il corteo che lo segue facendo con la destra un gesto imperativo — è l'alto magistrato che dirige i movimenti del corteo. In un afiresco che decora un santuario tetrarchico per il culto imperiale nel grande tempio di Luxor è pure presentata una processione, verosimilmente di alti dignitari dello stato (Tav. XXIIIa).

Uno di questi, che immaginiamo sia il dignitario incaricato di guidare la processione, tiene nella mano destra il bastone a fungo. Come il magistrato che guida la processione sulla Base dei Decennali, anch'egli si volge indietro verso la processione, probabilmente per dirigerne i movimenti. Gli affreschi, oggi scomparsi, sono riprodotti in acquerelli del Wilkinson del 1859, ritenuti fedelissimi (). (v. il supplemento alla fine della p. 104). Quando il bastone a fungo e il berretto tondo scomparvero del tutto con la tetrarchia, ciò accadde senza dubbio a causa dell'odio che in età cristiana colpì i persecutori e con loro tutto l’apparato rappresentativo, espressione del potere e della autorità dello stato tetrarchico. Orbiculi Il Cagiano

e altri de Azevedo

ornamenti

vestiari.

ha affermato che la ricca ornamentazione delle vesti con orbicoli,

così caratteristica dei mosaici di Piazza Armerina, non è usata prima di Costantino (). Ma non è difficile dimostrare il contrario. Già il Gentili ha mostrato degli evidenti paralleli agli ornamenti vestiari di Piazza Armerina nell'arte tetrarchica (°). Nei summenzionati affreschi tetrarChii a Luxor le vesti non sono soltanto decorate in modo ricchissimo con grandi orbiculi lussuosamente ricamati, ma questi presentano uno speciale schema di ricamo, che continuamente si ripete negli orbicoli dei vestiti di Piazza Armerina. Collochiamo qui, l'uno vicino all'altro, un particolare di vestiario degli affreschi di Luxor (Tav. XXIV a) e delle vesti ornate con orbiculi nella « Grande Caccia» di Piazza Armerina (Tav. XXIV c, d). Lo schema dell’orbiculus nel detto particolare a Luxor ha una figura principale che si può descrivere come una stella a otto punte ottenuta incrociando l'uno con l'altro due quadrati, entrambi con i lati leggermente incurvati: esattamente la stessa stella la troviamo a Piazza Armerina nell'orbiculus sulle spalle e sulle parti inferiori delle tuniche. La veste di Luxor mostra lungo l'orlo inferiore della tunica uno schema di ricami con rombi inseriti in campi rettangolari: anche questo schema si incontra nelle vesti di Piazza Armerina, come in un clavus della tunica nel mosaico dell’advenius (Tav. XXIV δ). Se si passa da tali particolari alla considerazione di tutto lo schema della tunica e si paragona la tunica di Luxor (Tav. XXIII a) colle tuniche delle figure di Piazza Armerina (Tav. XXIVc, d), si trova un'evidentissima corrispondenza generale nel disegno dei grandi orbiculi scuri sul fondo chiaro della tunica che, sia a Luxor sia a Piazza Armerina, è orlata da una larga banda a ricamo. I rilievi sull'Arco di Galerio mostrano almeno una volta, cioè in una raffigurazione di Galerio (4), un orbiculus plasticamente marcato nella tunica (5), mentre sull’Arco di Costantino

(a) U. Monneret de Villard, The Temple of the (3) Mosaici figurati, op. cit, pp. 68 sgg. Imperial Cult at Luxor, « Archeologia » 95, 1953, pp. (4) Kinch, L'Arc de Triomphe de Salonique, Tav. 85 sgg; Fr. W. von Bissing, Aifchristl, Wandmale- IV (zona superiore) e p. 13. reien aus Aegypien, «Festschr. zum 60, Geburtstag (5) L'orbiculus è collocato proprio dove normalvon Paul Clemens, Bonn 1926, pp. 181 sgg. mente ha il suo posto: cioè sopra il ginocchio, poco (2) Cagiano de Azevedo, op. cit, pp. 24, 26 sg. — sopra l'orlo inferiore della tunica. Perciò qui, senza 88

PIAZZA ARMERINA nemmeno un orbiculus è plasticamente riprodotto. Tanto nell’Arco di Galerio quanto nell'Arco di Costantino orbiculi e altri ornamenti delle vesti furono senza dubbio rappresentati originaria mente con colorazione. .

La datazione ed interpretazione proposte Azevedo per l'opera musiva di Piazza

da M. Cagiano Armerina.

de

Secondo il Cagiano de Azevedo i mosaici di Piazza Armerina non costituiscono un'unità, ma opera di parecchie generazioni, databile agli anni dalla seconda metà del 1v secolo all'inizio del

v secolo (360 circa-420 circa) (1). L'insigne studioso respinge dunque la nostra datazione dell'opera agli anni attorno al 300. Abbiamo già ribattuto uno dei suoi argomenti, cioè che la ricca

ornamentazione delle vesti con orbiculi non sarebbe usata prima di Costantino. Abbiamo anche potuto confutare una concezione che pare diffusa e della quale il Pace ()) come anche il Cagiano de Azevedo (*) sono sostenitori: cioè che il berretto tondo come insegna pubblica non sarebbe esclusivamente tetrarchico. Come abbiamo detto sopra e come ripetiamo, data l'impor-

tanza della cosa, emerge chiaramente dagli stessi monumenti che il berretto tondo nella vita pu-

blica civile è stato usato un'unica volta, cioè sotto la tetrarchia. Il «vecchio » con berretto tondo nell’ala destra della «Grande Caccia» (Tav. XXII δ), è, come abbiamo visto, un tipo puramente tetrarchico e in tutti i suoi tratti dominato dalla grande

immagine imperiale del tempo; gli si cercherebbe invano paralleli tra i ritratti della seconda metà del rv secolo. È dunque assolutamente escluso - e già il berretto tondo lo rende impossibile - che abbiamo dinanzi a noi un ritratto di Virio Nicomaco Flaviano, come vuole il Cagiano de Azevedo (1). Non voglio negare la possibilità che si tratti di ritratti, per esempio in un togato con palma

nella scena del Circo Massimo e nelle persone principali nelle scene di sacrificio e di banchetto nella « Piccola Caccia », ma non posso credere con il Cagiano de Azevedo che questi ritratti rappresentino necessariamente parecchie generazioni. Come sopra abbiamo mostrato, sia le acconciature femminili sia la foggia maschile dei capelli ¢ della barba mostrano le forme tipiche dell’epoca tetrarchica e della prima età costantiniana, Ritratti reali sono, verosimilmente, una matrona con i suoi due figli nelle Terme (5), nei quali il Cagiano de Azevedo (*) vede la consorte di Nicomaco Flaviano (Tav. XVI δ), con il figlio Appio Flaviano Dexter e la figlia Galla e ciò nonostante che la matrona abbia un'acconciatura inequivocabilmente tetrarchica. Anche se fosse possibile la datazione, l’identificazione sarebbe esclusa non trattandosi nel mosaico di figlia e figlio ma

di due maschi. AI Cagiano de Azevedo è sfuggito che entrambi i figli portano la tipica veste machile con la tunica corta e sopra questa la penula, Anché il Gentili, che tuttavia data il mosaico all'epoca tetrarchica, è incorso nella stessa svista. Cade pertanto la sua identificazione dei personaggi rappresentati con membri della famiglia imperiale — l'imperatrice Eutropia (la consorte di Massimiano) con i figli Massenzio e Fausta. Il Cagiano de Azevedo confronta i mosaici di Piazza Armerina con i mosaici di una casa di Mopsuestia in Cilicia, dei quali L. Budde ha dato qualche anticipazione in due articoli (€. La dubbio, si tratta di un orbiculus e non, come il Ca(3) Cagiano de Azevedo, op. cit., p. 18. giano de Azevedo pensa, di un medaglione di metallo — (4) Cagiano de Azevedo, op. cit., pp. 24 sg.

applicato. (5) Gentili, Mosaici figurati, Tav. IL 1) Cagiano de Azevedo, op. cit, pp. 15 seg. (6) Cagiano de Azovedo, op. cit, p. 25. (2 Pace, I mosaici, op. cit, p. 104, nota 23; p.117, (7) Die rettende Arche Noahs, « Rivista di Arch. nota 18. Cristiana », 32, 1956, pp. 41 gg; Die frühchrisichen 89

H. P. ORANGE

somiglianza di questi mosaici, che si datano tra il 392 e il 428 d. C., con i mosaici di Piazza Armerina secondo il Cagiano cè impressionante » ed egli pensa che la datazione tarda da lui proposta per i mosaici di Piazza Armerina trovi così una «lampante conferma » (1). Confesso che mi è impossibile seguire su questa via l'illustre studioso. A prescindere dal generale carattere tardo-antico, i mosaici di Mopsuestia, secondo il mio parere, non hanno nulla in comune coni mosaici di Piazza Armerina. Tutto il sistema di ornamento è diverso, la stilizzazione dell’acanto e dei girali ha un carattere del tutto differente e anche lo stile figurativo è sostanzialmente un altro: i mosaici di Mopsuestia parlano un linguaggio artistico di un'epoca molto posteriore, con una assai più progredita astrazione e semplificazione nel fogliame e nelle figure e una molto più avanzata riduzione dei valori plastici a piani e linee. Questo contrasto diventa in particolar modo evidente dove il tema figurativo è lo stesso a Piazza Armerina e a Mopsuestia 6 così, anzitutto, nelle rappresentazioni degli animali. Se collochiamo per esempio il leone nel mosaico di Noè a Mopsuestia (*) (Tav. XXV δ) a lato di quello della « Grande Caccia » (Tav. XXV a), non è secondo il mio parere la somiglianza, ma la fondamentale differenza che colpisce, Come nello spazio geografico, così anche nello stile c'è una grandissima distanza tra i mosaici di Piazza Armerina e quelli della cilicia Mopsuestia. Una chiara discrepanza stilistica io trovo anche tra l’arte di Piazza Armerina e quella delle altre più lontane province dell'Impero: si possono, per esempio, paragonare con Piazza Armerina i mosaici di Siria, di Gallia e Germania (?). Senza dubbio sarebbe più naturale cercare paralleli stilistici validi per la datazione nelle vicinanze della Sicilia, dove, come abbiamo visto, è disponibile un abbondante materiale di confronto. Come

nell'arte dei decenni intorno al 300 d. C. in queste regioni vicine, così anche a Piazza Armerina domina lo stile popolare d'Occidente che aveva il suo centro d'irradiazione in Roma, che ha anche

influenzato l’arte contemporanea dell’Italia del Nord, come si vede, per esempio, nei mosaici con scene di caccia a Oderzo in Provincia di Treviso (Tav. XXVII 6, c), recentemente pubblicati

da P.L. Zovatto (€), i quali presentano chiare somiglianze con i mosaici di Piazza Armerina (Tav. XXVII o, XXVI a), con quelli di Roma (Tav. XXVI δ) e dell’Africa del Nord. Questo stile, che si impone verso la fine del m secolo e ancora caratterizza i sarcofagi romani dell’epoca di Costantino, viene respinto dal nuovo classicismo tradizionalista che dà il tono alla seconda metà del rv secolo. Espressioni caratteristiche di questo nuovo stile sono in Roma il sarcofago di Giunio Basso e un gran numero di altri sarcofagi con scene della Passione, mentre a Costantinopoli lo sono i rilievi nel basamento dell'obelisco di Teodosio e la statua di Valentiniano ΤΙ ~ opere tutte queste di un modo stilistico totalmente diverso da quello di Piazza Armerina

e del nostro gruppo nordafricano. Tutto il carattere stilistico dei mosaici di Piazza Armerina s'inquadra assai male nel complesso dell'arte classicista e retrospettiva della seconda metà del 1v secolo. Cagiano de Azevedo fonda la sua tarda datazione dei mosaici di Piazza Armerina anche su di un'osservazione di E. Nash (5). «E. Nash notò, a proposito della riproduzione del Circo

Massimo, che l'obelisco non è al centro della spina, ma è spostato lateralmente verso la metà opposta ai carceres. Egli mise in relazione questo fatto con quanto dice Ammiano Marcellino Mosaihen von Misas-Mopsuestia in Kilikion, «Pantheon », 18, 1960, III, pp. 116 sg. (1) Cagiano de Azevedo, op. cit, p. 26. (2) Fig. da L. Budde, op. cit, Fig. 21 a p. 57. (3) D. Levi, Antioch Mosaic Pavements, Princeton, 1947; K. Parlasca, Die rimischen Mosaiken in Deutschland, Berlin 1959.

(4) P. L. Zovatto, Mosaici Opitergini con scene dell'aria aperta, « Critica d'Arte», 20, 1957, pp. 97 spi; Mosaici paleocristiani dele Venezie, Udine 1963. (5) E. Nash, «Rom. Mitt.» 63-64, 1056-57; Cagiano de Azevedo, op. cit, p. 26, cfr. Lugli, op. cit, PP. 67 sgg, 78.

PIAZZA ARMERINA

(XVI, 10, x7 e XVII 4, 12-16) narrando di Costantino che nel 326 aveva voluto collocare un secondo obelisco sulla spina e che per fare ciò aveva spostato da un lato quello erettovi da Augusto. Ma il trasporto dell'obelisco dall'Egitto si fermò ad Alessandria e fu portato a termine solo da Costanzo nel 357 (Platner-Ashby, Topographical Dict., s. v. ‘Obeliscus Augusti” e “Obeliscus Constantini’) ». Ciò dimostrerebbe, secondo il Nash, che il mosaico dovrebbe datarsi agli anni fra il 326 e il 357. Il Cagiano

de Azevedo,

accettando

questa interpretazione

dello spostamento

dell'obelisco,

lascia però valere soltanto il terminus post quem di questa datazione, cioè l’anno 326, non il

terminus ante quem, cioè l'anno 357, essendo questa data incompatibile con la sua datazione. Se il Cagiano de Azevedo, come il Nash, presuppone una tale rigorosa e minuta oggettività nella descrizione del mosaico di tutti gli elementi della spina, allora logicamente avrebbe dovuto concludere che i mosaici di Piazza Armerina sono anteriori al 357, l'anno dell'erezione dell'altro obelisco. Trovandosi, però, esempi tra i quadri del Circo Massimo databili post 357, in cui ci si accontenta di rappresentare solo uno dei due obelischi e si trascura l'altro, come ha osservato il G. Manganaro sul dittico dei Lampadi, il Cagiano de Azevedo non si sente legato all'anno 357 come terminus ante quem. Ma conseguentemente, mi pare, non dovrebbe sentirsi legato neanche all'anno 326 (l’anno dello spostamento) come terminus post quem. Infatti, proprio come il dittico dei Lam-

padi trascura il numero degli obelischi, così il posto centrale dell’obelisco d'Augusto viene trascu-

rato in numerose rappresentazioni del Circo Massimo databili a prima dell’anno dello spostamento. Come esempi riproduciamo due sarcofagi nel Vaticano (Tav. XXV c, d) e uno nel Museo Nazionale di Napoli (Tav. XXV e) dove Eroti fanno da aurighi o da cavalieri nella gara del Circo. In uno,

nella Sala della Biga, del rr o mt secolo (Tav. XXV c), la gara viene rappresentata due volte, così sulla cassa come sul coperchio: sul coperchio lo spostamento dell’obelisco è ancora più pronunciato che nel quadro di Piazza Armerina, sulla cassa lo è meno. Nell'altro sarcofago Vatica-

nense, nel Museo Chiaramonti, databile circa al 300 (Tav. XXV d), lo spostamento dell'obelisco ancora una volta è più marcato che a Piazza Armerina e lo stesso si dice per l'obelisco sul sar-

cofago del Museo Nazionale a Napoli del III secolo (Tav. XXV e).

La tipica rappresentazione del Circo Massimo ha sempre dovuto ridurre e semplificare il com-

plesso quadro dei monumenti, santuari, statue ed altri elementi della spina. Per l'artista l'intenzione essenziale era di mettere ben in rilievo la corsa e nello stesso tempo di indicare i monu‘menti più salienti e caratteristici sulla spina del grande circo di Roma,

l'ambiente in seriso topografico.

non invece di riprodurre

Anche A. Ragona () prende come punto di partenza per la datazione dei mosaici di Piazza Armerina l’obelisco del circo sostenendo che si debba trattare dell'obelisco eretto da Costanzo perché porta una fiamma sulla cima: di una tal fiamma sul detto obelisco (facis imitamentum... aureum... auro imbracteatum, velut. abundanti flamma. candentis) parla Ammianus.

Marcellinus (XVII, 4, 15), L'erezione dell’obelisco nel 357 darebbe dunque un terminus post quem al mosaico. A parte tutti i nostri dubbi riaguardo al fatto che si tratti qui davvero di una fiamma, dobbiamo, però, con A. Carandini, (*) accettare che una tale fiamma ap-

(1) A. Ragona, Un sicuro punto di partenza per la cioè una via di mezzo tra la tarda datazione alla datazione dei mosaici della Villa romana di Piazza fine del τν e al v secolo di Pace-Cagiano de Azevedo Armerina, Caltagirone 1961 © la datazione anteriore in epoca tetrarchica del Gen(2) A. Carandini, Ricerche sullo stile e la cronologia ili, Lavin e dell'autore di questo articolo. Il Carandini dei mosaici della Villa di Piazza Armerina, «Studi si basa in primo luogo sui ritratti e i tipi fisionomici, Miscellaneis, 7, Roma 1964, pp. 39 sg. Il Carandini sulle acconciature, sui vestiti e gli attributi. Mi pare, perviene ad una datazione dei mosaici di Piazza Ar- però, che le sue analisi formali abbiano un carattere merina nel cinquantennio compreso tra il 320-370, troppo generico per poter fissare distinzioni esatte. gr

H. P. ORANGE

posta all'obelisco non era riservata esclusivamente all'obelisco di Costanzo. Dice a questo proposito Isidoro (XVIII, 31) nella sua trattazione sugli obelischi: Summo obelisco superpositum est quoddam. auratum in modo flammae. [ormatum.. . INTERPRETAZIONE MONIALI.

DEI

MOSAICI

DELLE

DUE

RIPARTIZIONI

CERI-

Adventus Augusti. Con un significato ben chiaro la rappresentazione di un adventus si trova proprio all'ingresso (Pianta n. 2) della ripartizione cerimoniale (1) maggiore del palazzo: cioè nel pavimento del vestibolo che conduce al grande atrio (Tav. XXIX, XXX a, 5). Studiando nel 1952, durante una visita a Piazza Armerina, questo importantissimo mosaico insieme col prof. R. Bianchi Bandinelli, eravamo perfettamente d'accordo che doveva trattarsi d'una rappresentazione dell’adventus Augusti (?) e non già, come si è voluto, di una scena religiosa di misteri (). Uomini in vesti policrome da festa, con ricchezza di ricami, con corone di foglie e fiori sul capo, si fanno incontro ἃ chi entra: il loro sguardo fissa soltanto lui, i loro gesti sono a lui diretti, a lui essi porgono il saluto di benvenuto. Ma, come abbiamo mostrato (*), questo saluto si innalza a forme così solenni, ad un giubilo che non è concesso al comune mortale. Gli astanti reggono candelabri accesi o tendono

rami

di lauro

verso

colui che arriva, o, tenendo

davanti

al viso dei dittici,

innalzano inni ed acclamazioni cadenzate porgendogli rituali auguri di benvenuto. Solo ad un

imperatore spettano tali onori ed ovazioni. L'imperatore è ubique victor ed ha sempre diritto al ramo di lauro; è all'imperatore che vanno rivolti gli inni e le acclamazioni come quelli che qui si levano; è diritto dell’imperatore l'essere accolto con la luce cerimoniale dei candelabri. Dalla let-

teratura conosciamo assai bene il rituale imperiale dell’adventus e sappiamo che una parte es-

senziale dello stesso è proprio costituita dalla luce delle fiaccole, dall'offerta di lauro e fiori, dagli inni augurali. In relazione al nostro mosaico possono essere citate per esempio le parole di Ammiano Marcellino (*) sull’adventus di Giuliano l’Apostata a Sirmium: cum lumime multo δὲ floribus, votisque faustis... duxit in regiam. I giovani recitanti del nostro mosaico offrono, evidentemente, vota fausta, omina fausta dello stesso genere (9).

La

caccia globale per l'Anfiteatro di Roma omnia ex toto orbe terrarum animalia.

- La

cattura

di

Dietro il vestibolo con il mosaico dell'adventus si trova il grande atrio aperto con quadriportico (Pianta n. 3). I mosaici del quadriportico rappresentano protomi ferine in corone di alloro (Tav. XXXI ac; IL a-c). Tre scale, una media e due laterali, conducono dall'atrio attraverso la facciata colonnata

del nartece

allo stesso

(1) Sull'aspotto cerimoniale del palazzo di Piazza Armerina: Dyggve- L'Orange, «Symb. Osl. » 29, 1952, PP. 120 sgg. (2) L'Orange, The Adventus Ceremony, e Late Classical and Mediaeval Studies in Honor of A.M. Friend jr.», Princeton 1955, pp. 7 888. (3) Gentili, I mosaici della villadi P.A., op. cit, pp. 37 sgg.; Idem, La villa romana, op. cit., pp. 36

(Pianta

n.

4).

sg.; Pace, I mosaici, op. cit, Fig. XX a p. 115: Rito misferico. Adesso il Gentili accetta che si tratti di una rappresentazione dell'adventus Augusti: Gentili, La villa imperiale, op. cit, pp. 54 58. (4) Vedi nota 2. (5) XXI, 1o, 1. (6) L'Orange, The Adventus Ceremony, op. cit, PP. 7 SEE92

PIAZZA ARMERINA

Il mosaico con la «Grande Caccia» (Tav. XLV; XXXV) copre il pavimento del nartece, cioè il lunghissimo (ca. 60 m.) e strettissimo ambiente, che è posto come pronao davanti alla basilica (Pianta n. 5) e ha la fronte verso l'atrio. La basilica con il basamento di un trono nell'abside costituisce il culmine terminale di una successione assiale di ambienti, la quale abbraccia l'atrio con quadriportico, il pronao o nartece e la stessa basilica. L'asse di questa successione segue la linea media della scala centrale che porta dall’atrio al nartece e continua dal nartece lungo la linea media della scala che dal nartece porta all'aula basilicale. ΤΙ detto asse della successione spaziale segna il centro ideale e la parte media dell'enorme mosaico della «Grande Caccia»: una parte che, disposta sull'asse, occupa la superficie, approssimativamente quadrata, tra le due scale centrali (Tav. XXXV a). Questo centro ideale di tutto il quadro musivo non coincide, però, con il centro geometrico del nartece essendo l'ala destra alquanto più lunga della sinistra. Verso questo centro ideale gravitano

tutte le attività e tutti

i movimenti

delle due lunghissime ali del mosaico. Qui sboccano, come vedremo, tutti i rivi degli eventi, qui si risolve il dramma dell'azione e si concentra il significato e il pensiero del quadro totale, Questa parte centrale rappresenta una terra, che, ad entrambi i suoi lati, è chiusa da un mare, ma nello sfondo si continua in terra ferma: la chiameremo «la terra tra i due mari ». Che si tratti di mari a differenza delle immagini di laghi e fiumi che si trovano nelle due ali del mosaico, appare dalla fauna d’acqua con delfini e polipi. Verso questa terra tra i due mari, nel centro ideale del mosaico, si dirigono dalle due ali del

mosaico le navi col bottino delle cacce e qui il bottino viene scaricato e raccolto. Per rendere evidente questa operazione una nave è pronta sopra ognuno dei mari, cosicché le navi congiungono rispettivamente uno dei continenti laterali con la terra di mezzo e si possono vedere le belve, che da uno dei continenti laterali sono avviate o portate attraverso una passerella nella

nave per poi essere trasportate dalla nave, sempre attraverso una passarella, alla terra di mezzo.

Ma questo movimento verso la terra di mezzo non si limita al passaggio attraverso le passerelle

e le due navi, bensì decide anche della direzione e dell'attività complessiva che si svolge sui due

continenti. Nel continente di destra le belve catturate vengono condotte da destra a sinistra verso la nave e, al contrario, nel continente di sinistra, da sinistra a destra (Tav. XLV).

Parecchie delle belve sono comuni ai due continenti, altre no: soltanto sul continente a destra s'incontra la tigre, l'elefante, il cammello, il rinoceronte, l'ippopotamo ed il bisonte; e soltanto sul continente a sinistra lo struzzo e il cinghiale. Il continente a destra rappresenta l'Oriente: Asia, Arabia, Egitto; perciò troviamo su questo lato anche un grifone che era considerato un animale dell'Oriente (1). Il continente a sinistra, al contrario, rappresenta l'Occidente: in primo luogo il Nordafrica, E la terra tra i due mari che raccoglie il bottino di queste cacce nell'Oriente e nel

l'Occidente non può essere che l’Italia. È alle venationes della Roma imperiale che vengono portate le belve catturate, Il nostro quadro musivo vuole suggerire il concetto delle miriadi di belve che

vengono radunate da tutti gli angoli del mondo per le venationes delle grandi feste di stato di Roma. Tutto questo rico quadro di caccia è contenuto entro due mosaici a emiciclo alle sue due estremità. All'ala destra troneggia una personificazione femminile di pelle piuttosto scura, nuda con mantello rosso, che tiene in mano una zanna d'elefante (Tav. XXXII). Si tratta di una provincia orientale,

collocata entro la sua natura, fra una tigre femmina

e un elefante e con la

fenice, il favoloso uccello dell'Arabia e dell'Egitto a lato (). Nel mosaico corrispondente all'ala (1) Philostratus, Vita Apollonii IIT, 48. (2) Gentili, Mosaici figurati, op. cit, Tav. XXXVI. Si considera sempre luogo di origine della fenice un ‘paese asiatico, in genere l'India o l'Arabia, ma anche

la Siria o l'Assiria; e quando viene nominata l'Etiopia, è nel senso di Etiopia come l'antico paese leggendario connesso alle regioni asiatiche (A. Rusch, RE. XX, 1, col. 415, s. v. Phoinix). 93

H. P. L'ORANGE

sinistra, troneggia, conservataci soltanto in frammenti(), pure una personificazione femminile all’aperto, che porta una tunica a lunghe maniche, una cintura ornata di gioielli e ha una pelle di leopardo fissata sul petto con una fibula tonda (Tav. XXXIII). Come quella che le fa riscontro sull'ala destra essa porta al collo un cerchio verosimilmente con bulla e ha — sempre come l’altra - una grande chioma fermata da una benda e ricadente sulle spalle. Qui come là due belve si fronteggiano ai lati di tale personificazione (Tav. XXXIII ὁ, c): i frammenti mostrano le tipiche belve dell’ Africa del Nord, il leopardo e il leone che dominano tutto il continente dell'ala sinistra. Anche lo struzzo e il cinghiale hanno una parte importante in quest’ala: lo struzzo indica l'Africa, il cinghiale forse l'Europa. La provincia in pelle di leopardo sta per l'Occidente e specialmente per l'Africa, che nell'antichità era limitata all'Africa nord-occidentale, restandone escluso l'Egitto. I due mosaici a emiciclo con i quadri dei complementari rappresentanti dell'Oriente e dell'Occidente racchiudono e conglobano le cacce estese a tutte le regioni del mondo, Le zone di caccia in Oriente e in Occidente si estendono, per tal modo, all'intero orbis ferarum. È evidente che tali cacce, che raccoglievano masse di animali dalle più lontane province per metterle a disposizione dell’arena romana, potevano essere organizzate soltanto dallo stato romano, dall’imperatore. Sappiamo che le venationes imperiali nell'arena romana cercavano appunto questa piena totalità delle varie specie di animali: tutte le province dovevano dare il loro tributo, gli animali di tutta la terra — omnia in toto orbe terrarum (animalia) — dovevano essere presenti per dimostrare l'universalità dell'impero (ἢ). Nella Vita Pii ( si racconta che in certi spettacoli dati dall'imperatore, verosimilmente nei suoi decennalia o vicennalia, egli offriva elephantos ed corocottas et tigrides et rhinocerontes, crocodilos etiam atque hibbopotamos et omnia ex toto orbe orrarum cum tigridibus exhibuit ; centum etiam leones una missione exhibuit. A lato degli animali di tutta la terra è stato rappresentato nella « Grande Caccia» anche un animale celeste, cioè un grifone (ἢ), il quale si getta sopra una gabbia al cui interno sta un cacciatore (Tav. XXXIV a) (*, ©). W. Seston ha combinato questo grifone celeste della «Grande Caccia » cogli animatia celestia dei quali parla un editto di Caracalla. « Je me demande enfin » — mi scrive il Seston ()) — «si la présence du-grifion, animal «céleste » dans le défilé des bêtes destinées à ces venationes, ne marque pas le caractère impérial de tous ces animaux: je songe aux «animalia celestia » des forêts maurétaniennes dont parle un édit de Caracalla (L'année épigraphique 1948, n. 109) dont on s'est demandé s'ils étaient des éléphants ou des lions (cfr. J. Guey, Revue des études anciennes 49, 1947, pp. 248 sgg.) et qui

étaient, comme la mosaique de Piazza Armerina le suggère, toutes les bêtes féroces destinées à

une féte impériale ». In questo connesso dobbiamo sottolineare che, quando un’altra volta nella (x) Gentili, « Boll. d'Arte », 37, 1952, 39, Fig. 13. Cr. Neutsch, op. cit., pp. 578 seg. (2) Aymard, op. cit. pp. 184 sgg. (3) Script hist. Aug, Vita Pii, X, (4) Gentili, La villa imperiale, op. cit, Fig. 17.

in Oriente quanto in Occidente, l’animale sacro a Helios Apollo, del quale poteva trainare il carro, ed era perciò anche l’animale del re solare, che egli trasportava attraverso lo spazio celeste e di cui reggeva il trono (F. Cumont, L'aigle funéraire, pp. 94 sgg: Roscher,

(8) L'episodio si può paragonare con scene di are- Myth. Lev, s. v. Malachbelos 2, 2, col. 2300, F. W. na, come quella in cui un orso si getta sopra una Drexler; L'Orange, Iconography of Cosmic Kingship gabbia in cui sta un cacciatore, in un dittico di Areo- in the Ancient World, Oslo 1953, pp. 67 sgg. 72 S88. bindo, R. Delbrück, Consulardipiychen; Berlin 1929, 118 sgg. Ci sovviene del racconto della tarda anti. Tav. it. chit’ di come Alessandro nell'Estremo Oriente fece (6) Sia che si rappresentasse il grifone come un catturare uccelli giganteschi per il suo viaggio celeste. animale che vive realmente nell'Estremo Oriente (Phi- (L'Orango, Iconography, op. cit, pp. 118 sgg). lostratus, Vita ApolloniIT, 48) oppure come un essere (7) In una lettera del 2-12-1952. mitico, esso era comunque per il mondo antico, tanto 94

PIAZZA ARMERINA

tarda antichità un grifone compare in un quadro realistico di caccia, è cosa che succede nel palazzo imperiale di Costantinopoli () (Tav. XXXIV 5). Qui - non altrimenti che nella « Grande Caccia » di Piazza Armerina — il grifone imprime al quadro della caccia il suo alto segno: si tratta di animalia celestia, di belve destinate alle feste imperiali. Abbiamo dunque davanti a noi delle venationes nel grande stile imperiale come, mutatis mu tandis, esse si ripeteranno nell'arena romana. Ma di più: l’imperatore stesso è presente, se non di persona, nei tipici rappresentanti del suo governo. Tutte le catture sono poste sotto il controllo

dell'autorità statale, quell’autorita, cioè, che viene caratterizzata dal berretto tondo e dal bastone ἃ fungo, vale a dire l'impero tetrarchico. Il fattozche questi rappresentanti dell'impero tetrarchico siano collocati nel centro ideale del mosaico, nella terra tra i due mari, cioè nel centro del mondo, Roma, ha un significato speciale, non soltanto per l'interpretazione della « Grande Caccia », ma per tutta la nostra concezione della villa di Piazza Armerina, tanto più che essi compaiono in un gruppo di quattro, il numero simbolico della tetrarchia che rispecchia. il collegio imperiale e tutta l'amministrazione del regno. Che si tratti di un gruppo di quattro si può dimostrare

in modo

esatto, anche

miniamo qui i resti delle figure.

se il mosaico

è soltanto frammentariamente

conservato.

Esa-

Il gruppo si trova, come abbiamo detto, al centro ideale del mosaico con un gruppo di due a destra dell'asse medio e un gruppo di due a sinistra dello stesso asse. Il gruppo di destra è relativamente ben conservato (cfr. con Tav. XXXV a la Tav. XXII a e la Tav. XXXV d). Tutte e due le figure portano il berretto tondo e tengono nella mano sinistra il bastone a fungo. Di più la figura a sinistra tiene nella destra una frusta lunga e sottile, leggermente arcuata, fatta di una unica striscia di fesserae grigio-nere.

(La frusta manca

nel disegno

Gentili;

cfr. la nostra Tav.

XXXV d. La striscia nera dalla frusta in giù è una fesatura e non fa parte della frusta). È della stessa forma di quella frusta che abbiamo visto in mano ad un funzionario del continente occidentale (Tav. XXII c) che con essa sta punendo uno dei cacciatori. La frusta deve essere con-

siderata, quindi, come un attributo di autorità, poiché l’uso della stessa non è limitato agli ani-

mali ma è anche destinato alla punizione di subordinati. Immediatamente a sinistra della figura colla frusta, a sinistra dell'asse medio e all'altezza. del nostro gruppo destro di due, cioè sopra l'elefante (Tav. XXXV a), vi è un vasto campo di mosaico distrutto. In questo campo, che per la sua posizione e la sua forma darebbe posto ἃ un gruppo di due, in corrispondenza del gruppo di due da noi già esaminato, i frammenti conservati rivelano in modo indiscutibile che originariamente un tal gruppo si trovava qui. Della figura a sinistra di quella colla frusta, cioè della terza del nostro gruppo a quattro, è conservato un frammento della clamide sulla spalla sinistra (Tav. XXII a), oltre il piede sinistro che si trova all'altezza dei piedi delle due figure di destra (Tav. XXXV d) insieme con un piccolo lembo della clamide che pende dietro il piede. Il piede destro si nasconde dietro all'elefante che sta innanzi

(Tav. XXXV a). Della quarta figura del nostro gruppo di quattro rimane soltanto un piccolo frammento del suo attributo significativo, cioè della frusta (Tav. XXXV c, a destra, in alto;

‘manca nel disegno Gentili Tav. XXXV a, che però mostra il posto del frammento tra due campi distrutti del mosaico). La frusta è fatta di un'unica striscia di fesserae grigio-nere, esattamente

come la Un XXXV verso il

frusta sopra analizzata. La parte inferiore della figura doveva essere coperta dall’elefante. funzionario in subordine si trova ad ambedue i lati del nostro gruppo di quattro (Tav. 0). Posti più in basso nel mosaico, ambedue i funzionari, vivacemente mossi, sono rivolti gruppo al quale dirigono parole urgenti alzando la destra in gesto oratorio. Si tratta di

(1) Talbot Rice, Les Mosaiques du grand palais des empereurs byzantins à Constantinople, «La Revue des Arts» n. 3, 1955, 162, Figg. 6 e 166, 4. 95

H. P. L'ORANGE

ufficiali di ordinanza che consegnano prontamente i loro messaggi e attendono di ricevere nuovi ordini. Indicativo della funzione del messaggio è il rotolo che l'ufficiale di sinistra tiene nella mano e che originariamente, senza dubbio, era tenuto anche dall'ufüciale di destra. Tutto il gruppo di quattro, disposti identicamente nello stesso ordine, con gli ufficiali d'ordinanza ai lati, forma una composizione simmetricanel centro dominante di tutto il quadro della «Grande Caccia ». Come irradiati da questo centro organizzativo, funzionari statali sono inviati in ciascuno dei due continenti: ravvisiamo uno di questi funzionari nel personaggio anziano e specialmente distinto, protetto da due guardie sul continente orientale; egli porta il berretto tondo ed il bastone a fungo. Troviamo sull’ala del continente occidentale un altro funzionario, questa volta senza berretto, ma con bastone e frusta, il quale prende parte all’azione, in atto di infliggere una punizione sul luogo dell'imbarco. Una singolare rispondenza tra il rappresentante dell'autorità imperiale e lo stesso imperatore si nota nella sorprendente rassomiglianza dell'alto funzionario anziano al tipo dell'imperatore barbuto della vecchia tetrarchia (1): rassomiglianza che, al momento della scoperta mi condusse alla supposizione sbagliata che si trattasse d’un ritratto imperiale. Il Cagiano de Azevedo qui ha perfettamente ragione colla sua critica alla mia tesi. Così i quattro imperatori, nelle loro imperatoriae dispositiones, sono presenti nell'organizzazione della « Grande Caccia », non altrimenti che sulla Base dei Decennali nel Foro Romano gli imperatori appaiono nei quattro vexilliferi della processione dell'esercito (?). Così agiscono in tutti gli officia attraverso i loro funzionari: a vobis proficiscitur etiam quod per alios administratur dice il panegirico a Massimiano Augusto (*). Tutto l'impero quadripartito è dunque presente nel nostro quadro di cacce mondiali. La stessa struttura dell'amministrazione tetrarchica si manifesta nel numero quattro, il quale nella retorica dei panegirici contemporanei diventa un numero cosmico e fausto, simboleggiante la potenza divina degli imperatori regnanti, e viene derivato dalla struttura universale, dai quattro elementi, dalle quattro stagioni, dai quattro volti del cielo. Isto numinis uestri numero summa omnia nituntur et gaudent : elementa. quattuor et totidem anni uices et orbis quadrifariam etc. etc. (ἢ. Come abbiamo visto, ancora a Piazza Armerina l'uomo dal berretto tondo sovrasta le terre come figura dominante nel paesaggio dello stato e della realtà temporale. È una silhouette significativa che si disegna nel gruppo degli uomini col berretto tondo sul rialzo di terra tra i due mari, cioè nel centro di quell'orbis terrarum che si estende davanti a noi. L'autorità dello Stato torreggia su questo gruppo. Pochi anni dopo l’uomo col berretto tondo e tutto quello che egli rappresenta sarà scomparso da questo rialzo -- anzi da tutto il paesaggio dello Stato. I persecutori, dice Eusebio, sono precipitati nel nulla e sembra che mai neppure siano stati nominati — ὡς μηδὲ πώποτε ὠνομάσϑαι δοκεῖν (5).

L’Ercole Imperiale. La ripartizione cerimoniale minore a Piazza Armerina (pianta n. 10, 11) ἃ, come quella maggiore, una ripartizione assiale con un atrio aperto sul davanti e collegato a un triclinio monumentale sul fondo, Nella ripartizione minore non c' un nartece davanti al triclinio; per una scala, (1) Dopo la scoperta del gruppo a quattro al cen- dem Forum Romanum, « Róm. tro della « Grande Caccia », il «vecchio» deve es 14 sgg, Tav. 4 e Figg. 4. sere considerato rappresentante dell'impero tetrar(3) Mamertini Paneg. Max. chico, rispecchiando il tipo imperiale del tempo, non — tier, Paneg. Latins, 1, p. 34). però l'immagine dell'imperatore stesso come prima (4) Incerti paneg. Constantio pensavo (+ Symb. Osl. + 29, 1952, PP. 117 sgg). tier, Paneg. Latins, I, p. 85). (2) L'Orange, Ein tetvarchisches Ehrendenkmal auf (5) Eusebius, Hist. Eccl. X, 96

Mitt.» 53, 1938, pp. Aug. 11. (Ed. Galle Caesari 4. (Ed. Galle4, 15.

PIAZZA ARMERINA

attraverso un'arcata tripartita a colonne, si entra nel triclinio, la grande sala tricora (1). Anche il programma iconologico è, come vedremo, analogo in entrambe le successioni di ambienti. Il portico dell'atrio, che è di forma ellittica, reca mosaici ‘pavimentali i quali ripetono una infinità di protomi ferine entro girali di acanto (*) (Tav. XXXVI a, ὃ). I mosaici della grande sala tricora costituiscono il culmine della decorazione musiva delle ripartizioni cerimoniali, come la grande soluzione alla quale tendono le protomi ferine e le cacce: le fatiche di Ercole come simboli del governo dell'imperatore. Ripeto qui l'interpretazione da me proposta nel citato articolo in Symb. Osl. 1952 e alla quale anche il Gentili ha aderito (). La grande sala tricora consta di un ambiente quadrato con tre absidi, una a sud, una ad est e una a nord. La parte principale del triclinio, cioè il vasto ambiente centrale quadrato, è dominata da Ercole: la sua strage delle bestie nel mondo mitico corrisponde alla caccia degli animali nel mondo reale. I mosaici (Tav. XXXVII; I a) hanno conservato i cavalli di Diomede, il toro di Maratona, il leone nemeo, l'idra lernea, Gerione, il cinghiale erimanteo, la cerva. cerinitica (4) (trasformatasi in un cervo) e Cerbero. Soltanto le vittime sono rafigurate; non invece Ercole. Il fatto che la cerva cerinitica nel nostro mosaico si sia trasformata in un cervo, si spiegherebbe, forse, con l'identificazione della mitica caccia dell’eroe con una reale caccia dell'imperatore. In ogni modo la caccia al cervo occupa un posto importante nel mosaico della « Piccola Caccia », il quale pare che rappresenti scene di cacce imperiali (9. La parte più significativa del triclinio, cioè l'abside mediana, è dominata da Giove assistito da Ercole: la grande lunetta del pavimento rappresenta una scena della gigantomachia (Tavv. XXXVIII; 1), battaglia che gli olimpici hanno vinto con l'aiuto di Ercole. Anche qui non si vedono che le vittime della strage, i giganti feriti dalle frecce di Ercole, ma non si vede Giove né l'eroe stesso. Delle due absidi laterali, quella a nord è sotto l'insegna di Ercole, quella a sud sotto l'insegna di Dioniso: a nord assistiamo all'incoronazione di Ercole (Tav. XXXIX), a sud al trionfo delle potenze dionisiache su Licurgo (). Per tal modo la specifica teologia della tetrarchia si palesa nel triclinio: Diocleziano, il grande Giovio, assistito da Massimiano, il grande Erculio, dominano questa sala. La gigantomachia è il simbolo che riunisce le due dinastie imperiali della tetrarchia, la Giovia e l'Erculia: il nuovo Giove in terra abbatte, con l'aiuto del nuovo Ercole, tutte le potenze caotiche. Ma Ercole s'accompagna a Dioniso: sono i due conquistatori mitici del mondo, nelle cui immagini re ed imperatori ravvisano se stessi (5). Tutta la nostra interpretazione della decorazione del palazzo che si basa sul principio della similitas

fra

Ercole

e l'imperatore

Erculio

trova

(1) Tl posto centrale nella grande tradizione palazziale e sacrale della sala tricora (del #ricorium, fricorum, trichorum), durante l'Impero Romano e il Medioevo, e la sua prima associazione al culto imperiale, viene messo in rilievo da J. Lavin (The House of the Lord, «The Art Bulletin» 44, 1962 pp. 1 sgg). «1f we accept the attribution of the building to Maximianus Herculius, as I think we must at leastas a working hypothesis, then Piaz a Armerina would make very explicit the association in palace architecture of the triconch with the imperial cult» (Lavin, op. cit, p. ἢ). (9) G. V. Gentili, La villa imperiale, op. cit, PP. 49 sg. (3) Mosaici figurati, p. 31 e nota alla p. 83. (4) Gentili, La villa imperiale, op. cit, pp. 52 sg.

conferma

nella retorica contemporanea

della

e Fig. 32. Pace, I mosaici, op. cit, Tavv. V-VILa, PP. 55 566. © Fig. 11. (9) V. mia contribuzione alla Miscellanea in onore del prof. E. Arslan (im corso di stampa): Un sacrificio imperiale di caccia nei mosaici di Piarra Armerina. (6) Da Alessandro in poi, il conquistatore, sia esso re o imperatore, è accompagnato da Ercole e Dioniso (A. Aldi, Insignien und Tracht der rom. Kaiser, « Rom. Mitt.» 50, 1935, pp. 105 sg.; L'Orange, * Symb. Osl. », op. cit., p. 121). Sull'amicizia tra Εἰ cole e Dioniso: L. Preller, Griech. Mpthol. II, pp. 207 568. Cosi si spiegherebbe, secondo il Gentili, Mosaici figurati, op. cit p. 30, che Ercole nel mosaico della sua apoteosi porti invece della pelle di leone la pelle di pantera di Dioniso.

H. P. L'ORANGE

corte. Il panegirico di Mamertino recitato nel 289 a Massimiano Erculio pone a diretto confronto le vittorie di Massimiano e le fatiche di Ercole (). Dopo aver parlato delle imprese di Ercole, il panegirista si rivolge all'imperatore come praesens deus e toto orbe victor, il quale viene lodato per avere sottomesso un mostro molto più abominevole di quello tricipite abbattuto da Ercole Mamertino pensa probabilmente a Carausio — Giustamente il Gentili raffronta col mosaico un passo del detto panegirico, il quale rievoca le guerre sostenute in comune da Diocleziano e Massimiano con l'immagine della lotta di Giove ed Ercole contro i giganti (*). La duplice forma dei giganti — per metà uomini e per metà serpenti — corrisponde nel panegirico ai gallici Bagaudi nelle loro rivolte contro lo stato imperiale: per metà contadini, per metà soldati. Anche nel Panegirico di Mamertino dell'anno agr la gigantomachia viene rievocata come immagine dell'impero tetrarchico il quale abbatte tutte le forze ribelli (5). Nel mosaico i giganti sono tutti colpiti dalle frecce di Ercole; nessuno dai fulmini di Giove: e ciò sembra fatto con intenzione in quanto l'imperatore Erculio è in questo palazzo colui che dev'essere in primo luogo glorificato. SGUARDO

D'INSIEME

AL

PALAZZO

DI

PIAZZA

ARMERINA.

Si tratta dunque, secondo la nostra documentazione, di un complesso di palazzo dell’epoca della tetrarchia, In questo complesso spiccano due ripartizioni monumentali, le quali mostrano nel modo più evidente quello schema che il Dyggve ha rilevato nei palazzi imperiali di Spalato e di Costantinopoli e dimostrato essere tipico del palatium sacrum imperiale della tarda antichità e del primo medioevo (*): una disposizione che ha i suoi remotissimi antecedenti nei palazzi reali dell'Oriente e che in Occidente si annuncia già coi palazzi imperiali di Roma (). Questo palatium sacrum consta di due parti essenziali: un grande atrio o peristilio per la corte, la guardia ecc. e la sala del trono, il grande triclino per le udienze e le cerimonie imperiali. Generalmente un vestibolo conduce all’atrio ed un pronao o nartece dall’atrio alla sala del trono. Questa successione di ambienti cerimoniali, tipica dei palazzi imperiali, fu adattata al culto cristiano e, come ha dimostrato il Dyggve, si ritrova nelle grandi basiliche romane (6). Così pare che anche il nostro palatium sacrum di Piazza ‘Armerina sia stato, senza grandi alterazioni, trasformato in chiesa ed utilizzato come tale nel Medioevo (7. Nella maggiore delle due ripartizioni cerimoniali di Piazza Armerina gli scavi hanno liberato al vasto triclinio, un'aula absidata di circa 30 m. di lunghezza (Pianta n. 5), dal Gentili pubblicato e interpretato come salone del trono

(8) e che si trova ad

est della

galleria

della

«Grande

Cac-

cia» (Pianta n. 4) la quale fa da pronao o nartece al triclinio (). Il nartece, come abbiamo

(1) Mamertini Paneg. Max. Aug. 2. (Ed. Galletier, Paneg. Latins, L, p. 25). (2) Mamertini Paneg. Max. Aug. 4. (Ed. Galletier, Paneg. Latins, I, p. 27): Praecipitanti Romano nomini iusta principem subisti eadem scilicet auxilii opportunitate qua tuus Hercules Jovem vestrum quondam terrigenarum bello laboraniem magna victorias parle iuvit probavitque se non magis a dis accepisse caeTum quam eisdem reddidisse. An non illud malum simile monstrorum. biformium in hisce terris fuit quod tua, Caesar, nescio utrum magis fortitudine vepressum sit an clementia. mitigatum, cum militaris habitus ignari agricolae appetivorunt, cum arator peditem, cum pastor aquitem.. imitatus est? (3) Mamertini Paneg. gonethl. Max. Aug. 3 (Ed. Galletier, Paneg. Latins, T, pp. 52 st).

— (4) E. Dyggve, Palatium Sacrum Ravennatum, e, «Symb. Osl. », op. cit., pp. 122 sgg. — (s) A. Boethius, The reception Halls oj the Roman Emperors, «Annual οἱ the British School at Athens», 46, pp. 25 sag. L'Orange, Domus aurea — der Sonnen palast, Serta Eitremiana, « Symb. Osl. » Fasc. suppl. 11, 1942, PP. 92 98g. — (6) E. Dyggve, Doedekult, Keiserkult og Basilika, Copenaghen 1943, passim. — (7) Gentili, I mosaici della villa di P. AL, op. cit, p. 34, Fig. 5 e nota 12. Cfr. l'obiezione di Pace, op. cit, pp. 118 sgg. — (E) La villa imperiale, op. cit, pp. 4o sgg. (5) L'Orange, Aquileia e Piazza Armerina, «Studi Aquileiesi offerti a Giovanni Brusin», 1953, pp. 187 sgg.; The Adventus Ceremony, op. cit. pp. 7 Sg. 98

PIAZZA ARMERINA

visto, si apre da un lato verso l'atrio, sopra il quale s'innalza con tre scale e dall'altro si apre, con una larga scalea, sul triclinio. Questo viene perciò a trovarsi ad un livello assai più alto e, quindi maestosamente elevato sopra la galleria del nartece e sopra l'atrio ed il vestibolo (1). Come abbiamo veduto, vi è nel complesso di Piazza Armerina un altro ambiente minore, disegnato secondo lo stesso schema del palatium sacrum (Pianta nn. 10, 11) e da noi, nel citato articolo in Symb. Osl. 1952, interpretato come tale: cioè la sala tricora col suo atrio che fu scavato nel suo insieme prima della grande ripartizione centrale (), Troviamo cioè un atrio aperto con portici e sull'asse longitudinale di questo la sala culminante per cerimonie, il triclinio.

In queste due ripartizioni cerimoniali l’ornamentazione musiva non è disposta a caso nei singoli ambienti ma si ricollega in modo assai significativo al luogo che decora. Come abbiamo

visto, la rappresentazione di un adventus Augusti, con un suo significato ben distinto, sta proprio

all'ingresso della maggiore ripartizione cerimoniale: cioè nel vestibolo che conduce al grande atrio. Dal vestibolo entriamo in questo atrio con portici a colonne su tutti i quattro lati. Verso est, sull’asse longitudinale dell'atrio, il portico si apre verso la scalea mediana che conduce alla galleria del nartece ed al triclinio, I pavimenti dei porticati sono coperti da ricchi mosaici (). Larghe fasce, con trecce policrome, danno origine ad una doppia teoria di riquadri in cui si trovano dei protomi di animali dentro corone di alloro (Tavv. XXXI a-c; Π a-c). Vediamo qui ripetuti l'elefante, il leone, la pantera, l'orso, il cinghiale, l'onagro, il toro, la gazzella ecc. e anche un animale

alato, cioè lo struzzo. Prima di tutto troviamo,

dunque,

le belve dell'anfiteatro; così lo

struzzo è il tipico uccello dell’arena (9). Le protomi preludono al mosaico della « Grande Caccia » nella galleria del nartece, che davanti ai nostri occhi fa rivivere per intero lo stesso mondo ferino. Come nella « Grande Caccia », anche nelle protomi le più feroci tra le belve sono rappresentate con la caratteristica espressione della furia e bestialità della loro specie. L'uomo dell'antichità riconosceva qui tutto quel ruggente e sanguinoso tumulto dell'anfiteatro, che appunto peril suo

orrore attraeva, eccitava ed appassionava e che per questo fu uno dei temi prediletti nell'arte figurativa dell'epoca. Già nelle protomi è evidente l'intenzione di radunare le varie specie di animali, omnia. ex toto orbe terrarum animalia, così che tutte le province dell'impero siano rappresentate. Si tratta dunque di venationes imperiali — con l'arena gremita di animali di tutto il mondo - che le protomi, non altrimenti che le belve della « Grande Caccia » intendono evocare. Come abbiamo visto (v. sopra p. 94), l'enorme mosaico della «Grande Caccia», che alle sue due estremità è chiuso da un mosaico a emiciclo, l'uno con la personificazione dell'Oriente, l’altro con una personificazione dell'Occidente, presenta una caccia globale: il campo della battuta si estende all'orbis ferrarum. Le cacce si riferiscono, dunque, alle venationes imperiali. L'imperatore stesso, anzi, è presente alle cacce, nel tipico rappresentante dell'autorità statale dell'epoca,

cioè nel cattura l’Italia, animali

funzionario tetrarchico dal berretto tondo e dal bastone a fungo, che sopraintende alla delle belve, al loro trasporto attraverso i due mari e alla loro adunata nella terra di mezzo, dove saranno impiegate nell'arena romana. Un grifone presente alle cacce conferisce agli il carattere di animalia caelestia, cioè animali destinati alle venationes imperiali. ΑἹ por-

(1) Contro la critica del Cagiano de Azevedo deve essere sottolineato che non è l'aula absidata considerata a sé la quale indica che nel complesso di Piazza Armerina si tratta di un palazzo imperiale: bensì la disposizione totale della quale l'aula absidata è la parte culminante. (2) Dyggve e L'Orange, op. cit., pp. 122 sg.

(3) Gentili, La villa imperiale, op. cit, p. 29 e. Figg. rr sgg.; Pace, I mosaici, op. cit, Figg. 12-17. (4) Herodianus, I, 15, 5. Steier, RE IV A col. 346 sg. (s. v. Strauss); Aymard, op. cit, p. 195; Pace, I mosaici, op. cit p. 64; J. A. Jennison, Animals for Show and Pleasure im Ancient Rome, Manchester 1937. 99

H. P. L'ORANGE

tico destro dell'atrio si apre la sala della cosidetta « Piccola Caccia » (Pianta n. 7) con la sua significativa rappresentazione di un sacrificio di caccia di tipo imperiale (p. 97, nota 5). Procedendo dalla galleria del nartece verso l'aula basilicale (Pianta n. 5), si oltrepassa il portale d'ingresso incorniciato da due grandi colonne di granito rosso e si entra in un vasto ambiente rettangolare in fondo al quale è un'abside. In origine le pareti della sala erano rivestite d'incrostazioni marmoree, mentre il catino dell'abside era decorato in opus musivum ed il pavimento in opus sectile con largo impiego di porfido. Nel fondo dell'abside è ricavato un nicchione per una statua di dimensioni colossali. « ΑἹ di sotto di questa conservasi la traccia dello stallo in muratura entro cui aveva sede il trono preceduto da un tappeto a tarsia marmorea assai fine» (!). Gli scavi hanno portato alla luce una colossale testa marmorea (Tav. XL a) la quale, secondo il Gentili, è di una statua che doveva trovarsi nel nicchione dietro il trono () Prima il Gentili ravvisb in questa testa un ritratto di Massenzio, figlio di Massimiano (*); ora vi vede un Massenzio eroizzato nel tipo d'Ercole (‘). Io penso che si tratti di una testa puramente ideale, senza il minimo tratto d'un individuo mortale (5). Ha l'alta corona di capelli, tipica degli dei, sopra la fronte e l’enorme collo taurino, espressione di forza sovrumana. Senza dubbio non è Massenzio che ci si presenta in questa statua, ma è Ercole stesso — un Ercole imberbe, dalle corte fedine e dai capelli rialzati sulla fronte e sulle tempie, come lo si vede, per esempio, nelle celebri statue bronzee del Museo dei Conservatori e del Vaticano ed in una bella testa del Ny Carlsberg Glyptotek ἃ Copenaghen () (Tav. XL δ). L'Ercole tardo-antico di Piazza Armerina è di lavoro rozzo e grossolano e mostra le forme violente ed esagerate, l'espressione agitata ed aizzata, tanto caratteristiche della tetrarchia, Così il nuovo Ercole terreno troneggia ai piedi dell’Ercole Olimpico e dagli adoranti dinanzi al trono è visto come una sola cosa con questo. Né si potrebbe immaginare una collocazione più significativa del trono sotto l'aspetto dell'ideologia politico-religiosa del tempo. La stessa ideologia erculia, con lo stesso crescendo simbolico ed iconologico nel passaggio dall'atrio all'aula, si presenta ai nostri occhi nell'altra, minore ripartizione cerimoniale di Piazza Armerina. Come c'erano protomi animali in corone d'alloro nei portici dell'atrio della ripartizione maggiore, cosìin quella minore s'incontrano protomi animali in girali di acanto nei portici

dell'atrio (7: nei grandi girali sono racchiusi torsi ferini — tigri, leoni, lupi, sciacalli, ecc. e anche uccelli, per es. lo struzzo (Tav. XXXVI). E come nel triclinio della ripartizione maggiore ci trovavamo di fronte, nella sua statua colossale, Ercole stesso, così vediamo questo nei mosaici pavimentali del triclinio della ripartizione minore. Salendo dall'atrio i quattro gradini che portano al triclinio (*), incontriamo nei mosaici del pavimento di questa sala la piena orchestrazione del

tema a cui i mosaici degli atri e la « Grande Caccia » preludevano. Qui, come abbiamo visto, le

venationes imperiali si esaltano trasformandosi nelle cacce e nei combattimenti di Ercole. Tutta

(2) Gentili, Le villa imperiale, op. cit, pp. 40 88. (2) Gentili, op. cit, pp. 17, 25 sg, Fig. 37; Iconografie, op. cit, pp. 47 SE. Ὁ) I tratti di Massenzio ci sono noti dalle monete e sono assai diversi da quelli della testa colossale. L'Orange, Studion, op. cit, p. ros, Fig. sod, cf. Figg, 137-140. La statua togata del Musco di Ostia che si xoleva identificare con Massenzio (R. Calza, «Bull della Comune di Roma» 72, 1946-48, pp.83 sgg. ; Gén til, Iconograji, op. cit, p. 49) mi pare lavorata in un'epoca anteriore alla tetrarchia. (4) G. V. Gentili, Iconografie, op. cit., pp. 47 Sg.

(Ὁ La mancanza di indicazione plastica d'iride ὁ di pupilla è tipica delle sculture ideali, ma soltanto eccezionalmente si riscontra nei ritratti di quest'epoca. (6) Ny Carlsberg Glyptotek n. 256. L'Ercole di Scopa, ultimamento trattato da B. Schweitzer, « Oestarreich. Jahreshefte », C. Praschniker gewidmet, 30, 1952, pp. Tor sgg., introduce questo motivo dei capelli rialzati sulla fronte © sulle tempie nell'iconografia dell'eroe. (ΟἽ Gentili, La villa imperiale, op. cit, pp. 49 58. (8) Gentili, op. cit, pp. 52 sg. © Fig. 32; Pace, I mosaici, op. cit, Figg. V-VII a pp. 55-57e Fig. 11 100

PIAZZA ARMERINA la decorazione figurativa del palazzo raggiunge il suo τέλος conclusivo nell'inno in lode di Ercole.

Egli è l'eroe e il genio di tutto il palazzo. In lui si manifesta l’imperatore regnante, che appare

trasfigurato nell'immagine di Ercole.

Il Seston ha rilevato, nel Panegirico di Mamertino a Massimiano Erculio dell'anno 29r

d.C. (1), un passo che fa il parallelo delle imprese del venator sull'arena con le cacce di Ercole e che illustra, in modo convincente, il riesso simbolico tra le venationes della «Grande Caccia » e la strage di Ercole nel triclinio (#). Il Panegirico illustra dapprima l'infinita attività salvatrice di Ercole, che libera il mondo da tutti i mostri malefici; poi Ercole come l'etérno difensore e rafforzatore della virtus che conforta nelle loro lotte gli uomini buoni; da ultimo, Ercole che incita i partecipanti ai giuochi dell’arena affinché compiano gesta le quali assomiglino alle sue. Qui dunque vengono messe direttamente a confronto le imprese dell'arena e le fatiche di Ercole. «Le ch, ΠῚ du panégyrique de 297, qui fut, je crois l'avoir prouvé, prononcé devant Maximien è l'occasion de la fête de l''epifanie' des ovii et des Herculii, donne

un commentaire

excellent

des

mosaïques de l'ambulacro (cio? la galleria della « Grande Caccia ») comme de celles du quadrato centrale du triclinium (cioè della ripartizione minore); ici rappel de l'oeuvre d'Hercule pacator orbis la défilé des bêtes destinées aux jeux du cirque. Le panégyrique fait suivre ce rappel de cette

phrase curieuse: His quidem certe diebus quibus immortalitatis origo celebratur instigat (sc. Hercules)

ut videmus illos sacris certaminibus accitos ut pertinaci animositate certandi multa faciant ipsius

similia Victoris. Ces combats qui sont au programme de la féte imperiale sont de toute évidence, parce qu'ils rappellent de quelque manière ceux d'Hercule, des venationes dans l'amphithéatre ». Secondo Lampridio fu come venafor dell'arena che Commodo si guadagnò il nome di Ercole. Apellatus est etiam Romanus Hercules, quod feras Lanuvii in amphitheatro occidisset (2). La somi glianza tra le venationes dell’arena e le cacce erculee nella libera natura tanto più colpivano in quanto l'arena nelle grandi feste veniva trasformata in un vivo paesaggio con rocce, boschetti, alberi e cespugli. Si intendeva appunto provocare l'illusione di cacce nel paesaggio di natura (*).

Nello stesso modo il venator della vita reale viene paragonato ad Ercole su sarcofagi del rr

secolo dove le imprese del venator, in una zona inferiore, corrispondono alle cacce mitiche del-

Veroe nella zona superiore (Tav. XL c). Il venator diviene un nuovo Ercole. Così viene rappresentato in ritratti dell’epoca nei quali il cacciatore appare nel tipo di Ercole. In questa imma-

gine lo vediamo, coll'elmo leonino, il cane al suo lato destro, l'arma -- lancia o arco ~ nella sua mano sinistra, in una impressionante statua nel Museo del Bardo degli anni attorno al 250 d. C. 6) (Tav. XLI, XLII). Così come l'interpretazione erculia fa rivivere i mosaici di Piazza Armerina, il palazzo ha

indotto il Dyggve a pensare alla «basilica Herculis» che Teodorico, secondo Cassiodoro, desi-

derava erigere a Ravenna secondo un antico costume: un tipo di edificio, si dice, che fu nel mondo antico venerato ed ammirato in particolar modo. Basilicae Herculis amplum opus aggressi (sc. nos, Theodoricus) cujus nomini antiquitas congrue tribuit, quicquid in aula praedicabili admiratione fundavit. «Qui abbiamo senza dubbio la spiegazione: l’edificio a Piazza Armerina è precisamente un ricco esempio ben conservato di una tale basilica erculia tradizionale, cioè, come si vede ora, un'aula monumentale a scopo di glorificazione ... » (9). (1) Mamertini Paneg. Genethl. Max. Aug. 3 (Ed.

(5) Inst. Neg. 61.615-617. Merlin-Poinssot,. Guide,

Galletier, Paneg. Latins, 1, p. 53) op. cit, p. 36; 19. Ringrazio il dott. H. Sichtermann (2) In una lettera del 2-12-1952. per il gentil permesso di riprodurre le fotografe. (3) Script. hist. Aug, Vila Commodi, VIII, 5; ^ (6) Dyggve, «Symb. Osl.», op. cit, p. 126; «Atti

Aymard, op. cit, pp. 194 96: 467 UE; 537 UE. (4) Aymard, op. cit, pp. 184 sgg 189 sug.

del IV Congr. archeol. ctis», pp. 415 sg; Excursus

ror

H. P. ORANGE

Se è giusta la datazione del palazzo di Piazza Armerina all'epoca tetrarchica e l’interpretatazione dello stesso come palazzo imperiale, sembra evidente che non si possa pensare ad, altri che Massimiano Erculio come proprietario del palazzo (ἢ). Ci troviamo di fronte ad un palazzo erculio il quale nel suo insieme come nei particolari è intonato ad Ercole imperatore. Dev'essere osservato a questo riguardo che anche la residenza romana di Massimiano, «la prima sede della sua divinità » (*), cioè lo stesso palazzo imperiale sul Palatino, nel: Panegirico di Mamertino a Massimiano dell’anno 289 compare come palazzo» erculio (*). È stato, dice Mamertino a Mas-

simiano, Ercole a consacrare il palazzo sul Palatino. Verum est... sanctum illud venerandumque

dalatium... Herculem hospitem. consecrasse. È del tutto sicuro, continua, che questo eroe, «il

primo della tua razza e del tuo nome », arrivò sul Palatino glorioso di vittorie e fu ricevuto nella piccola antica reggia e lì gettò le fondamenta della « grandezza futura del palazzo », ut esse posset domus Caesarum quae Herculis fuisset hospitium. Lo stesso palazzo imperiale sul Palatino diviene dunque nella retorica contemporanea un palazzo di Ercole e può essere sotto questo aspetto paragonato col Palazzo a Piazza Armerina.

sulla «Basilica Herculis» ricordata. da Cassiodoro, «Corsì di Cultura sull'arte ravennate © bizantina v IV (D2,1957, 75 588. più o meno convincenti, Nuovi PP. argomenti, sono apportati dal Gentili a suffragio dell'assunto che il complesso monumentale di Piazze Armerina sia veramente un palazzo di Massimiano Erculio. Un'epigrafe molto frammentaria su un ploteo mar. moreo «evidentemente di. uno degli intercolumni dell'ambulacro della Caccia», dal Gentili, nella sua ultima revisione, viene ricostruito Hercileus) V(iztr) GALLO(vum liberator o pacificator (?)): Gentili, eo. nografie op. cit, p. 36 e Fig. 4. Secondo la fotografia riprodotta dal Gentili (op. cit, Fig. 4) la lettura del Pace (I mosaici, op. cit, p. 37) che sostituisce N ad H è sbagliata,Le lettere e V interpretate come Herculeus Victor, portano il Gentili a questa congettura. Alle dette due lettere si riallaccia, secondo il Gentili, con lo stesso significato, l'edera cho vediamo largamente profusa per ogni dove sulle superfici di mosaico dei portici e delle sale: perché l'edera, nella forma latina edera, ci da l'iniziale stessa del divino cognomen dell'imperatore Hereuleus ὁ perché la foglia dell'edera viene speso sorretta da due volute a V. Si aggiunge che l'edera ὃ, nello stesso tempo, un attributo dionisiaco ὁ fa pensare a Dioniso, il quale, come abbiamo visto (sopra p. 97), mantiene strette relazioni con Ercole. Si potrebbe forse aggiungere che Costantino, il quale pure apparteneva alla tradizione dinastica degli Erculei, è rappresentato in un rilievo dell'Arco di Costantino seduto su un carro decorato di una serio di foglie di edera ὁ con i dionisiaci coni di pigna (L'Orange - von Gerkan, op. cit, Tav. 12A e p. 75). = Con la nostra interpretazione del palazzo si aprono muove possibilità. per la identificazione dei ritratti dei mosaici. Non è possible, però, seguire il Gentilé nelle singole identificazioni. Non posso riconoscere le persone storiche della tetrarchia nelle figure puramen-

to ideali di dei, eroi e personificazioni nel mosaico della incoronazione d'Ercole (Gentili, Iconagrafie, op. cit, pp. 11 sgg) e nella testa colossale di Ercole (sopra p. 100). Il « busto bronzeo di Costanzo + a Monaco (op. cit., Fig. 36 e p. 53), che gli serve per l'iconografia del detto imperatoro, non è antico. In una scena del frigidarium delle terme, Fausta, secondo il Gentili, sarebbe rappresentata « nell'atto di svestirsi» (Gentili, La villa imperiale, op. cit, Fig. 7 e p. 25; Iconografie, op. cit., pp. 50 sgg. con Tav. e Fig. 30). Che una imperatrice o una principessa imperiale sia stata raffigurata e nell'atto di svestirsi ν, è, secondo il mio concetto dell'arte romana, impossibile. Inoltre la ‘tunica corta con segmenta, orbicoli, ecc. esclude che questa figura sia una figura femminile. Vediamo il vestito maschile tipico del tempo, come appunto è ben dimostrato dai mosaici di Piazza Armerina. La ‘tunica di « Fausta » può essere particolarmente confrontata con quella del lampadoforo in una tomba a Gargaresh in Tripolitania, datata dal P. Romamelli al iv secolo (Notiziario Archeol. Ministero delle Colonie v, 3, 1922, fig. 8; si noti il candelabro, che è della stessa forma di quello del lampadoforo nella scena dell'adventus a Piazza Armerina, Tav. XXXa. I due bambininel mosaico del vestibolo trapezoidale delle terme (Gentili, op. cit., pp. 27 sg. e Fig. xo; Iconografie, op. cit., pp. 43 Sg. e Figg. 15; 16) portano entrambi la tunica maschile e la penula, perciò è escluso che rappresentino Fausta e Massenzio. La testa marmorea +forse di Costante » (Gentili, La villa imperiale, op. cit., Fig. 38 e pp. 17 sgg) eche ha i caratteri classici propri dell'arte costantiniana », è senza dubbio, come era anche l'opinione di B. Pace, del i sec. d. C. (2) Mamertini Paneg. Maximiano Augusto x. (Ed. Galletiér, Paneg. Latins, I, p. 24). ᾿ς (8) Mamertini Panog., loc. cit. 102

PIAZZA ARMERINA

Contro S. Mazzarino (), il Cagiano de Azevedo () ed il Lugli (?), devo affermare che forti ragioni storiche parlano per Massimiano Erculio come il padrone del Palazzo di Piazza Armerina (4). Infatti, dalle fonti storiche sulle quali Kühler ha richiamata la mia attenzione, risulta che la casa imperiale di Massimiano possedeva grandi praedia in Sicilia, Il Liber Pontificalis narra che Costantino

portò alla Basilica

Lateranense

dei contributi dalle sue pro-

prietà del Lazio e della Sicilia, dunque dalle proprietà che, dopo la vittoria su Massenzio nel 312, passarono a lui e che, necessariamente, erano appartenute a Massimiano Erculio (5) Parimenti

Costantino, quando

poi eresse la chiesa di San Pietro, vi recò dei contributi

dalle

sue proprietà in Oriente, che dopo la vittoria su Licinio nel 324, erano passate da Licinio a lui (9. Tra le proprietà di Costantino in Sicilia il Liber Pontificalis ricorda massa Castis, ter-

ritorio Catinense e massa Trapeas, territorio Catinense. Catina è Catania, e da Catania l'antica via publica portava direttamente attraverso la regione di Piazza Armerina fino ad Agrigento. La designazione territorio di Catania viene così ad indicare esattamente la situazione topografica di Piazza Armerina. Se muoviamo dalla premessa che le effigi dell'antichità dovessero esprimere un concetto per l'osservatore, e che codesto concetto dovesse trovarsi in rapporto con gli ambienti che quelle adornavano, si deve concludere che il palazzo di Piazza Armerina dovette essere un palazzo erculio eretto nel segno di Ercole. Esso si differenzia dunque nel modo più assoluto da quel tipo di ofia siciliensia che venivano eretti nel segno della filosofia a cui lo collegano il Pace (ἢ ed il Mazzarino (5). La vita filosofica e contemplativa viene espressa nell'arte figurativa dell’epoca mediante un'iconografia distintamente determinata, non meno

di quanto lo sia la vita erculia ed

entrambi questi modi di vita si rispecchiano in un ricchissimo materiale di monumenti conservati. Nel ΠῚ e nel rv secolo che, con tanta enfasi, collocano in primo piano l'antico concetto della filosofia come norma spirituale e come simbolo d'immortalità, il filosofo compare, per esempio, sui sarcofagi pagani e cristiani, in una serie infinita di rappresentazioni di studio, di letture, di dissertazioni © di discussioni, in rappresentazioni dei Sette Savii, delle Nove Muse ecc., nelle quali i cultori della saggezza si mostrano sempre ravvolti nei loro pallii e con le caratteristiche

(a) Sulltium. di Massimiano Ereulio dopo l'abdicazione, «Rend, Lincei», ser. VIII 8, 1953, PD. 417 588 (@) I propriciari della villa di Piazsa Armerina, «Scritti di Storia dell'Arte in onore di Mario Salmi Roma 1961, pp. 15 588. 73 sex. in II Palasso di Massi (Ὁ(3) Op.Cir. cit, le miepp. ragioni miano Ereulio di Piassa Armerina, «Studi in onore di Aristide Calderini e Roberto Paribeni», MilanoVarese 1956, ΠῚ, pp. 598 sgg (5) Liber Pontifealis, ed. L. Duchesne, L pp. 173 58 16) Liber Pontifalis, 1, p. 78 sg; 177 sg. Cf. A. Piganiol, L'empereur Constantin, Paris 1932, PD. Iia sgg W. Seston, Dale de Saint-Pierre de Rome, 4 Cahiers Archéol », 1, 1945, DD. 153 586. (7) Pace, I mosaici, op. cit. La mia risposta alla critica del Pace, op. cit, pp. rrr sgg, si trova mel mio articolo negli «Studi în onore di A. Calderini e di R. Paribeni n, IIT, 1956-57, pp. 596 sgg (8) S. Mazzarino, Sull’otium di Massimiano Erculi dopo l'abdicazione, «Rend. Lincei», VIII, 8, 1953,

PD. 417 sgg. Mazzarino, partendo dalla tradizione toponomastica della contrada di Casale, pensa che Sabuggio possa derivaro da Sabucius o cho qualche latifondista di tale celebre nome sia stato il proprietario del palazzo (G. Giannelli e S. Mazzarino, Trattato di Storia Romana, Roma, 1956 II, p. 417). Questa ipotesi per essere corta dovrebbe essere confermata in qualche modo dalle rappresentazioni musive. Non penso, però, che le interpretazioni di G. Manganaro abbiano potuto dare questa conferma dell'idea del Mazzarino (Aspetti pagani dei mosaici di Piazza Armerina, 1 Archeologia Classica» 11, 2, 1959, pp. 241 sgg). Il Manganaro paragona il grande mosaico di Ercole nella Sala Trichora con una scena su un «contorniato » di Sabucius Pinianus, sostenendo che ambedue rappresentano Ercole nel giardino delle Esperidi, mentre in realtà l'Ercole del mosaico si trova non sotto l'albero delle Esperidi, ma sotto il pino sacro ad Ercole, le cui pigne — e non mele | — sono ben visiDili, e non lo si vede, come pensa il Manganaro, in compagnia delle Esperidi, ma nel momento della sua apoteosi quando viene incoronato da Giove. 103

H. P. ORANGE

lunghe barbe filosofiche (1). Ma non è codesto il mondo che noi troviamo nei mosaici di Piazza Armerina. Non vi sono qui scene di vita contemplativa, di lettura, di discussione, dei Savi e delle

Muse, non filosofi dalla lunga barba ed in pallio, ma scene di vita pratica ed attiva, uomini indossanti la tunica con clamide, la veste soldatesca, i paramenti dei funzionari pubblici e secondo la moda del tempo, imberbi o con la corta barba militare. Allo stesso settore del mondo umano appartengono le forme architettoniche e di spazio, decorate dalle predette raffigurazioni. Noi non passeggiamo qui per portici filosofici o per ambienti del genere, bensì attraverso un complesso totalmente dominato da quel disegno cerimoniale della corte che già conosciamo dai palazzi imperiali.

Un durre le dott. H. ziamenti

.

ringraziamento speciale vada all'Istituto Archeologico Germanico per il permesso di riproseguenti fotografie: Inst, Neg. 65.528, 55.740, 61.514, 61.615.617. Siamo molto grati al Sichtermann per il suo gentile aiuto nella ripartizione fotografica del detto istituto. Ringrasentiti per fotografie anche al dott. M. Grbié, al dott. Hj. Torp e al prof. P. L. Zovatto.

(1) F. Cumont, Le Symbotisme funéraire, op. cit, ‘pp. 2538g8.,26386g. e passim; G. Rodenwaldt, e Jahrb. d. Deutschen Archáol. Instituts» 51, 1936, pp. 1008gg.;

F. Gerke, Die christl. Sarkophage der vorkonstantin. Zeit, pp.246;271 sgg.e passim; L'Orange, Iconography οὕ Cosmic Kingship in the Ancient World, pp. 188 sgg.

Finito questo articolo, ho trovato nel Museo di Aquileia un gruppo di rilievi sepolcrali di militari della fine del III secolo (secondo Giovanni Brusin) con berretto tondo e bastone a fungo, importantissimo per gli studi qui presentati. Questi militari appartengono alla legione XI Claudia, che era stata reclutata nellIlliria e nella Moesia, cioè proprio nelle provincie dalle quali provengono gli imperatori tetrarchici. V. Giovanni Brusin, Aguileia e Grado, Padova, 1964, p. 202, fig. 122; W. Kubitschek, Legio, RE XII, 2, coll. 1960 sgg., 1699 sg. I rilievi saranno da me trattati nel prossimo vol. di Symb. Osl. (1966).

104

H. P. L'Orange, Piazza Armerina

a, 8) Piazza Armerina, particolari della «Strage di Ercole» e della « Gigantomachia v.

TAV. I

TAV. II

H. P. L'Orar

ac) Pazza Armerita medaglioni a protomi nel grande atrio.

H. P. L'Orange, Piazza Armerina

Tunisi, Museo del Bardo, medaglioni a protomi dal tepidario delle terme di una villa in Thuburbo Maius (Taasia).

H. P. L'Orange, Piazza Armerina

TAV. IV

4,D) Piazza Armerina, dettagli del mosaico dal » Cubicolo con scera erotica ». 5) Tunisi, Museo del Bardo, mosaico con protomi ferin-.

H. P. L'Orange, Piazza Armerina

TAV. V

a) Hippo Regius, dettaglio d'un mosaico. 2) Tunisi, Museo del Bardo, dettaglic d'un mosaico da Thuburbo Maius. 6) Piazza Armerina, dettaglio della « Grande Caccia s.

H. P. L'Ocange, Piazza Armerina

Tunisi, Museo del Bardo, mosaico del « Corteo di Nettuno +, dettaglio.

TAV. VI

H. P. L'Orange, Piazza Armerina

TAV. VII

a) Tunisi, Museo del Bardo, parte superiore d'un mosaico. 8) Piazza Armerina, dettaglio d'un mosaico. £) Thuburbo Mains, dettaglio d'un mosaico.

H. P. L'Orange, Piazza Armerina

TAV. VIE

ae) Piazza Arms-ina, dettagli di animali ir girali di acanto. 4) Thuburbo Maius, dettagli di animali in corone di alloro (cf. Tav: III).

H. P. L'Orange, Piazza Armerina

a-d) Piazza Armerina, rappresentazioni di elefanti nei mosaici.

TAV. IX

H. P. L'Orazge, Piczza Armerina

a,b) Roma, S. renzo fuori le coperchio d'en

TAV. X

erchio d'un sarcofago. d) Ostia, elefanti sul 0 dei Lampadi. Dettaglio da R. Delbrueck, sadriga di elefani Consulardiptychen 50.

H. P. L'Orange, Piazza Armerina

TAV. XI

Oudna, elefante, dettagli

H. P. L'Orange, Prazza Armerina

TAV. XII

ai Piazzz Armerira, dettaglio del mosaico della « Piccola Caccia ». 2) Piazza Armerina, dettaglio del mosaico della «Grande Caccia».

H. P. L'Orenge, Piazza Armerina

a,b) Tunisi, Museo del Bardo, mosaici con scene di caccia.

TAY. XIII

H. P. L'Orange, Piazza Armerina

a,b) Hippo Regius, Algeria, mosaico con trappola di caccia.

TAV. XIV

H. P. L'Orange, Piazza Arvcerina

TAV. XV

Fiazza Armerina, «Grande Caccia», dettegli: a) cacciatori in agguato. è) gabbia a trapoola; c) capra szozzata,

H. P. L'Orarge, Piazza Armerina

TAV. XVI

a) E.U.R., Roma. Dettaglio d'un mosaico dall'Esquilino, cfr. Tav. XXVI b. Trappela di caccia. δ) Piazza Acmerina, Dettaglio d'un mosaico nelle terme, Ritratto di una signora. v) Piazza Armerina. Dettaglio d'un mosaico nelle terme, Testa di una nereide. d) Piazza Armerina. Dettaglio d'un mosaico nel cubicolo con scena erotica, Testa di stagione. e) Magnia Urbica. fg, δ, i) Galeria Valeria. j) Fausta. 4) Helena.

H. P. L'Orange, Piazza Armerina

ai) Piazza Armerina, dettagli del mosaico di Arione. Teste di Nereidi.

TAV. XVII

H. P. L'Orange, Piazza Armerina

TAV. XVIII

dettaglio 4) Piazza Armerina dettaglio del mosaico della « Grande Caccia ». δ) Roma, Arco di Costantino, testa st un fzezio ὁ, storico Jegoslavia, f,g) Tetrarcai. dei ritratti con porfido di gruppi dei teste tre Marco, San Venezia, d) c, della Profectie Consiantini. Nis, frammento d'ua ritratto di porfido d'un Totrarca.

H. P. L'Orange, Piazza Armerina

TAV.

a-d) Solin (Salona), due erme doppie di Tetrarchi e dei fuvieli provenienti da Spa.ato. £j Spalato, medaglione-ritratto di Diocleziano nel suo mauscleo, gesso.

XIX

H. P. L'Orange, Piazza Armerina

TAV. XX

a) Salonicco, rilievo dell'Arec di Galerio dettaglio del rilievo c. b) Treviri, museo, rilievo. e) Salonicco, rilievo dell'Arco d: Galerio, dettaglio, cf. a. 4) Roma, Arco di Costantino, dettaglio dalla Profestio Constantini.

H. P. L'Orange, Piazza Armerina

TAV. XXI

Pasi FEDAVARN

H. P. L'Orange, Piazza Armerina

TAV. XXII

3e] Piazza Armerine, dettagli del mosaico dalla « Grarde Ceccie ν, alti funzionari col bastone a fungo.

H. P. L'Orange, Piazza Armerina

TAV. XXIII

a) Affresco in un santuario tetrarchico nel grande tempio a Luxor, copia in acquarello del Wilkinson del 1859. δ) Roma, Foro Romano, rilievo dei Suovetaurilia sulla Base dei Decennali

H. P. L'Orange, Piazza Armerina

TAV. XXIV

XII a. δ) Piazza a) Dettaglio dell'affresco nel santuario tetrarchico nel grande tempio a Luxor, cir. tz Armerina, dettaglio dell'aduentus, cfr. tavv. XXIX, XXX. cd) Piazza A rina, dettagli della «Grande Caccia s.

H. P. L'Orange, Piazza Armerina

TAV. XXV

a, 0) Paragone dei differentissimi leoni dei mosaici di Piazza Armerina (4) e Mopsuestia (D). c, d) Vaticano, Museo, sarcofagi romani con rappresentazioni dei giochi nel Circo Massimo. 6) Napoli, Museo Nazionale, sarcofago romano, rappresentazione dei giochi nel Circo Massimo,

H. P. L'Orange, Piazza Armerina

TAV. XXVI

=) Piazza Armerna, dettaglio della «Piccola Caccia», scena di caccia a: cinghiele. δ᾽ Roma, scena di caccia al cinghiale, dettaglio del mosaico di caccia dall'Esquilino, cfr. Tev. XVI a.

H. P. L’Orange, Piazza Armerina

TAV. XXVII

a) Prazza Armeria, dettaglio della «Piccola Caccia », scene di caccia alla volpe. b,x) Odarzo, Museo Archeologico, fammenti di mosaici pavimentali con scone di caccia.

H. P L'Orange, Piazza Armerina

TAV. XXVIII

Sclonicco, dettagli dell'Arco €- Galerio: a) testa di Galerio i] celco d'un frammento apreriznente al viso è Sssato alla testa), ) testa di Massimiano. ὁ, Δ) testo anonime. 4) Spalato, Museo Archeologico, frammento di rilievo.

H. P. L'Orange, Piazza Armerina

TAV. XXIX

Piazza Arwerirs, mosaico nel vestibolo che conduce al grande atrio, adventiss August.

H. P. L'Orange, Piazza Armerina

a,b) Piazza Armerina, alvenus Augusti, dettagli, cfr. Tav. XXIX.

TAV. XXX

H. P. L'Orange, Piazza Armerina

TAV. XXXI

a Piazza Armerina, grande atrio, dettagli del mosaico, mecagioni con protomi di animali.

H. P. L'Oraage, Piazze Armerina

a,b) Piazza Armerina, l'emiciclo di destra della 4 Grade Cacciax

TAV. XXXII

H. P. L'Orange, Piazza Armerina

TAV. XXXIII

ac) Piazza Armerina, frammenti del mosaico dell'emiciclo di sinistra della «Grande Caccia s.

H. P. L'Orange, Piazza Armerina

TAY. XXXIV

a} Piazza Armerina, grifone nel mosaic delle «Grande Caccia», 0) Costantinopoli, grifone al mosaico d: saccia nel palazzo imperiale.

H. P. L'Orange, Piazza Armerina

TAV. XXXV

Piazza Armerina, «Grande Caccia ». a) La terra tra i due mari secondo il disegno del Gentili. bud) Dettagli delle figure sulla terra tra i due mari: δ) uficiale d'ordinanza a sinistra del gruppo a quattro sulla terra tra i due mari; c) sotto, a sinistra, la testa dell'ufficiale d'ordinanza riprodotto in b; sopra, a destra, un frammento della frusta nella mano della figura sinistra del gruppo a quattro; la striscia arcuata, larga. e scura cho continua a sinistra dietro l'uficiale d'ordinanza in à, rappresenta la linea dell'orizzonte da cui spuntano delle piante; d) la parte inferiore delle tre figure a destra nel gruppo a quattro sulla terra tra i due mari, Cîr. a.

H. P. L'Orange, Piazza Armerisz

TAV. XXXVI

a,b) Piazza Armerina, dettagli de> mosai> nel partico dell'atrio nella r-partisione cerimeniale minore, geotomi ferine in girel. di acanto.

H. 2. L'Orange, Piazza Armerina

a,b) Piazza Armerina dettagli del mosaico della «Stregs 3i Ercoleν.

TAV. XXXVII

H. P. L’Orange, Piazza Armerina

TAV. XXXVIII

Piazza Armerina. a) Mosaico della gigantomachia ; 5) dettaglio dello stesso mosaico

H. P. L'Orange, Piazza Armerina

TAV. XXXIX

Piazza Armerina, dettaglio del mosaico dell'apoteosi di Ercole: testa di Ercole incoronato da Giove.

H. P. L'Orange, Piazza Armerina

TAV. XL

a) Piazza Armerina, testa colossale d: Ercole. b) Copenaghen, Ny Carlsberg Glyptotel, testa di Ercole, 2) Roma, Museo Borghese, sarcofago.

H. P. L’Orange, Piazza Armerina

Tunisi, Museo del Eario, cacciatore come Ercole, er. Tav. XLII,

TAV. XLI

H. P. L'Orange, Piazza Armerina

Tunisi, Museo del Bardo, cacciatore come Ercole.

TAV. XLII

H. P. L'Crange

Piazza Arrierina

TAV. XLII

Capitelli Istrarchici di produz.cne a serie. a) Falazzo di Piazza Armerina. ὦ, e) Siracasa, Museo. /, 4) Roma, Museo delle Terme di Diocleziano. 1.7} Palazzo di Diocleziano a Spaleto.

H. P. L'Orange, Piazza Armerina

PIANTA.

TAV. XLIV

s Pana Armerina

i

PE:

SG

Φ

"e:

RX

" DEOR es

T

=

" E

* Fa

i

CIRIE

ιν

ταῖνἐς =

m

SEO

È

FA

4

L2

r

Ὧν

^

dat

En

pu

os

PIANTA. T.

HJALMAR TORP

undercut outlines that detach the figures from the background, and a similar tendency to break the flow of movements and gestures into staccato rhythms (). Furthermore, one can find in many of these reliefs from Bawit the same sharply chiselled edges and firmly drilled grooves defining details of figures and garments (). The sculpture of this style found at Bawit may be dated around the middle of the sixth century (*), and most probably is the product of a workshop active in the construction of the South Church (). The significance of the figure sculpture from Bawit within the larger frame of Coptic art will be discussed in a second article along with a study of the problems of Old Testament iconography raised by the scenes represented on reliefs I and II. In the present article we will describe and attempt to explain these scenes.

Reus I.

Assuming that the original length of this broken slab, like the height, corresponded to that of

relief IT, approximately a quarter of the relief would be missing, comprising two, or at most three,

figures. The surface has suffered rather badly. The heads especially have been severely damaged, evidently deliberately, and probably by Moslems, either when the monastery was abandoned and fell into decay during the twelfth century or when the slab was found and transferred to the near-by village of Dahshlout. As several of the figures are indistinct due to the bad state of preservation, it is necessary to start with a detailed description of the relief.

In the first scene from the left a haloed person sits on a cushioned throne, flanked by two soldiers of the bodyguard (Pls. I 4; II). All three figures face the onlooker. Like the rest of the figures in the relief, they have short, curly hair, and the enthroned person, in addition, is bearded. His right hand is raised, the fingers expressing the attitude of speech (). The left hand displays an unrolled scroll with illegible fragments of Coptic letters painted in black. The figure is dressed in a long tunic and a chlamys. The tunic has an ornamented band running down in front, and wide, short sleeves like those of a colobium (thus indicating that underneath the person might wear a tunica interula with long, tight sleeves no longer visible in the sculpture), The chlamys is fastened, as was the custom, on the right shoulder, while the back of it is draped across the lap in a rather unusual way, almost as if it were a pallium. The head of the seated

figure is crowned with a diadem, badly preserved, which consisted of a narrow band enci cling the head and supporting upright elements of alternating gabled and arched shapes. A similar type of diadem appears, better preserved and more readily distinguishable, in relief II, cf. below, p. 117, and PI. IV. ‘The bodyguards are dressed in the uniform of late Roman and Byzantine legionary soldiers: a short tunic with long, tight sleeves, a cuirass consisting of a short leather jacket with decorated vertical bands and pendant straps to protect the upper parts of the arms and the thighs, a belt with a round buckle (seen on the soldier to the left), and, finally, high boots or, perhaps, trousers and normal footwear. Around the neck of the left-hand soldier one sees the torque

(1) Op. cit. pls. LXX (cf. our PI. VI δ), LXXXVIII (cf. our PL. VI à, LXXXIX, 2; G. Duthuit,La sculpture cople. Statues — bas-reliefs — masques, Paris 1931, pl. XIII, a and c; the first (cf. our PI, VI a) is from Bawit, the second is of uncertain origin but probably also from that monastery. (2) Cf. preceding noto.

worn

(3) E. Kitzinger, Notes on Early Coptic Sculpture, « Archacologia s, LXXXVIT, 1937, pp. 189 ff (4) This problem will be examined more closely in another connection. (5) The figure does not seem to have held an object, although the photographs may give that impression. 106

TWO SIXTH-CENTURY COPTIC STONE RELIEFS

by the members of the imperial guard; he holds a spear or a lance in his right hand, while his companion is armed with a sword in a broad ornamented scabbard carried on his left side and supported by a bandoleer across the right shoulder. Both soldiers lean upon oval shields decorated with a great cross, Small holes bored in the shields and in the legs and 'cushion of the throne probably are meant to indicate a jewelled decoration (there are not vestiges of glass paste left in the holes). The entire group is highly reminiscent of similar representations of fifth- and sixth-century imperial art which, evidently, were the ultimate source of the scene (!). Most of the individual features (dress, weapon, shields, throne) () can also be traced back to this source even if some are of a different origin. Thus the diadem of the seated person is not of imperial Roman or Byzantine origin, but must be derived from an Oriental prototype, as will be shown in the second article on the reliefs. Regardless of its origin, however, the diadem undoubtedly is a royal insignia and certainly indicates that the person wearing it is of princely rank. The next scene shows a person moving toward the right. Since he wears the same clothes as the seated figure in the preceding scene and is crowned with the same diadem, he his, no doubt, intended to be the same princely person, The stafiin his right hand therefore is probably a sceptre; with his left hand he pulls the back of his chlamys across his body. He steps towards, and by the movement of his head seems to address, a young beardless figure (Pl. III). This figure is clothed in a costume which in early Christian and medieval art characterizes a per-

son as a Persian or as a Hebrew in the Babylonian diaspora. The costume is regularly seen in representations of, for instance, the Magi, He wears a tunic (or rather chiton), trousers, boots and a wide cloak clasped in front and covering his back. The chiton is wide, has long sleeves and is girdled below a very long over-fold. It tapers at the waist line, is slit at the sides and fastened up at the thighs in order to leave the legs free when

riding a camel.

The trousers or

anaxyrides are decorated with a vertical series of small roundels, each enclosing what seems to be a floral motif, It is uncertain whether or not the figure represented in the relief wears the Phrygian cap which belongs to the costume. With both hands the person in Oriental clothing holds up a partly preserved diadem or circlet in front of a small figure mounted on a columnar pedestal. The figure on the column has broken arms and, except for a cloak hanging down his back, he is naked, The column, which is complete with base and capital, is decorated with a bow of ribbon. The latter seems to be a votive offering, and the figure on the column must be interpreted as the statue of a deity,

To the right of the statue and facing it, the prince, dressed in his usual apparel, but, it

seems, without his diadem, reappears for the third time. He is shown in a strange attitude fall-

ing to his knees, legs spread, arms waving wildly in the air as if expressing despair or surprise. Between the prince and the column there is a pair of severed forearms and hands. In spite of their disproportionate size they must be the limbs of the idol, broken and falling to the ground. The relief in its present state terminates with the back of a figure turned to the right and

wearing Persian costume.

‘The key to the meaning of the relief may be found in the scene with the naked statue on its column. In East Christian art the statue on a column was the most common and typical pictorial expression of the pagan deity. It was the image par excellence of the idol (9). As an

() R. Delbrueck, Die Consulardiptychen und (2) CL. preceding note. veruandie Denbmiler, Berin-Leipzig 1929, pls. 2 (3) W. Haftmana, Das ialimische Sduleumonu(upper fiere) 57, and 62, ment, Teiprig-Berlin 1939, pp. 52 fl; K. Weitz: 107

HJALMAR TORP

example, we might reproduce a twelfth-century Byzantine miniature of the adoration of Aphrodite whose idol stands on ἃ column (PI. VIL). The broken limbs, in their turn, signify the overthrow and destruction of the idol, as can be seen in numerous pictorial representations. It may suffice to present (PI. VII c) only one example, the illustrationin the Books of Kings, Rome, Bibl. Vaticana, MS. Vat. gr. 333 (1),of I. Sam. 5,3, the Fall of Dagon, The miniature illustrates the passage: ‘...behold, Dagon was fallen upon his face to the ground before the Ark of the Lord’, and shows the idol of the Philistines twice, first on its columnar pedestal, next broken at the base of the Ark of the Covenant. The idol together with the figure in Persian dress suggests that the relief illustrates the adven-

tures in Babylonia of Daniel and his friends Hananja, Misael, and Asarja (Daniel, chapter iii).

To my knowledge, in no other context in Christian iconography does a figure in Persian costume

appear in association with an idol. The scenes of the Coptic sculpture, however, only in part correspond to the known iconography of Daniel. Pictorial illustrations of Daniel, chapter iii, usually start by depicting King Nebuchadnezzar ordering ‘all the world ' to gather to worship the immense golden idol he had erected in the ‘valley of Dura in the landscape of Babel ’. The king is often shown seated and attended by one or more of his bodyguard. The subject is found alreadyin early Christian art of about 300 (?). The attendant soldiers may be in armour, as in the relief, and the king ordinarily wears the

diadem, althouglr more often than not he is dressed in the pallium instead of the chlamys; when the king wears the chlamys it is often draped across the lap as in the Coptic relief (PI. VII a) 6). One might perhaps object to the halo and the scroll as attributes of Nebuchad-

nezzar, but both may be taken as mere symbols of kingship without specific Christian significance, The halo, in fact, is worn by Nebuchadnezzar himself in an representation on an early Christian

lamp (*), while the scroll as.a symbol of royal dignity may be seen in the drawing on fol. zr" (PI. VIII e) of the Utrecht Psalter which illustrates the words: ‘Thou settest a crown of fine gold

on his head’ (Psalm 21). In the drawing an angel is depicted crowning a king (David ?) while at the same time handing him an unrolled scroll (5). Seated on a cushioned throne, wearing a nimbus, his right hand raised and his left holding an

inscribed roll, the King of Babylon in the relief recalls above all certain images of Christ (*)

mann, Grech Mythology in Byzantine Art, Princeton 1951, PP. 37, 42, 49, 55, 64 f, 7% ff, figs. 35, 45 1, 58, 66 (cf, our PI, VILd), 771, 881, etc. Western ‘examples, E.T. DeWald, The Stuttgart Psalter, Princeton 1930, fols, 45%, 94°, 122, and 150". (2) J. Lassus, Les miniatures byzantines du Livre des Rois d'après un manuscrit de la Bibliothèque Vaticano, Mélanges d'archéologie et d'histoire», XLV, 1928, pp. 38 f . Western examples, P. Toesca, Gli affreschi della Cattedrale di Anagni, «Le Gallerie nazionali italiane +, V, 1902, pp. 131 f. (vaults IX and XD), pl. IV. (2) G. Wilpert, I savcofagi cristiani antichi, II, Roma 1932, pp. 259 ff, pls. CLXXXXIX fL; ibid, IIT, 1936, pp. 23 £ pl. CCXXVII, 1; F. Gerke, Der neugefundene altchristiche Friessarkophag im Museo Archeologico su Florens und das Problem der Entwicklung der altesten christlichen. Friessarhophage, « Zeitschrift £. Kirchengeschichte», 54, 1935, PP. 18 fh, 24 ἢ,

(3) Nebuchadnezzar with chlamys and diadem and soldier in armour, Wilpert, op. cit, IT, pls. CLXXXXI, 6; CC, 5 (cf. our Pl. VIL a). King N. with chlamys draped across the lap, ibid., pls. CLXXXXIX, 3; CC, 4,5 (cf. our PI. VII a); ibid., IIT, pl. CCLXXXVII, 1. (4) A. Toulotte, Le roi Nabuchodonosor sur les monuments africains, «Nuovo Bullettino di archeologia cristiana +, VI, 1900, pp. 113 fL; Dict. d'arch. chedt. e de liturgie, XII, cols. 579 fi. (H. Leclercq). (5) Utrecht, University Library, MS. 32; E. T. DeWald, The Illustrations of the Utrecht Psalter, Princeton 1933, p. 12, pl. XVIII. (6) F. Gerke, Christus in der spatantiken Plastik, Mainz 1948*, pp. 35 ff, figs. 49, 52 £3W. N. Schumacher, ‘Dominus legem dat’, « Romische Quartalschrift f. christliche Altertumskunde und Kirchengeschichte ν, 54, 1959, pp. τ f; A. Xyngopoulos, Τὸ Καϑολικὸν τῆς Μονῆς Λατόμου ἐν Θοσσαλονίκῃ καὶ τὸ ἐν αὐτῷ, ψηφιδωτόν, «Api, Ache. v, 12, 1929, pp. 161 108

TWO SIXTH-CENTURY COPTIC STONE RELIEFS

that, like the one of Nebuchadnezzar, are based upon imperial models (). Indeed, Hebrew exegesis leads one to suspect that ideas similar to those underlying representations of the Christian All-ruler may also have influenced the iconography of Nebuchadnezzar in the Coptic relief. Nebuchadnezzar himself was thought of as a World-ruler, a Cosmocrator (*), who, according to a relatively late tradition, was the son of Solomon and the Queen of Sheba (*). He was thought to have been the only Gentile ruler, with the exception of Cyrus, found worthy to sit on the throne of Solomon -- a satisfaction which cost him a broken leg (4). Ephraim the Syrian praised the penitent Nebuchadnezzar and exalted him as a model for the Christians (); Augustin looked upon him as a typos of the Roman emperors who, after having persecuted Christianity, later recognized it (*). According to the Bible, King Nebuchadnezzar was a God-fearing man. After the miracle of the three young Hebrews who came unscathed out of the blazing furnace, the king exclaimed: ‘Glory to the God of Sadrak, Mesak and Abed-Negod (the names given by the king to the three friends of Daniel) ..." (Dan., iii, 28). The scroll in the king’s hand may be an allusion to his law-making power and the fragmentary inscription may consequently have reproduced the king’s command to worship the idol. When seen in the context of the message contained in the relief as a whole, however, it seems more probable that the words of the inscription were his eulogy ‘ Glory to God ' (). A similar ‘ positive ' conception of Nebuchadnezzar as 2 ‘ righteous king ’ may also explain the crosses on the shields of his bodyguard, although these may be merely and addition by the Coptic craftsman who would have known such emblems from the shields actually carried by Byzantine troops in Egypt

as well as from pictorial representations (). Nothing, then, in the first scene of the relief precludes the interpretation of the seated figure as Nebuchadnezzar. In the following scene it is that same king, now standing and with a sceptre, who addresses the figure in Persian costume next to the idol. Nebuchadnezzar portrayed in this attitude is known from early Christian renderings of the king calling upon the three Hebrews

to worship the idol (PI. VII 5) (€). On early Christian sarcophagi of Western origin the idol

often was made to resemble the king (PI. VIL a), whereby the refusal of the three youths cleverly came to represent a pertinent paradigm for Christian refusal to worship temporal authority (9). In East Christian biblical illustration, however, the idol erected by the King of Babylon in conformity with the words of the Scriptures is depicted not in the likeness of the king, but fL, fig. 26, p. 167. Cf. also our PI. VI a, although in this relief Christ holds a book and not a scroll. Christ,

Cross) seems, however, to have been the most common type of Christian symbol on soldiers’ shields. That

of course, in all these representations does not wear type goes back at least to the Theodosian period, the chlamys, but the pallium. J. Kollwitz, Ostrbmische Plastik der theodosianischen (1) A. Grabar, L'Empereur dams l'art bysantin. Zeit, Bertin r9gr, Beil. 5 f. Recherches sur l'art officiel de l'Empire d'Orient, Paris (9) Wilpert, op. cit, II, pls. CC, 8 (ef. our Pl. 1936, pp. 200 fi. VIT 8), CCII, τ; Dict. d'arch. chré. et de liturgie, VI, Ὁ) L. Ginzberg, The Legends of the Jews, IV, figs. 5605 £, cols, 2156 f. (H. Leclercg) (10) E. Becker, Protest gegen dem Kaiserkult und Philadelphia 1913, pp. 275 f, 333; ibid., VI, 1928, n. 83, p. 368; n. 96, p. 422. Verherrlichung des Sieges am Pons Milius in der christlichen Kunst der honstantinischen Zeit, Konstan(3) Ibid., IV, p. 300; VI, n. 21, pp. 389 f. (4) Ibid, VI n. 7, p. 433. tin der Grosse und seine Zeit (-Supplementheft (5) E. Beck, Ephraems Hymnen über das Paradies. XIX of the « Rômische Quartalschiîft ἡ, 1913, pp. Ubersetsung und Kommentar, «Studia Anselmiana v, 155 fl; H. Kruse, Studien sur ofisillen Geltungdes. 26, 1951, pp. 143 ff; cf. below, p. rtr and n. 2. Kaiserbildes im romischen Reiche, Paderborn 1934, pp. 84 f; Wilpert, op. cit, ΠῚ, p. 259; F. Gerke, (6) Toulotte, op. cit, p. 117. Die chrstichem Sarkophage der vorkonstantinischen (7) Ct. below, p. rrr and m. 2. (8) The Chi-Rho monogram (and not the simple Zeit, Berlin 1940, p. 187. 109

HJALMAR TORP

as an ordinary cult statue, An example of this is provided by a miniature in the Psalter Bibl. Vaticana, MS. Vat. gr. 1927, which in the upper zone shows King Nebuchadnezzar enthroned, attended by a soldier, and, before him, a large, gold idol standing on a column (1). The Bible tells how the three faithful companions of Daniel refused to worship the idol and therefore were thrown into the fiery furnace, But even if the three young Hebrews refused to fall down and adore, the Israelites in general, as did ‘all natiohs’, obeyed the royal request. Consequently, in the relief, the figure in Persian costume which holds up a diadem or a circlet before the statue may be intended to represent Israel's worship of the idol, and we might construe the scene to be an abbreviated version of the worship of ‘all nations’ as it is amusingly rendered in a Spanish illumination of the eleventh century (PI. VIL e) (). The miniature shows the king, with crown and sceptre, calling upon a group of persons to worship, and the cult of the idol, with the believers doing proskynesis accompanied by musicians and jugglers. In the Roda Bible the worship scene is placed between scenes showing the refusal and the accusation of Daniel's companions, thus contrasting their unshakable observance of Hebrew law with the idolatry of those who have accepted the rule of Nebuchadnezzar. In medieval art in general, this part of the biblical story of Daniel is represented in order to glorify those of the diasporà who remained faithful to the law; why, then, should the Coptic relief illustrate the fall of the unfaithful Israelites ? Moreover, if the figure in Persian dress is an Israelite offering the diadem to the idol in compliance with the royal command, why then the king's consternation ? The text of the Bible hardly suffices for the elucidation of this whole section of the Coptic relief, But a narrative in Hebrew legendary literature that describes Daniel's adventures in Babylon may explain its meaning. L. Ginzberg's summary of the story runs thus (*): ‘The Babylonian king

felt so ardent an admiration for Daniel that he sent him from the country when the time arrived to worship the idol he had erected in Dura... In.spite of all this, Nebuchadnezzar endeavoured to persuade Daniel by gentle means to worship an idol. He had the golden diadem of the high priest inserted in the mouth of an idol, and by reason of the wondrous power that resides in the Holy Name inscribed on the diadem, the idol gained the ability to speak, and it said the words: “Tam thy God”. Thus were'many seduced to worship the image, But Daniel could not be misled so easily. He secured permission from the king to kiss

the idol. Laying his mouth upon the

idol’s, he adjured the diadem in the following words:

“I am but flesh and blood, yet at the

same time a messenger of God. I therefore admonish thee, take heed that the Name of the Holy One, blessed be He, may not be desecrated, and I order thee to follow me. So it happened. When the heathen came with music and’song to give honour to the idol, it emitted no sound, but a storm broke loose and overturned it”. The king reaction to this miracle is not

mentioned, but in connection with the miraculous escape of Daniel's three companions we read

that he was ‘stunned with fright’ (ἢ, On that same occasion, the story continues, six other

miracles occurred, ‘ all of them driving terror to the heart of the king ’. One of these miracles was the fall of the idol which he had erected in the valley of Dura (9). The king’s persuasion of Daniel to worship, the miracle of the diadem performed by Daniel, and the consequent fall of the idol to the fright and terror of the king, these apparently are the (1) Fol. 270°; E. T. DeWald, Vaticanus Graecus alispanische Buchmalerei, Bonn-Leipzig 1922, pp. 10 1927 («The Illustrations in the Manuscripts of the f, 89 fL, pl. 32, fig. 98 (cf. our PI. VII»). (3) Op. cit, IV, pp. 337 f; VI, notes rog-rir, Septuagint», ΠῚ, i, Princeton x041,pp. 47 £, pl. LXX. (2) Roda Bible; Paris, Bibl. Nat, Cod. lat. 6, P. 427: vol. ti, fl. 64"; W. Neuss, Die hatalanische Bibelillu(4) Ibid, IV, pp. 328 fi. stration um die Wende des ersten Jahrtausends und die (5) Loc. cit. x10

TWO SIXTH-CENTURY COPTIC STONE RELIEFS

events shown in the Coptic sculpture. I know of no parallel in art to this sequel of episodes, while, as mentioned above (p. 107), normal iconographic tradition is observed in representing the figure of Daniel (1). The message of the little cycle is obvious and may be stated by citing a passage from a sermon by ‘ Pseudo-Demetrios ἡ, probably a late fourth-century archbishop of Antioch (Flavian ?). Thestory in Dan. iv, 30 fi. offers him the opportunity like Ephraim to praise Nebuchadnezzar’s penitence: “ Then the King Nebuchadnezzar knew that his power was nothing beside the power of God, and he repented; he confessed his impotence; he issued a proclamation to all the world so that all should submit to the God of the Heavens, and he said thus: “Iam the King Nebuchadnezzar; I praise, I exalt, I glorify, and I confess the king of the Heavens because there is no other God in the Heavens or on the earth who can heal thus among men; and it is He who humiliates those who walk in pride”’ (). If it is based on Hebrew legendary litterature, the subject matter of the last, incomplete scene of the relief would be difficult to ascertain, If, on the other hand, it is one of the better known biblical stories some conjectures may be offered. The figure whose back still remains in the relief may be one of the three Hebrews in the furnace, But since one would then expect to see flames, which are not visible, that theme, common in Coptic art ('), can probably be excluded, Instead, the scene may have represented Daniel in the lions’ den. The fragmentarily preserved figure in Persian costume would probably then be Habakkuk bringing food to the prophet (Dan, xiv,o f) (ἡ. Finally, the scene could be that of Daniel and the Dragon (Dan., xiv,26) or Daniel and Susanna (Dan, xiii,x fl), the back of the figure being either that of Daniel himself or of a companion often shown with him in pictorial representations of these incidents (). The incompleteness of the relief does not seriously impair its value as a document for iconographic study. From this viewpoint the significance of the slab lies primarily in those portions inspired by Jewish legendary material. As we have seen, the portrayal of Nebuchadnezzar in the guise of Cosmocrator may be based on a source of that type. In the scene centering upon the idol, Hebrew influence is paramount.

Embodied in these

few figures is a sequel of scenes that

originally may have formed at least three independent images: (1) Nebuchadnezzar who ‘ by gentle means’ invites Daniel to worship an idol; (2) Daniel before the idol, holding the diadem and probably accompanied by the king; (3) the fall of the idol in the presence of Daniel, the terrified king, and perhaps some of the heathen who ‘came with music to give honour to the

idol ().

(x) Ci. Berlin, Staatliche Museen zu Berlin, No.

242 (I. 3019), wood corbel from Bawit; O. Walf, Altchristliche und. mittelalteriche byzantinische und italienische Bildwerke, Berlin 1909, pp. 79 f. (ig); this relief most certainly belongs to the sixth-contury sculpture from the South Church at Bawit, although not to the same style group as our reliefs I and ΤΙ. (2) Sermon on 15, 16, 17, preserved in Rome, Bibl. Vaticana, MS. Copt. Vat. LXVII, fole, rot f. The citation is from fol. 137%, taken from the (French) translation of H. De Vis, Homélies copies de la Vaticane, Hauniae 1922, p. 193. On the possible identification of this Demetrios with Flavian, ibid., Pp. 127 f. (3) O. M. Dalton, A Coptic Wall Painting from

Wadi Sarga, « Journal of Egyptian Archaeology», II, 1916, pp. 35 fl; Dict. d'arch. chrét et de liturgie, VI, cols. 2107 fl. (H. Leclerea); cf. J. Muyser, Le culis des Trois Saints Jeunes Gens chez les Coples, « Cahiers Coptes », 6, 1954, pp. 17 f. (4) Diet. d'arch. chrát et de liturgie, IV, cols. 221 8. (5) Loc. eit; ibid. III, cols. 3004 f, fig. 3333; ibid., XV, cols. 1742 f; Wilpert, op. cit, I, p. 40, pis. XXIX, 3; CXV, 2; H. Stern, Les peintures du mausolée «De l'Esode» à El-Bagaouat, © Cahiers archéologiques s, XI, 1960, p. 112; J. Schwartz, Nouvelles études sur les fresques d'El-Bagauat, ibid., XIII, 1962, pp. 4, 7: (6) The possible sources of the Jewish elements in the relief will be discussed in the second part of rr

HJALMAR TORP Rem IL

The entire length of the slab is preserved, but it is somewhat battered at the left end. All the figures have suffered and the heads have been mutilated, perhaps wilfully, in the same manner as in relief I. The scenes on this slab are not difücult to interpret. In dense sequence are represented, from left to right, episodes from the life of David as related in the first Book of Samu Samuel anointing David (x Sam., xvi, 13); David going to King Saul (1 Sam., xvi, 20); David playing before King Saul (x Sam., xvi, 23); David meeting Goliath (x Sam., xvii, 40); Goliath prostrate (x Sam., xvii, 51). In all the scenes David is a young man, beardless, with short, curly hair, dressed in a short-sleeved, girdled tunic reaching to the knees and a maud (possibly the woolen, so-called Jane) over his shoulders. In scene si the cloak hangs down straight behind his back. In scenes dii and iv, and possibly also in scene $, the cloak after the Egyptian manner is flung across his left shoulder and tied, almost in the guise of a broad belt, around his waist with a knot in front, the free end hanging down. In the art of late Antiquity the short, girdled tunic and the maud is the dress typical of common people, and is worn by David as a shepherd. His footwear is undistinguishable in all scenes. Scene i (PI. ID). David receives the unction with a deferential bow before Samuel. In his right hand he holds a staff which, however, is not a shepherd’s crook, the sheephook (pedum), but a straight stick (baculum; cf. 1 Sam, xvii, 40), reminiscent of King Nebucchadnezzar's sceptre in relief I. David extends his left hand toward Samuel, The prophet is curly-haired, like David, and bearded. He wears a tunic reaching to his feet, and a pallium. With his right hand Samuel extends the horn of anointment over David's head; with hisleft hand he holds up a wreath of leaves. The two figures conform in general with the normal pictorial representations of the Anointing in East Christian art. Thus David’s attitude of modesty regularly belongs to the iconography of the scene and is sometimes emphasized by introducing the personification of Humility (cf. below, p. 119, and PI. VIII). The staff. however, is rarely found in this particular scene. The Bible does not specifically state that David was carrying it at the moment of his anointing, but even so, its presence in the scene does not necessarily presuppose a different literary source.

On Samuel’s request, Jesse had David summoned in haste from the fields where he was keeping the flocks: ‘For we will not sit down till he come hither’ (x Sam., xvi, 11). The staff in David's hand may be intended to suggest that he has come directly from his flocks. To my knowledge, the only other East Christian work depicting David with his staff at the moment of the Anointing is the carved panel of that scene on the tenth-or eleventh-century ivory casket in the cathedral of Sens (PI. VIII a) (). In this panel the bucolic setting of the scene is this study to be published in Acta IIL On the problem of Hebrew motifs in Christian art, see H. L. Hempel, Jüdische Traditionen in frithmittelalterlichen Miniaturen, + Beitrüge zur Kunstgeschichte und Archiiologie des Frihmittelalters » (= Akten zum VII. int. Kongress für Frihmittelalterforschung), Graz-Küln 1962, pp. 53 fi, with bibl. notes; K. Weitzmann, Zur Frage des Finflusses jülischer

Bilderquellen auf die Illustration des Alten Testamentes, 4 Mullus è (= Festschr. Theodor Klauser), Münster Westf. 1964, pp. 401 fl () A. Goldschmidt, K. Weitzmann, Die byzantinischen Elfenbeinshulpturen des X. bis XIII. Jahrhunders, I, Berlin 1931, p. 65, pl. LXXII, ἃ (ef. our PI. VIII a). A Western example is Moulins, Musée Municipale, Bible of Souvigny, fol. 93%. axe

TWO SIXTH-CENTURY COPTIC STONE RELIEFS

emphasized by the presence of David's animals grazing behind him. On the casket the Anointing is preceded by a scene showing Samuel choosing among David's elder brothers and by scenes of one or two incidents from his life as a shepherd (). The pastoral details in the scene of the Anointing — the animals and the staff — stress its connection with the representations of David the shepherd preceding it. Consequently, David's staff in scene é of the Coptic relief may be an indication that this episode was preceded by one or more scenes from the life of David as a shepherd, It is not possible to ascertain whether such scenes were actually carved by the sixth

century Coptic craftsman on another slab, now lost but originally preceding relief I, or were only to be found in the iconographical prototypes. The wreath in Samuel's hand poses different problems. This attribute is unique in representations of the Anointing and is not explained by the biblical story of David; nor have I been successful in tracing any other written source, whether apocryphal or legendary, that directly accounts for its occurrence in a representation of the Anointing of David. The wreath is, however, intended for David. This is borne out by the next scene (see below, scene 5i) where David actually wears a wreath of heart-shaped (ivy- ?) leaves on his head. One might interpret the wreath (ἢ) as a crown of victory, David's due for his future defeat of Goliath, shown in the concluding scene of the relief. An allusion, at the moment of the Anointing, to his future victories is contained in, e.g., Josephus, Jewish Antiquities, vi, 163 fl: ‘ Then, in the sight of David, he (Le. Samuel) took the oil and anointed him and spoke low into his ear, explaining that God had chosen him to be king. . .; he would subdue the Philistines and, victorious and triumphant over all nations with whom he might wage war, he would in his lifetime attain glorious fame. ..’, An interpretation in this sense, as an anticipated crowning of the future victor, thus seems plausible. On the other hand, it would probably be only an incomplete explanation

of the significance of the wreath, just as the act of anointing, as explained by Josephus, also implies more than a promise of military victories (?).

‘An interesting variant of the motif of the crown is found in a drawing from the Utrecht Psalter illustrating the apocryphal 151st Psalm (PI. VIII j) (). In the Septuagint this additional Psalm (which seems to have been translated from a Hebrew original) is appended to the Psalter with the following title: ‘ This Psalm was written by David with his own hand (and it is outside the number) when he fought in single combat with Goliath’. Its verses summarize the early story of the hero (): I was little among my brethren,

and the youngest in my father’s house; I fed my father’s sheep. My hands made a harp, And my fingers tuned a psaltery.

And

who will declare unto my

Lord?

(ἡ Goldschmidt, Weitzmann, 1. c. pl. LXXII, a,b. angel's right hand cannot readily be distinguished in (a) On the meaning of the wreath in general, ‘he reproductions but its identification seems certain seo K. Baus, Der Kranz in Antike und Christentum. to DeWald, le.’ ... an angel of the Lord descends Eine religionsgeschichtliche Untersuchung mit beson- from heaven fo place ἃ wreath upon his (e. David's) head" derer Berüchsicigung Tertullians, Bona 1940. (3) D. Lys, L'onction dans la Bible, « Etudes théoand the quoted on Peslm notes from (s) Theof itabove logiques et religieuses ν, 20, 3, 1054, pp. 3 f. version are talea G, P. 15%Huntington, H. A. (4) Fol. or"; DeWald, The Utrecht Psaltr, op. cit, ‘Metcalf, Treasury of the Psalter, New York 1892, pL CXLIV (cf. our PL. VII f), p. 72. The object in the PP. 575 f. 113

HJALMAR TORP

He is the Lord, and He heareth me. He sent His angel and took me from my father’s sheep, And anointed me with the oil of His anointing. My brethren were comely and tall, And in them the Lord had no pleasure. I went out to meet the Philistine, And he cursed me by his idols; But I drew the sword from his side, and beheaded him, And took away the reproach from the children of Israel. In the drawing from the Utrecht Psalter an angel is shown descending from heaven, carrying a wreath (cf. p. 113, n. 4) for David who guards bis flock; to the right, in the succeeding scene, David is shown as the victor, brandishing the head of his slain foe. He does not, however, wear the angel's wreath, which, therefore, can hardly be specially intended for David as the slayer of Goliath. Since the angel appearing to the shepherd obviously illustrates the verse ‘ He sent His angel and took me from my father's sheep’, the attribute carried by the angel should instead, it seems, be explained in the light of the subsequent verse: ‘ And anointed me with the oil of His anointing’, that is, as ἃ symbol of ‘ spiritual unction’. If this is so, the illustration of Psalm 151 in the Utrecht Psalter assimilates the angel’s wreath with the anointing in a manner that corresponds to the relief where Samuel displays the wreath in balanced juxtaposition to the horn of unction. Correspondingly, David’s coronation with the wreath may be conceived as a complementary expression of his being anointed. As Josephus states, Samuel's anointing of David first of all means ‘ that God had chosen him to be king”. Samuel's simultaneous crowning of the boy, accordingly, may be explained as an anticipation of David's coronation with the royal diadem at the moment of his elevation to the throne, after the death of King Saul. Such an interpretation is consistent with the meaning given to Samuel's anointing of David in 1 Chr., xi, 3, as a ‘ word of the Lord by Samuel’ which became fulfilled at the moment of David's final exaltation to kingship (!), and is also supported by the wall-paintings of the Synagogue at Dura-Europos (245-256 A. D.). In the panel of the Anointing, David, in contrast to his brothers, wears a royal purple himation, a rather obvious allusion to his future kingship (*). Perhaps the sceptre-like form of David's staff in the Coptic relief and other medieval works of art may be similarly explained as alluding to his future reign;

this, however, is less certain and would be difficult to prove. ‘The anointing of David was a ‘ word of the Lord by Samuel”, a promise of victories and kingship. But the ceremony also, and above all, was a sacred rite of consecration to the Lord: ‘ Then Samuel took the horn of oil, and anointed him in the midst of his brethren: and the spirit of the Lord came mightily upon David from that day forward’ (x Sam., xvi, 13). To be anointed means to be seized by the spirit of God, to become His servant. The presence of the spirit of the Lord in the anointed one inspires the realization of the tasks which God has charged him to accomplish

for the benefit of His people (*).

‘The actual presence of the divine spirit in David after his coronation as king is illustrated in one of the ninth-century miniatures in the Paris Byzantine Psalter (PI. VII f) (‘). The king

(1) S. Amster, David, Roi et Messie. La tradition 123, 1941, pp. 173 f; C. H. Kraeling, The Symadavidigue dans l'Acien Testament, Neuchâtel 1963, gogue (¢ The Excavations at Dura-Europos. Final P. 30 with n. 1. Report #, VIII, 1), New Haven 1956, pp. 165 ff. (@) A. Grabar, Le thime religione des fresques de la (3) Lys, op. cit, D. 34. synagogue de Doura, « Revue do l'histoire des religions» (4) Paris, Bibl, Nat, MS. gr. 139, fol. 7. 14

TWO SIXTH-CENTURY COPTIC.STONE RELIEFS

stands between personifications of Wisdom and Prophecy and is depicted with the Dove of the Holy Spirit above his head. It would seem that both the Coptic relief and the Carolingian illustration in the Utrecht Psalter by the very choice of scenes intend to express that it is through the presence of the divine spirit in the anointed shepherd that Goliath is defeated. Indeed, it appears possible to argue in favour of an interpretation of the wreath similar to that of the Dove above David's head in the Byzantine miniature, In Antiquity the crowning with a wreath or a garland often formed part of the rifus consecrationis (1), and the expression to ‘receive the crown’ in religious texts was commonly used to indicate the vision of or complete mystical possession of the deity. The metaphor is found frequently in Hebrew and Christian writing and often occurs in the Bible, as in Psalm 21: ‘For thou shalt prevent him with the blessing of goodness and shalt set a crown of pure gold upon his head’, or Wisd., v, 15, 16: ‘ The righteous live for evermore; their reward also is with the Lord, and the care of them is with the most High. Therefore shall they receive a glorious kingdom, and a beautiful crown from the Lord's hand.’ The literary image of the crown was especially favoured by the unknown author of the so-called Odes of Solomon, written perhaps in Syria (în Greek or Syriac) about the first century A. D. (). It is a disputed question whether the Odes are of Hebrew or Christian (or Gnostic) origin, but in any case they were very soon admitted by the Christians (). They were known in Coptic translations as well as in Greek to the Gnostics and the Christians of Egypt (8). In the English version by R. Harris the opening lines of Ode i are:

The Lord is upon my head like a crown; And I shall not be without Him. The crown of truth was woven for me; In Odes v, ix, xvii, and xx the image of the crown is adopted in related but always varied forms. Thus, in Ode xvii:

I was crowned by my God: And my crown is living; And in Ode xx:

But put on the grace of the Lord without stint; And come into His Paradise and make thee a garland from His tree,

And put it on thy head and be glad; (1) Baus, op. cit, pp. 66 ff, 124 ft, etc. (2) R. Harris, A. Mingana, The Odes and Psalms of Solomon, TI, Manchester 1920, pp. 61 f., 138 f.

(3) J. Carmignac, Un Qumrénien converti aw Christianisme : l'auteur des Odes de Salomon, «Qumran- Probleme. Vortrige des Leipziger Sympo-

sions über Qumran-Probleme », Berlin 1963, pp. 75 £. (4) Harris, Mingana, op. cit, p. 186. Odes i,

ν, vi, xxii, and xxv are known in Coptic, Odes iii (from the middle) to xlii in Syriac, Ode xi in Greek, cf. Carmignac, op. cit, p. 104, n. 6. us

HJALMAR TORP

Odes v and ix are of particular interest, because in them the metaphor of the crown appears in contexts that bring to mind David before the fight against Goliath and the illustration of Psalm 151 in the Utrecht Psalter. Ode v: And He is as a garland on my head, And I shall not be moved;

Even if everything should be shaken, I stand firm: In Ode ix, finally, the connection with David is the more obvious, as the poem is composed on the basis of Psalm 20: An everlasting crown is Truth, Blessed are they who set it on their heads: Put on the crown in the true covenant of the Lord; And all those who have. conquered shall be inscribed in His book.

For their inscription is the victory, which is yours; And she (Victory) sees you before her, and wills that you shall be saved (1). Thus, in the Odes the metaphor of the crown successively takes on all the various meanings which were adduced above for the wreath of David-in the Coptic sculpture (?): a crown of victory, ἃ crown of kingship, and, especially, the meaning which embraces all the others, the wreath as a poetic expression of the religious concept ‘to have God’, Θὲον ἔχειν (). More specifically, the wreath in the relief expresses the intervention of Jahveh in the events of David's ascent to the throne (ἢ). Some of these events are depicted in scenes ii-iv. Scene $i (PI. IV). In his shepherd's dress and with the wreath of heart-shaped leaves on his head, David is shown moving to the right, on his way to King Saul. According to the Bible, Jesse ‘took an ass laden with bread, and a bottle of wine (i. e. a wineskin), and a kid, and sent them by David his son unto Saul’ (τ Sam., xvi, 20). David carries a basket over his right arm and the kid around his neck. It is uncertain whether he held the bag of wine, or perhaps the staff, in his damaged left hand which is extended towards the king to his left. Neither seems likely; the ass, at any rate, has been omitted,

The scene is extremely rare in medieval art. In fact, the only possible instance of the episode registered by the Princeton Index of Christian Art is a small miniature in the Byzantine Psal(1) Cf. the ‘Angel of Victory’ represented in the David-Goliath panel of the Milan doors (infra,

P. 117, n. 3); C. O. Nordstrôm, Some Jewish Legends in Byzantine Art, « Byzantion s, XXV-XXVI-XXVII (1957) (= Mélanges Ejnar Dyggve), pp. 495 ff (2) Harris, Mingana, op. cit., pp. 207 ff. (3) On this concept, seo H. Hanse, ‘Gott Haben" in

der Antike und im frühen Christentum. Eine veligionsund begriffsgeschichiliche Untersuchung, Berlin 1939. (4) C£. Amsler, op. cit, pp. 31, 39 f. The use of he consecration wreathin Egyptian art of the GrecoRoman and Coptic periods will be discussed in the second article in connection with the study of the iconographic origins. 116

TWO SIXTH-CENTURY COPTIC STONE RELIEFS

ter of the Vatican Library, MS. Vat. gr. 752 from about roo (PI. VIII c) (1). It depicts David marching through a mountainous landscape, probably on his way to Saul. He carries a knapsack fastened to the staff over his left shoulder and a kid under his right arm; the ass again is omitted. Although the artist has taken certain liberties in his selection of the gifts for King Saul, there can be little reason to doubt that the scene in both the Coptic relief and the Byzantine miniature is based upon the biblical text. Each may, however, be derived from prototypes that illustrate the text more faithfully and completely (?). Scene iii (PI. IV). King Saul, seated, and in a posture implying distress, leans heavily upon the sceptre in his left hand. He is turned towards David, who is playing ἃ lyre at his right and whom he seems to address since the fingers of his (much damaged) right hand express the attitude of speech. Saul appears to be bearded. He is dressed in tunic and chlamys, like King Nebuchadnezzar in relief I; in fact, his throne — cushioned and with small holes imitating a jewelled decoration — his sceptre, and his diadem with its upright, curved and gabled elements are all similar to those of the Babylonian king. David is shown playing his lyre, standing in a relaxed position with his left leg crossing the right one; he no longer wears the wreath upon his head. In the Middle Ages artists preferred to represent David playing the lyre before Saul according

to x Sam., xviii, τὸ f,, or 1 Sam., xix, 9 f., thus combining the playing and Saul's threat to kill the boy with his spear (. When, as in the Coptic relief, 1 Sam., xvi, 23 is represented, Saul usually reclines on his bed with David near-by (9. However, a composition similar to that of the Coptic relief exists in one of the wooden panels dating about 400 from the old doors of St. Ambrose in Milan (). In it, David, playing the lyre, stands to the left of King Saul who twists unhappily on his throne. Scene iv (PI. V). David (without the wreath) is moving energetically towards the right to confront Goliath who rushes forth from the border of the relief. With his left hand David holds his staff protectively in front of him. A small bowl-shaped container for his stones hangs down

from his elbow. The sling loaded with a stone swings from his right hand; David is ready to hurl the fatal missile against the Philistine. Goliath wears a helmet, a cuirass with overlapping scales and belt, and high boots. A military cloak, the sagum, covers his upper left arm and flares out in a homlike fold behind him. The giant protects himself with his shield on his left arm while threatening David with a spear lifted horizontally above

his head,

the

point

touching

David's staff. Between, and partly hidden by the two bouncing antagonists, the fallen, though

(x) Fol 1°; E. T. DeWald, Vaticanus Graecus 752 (« The Illustrations in the Manuscripts of the Septuagint », ΠῚ, ii), Princeton 1942, p. 3, pl. II. (2) In this connection it is worth noting that on the ivory casket at Sens the scenes that follow the Anointing depict Samuel (or Jesse ?) asking David to go to Saul and the presentation of David to Saul, Goldschmidt, Weitzmann, 1, c., pl. LXII, e, f. In View of the rarity of scenes illustrating 1 Sam., xvi, 20, it is worth mentioning that the Gothic Tickhill Psalter, New York, Public Library, MS. Spencer 26, fol. οἵ, contains a miniature of the preparation for David's departure, including the ‘ ass laden with bread... *. (3) Examples: Bawit, wall-painting, J. Clédat, Le monastère et la nécropole de Baouîl, «Mém. de l'Inst. fran. d'arch. orientale», XII, 1, 1904, pp.

19 £., pl. XVI; Rome, Bibl. Vaticana, MS. Vat. gr. 333, fol. 25%; etc. (4) Examples: Milan, Bibl, Ambrosiana, MS. E. 49-50, p. 778, A. Grabar, Les miniatures du Grégoire de Nazianze de l'Ambrosienne, Paris 1943, pl. LXXIV; Rome, Bibl. Vaticana, MS. Vat. gr. 752, fol. 27, DeWald, op. cit.,p. 4, pl. ΠῚ; Rome, Museo di Palazzo Venezia, Ivory Casket, Goldschmidt, Weitzmann, op. cit, p. 64, pl. LXXI, c; etc. (5) A Goldschmidt, Die Kirchenthür des M. Ambrosius in Mailand, Strassbourg 1902, p. 13, pl. ΠῚ cf. pl. VI for the restaurations. See also F. Reggiori, Cimeli © capi d'arte della Basilica Ambrosiana poco noti o mal noti, « Ambrosiana. Scritti di storia, archeologia ed arte pubblicati nel xvi centenario della nascita di Sant'Ambrogio CCCXL-MCMXL», Milan 1942, pp. 168 ff, pl. XVIII. E]

TWO SIXTH-CENTURY COPTIC STONE RELIEFS

emblematic character of the relief is made clear by comparing the Anointing (PI. ID) with the parallel scene on fol. 22” in Bibl. Vaticana, MS. Vat. gr. 333 — the only illustrated volume of the Books of Kings which is known (PI. VIIL) (). The miniatures includes all the persons mentioned by the biblical text: the two principal persons, Jesse, David’s seven brothers, and also a woman with a veil over her head who points to David ~ obviously a personification, probably of πραότης, Humility. Behind the figures there are parts of a building to suggest the house of Jesse as the setting for the scene. The Coptic craftsman has carved a series of figured symbols rather than of narrative scenes. This art of symbols is characteristic of most early Christian painting and sculpture as we know from the catacombs, the Baptistry at Dura-Europos, the sarcophagi, and the decoration on furniture and on objects used in the cult. Such representations, even when more scenes from the Old or the New Testament occur together, were not primarily seen as component parts of a narrative but as individual paradigms of faith and salvation (ἢ Several scenes from the life of David can be found in this symbolic art from the third to the fifth century (). Among these is the encounter between David and Goliath, which is also rendered in the relief (8). Yet the relief, for several reasons, does not appear to have been composed of symbolic paradigms, In the first place, the scene of encounter is, as we have demonstrated, a conflation

of two

originally distinct

representations

of successive moments

in the combat.

Secondly, there is the possibility that one or more episodes originally preceded scene i (cf. above, Ρ. 113). Notwithstanding the laconic conciseness of the pictorial expression, the sequence of scenes carved by the Coptic craftsman, instead of being ἃ series of symbolic paradigms would seem to be summarized from a more extensive cycle whose individual scenes have been abbreviated and, in the case of scene iv, combined (). This conclusion is consistent .with that reached above, in the analysis of relief I; both reliefs are iconographically derived from lengthier cycles.

The very ruins of the monastic settlement of Bawit provide material for developing the preceding observations on relief I and IL This imaterial, in conjunction with the various elements of Hebrew, Oriental, and early Christian origin pointed out above, also will help to trace the iconographic sources for the Old Testament scenes carved on the two Coptic reliefs, ven 1934, pp. 275 fL,

pl. XLVII, 2. Cf. also an early

ighth-century wall-painting in S. Maria Antiqua, Rome, where David, in the attitude of a victorious ‘emperor, places one foot upon the prostrate enemy, J. Wilpert, Die romischen Mosaihen und Malereien der kirchlichen Bauten vom IV. bis XIII. Jahrhundert, IV, Freiburg 1917, pl. 178. (2) Lassus,op.cit., pp. 57 fl, fig.4 (ci. our PI. VIII). (2) This view probably holds good even in such cases where separate scenes illustrating consecutive moments of a narrative were juxtaposed: to form a ‘micro-cycle’, as on the lost sarcophagus, Wilpert, op. cit, I, p. 57, fig. 24. On it, the central clipeus with the image of Christ is flanked to the left by the combat of David and Goliath and to the right by David bringing King Saul the head of the slain foe. On the religious significance of this ‘symibolic’ art, see, f. inst, E. Stommel, Beilrlge sur Thonographie der honstantinischen Sarkophagplastib, Bonn 1954. PP. 30fî, 53 ff, 58 fl.

(3) Cf. above, p. 118, n. 3. (4) Above, Lc. (5) On the problem of the mutual relation of the ‘symbolic’ and the ‘narrative’ trends in early Christian art — the question, that is, of whether the former is ἃ ‘theological’ condensation of the latter or, rather, whether it represents a distinct iconographic tradition -- see K. Weitzman, Die Illustration der Septuaginta, «Münchener Jahrbuch der bildenden Kunst s, TIL/IV, 1952/53, pp. 96 ff., 107; Kraeling, op. cit, pp. 361 fh, 385 fL, 398 f. Abbreviations of narrative representations in the manner of the David scenes of our Coptic relief II are, of course, a most normal procedure in medieval art. Thus, abbreviated narrative scenes joined to form sequences comparable to those of the two Coptic reliefs were found in the excavations of the Martyrion at Seleucia Pieria, near Antioch (Antioch où the Orontes, III, Princeton 1941, pp. 124 fh, 135 ff). 119

Hj. Toro, Two sixth-century Cobtic stone riliejs

PLATE I

ÀH E E $ἕξ £ ὅ

PLATE IT

Detail of Plate La.

Hj. Torp, Two sixth-ceninry Coptic stone reliefs

Hj. Torp, Two s

plie stone reliefs

PLATE IN

PLATE IV

Detail of Plate 14.

Hj. Torp, Two sixth-century Coptic stone reliefs

Hj. Toro, Two sixth-century Coptic stone reliefs

PLATE V

Hj. Terp, Two sixth-century Coptic stone reliejs

PLATE VI

&, 9) Cairo, Coptic Museum, Nos. 37798 (7110) and 35838 (7135): Reliefs from Bawit. o) Paris. Louvre; Relief from Bevit.

Hj. Torp, Two sixth-century Coptic stone reliefs

PLATE VIL

a) Rome, S. Sebastiano, Museum: Fragment of sarcophagus (detail). 2) Rome, Cemetery of S. Ciriaca: Fragment of sarcophagus (detail). c) Rome, Bibl. Vaticana, MS. Vat. gr. 333, fol. 9 v.d) Mt. Athos, Panteleimon, ΜΒ, 6, iol, 164 r (detail). e, f) Paris, Bibl. Nationale, MS. lat. 6, vol. it, fol. 64 v; MS. gr. 139, fol.7.

Hj. Torp, Two sixth-century Coptic stone reliefs

χε (ee

PLATE VII

j

NPUCNAG

JNrhicpsEiN SE

^

E

MANIEPORSCRIPET

a) Sens, Cathedral: Ivory casket (detail). b, z, 2) Rome, Biol. Vaticana, MSS. Vat. gr. 333, fol. 22 v. Vat. f) “trecht, University Library, MS. 32, fob mr v fol, 254 gr. 752, fol. rv (detail; Vat. gr. 1927, and Dit v9. €,v (details.

The mosaics of John VII (705-707 A.D.) The mosaic fragments and their technique PER

JONAS

NORDHAGEN

The mosaies from the Oratory of John VIL in the Old Basilica of St. Peter have long been known to students of early Christian art. The preserved mosaic fragments as well as Grimaldi’s drawings and descriptions of the mosaics were published by E. Mintz () and R. Garrucci (?) in'the last quarter of the nineteenth century, and shortly before 1900 the subject was again discussed by G. B. de Rossi (*). A new fragment came to light a little later and was described by A. Bartoli (ἢ. In W. de Griineisen’s work on S. Maria Antiqua, the entire material was for the first time published in photographical reproductions (*), while in J. Wilpert’s Die rümischen Mosaiken und Malereien, where the Oratory and its decoration are discussed at length, some of the fragments were reproduced in colour (*). Thé mosaic fragments were described and their

restorations briefly listed in the study on early medieval painting in Rome by E. Kitzinger (7. Because of their well-documented date (*), the mosaics of John VII's Oratory constitute one of the fixed points in a period where much uncertainty still reigns as to the chronology of the preserved monuments. Yet, despite the extensive literature on the subject, much remains to be investigated concerning the technique, style and iconography of these mosaics. In the present paper, the author will re-examine the mosaic fragments, and the aims of this study are three-fold. First, we shall delimit all areas that have been subject to restorations in order to exclude these

parts from future discussions of the mosaics, Secondly, we will present a detailed description of the original mosaic surface, according to the principles for mosaic analysis laid down in T. Whittemore's reports on the mosaics of Hagia Sophia at Istanbul (*). Solely through a description of this kind can full knowledge of the technical and aesthetic characteristics of the mosaics be obtained. Thirdly, we shall conduct a comparison, based on technical qualities, between the mosaics of John VII and other mosaics, Western and Byzantine, of the Early Middle Ages, in an effort better to establish the position of the Oratory mosaics within the interplay of artistic tra-

ditions of the period (9).

(1) E. Mantz, Notes sur les mosaïques chrétionnes de Malis, IV. L'Oratoire du pape Jean VII, « Revue Archéologique», 34, 1877, PP. 145 fi. (2) R. Garrucci, Note del sig. Mint sul musaico di papa Giovanni VII, à Civiltà Cattolica », serio X, vol. V, 1878, pp. 339 fi; by the same author a more thorough publication in Storia dell'arte italiana, I, Prato 1881, pp. 581 fi; IV, Prato 1877, pp. 97 ff. (3) G. B. de Rossi, Musaici cristiani, Rome 1899, fase. XXIII, pp. 1 fi (4) A. Bartoli, Un frammento inedito dei musaici vaticani di Giovanni VII, « Bollettino d'arte», 1, 6, 1907, pp. 22 ff.

(5) W. de Griineisen, Sainte Marie Antique, Rome 1911, pp. 280 fi. (6) J. Wilpert, Die rómischen Mosaihen und Maleveien der kirchlichen Bauten vom 4. bis sum 13. Jahrhundert, Freiburg 1916, I, pp. 388 f . (text); ΠῚ, pls. 113, 1-3, and 114, 2. Ὁ) E. Kitzinger, Romische Malerei vom Begin des 7. bis sur Mitte des 8. Jahrhunderts, Munich 1934, PP. 15 Í, 20 and 39. (8) Seo Wilpert, op. cit., I, pp. 388 f. (9) T. Whittemore, The Mosaics of St. Sophia at Istanbul, I-IV, London 1933-52. (10) Some results of the author's studies were 12r

n

PER JONAS NORDHAGEN ‘As for thé comparative material, the author has drawn upon the published information known to him concerning the technique of early medieval mosaics (). In a limited number of cases, where the mosaics are within easy reach from the ground, he has himself been able to carry out an examination of the monuments (*), A number of mosaics that are inaccessible without special scaffoldings, have been studied through good photographic enlargements (). The author's work has been greatly facilitated through the kind assistance of Prof. Filippo Magi, Reggente of the Vatican Museums, under whose supervision a considerable part of the material relating to the Oratory of John VII rests. Dr. Francesco Vacchini, of the Reverenda Fabbrica di S. Pietro, gave valuable help during the study of the fragments in the Vatican Grottoes. In the churches where some of the mosaic fragments are kept, in Rome, Florence and Orte, all possible assistance was rendered by their clergy. Prof. Stylianos Pelekanides gave the author access to the mosaics in the church of Hagios Demetrios in Thessaloniki, and due to the kind permission of His Beatitude Athenagoras I, Patriarch of Constantinople, the important mosaic fragment in the church of Hagios Nicolaos in Istanbul could be included in the comparative material. submitted in Nuove constatasioni sui rapporti. artistici ἵνα Roma e Bisanzio sotto il pontificato di Giovanni VII (705-707), « Atti del 3° Congresso Internaz. di studi sull’alto medioevo », Spoleto 1959, pp. 445 fi. (1) The technical aspects of mosaic art have not aroused the same interest among scholars as have style and iconography. However, some information on the technique can be gathered from general works on mosaic art like E. Mintz, La mosaique chrétienne, Paris 1895; E. W. Anthony, A History of Mosaics, Boston 1935, and A. Frolow, La mosaïque murale byzantine, «Byzantino-Slavica., 1951, pp. 180 fi. Of the countless monographs on Early Christian and Byzantine mosaics, almost none contain technical studies; exceptions are Whittemore's publications of the mosaics in Hagia Sophia at Istanbul (cf. n. 9, p. 121), followed by other studies on the Hagia Sophia mosaics; cf. P. A. Underwood and E. Hawkins, The Portrait of the Emperor Alexander, «Dumbarton Oaks Papersν, 15, 1961, pp. 189 ff. Other sources for the study of mosaic technique are G. Bovini's many articles on the discoveries made during the cleaning and consolidation of the Ravenna mosaics, e. g., his Nuove constatazioni sulla tecnica ὁ sui mosaici di S. Apollinare Nuovo di Ravenna, « Atti del 1° Congresso Naz. di Archeologia Cristiana », Rome 1952, pp. 101 fi (2) The following monuments have been examin cd by the author: In or near Rome : 1) Mosaics in Mausoleum M under St. Peter's; 2) the Silvanus mosaic in the Lateran Museum; 3) the Cappella di S. Venanzio, Lateran Baptistery; 4) the Cappella di 8, Zenone, in S. Prassede; 5) the fragments from Leo III's triclinium ; 6) the St. Sebastian fragment in S. Pietro in Vincoli; 7) the mosaic at Narni. Ravenna: 8) the Mausoleum of Galla Placidia

(lower parts); 9) the panels with the Crist cycle from S. Apollinare Nuovo (exhibited at Ravenna 1957-58); 10) the Reparatus mosaic at 5. Apollinare in Classe. Thessaloniki: 11) the panel with St. Demetrius between two donors, in Hagios Demetrios. Istanbul: 12) the fragment in the church of Hagios Nicolaos; 13) the panel with Emperor Leo VI in the narthex of Hagia Sophia (studied 1959 from the scaffolding of the Byzantine Institute of America); 14) the panel with Emperor Alexander (studied 1959 from the scaffolding of the Byzantine Institute). (3) For the mosaics in Rome, the author has had the opportunity to study the photographs made by the Soprintendenza dei Monumenti of the following works: 1) SS. Cosma e Damiano; 2) S. Lorenzo fuori le mura; 3) S. Agnese fuori le mura; 4) 5. Stefano Rotondo; 5) SS. Nereo e Achilleo; 6) S. Prassedo; 7) S. Maria in Domnica; 8) 8, Cecilia in Trastevere; 9) 8. Marco. ‘Anderson, Rome, has provided the photographs regarding: 10) S. Pudenziana, and, as to 11) S. Maria Maggiore, the files of the Archivio Fotografico of the Vatican Museums supplied the necessary material. The mosaics of Ravenna have been photographed by the Ditta Trapani for the Soprintendenza dei Monumenti; photographs from this firm have served the author for the study of 12) the Orthodox Baptistery and 13) the Imperial panels in S. Vitale. As to Thessaloniki, excellent photographs of 14) Hagia Sophia were furnished by the firm Lykides. The large photographic collection which covers 15) the mosaics of Hagia Sophia at Istanbul is now located in the Dumbarton Oaks Research Library and Collection, Washington D. C., and was consulted when the author had a research scholarship to that institution. 122

THE MOSAICS OF JOHN VII (705-707 AD.)

‘The author is indebted to the University of Oslo, the Italian Ministry of Foreign Affairs and the Norwegian Research Council for Science and the Humanities for scholarships which have made possible his studies in Italy, Greece and Turkey. The first of the article’s three sections (pp. 124-144) comprises a catalogue of the mosaic fragments, The second section (pp. 145-153) concerns the technique employed in the mosaics. The final section (pp. 154-164) will discuss the relationship of the Oratory mosaics to early medieval mosaic art in Italy and in Byzantium.

123

CATALOGUE OF THE MOSAIC FRAGMENTS Cat. no. The Virgin represented as Maria Regina Orans, fragment mounted as altar picture for the Ricci altar, S. Marco, Florence.

Measurements, Total height of mosaic panel: 349 m. width: 1.36 m. Height from Virgin's left shoe to highest point of nimbus: 2.70 m. Diameter of nimbus: 0.72 m. Distance from top of diadem to line between throat and jewelled collar: 0.55 m, from browto chin: 0.21 m. Greatest width of face: 0.18 m. Restorations (text fig. 1). Gold background surrounding figure, and greater part of podium restored. Nimbus with its outline is original; an area of original gold ground seen around shoes. Hands and lower parts of both arms restored, as are also some of upper parts towards shoulders. Lower part of figure has been broadened by addition of restored zones to both sides of original part; large restored area cuts across the figure below the waist. Smaller restorations occur here and there in dress, In halo, to left of head, a broad crack in the plaster has been repaired with new gold tesserae. Through diadem and face runs a series of fissures, all filled in with new tesserae. For detailed discussion of restoration of hands and face, see below, [US General description (PI. 1), The Virgin stands, frontally, her right leg as the Standbein. She wears Byzantine court costume and diadem. Arms, restored, reflect original gesture (p. 127), that of an orant. Head inscribed in golden nimbus. Wide open eyes with strongly articulated shadows and lines. Nose long, narrow, sharply pointed. Mouth relatively small and thin. Dress consists of tunic, dalmatica and #rabea. Tunic, tunica interior, seen at lower parts of arms, is purple. It has a manica (restored, but evidently faithful to original) of golden stripes and rows of pearls. Over ‘tunic hangs long, purple dalmatica, sleeves of which are seen at elbows. Lower part of dalmatica has pattern of dark wavy lines, and is decorated with high, white and blue rectangular fabula under Virgin's left. knee. Edge of dalmatica curves over left shoe and lies as broad fold on right shoe. Over dalmatica she wears purple trabea, whose long hanging lap, bordered by pearls, is seen against the dalmatica on lower part of body. Trabea : from hanging lap goes up front of figure, over left shoulder around to back. Reappears on right shoulder, and from here — under fold of its previous tum — runs across breast towards left hip. From here it again circles body, and when reappears, runs obliquely across lower part of body from lower left upwards to right where it is fastened at belt (). Long, sloping, pearl-studded border seen across lower part of figure belongs to last turn of the traba. However, border that should run across the breast, from left upwards to right, not reproduced in mosaic, although clearly visible on nearly identical Maria Regina figure of contemporary icon from S. Maria in Trastevere (). To left, on lower part of figure, long vertical fold lines mark Standbein; to right concentration of short oblique folds where knee of Endlastungsbein presses against dress. In breast area series of vertical folds runs to belt. On shoulders are large, orbiculi of red and blue jewels framed in gold, surrounded by one row of pearls. Relationship of orbiculi to costume is not clear: usually shoulder segments belong to decoration of dalmatica (), (1) On the draping of the #rabea, see M. G. Hous(3) Cf. the Empress Theodora in S. Vitale ton, Ancient Greck, Romam and Byzantine Costume — venna; A. Grabar and C. Nordenfalk, Peinture and Decoration, London 1947, pp. 142 fi. (on p. 144 tine, Geneva 1953, illustration p. 64 (the right particular mention is made of the figure from the er). In the dress of tho EmpressZoe in Hagia

present fragment). (6) C. Bertelli, La Madonna di Santa Maria in Trastevere, Rome 1961, fig. 1.

at RaByzanshouldSophia

at Istanbul (eleventh cent), the orbiculi, however, may be part of the outer tvabea, cf. Grabar, op. cit, illustration p. 98.

T4

PER JONAS NORDHAGEN

tesserae. Shadow between eyebrows, immediately above nose, formed by three cubes of light grass-green glass. Lines of eyebrows, of which the right continues uninterrupted as outline of nose, are black glass. Under eyebrows lines of green transparent glass continue down to right as outer shade line along nose contour, and to left — changing from transparent to light yellow-green - as outline of nose. Tip of nose outlined in black, at point two dark orange tesserae. Shadow under nose of two transparent green and two light green opaque tesserae, To left of nose line of dark orange glass leads out on cheek. On right cheek line of dark orange glass runs parallel to cheek contour. Eyelashes set in purple-black; above them row of pale marble tesserae indicates eyelids, outlined by row of black glass (the last tessera on both sides in purple-black). Pupils single large black tesserae. Iris three grey marble tesserae surrounded by circle of rustbrown tesserae. Eyeball groyish-white marble and light ultramarine glass tesserao, lower outline purple-black. Curved line under eyes transparent and light green opaque tesserae. Shadow stripe under mouth and line dividing lower lip from upper, black. Lower lip dark orange glass, upper lacquer-red glass. The throat. Lower contour towards collar green transparent glass. Marble of many different types in flesh area of throat, e. g. ἃ rather dark colour in line directly under chin. On upper right part of throat shaded area of same marble, Far left on throat a series of curved rows with marble tesserae slightly larger than ordinary size. To lower right on throat a dark shaded area of purple glass tesserae of varying lighter and darker values. The nimbus. Gold tesserae of fairly small size. Outline separating nimbus from golden ground composed of inner circle of white glass, three circles of dark azure-blue. The diadem (PI. II). Purple-black contour over Virgin's brow continues to both sides as division line between diadem and pendant pearl bands. Above, golden band of double row of gold tesserae divided into six small fields by lines of transparent green glass. Next, row of nine pearls, each pearl made up of one central, large tessera of white glass, surrounded by circle of azure-blue glass tesserae with outline of ultramarine. Then zone of six blue jewels, each jewel has a core of azure-blue tesserae with outline of ultramarine, surrounded by frame of gold tesserae, Each of three high points has triangular jewel surrounded by pearl rows of type already described. Jewels, framed by gold tesserae, are set with dark orange tesserae in side points, while point in middle set with azure-blue. Rubies topping side points dark orange glass outlined by tesserae in rust-brown. Trefoil on middle point consists of pearls of usual type, central pearl having extra large, white tessera of coarsely grained marble, about 20 x 15 mm. Diadem outlined against gold of nimbus by dark ultramarine tesserae, Farthest perpendulia (pearl bands hanging from diadem) to left white marble for centres of pearls. Others all have white glass. Between pearl bands and area of face and throat run some vertical rows of alternating gold and purple-black tesserae. The collar (PI. IV). Under neckline runs row of pearls, followed by zone of blue jewels and new row of pearls, all made as described above. Large jewel pendants hang from single pearls lying along edge of collar. Four red pendants have rust-brown outline around core of dark orange tesserae, three white have ultramarine outlines surrounding core of white glass tesserae. From each pendant hangs small white pearl, of ultramarine around one white tessera, and, below, single tessera of gold. The orbiculi. Orbiculi on shoulders surrounded by row of pearls, inside of which runs line of gold tesserae, from which short rows of gold tesserae enter area of jewel proper, imitating prongs securing precious stones in real jewellry. Central jewel of each orbiculus consists of outer border of dark orange, inner border of lacquer-red surrounding square containing gold tesserae mixed with azure-blue. The belt. Strap of two rows of pearls, each pearl with centre tessera of white marble surrounded by azure-blue glass tesserae and outer zone of gold. Pearls of same type edge belt disc of dark orange centre surrounded by double line of rust-brown.

The costume. Garments all set with purple tesserae, caput mortuum of different values. For darkest shade-lines black used, as well as purple-black and some brown-black. Many intermediate values of purple also used in modelling, while lightest part of drapery indicated with very light purple. For the principle of modelling, see p. 152. Long pearl border on lower part of figure feature pearls of same type as those of diadem and collar. Where this border reaches restored vertical zone to right of figure, some original gold tesserae visible; these must be part of original gold background. Wave pattern of dalmatica purple and black tesserae. Tabula has one zone of blue glass tesserae, one-of white, light grey and light blue. Above this is larger, triangular field, close to pearl border, of light grey and white glass tesserae. The shoes. Mainly dark orange, with outlines of rustbrown. 126

THE MOSAICS OF JOHN VII (705-707 A.D.) Sources for the provenience and original state of the fragment The provenience of the fragment is documented by the following sources. A) The marble inscription placed below it states that the image of the Virgin comes from the Oratory of John VII and was brought to Florence from Rome in 1609 (). B) In Grimaldi’s own notary public record from February 22, 1609, he testifies that on that day the fragment was taken out of the walls of the Basilica of St. Peter to be sent to the church of S. Marco in Florence as a gift from Cardinal Paltotto to Antonio de Ricci, Archbishop of Arezzo (+). C) In Grimaldi's manuscripts, he, both in the general description of the Oratory as well as in his com‘mentary on the drawings of the mosaics, refers to the transport of the fragment to Florence and even quotes the marble inscription which still can be seen below it (see A) (5. D) From a letter (referred to by Garrucci) written September 11, 1609, by Archbishop de Ricci we learn that the possibility of displaying only half of the fragment on the altar was discussed (4. On very safe grounds Mintz and Garrucci identified the fragment with Maria Regina of the east wall mosaics of the Oratory (’). Its exact position can be seen from drawings of the mosaics in Grimaldi’s manuscripts (Pl. XVIII). Furthermore, the details of its original appearance can be studied in the pencil drawing from the ‘Album’ in the Vatican Library (PL. XIX 6). ‘Maria Regina stood on a low podium which was decorated with diamonds and pearls. To her right was the figure of Pope John VII (cf. cat. no. 2) who, on his veiled hands, offered her a model of the Oratory. Her pose was that of an orant, with both hands lifted, palms upward. The drawing further shows that before the restorations gave her her present baroque pondus, her body was supple, tall, almost columnlike. A close comparison between the drawing and the mosaic panel shows that the original figure is still preserved completely intact between the restored zones added to both sides. The original outlife of the body must have run exactly where the border between restored and unrestored areas is today (cf. text fig. 1). The restorations took place before 1750, for on a print in Richa's Notizie storiche delle chiese fiorentine, 1754-62, the extension of the lower part of the body is already visible (*). Probably the restorations took place when the fragment was set up over the altar in S. Marco in 1609. The break in surface below the waist, where there is a large restored zone, probably indicates the point where the fragment was cut in two before being transportedto Florence: hence the discussion whether only half of it should be used (see the sources above). Cat. no. 2. Pope John VII represented as the donor of the Oratory, fragment în Room VI, Vatican Grottoes, Rome. Measurements. Height of the fragment: 0.78 m, width: 0.57 m. Restorations (text fig. 2). Only brow is original, including eye to right. For detailed discussion of restoration of face see p. 149. Colour and material The head (Pls. V, VI a). In original areas flesh of finely grained marble tesserae, Wrinkles of brow greyish purple marble, Between dark wrinkle lines are highlight stripes of white marble tesserae. Eyebrow (1) Full text of the inscription: VETUSTA HAEC DEI GENITRICIS IMAGO IN VATICANA BASILICA SUPRA PORTAM SCAM ORATORIO OLIM A IOHANNE VII PONT. MAX. SAL. ANNO Decmi (sic) CONSTRUCTO DIU SERVATA ATQ AD HUNC DIM RELIGIOSISME CULTA, CUM IAM TEMPLUM ILLUD IN AUGUSTIOREM FORMAM REDIGENDUM DETURBARETUR, ® RUDERIBUS ERUTA EST ET NEQUA AUT SALTEM MINIMA IN EAM DEVOTIONIS IACTURA FIERIT IN HANC ARAM ROMA TRANSLATA MDCIX

(@) Grimaldi, MS Bibl. Vat, Arch. Bas. S. Pietro G. 13, fol. 124 fi. (6) For instance in Grimaldi, MS Florence, Bibl. Nar, ILILATS, fol. 17 and 102 (4) Garrucci, Storia, IV, p. 98. (5) Mintz, L'Oraioire, p. 158; Garrucci, Storia, IV, p. 98.

(6) G. Richa, Notisie storiche’ delle chiese Fio-

ventine, Florence 1754-62, VIII, illustration to D. 138. 127

PER JONAS NORDHAGEN

over eye to right brownblack and purple-black glass. Upper line of eye brown and purple glass, pupil black glass. White of eyeball rendered solely by white, unset plaster ground which is surrounded by circle of brown glass tesserae, To left of brow, between flesh and hair, runs row of transparent green glass tesserae, Hair, largely repaired, glass tesserae of black, dark sienna, dark purple. Tonsure (restored?), marble of grey and pink. Sourcesfor the provenience and original state

A) Grimaldi constantly refers to this fragment in his description of the Oratory of John VIL. He tells where it was located in the Grottoes after the destruction of the Oratory, and also includes a special drawing of it in many of his manuscripts (see below) (1). B) Torrigio describes the fragment in his Sacre Grotte Vaticane from 1635; at that time it was located in the chapel of Maria Praegnantium close to the circular corridor of the confessio, and had the following inscription on a tablet: IMAGO IOANNIS VII. P.M. EX SUO SACELLO HIC REPOSITA ANNO M.DCIX (ἢ. About 1924 the fragment was moved to the Museo Petriano in the Vatican 6), but after the rearrangement of the Groftoes, 1040-50, it was brought back. It now stands without any accompanying inscription,

Mintz and Garrucci identified the fragment as the upper part of the figure of Pope John VII from the donor panel in the east wall mosaics of the Oratory (ἢ. The original form of the figure and its place within the panel can be seen on the detailed pencil drawing from the ‘Album’ in the Vatican Library (Pl. XIX 5). John VIL, with the square halo, stood below the podium of the Maria Regina figure, on her right, and so close that he actually was below her outstretched right hand. He was turned to the right with his face frontal, and was dressed in a aemula worn over a long tunic, Part of the paenula covered his hands, which held a model of the Oratory, depicted as a small chapel with an arched door and a cross on the roof. On his shoulders could be seen the papal Pallium, which also emerged from under the model on his hands and hung below it. This lower part of the pallium was decorated with a cross. Except for the last item, all the details mentioned above have been quite correctly reproduced in the restored parts of the fragment. The artist who made the pencil drawing represented the Pope barefooted, but with loosely sketched sandal bands. The elongated ‘toes’ must be a misunderstood rendering of the embroideries on the sandal points (). The coloured drawing of the fragment, which Grimaldi included in some of his manuscripts, shows that in the restoration the colour of the dress is reproduced rather accurately and that the paonula was originally of a yellow brown corresponding to the colour of Pope John's paenula in the frescoes in S. Maria Antiqua (7. The square halo, however, which in the fragment has been restored as golden, must originally have been blue, as can be seen both from Grimaldi's drawing (PI. VI δ) and from the Pope's donor figure in S. Maria Antiqua. ‘A photograph of the fragment, taken at the beginning of this century, shows the head of the Pope as it looked in its original condition (PI. V). The face had aregular oval form, into which cut the curved hairline of the brow. Under the double wrinkle of the brow were the strongly accentuated, curved eyebrows, the lines of which continued down as the contours of the long, narrow and sharply pointed nose. Between the eyebrows one short wrinkle went up into the brow. The eyes had a line marking the upper edge of the eyeball, followed by the line of the eyelashes, and the pupil with its surrounding iris, and the huge cornea which gave the eyes their wide-open stare, From the wings of the nose sharp lines led to the edges of the mouth, which was dominated by a strongly down-curved line. Between the lower lip and the chin was a short, dark shade line, on the chin and along the lower part of the cheeks were dark stripes indicating the beard. The muscles of the throat were indicated by means of a broad U-shaped outline of dark glass tesserae in the marble flesh area. The ear to the right was still intact, and projected curiously from the head; the ear to the left was already destroyed by an earlier restoration. ‘The fragment must have been restored more than once in its history. ‘The gold ground, halo and body seem to have been restored rather early. Although, according to Cascioli, this restoration took place in the beginning of the eighteenth century (7, it seems to date from the middle of the nineteenth; judging from a

(2) Grimaldi, MS Florence, Bibl, Naz, II-III(4) Mintz, L'Oratoire, p. 159 Garrucci, Storia, 173, fols. x7, 97 and 102. The drawing is on fol. rro, IV, p. 98. (2) F.M. Torrigio, Le Sacre Grotte Vaticane, Rome (s) Cf. the sandals of John VII in the frescoes of

1635, p. 82. () G. Cascioli, Guida illustrata al Nuovo Museo di San Pietro, Rome 1924, p. 31.

S. Maria Antiqua; Wilpert, op. cit, TV, pl. 154 (left). (6) Ibid, same pl. (7) Cascioli, loc. cit

128

THE MOSAICS OF JOHN VII (705-707 A.D.)

10

20

30 Cm.

Fic. 2. - John VII as donor, cat. no. 2. Diagram of the restorations. print in Dionisio's Sacrarum Vaticanae Basilicae Cryptarum Monumenta, the fragment did not yet seem, in 1828, to resemble its present state (PI. XXII a), cf. the obvious disintegration of the church model and the Pope's hands. The restoration of the lower part of tlie head must be of a more recent date. It was probably made about 1940-50 in connection with the return of the fragment to the Grottoes (cf. the modern conception of a ‘stylistically correct’ restoration, p. 149). 129

PER JONAS NORDHAGEN Cat. no. 3. The Virgin, from the Nativity. Fragment, Duomo, Orte, Northern Lazio.

Measurements, Height of the fragment: 0.58 m, width: 0.48 m. Restorations (text fig. 3). Gold background and blue frame line along edges of fragment restored. Black outline inscribing figure mainly restoration.

ο

10

20

30Cm.

Fic. 3. - The Virgin, cat. no. 3. Diagram of the restorations.

*

General description (Pl. VII a). Fragment contains upper part of figure of Virgin, shown half turned toward right, her head deeply inclined, her hands crossed below breast, right hand on top of left. Head surrounded by golden nimbus. She wears blue falla, with small golden emblems on shoulders, breast and front of maphorion. Edge of dress around face has outline in gold, gold stripes also set along dark fold lines of maphorion. At wrist of right arm sleeve of purple tunic with manica of gold. Strongly marked eyebrows. Short vertical line leads from eye to eyebrow, perhaps to create expression of suffering (). Strong vertical, ‘drawn’ lines at either side of mouth probably serve same function, (1) Cf. the good reproduction of the fragment in Wilpert, op. cit, ΠῚ, pl. 114, 2. 130

THE MOSAICS OF JOHN VII (705-707 A.D.) Colour and material

The face. Face and throat set with marble tesserae, whose colour can scarcely be determined because of dirt accumulated on surface. Dark type of stone used to right of nose, on right (shaded) part of throat and high up on cheek to left. Outlines of face mostly black glass, as are line separating it from nimbus, contour of chin and cheek, lines around eye -- including short vertical lines going from eye to eyebrow ~ and shade under lips. Pupils black, surrounded by white, bare setting-bed which alone renders colour of eyeball. Lower contour line between face and throat green transparent glass, as is upper contour line along cheek and lower contour line of eye, Dark orange in: 1) highlight stripe on cheek; 2) tip of nose; 3) lips; upper lip of one large tessera, lower lip of two small tesserae. Line between lips brown with outer tesserae of row in black. Flesh areas of hands marble. Contours of hands are not set (cf. p. 147). The costume. Palla contours and fold lines black and dark ultramarine glass. Palla modelled with variety of glass in blue tones, ultramarine, cobalt and many values of blue grey. Plasticity of upper arm achieved by modelling from black of shaded folds through dark blue to lighter values in centre of arm. Black fold lines in maphorion edged with rows of gold tesserae. Row of gold tesserae encircles hem around face. Each of small golden emblems on palla made of four golden tesserae set together in form of cross. Tunic, seen at wrist, purple coloured glass (caput mortwum) interspersed with gold tesserae. Along contour of body towards hand runs uninterrupted row of gold. The nimbus. Original, rather small gold tesserae preserved. Outlined by inner circle of white coarse grained marble and two outer circles of dark azure-blue glass.

Sources for the provenience and the original state Inserted in the restored golden ground is a tablet of wood which bears the inscription: IMAGO DEIPARAE VIRG. EX SACELLO IOHANNIS VII SUPRA PORTAM SANCTAM VETERIS BASIL. S. PETRI ACCEPTA A.M.DC.IX. A. Bartoli, who first published the fragment, informs us that the fragment was, up to 1900, kept in the Benedictine convent at Orte (). Though no documents relating to the fragment could be found either in Orte or Rome, Bartoli suggests that it was brought to Orte by the Augustine monks, who held the monastery before the Benedictines. A total of at least three other fragments from the Oratory mosaics were given to the Augustines in Rome when St. Peter's was demolished (cf. cat. nos. ro, 11 and 12); and in the year 1609 which is mentioned in the inscription on the fragment, the bishop of Orte was Ippolito Fabiani, of the order of the Augustines (Bishop 1607-21). Possibly through him the fragment was donated to the monastery at Orte (). Bartoli showed that the fragment is part of the reclining figure of the Virgin from the Nativity scene in the east wall mosaics (PL. XX a), and he based his identification on the detailed pencil drawing in the ‘ Al‘bum’ of the Vatican Library. The drawing shows the Virgin reclining on her couch in the middle of a rocky landscape, her head, wearily inclined, turned in the direction of the Child’s washing. The restoration of the fragment probably took place soon after it had been taken out of the wall of the Basilica. The detached part was made into a self-contained icon, and to achieve this, part of the couch and the landscape background which might have been included in the fragment, were eliminated and substituted with gold. Further, the reclining aspect was changed by cutting away the part where her body swung towards the side (text fig. 3). Cat. no. 4. Salome and the Christ Child, from the Nativity. Fragment in Room IV, Vatican Grottoes, Rome. Measurements. Height of the fragment: 0.60 m, width: 0.54 m. Restorations (text fig. 4). Only small parts of fragment undergone complete restoration, δ. g. area around Salome's head. But throughout one finds extensive repairs. Original tesserae have been fastened to surface with new cement. Some of repairs discussed on p. 148. General description (PI. VII t). Salome and Child appear against strongly differentiated background, green on both sides of woman, yellow in small area to left of Child, deep red in upper left corner. Background is part of landscape from Nativity scene, with red area probably part of Virgin's couch. Child nude, stand(1) Bartoli, op. cit, p. 22.

(2) Ibi 13r

Ῥ. 23.

PER JONAS NORDHAGEN ing half submerged in golden basin. Body almost frontal, arms held towards right grasping arm of Salome. Head, turned to the left, surrounded by nimbus with cross. Between Him and left edge of fragment is seen mouth and part of neck of jug, from which water pours down into basin, Salome sits on rock to right of

Fic. 4. — The Washing of the Child, cat. no. 4. Diagram of the restorations, Child, is turned towards Him and, leaning slightly forward, holds Him with outstretched arms. She wears long yellow chifon, which leaves both arms bare from shoulders and is fastened with band under breast. ‘Two dark fold lines run down body from shoulder to lap, another fold leads from high above hip-down to lap. Head covered by bonnet which has vertical white and red stripes. In upper right comer, in a light field, appears a white oblong form (a foot? part of a sheep ?) with dark outlines, Colour and material The Child The head (PL VIII). Face no outline against gold of nimbus. Rest of head outlined with black and purple-black glass. Glass of same colours used for hair-line and facial features. Hair rendered by juxtaposition of green transparent, olive-green and light khaki glass tesserae. Marble tesserae for flesh in face and 132

THE MOSAICS OF JOHN VII (705-707 A.D.)

body. White, pink and light yellow marble in face, with row of dark greyish purple marble along chin and similar row inside contour to right of throat. Single dark orange tessera on tip of nose and one on point of chin. The body. Outlines of body dark grey, dark red and blue-black glass. Breast and stomach outlined in rust-brown glass, Modelling of flesh, done with marble of four colours, best studied in arm. Here, in successive rows above lower contour line (of black glass), are used: 1) dark greyish purple, 2) dark pink (two rows), 3) light pink, 4) white (one row) marking the highlight, 5) one row of pink. The basin. Rim gold with upper and lower outline of ultramarine and blue-black. Corpus set with vertical rows of glass tesserae of different colours, reading (from ultramarine outline) towards right: rustbrown, azure-blue, gold, ultramarine, gold, dark orange, ultramarine, gold — and ultramarine outline. Probably meant to imitate richly decorated vessel with gilded grooves. Water in basin rendered: by mixture of azureblue, sea-green and ice-green glass tesserae, Below upper, azure-blue contour of water is horizontal stripe of flesh-coloured marble, creating illusion that part of body is seen through water. The jug. Jug mouth marked by circle of white marble tesserae of coarse crystals. Within circle is an inner row of ultramarine around core of large azure-blue glass tesserae. Corpus of jug seems to haveundergone restorations, Water running down into basin outlined by ultramarine and dark azure-green stripes, and consists of greyish tones, all glass: olive-green, mauve, and lavender, with some tesserae of azure-blue. Salome

The head (Pl. IX). In headdress, single rows of white marble alternate with rows of red glass. Upper tesserae in each red stripe rust-brown, lower ones dark orange, Contour of headdress dark rust-brown at brow, dark purple along temples and neck. Flesh of face set with marble tesserae: around eyes, at nose and at mouth white marble used; on cheek, chin, brow pink marble, in periphery of face, along temples, ear and down along right side of throat yellow marble employed. To left on throat (lighted part) some rows of white marble. Outline of face dark rust-brown glass. Same colour for lower lip, lower horizontal contour of nose. Other lines in face black glass. Dark orange in small concentrations on tip of nose, both lips, chin and as row on check. The body (PI. ΨΙΙ δ). Arms set with tesserae of white and pink marble, contour rust-brown. Chifon light yellow marble tesserae of large size. Contours and fold lines green glass, mainly of transparent tesserae, but along some folds and as contour of back are rows of opaque glass, light green and olive-green. Band holding chifon together dark rust-brown glass. The background. Red field in upper left corner: oblique parallel lines (folds) black glass; red field brown, red-brown (umbre) and rust-red tesserae. Blue at head of Salome restored. Yellow area to left of Child's head large-sized yellow marble tesserae, of kind used in Salome's head. Green field to both sides of Salome's body in various tones of green glass: to left of Salome lighter and more yellow tones, sea-green, olive-green, green ochre and grass-green; to right general effect more khaki, with grass-green, emerald-green and sea-green plus some tesserae of more greyish olive-gréen. Rock on which Salome sits set with marble tesseTae, light grey with darker spots, its outline dark greyish purple marble. Field itself set with glass tesserae, light azure and green and some of white to light grey. Oblong object white and pearl-grey glass, with outline of dark blue glass to right, followed by line of dark red-brown. Sources for the provenience and the original state

The fragment was in the Museo Cristiano in the Lateran until 1924, when it was moved to the Museo Petriano in the Vatican (). From here it was finally brought to the Vatican Grottoes, probably ca. 1940-50. No sources establish its history between the destruction of St. Peter's and its incorporation in the Lateran Museum. However, Garrucci suggests that the fragment is one of two pieces of mosaic from the Oratory which, according to Torrigio's Sacre Grotte Vaticane (ed. 1635), were brought to the church-of S. Filippo Neri in Via Giulia in 1633 (). The church was rebuilt and restored about 1850, at the time when the Museo Cristiano was established by Pius IX (1854). Probably at that time the present fragment and the other fragment of the Oratory mosaic belongingto the Museo Cristiano (Cat. no. 6) were brought to the Museum. No ‘mosaics are preserved in the church of S, Filippo Neri today. Mintz was the first to show that the fragment is part of the scene in which the Child is washed by the two midwives, in the lower right comer of the Nativity panel (). This identification is confirmed by the (1) Cascioli, op. cit, p. 34. (2) Garrueci, Storia, IV, p. ror.

(3) Müntz, L'Oraloire, pp. 150 f. 133

PER JONAS NORDHAGEN

pencil drawing of the panel (PI. XX a), which shows the two women and the Child to the right below the reclining Virgin. While the woman to the right sat, holding the Child, the woman to the left stood, pouring water into the basin from a jug. The basin stood on a high, thin foot, like a chalice. The woman to the right, preserved in the fragment, was identified by Garrucci as the doubting Salome mentioned in apocryphical scriptures. She was originally seen twice in the same panel, as is proved by the drawing of a now lost fragment, cat. no. 9 (Pl. XXI D): wearing the identical red and white bonnet and Jong chiton, she knelt at the Crib, with her withered hand outstretched. In the present fragment the following sections of the Nativity scene are preserved (besides the two figures): 4) the jug held by the second midwife, B) the couch of the Virgin — the red drapery in the upper left corner, C) (possibly) a part of the Annuncia tion to the Shepherds — the white field in the upper right hand corner, D) the rocky landscape background — the green, yellow and grey areas. Cat. no. 5.

Fragment of the Adoration of the Magi, Sacristy of S. Maria in Cosmedin, Rome. Measurements. Height of the fragment: 1.00 m, width: 1.06 m. Restorations (text fig. s). Fragment in good state of preservation. Right part of angel's nimbus restored. Gold background seems largely reset. Goneral description (PI. X), Against gold background — from left to right — Joseph, enthroned Virgin with Child, angel and part of first of Magi. Joseph, standing behind throne, half covered by its back, is turned to right, his head slightly inclined. He wears white Pallium which goes high up behind his neck in fold and is seen in broad fold on right side of breast. Strong, vigorous face, broad neck, heavy jaw. Beard and hair articulated by short hatched lines. Virgin sits, turned to right, on green cushion on golden, jewel-encrusted throne. She wears purple tunic, and blue palla with small golden emblems on front of maphorion, on breast, on her right shoulder. Between "headdress and face part of veil can be seen. Head surrounded by golden nimbus. Face has rounded and rather large features. Her glance, directed towards advancing Magi, is serene. Upper part of body an oblong, stylized shape. Palla draped in horizontal or oblique folds over breast, short folds run across arm with part of palla hanging vertically from wrist. Tunic has series of densely set vertical folds at knees. On Virgin's lap sits Child. | Virgin holds (or steadies Him) with her right hand on His hip, her other hand holds His leg. Child is turned towards right, leans forward in eager movement, with His right hand extended towards box of gifts offered by King. Left hand, held under breast, clutches scroll. Child wears tunic, with red clavi, and pallium, both of gold; nimbus with cross encircles head. To right of Holy Family, angel stands in frontal position, head turned to right, towards Magi, With his right hand, liftedin front of his breast, he makes a gesture of introduction or indication. His left hand holds staff, which passes over shoulder and across unrestored rim of blue halo surrounding his head. Angel wears white pallium and tunic, both with greyish-red clavi. Tunic, with wide sleeves, falls in broad folds at wrist and from its wide opening at the neck a few short folds fall to breast. Pallium makes broad turn across body at waist, passes behind angel's right shoulder and come in broad vertical folds down left shoulder. Angels wings aro extended against gold background. Angel's head, turned to right, slightly inclined, has rich hair in locks held by white band. Face full features of classical regularity. Its vigorous turning movement brings out lines of muscles at throat, AIL that remains of first of Magi are two outstretched arms holding box of gifts toward Child. One hand is covered by orange red cloth with white arrow pattern, probably part of King's chlamys, Other hand not covered. Square box, drawn in perspective, contains small golden balls. ‘At upper edge of fragment, directly above Child, is star, consisting of golden half circle with outline, from which three broad ray emerge.

Colour and material

|

Joseph

The costume. Pallium white marble tesserae of coarse grain, interspersed with a few scattered cubes of white glass, Contours datk ultramarine and cobalt The head (PL XI). Face and throat marble: white, light yellow, pink and greyish brown: Lightest ‘used as highlights in row under eyes, at temples, on tip of nose and upper lip. Highlight on point of nose 134

THE MOSAICS OF JOHN VII (705-707 A.D.) white glass. Outlines of eyes, nose, mouth black glass. Lips dark orange, Cluster of four dark orange tesserae makes concentrated red spot on cheek. Between lips dark line of two brown glass tesserae flanked by two black. Contour encircling face -- from chin, past ear up into brow — green transparent glass, with occasional, turned gold cubes. Beard alternating rows grey-brown marble and green transparent glass, tesserae

Fio. 5. -- The Adoration, cat. no. 5. Diagram of the restorations. of latte: material forming dark short lines which articulate beard along jaw. Hair similar structure, but with stripes made either of green transparent or light purple glass. The Virgin The costume. Palla outlines and fold lines in black and blue-black glass, modelling of lighter parts ultramarine, cobalt, many values of blue-grey and lavender. Tunic, seen at wrist and on lower part of body, has contours and fold lines in black, dress colour purple and light greyish red, mostly of light values. Gold emblems of palla each consist of four gold tesserae. 135

PER JONAS NORDHAGEN

The nimbus. Outline against gold ground is single row of white, coarse-grained marble and double row of azure-blue glass. The head (PL XII). Veil between maphorion and face greyish brown marble with short stripes of transparent green and opaque red-brown glass. Line between veil and face not set, though space reserved for it in plaster (p. 148). Towards nimbus face has black contour, while lino dividing face and throat is of green transparent glass. As in other faces, lines of features black glass. Dark orange on point of nose, lips, cheek, chin. Line between lips rust-brown. Flesh pink and brownish marble, white marble tesserae highlights in comer of eye, high up on cheek, over eyebrows, along ridge of nose, on lips (text fig. 6). Series of rows of light yellow marble at left on throat; at right side concentration of darker marble colours: The hands. Flesh pink and yellow marble. Contours of right hand black glass, while left hand, emerging from behind Child, unfinished, outlines lacking (p. un. The throne. Alloutlines of throne ultramarine. Back and seat densely set with gold tesserae. Jewels on side posts and front of seat: blue square jewels have core of azure-blue glass surrounded by frame of gold tesserae; circular rubies have core of dark orange surrounded by ultramarine contour. Jewels set against background of green transparent glass, each being flanked by pearls of single white marble tesserae. Conic knobs crowning side posts set with white marble and blue glass tesserae, Side of cushion grass-green and emerald-green glass with pattern of widely set dark marble tesserae — dark greyish purple or greyish brown. Upper and lower part of cushion, divided from sides by golden ‘seams’, set with greyish blue and ultramarine tesserae, the latter marking dark fold lines The Child The costume. Tunicand pallium gold tesserae. All contours and fold lines opaque, yellow-green glass — yellow Khaki, olive-green, green ochre. Clavi on tunic vermilion-red. The nimbus. Gold, with cross arms and contour in white marble and azure-blue glass. The head. As in other heads, flesh marble, face lines black glass, while outline between face and ‘throat transparent green glass. Hair mixture of green transparent, yellow, yellow-beige glass and some gold tesserae, Hands and feet marble, lack contours although space reserved for these. Scroll black outline, within which runs one row of white marble, one of light lavender-blue glass. The angel The costume. Tunic and pallium white marble, tunic with faint greyish tone. In pallium, especially shoulder, rows of very light blue and grey glass tesserae inserted among marble ones, Contours mainly of dark ultramarine and cobalt, but with green transparent gloss for some fold lines, especially on angel's left shoulder and in vertical turn of Pallium across body. Clavi on tunic and pallium dark greyish purple marble. Baculum consists of one row of greyish purple marble and one of dark blue glass. The nimbus. Unrestored part to left of head lavender-blue glass in many different values. The head (PI. XII). Flesh of face white, pink, yellow and brownish marble, with highlight modelling as in other faces of fragment (text fig. 7). Throat light pink-brownish colour, plasticity suggested by white highlights and by outlining muscles in brown and purple. Hair has outlines of locks in black glass; lighter areas yellow-grey marble and green transparent glass, Fillet through hair one line brownish marble and one cream-colonred, The hands. In angel left hand only marble ‘ribs’ for fingers set, while contours are lacking. Right hand ultramarine contour along wrist and on lower side, but upper side lacks contour. This hand offers fine example of modelling with tharble colours (text fig. 8): flesh set with pink and light brown marble, white Fic, 6. -- Head of the Virgin in the Adoration fragment, Tracing showing the distribution of the white highlight cubes,

136

THE MOSAICS OF JOHN ΨΙΙ (705-707 A.D.) highlights of marble tesserae at wrist, on thumb, from here out to index finger, and finally in short row on back of hand. The wings. Contours dark ultramarine, some black tesserae, Upper narrow edges of wings mixture of dark green, ultramarine and gold tesserae, then dark blue and black contours. Upper row of long feathers has golden zone, then stripe of light green, and in feather tips dark greenish blue with ultramarine contour. Lower row: each feather made of one zone containing ice-blue, lavender-blue and grey glass between white marble rows.

The King. Tunic seen on arm ultramarine, dark green-blue, sea-green, ice-green glass, with lighter values concentrated along upper contour. gold of manica divided by rows of deep redStripes lacquer.of Bare hand marble, lacks contours. Lap of chlamys, covering second arm, dark orange tesserae with pattern of small arrow-shaped emblems of white marble; contours and folds of rust-brown glass. Sides of box of gifts dark blue outlines. Front set with horizontal rows of white marble alternating with rows of white opaque glass; right side set with rows of green transparent glass alternating with light purple. Box contains round ‘golden balls", each consisting of two to four gold tesserae within circular contour of rustbrown glass. The star. Circular contour and rays marble. Rays have streaks of green transparent glass.

Sources for the provenience and the original state A) The marble inscription underneath the fragment says that it comes from the Oratory of John VII in the Vatican Basilica, and that it was put up in the church in 1639 as a gift from Ant, Ghezzius, Canon of S. Maria in Cosmedin (1). B) In a letter dated September 2, 1636, Ant. Ghezzius writes about his donation of the fragment to the church and states that the mosaic comes from the Oratory of John VII at the Porta Sancta. This letter, pub. Fic. 7. — Head of the angel in the Adoration fraglished by Crescembini in his book on S. Maria in ment. Tracing showing the distribution of the white Cosmedin from 1715 (?), also contains the information highlight cubes, that after the demolition of the Oratory the fragment was kept for many years by the Chapter of St. Peter's before being given to Ghezzius, who had it put up over the door of S. Maria in Cosmedin (). The fragment must have been moved to the Sacristy from its original place over the door some time be tween 1860 and 1870, When Barbet de Jouy mentions it in 1857, it is still over the door (*), while Mintz

reports that it was to be found in the Sacristy (9).

Grimaldi gives no information about this fragment. However, from the drawings in Grimaldi's manuscripts, Mintz could deduct that it formed part of the Adoration panel on the east wall of the Oratory (1) Text of the inscription: URBANO vanpar. VETUSTISSIMAS HAS MUSIVAS IMAGINES IN ORATORIO 5. DEI GENITRICIS INTRA D. PETRI BASILICAM A IOANNE VII AD, ANNUM DCCV EXAEDEFICATO OLIM EXTANTES ET IN EIUSDEM BASILICAE SUB PAULO V AMPLIFICATIONE PIE SERVATAS HIC AD PERPETUUM REI SACRAE MONUMENTUM IO. ANT. GHEZZIUS ROM. HUIUS DIA-

CONIAE CANONICUS DONAVIT ET AFFIGENDAS CURAVIT AN. SAL, MDCXXXDX. (2) G. M. Crescembini, L'Istoria della basilica diatonale, collegiata parrocchiale di S. Maria in Cosmedin di Roma, Rome 1715, pp. 146 fi. (3) Loc. cit. (4) B. de Jouy, Mosaiques chrétiennes, Paris 1857, Poe (5) Mintz, L'Oraloire, p. 152. 137

PER JONAS NORDHAGEN (Pl. XVIII) (). The drawing shows that the first King was Imeeling or deeply bowing before the Child while presenting his gifts. The two other Magi were standing, the second with his head turned towards the third (9). Cat. no. 6.

Christ, from the Entry into Jerusalem, fragment in Room IV, Vatican Grottoes, Rome. Measurements. Height of fragment: 0.62 m, width: 0.54 m. Restorations (text fig. 9). Zones along right and left edge of gold background restored. Large restored zone in dress, at Christ's right elbow. Smaller restored areas here and there elsewhere in dress. Left eye of Christ restored, but apart from this only smaller repairs seen in head. General description (PL XIV). Frontal halfdength figure of Christ seen against gold background. Head, surrounded by golden nimbus with cross, half turned to right, slightly inclined. Right hand raised in front of breast in gesture of speech; left, held below right, clutches scroll. Long hair and beard. Purple-red tunic and purple-blue allium, both with golden clavi. Colour and material

The costume. Tunic, on Christ's right shoulder and breast, caput morkwum glass of different values. Folds black, some blue-black tesserae. Clavi gold tesserae. Pallium, draped in vertical folds on Christ's left shoulder and in horizontal turn acrossthe body, purple-blue colour made up of following components: dark folds black, dark ultramarine and dark purple glass; lighter stripes light purple and lavender-blue. Clavi gold. ‘The nimbus. Disc and cross arms gold, white outlines of both nimbus and cross arms glass, with ‘io, 8. - Right hand of the angel in the Adoration outlines azure-blue glass. fragment. Tracing showing the distribution of the The head (Pl. XV). Flesh marble of three to white highlight cubes. four tones, from white through pink to brown. In face and throat, highlights, now somewhat faded, model features: white tessera rows under eye, along ridge of nose, on chin, in white lines along U-shaped, opaque greyish green shade line which describes anatomy of ‘throat.’ On both cheeks short lines of dark orange tesserae, same colour on tip of nose (two extremely small Cubes) and lips. Lines of nose and eyes black or purple-black glass,as is contour against nimbus. Long locks of hair set with alternating rows of black and purple glass; among the black, opaque tesserae some dark transparent green. Between area of hair and beard and face runs continuous outline of light green transparent glass. Beard same colour as hair, but greater number of transparent tesserae. The hands. Contours black glass, flesh light marble, The scroll. Dark rust-red tesserae outline lighter field of one row white marble, two rows light bluegreen glass. Sources for the provenience and the original state

The fragment has had the same fate as cat. no. 4, see p. 133. ΤῈ was probably kept in the church of S. Filippo Neri in Via Giulia until about 1850 (*), and then brought to the Museo Cristiano of the Lateran Museum, where it remained until 1924. That year it was movedto the Museo Petriano in the Vatican (9), and, finally, after World War II, to its present location in the Grottoes. (ἡ Loc. cit. (a) Torrigio gives the colour of the faces as white, atutti tro carnagione bianca», op. cit, p. 12.

(3) Ct. note 2, p. 133. (4) Cascioli, op. cit, p. 34.

138

THE MOSAICS OF JOHN VII (705-707 A.D.) Grimaldi does not once mention this fragment. Müntz, who was the first to study it, believed it to be a fragment of the figure of Christ from the scene with the Healing of the Hemoroissa and the Two Blind Men (1). Garrucci, however, by referring to the detailed pencil drawings of John VII's mosaics in the Vatican ‘Album’, could show that the fragment came from the scene with the Entry into Jerusalem (Pl. XX δ) ().

o

10

20

30 Cm.

Fic. 9. ~ Christ, cat. no. 6. Diagram of the restorations. Here Christ, head inclined, with same gestures as in the fragment, sits sideways on the ass which moves to the right, towards the crowd outside Jerusalem, ‘The original gold areas which surround Christ on the fragment prove that the mosaics of John VII had a gold background. The hatching of.the background seen on the drawing of the Entry panel, as well as on the pencil drawings of other panels, is therefore probably meant to indicate this gold ground. (1) Mantz, L'Oratoire, p. 156.

(2) Garrucci, Note, p. 339. 139

PER JONAS NORDHAGEN

"The fragment seems to have been restored in two different periods. The outer zones of the gold ground were probably patched at an early date, while the restoration of the left eye must have been done more re cently. In a photograph from: about 1900 this latter change does not yet appear (1). Cat. no. 7. The Virgin and Longinus from the Crucifixion, fragment kept at the Scuola del Musaico, Vatican, Rome.

Measurements, Height of fragment: 0.58 m, width: 0.52 m. Restorations. Only face of Longinus is original. Colour and material (PL. XVII). Face of Longinus, upturned, in profile to right, pink and white marble. Eyebrow, nose contours black glass. Along lower contour of point of nose two dark orange glass tesserae. Short row of same colour on check running down from eye. One large tessera of orange as upper lip, black line of glass for shade between lips, two dark orange tesserae as lower lip. Shadow under lips black glass. Black line, evidently part of beard runs from edge of mouth to jaw. Sources for the provenience and the original state

In his Sacre Grotte Vaticane, Torrigio states that the present fragment, together with other relics from the Basilica, was located in the circular corridor of the confessio, and that it bore the following inscription: HAEC B VIRGINIS MUSIVA IMAGO ERAT AD SACELLUM IOANNIS PAPAE VÍI (". The inscription was stil in sifu above the fragment up to this century, as old photographs show (). In 1924 the fragment was moved to the Museo Petriano (0), and from there it was broughtto the Scuola del Musaico where it is still kept. Grimaldi makes no mention of the fragment. However, from the drawings of the cast wall mosaics in his manuscripts, Mintz identified the fragment as a part of the Crucifixion scene in the lower right panel (Pl. XVII) (). Here the Virgin and Longinus stand to the left of the Cross. The Virgin lifts both hands, which are veiled in her palla, before her face in sorrow. Longinus, head lifted, turned in profile towards the Cross, pierces Christ's flank with his lance. ‘The appearance of the fragment immediately after it had been taken from the Oratory walls can be studied in a watercolour from about 1630 (PL. XXI a) (). Of particular interest is the figure of Longinus, which was later completely transformed by the restorations. As seen here, he was originally bareheaded, with long hair closely curled at the neck, and had a pointed beard. He wore a long-sleeved tunic, gathered by a band at the waist. "The first thorough restoration of the fragment took place some time before 1840. In a print published in Dionisio's 1828 work of the Grottoes, the figure of Mary is still clearly visible, although there is no trace of Longinus (PL ΧΧῚ ὁ) (1, while in Sarti and Settele's Appendix, edited in 1840, Longinus appears completely restored in the form he has today (PL. XILd) (). ‘A photograph of the fragment at this stage, taken ca. 1900, shows that in spite of the heavy repairs it had undergone, the upper part of the body and the head of the Virgin were still completely intact (Pl. XVI c) The original part included the head with its nimbus (minus the outer contour of the latter), the shoulder and the higher part of the veiled lifted hands (text fig. ro). An enlargement of the photograph makes it possible to give the following description of these parts, which disappeared after the last restoration (PI. XVI a). The Virgin, facing right, raised head upward (towards the Cross). The nimbus. Original gold tesserae, inner contour of white marble still intact, () O. Marucchi, I monumenti del Museo Lateranonse, Milan 1910, pl. 7. (2) Tortigio, op. cit, p. 39. (3) Phot. Anderson, No. 20319. (4) Cascioli, op. cit, p. 84. (5) Mintz, L'Oratoire, p. 157. ‘6) MS Bibl. Vat.. Cod. Barb. lat. 4s.

fl. 11.

(7) T. L. Dionisio, Sacrarum Vaticanae Basilicae Crypiarum Monumenta, ed. 1828, Rome, pl. LXXV, mo. r. (8) A. Sarti and G. Settele, Ad Philipii Laurentii Dionysii opus de Vaticanis cryplis Appendix, eto, Rome 1840, pl. XIV.

140

THE MOSAICS OF JOHN VII (705-707 AD.) The head. Flesh very small cubes evidently of marble, outline towards nimbus in dark glass. Contour between face and throat, and line along right side of latter seem to have been transparent glass. Along outline to right side of face shade line in dark marble or glass. Pupil of eye surrounded by only very few tesserae, white of eyeball mainly white plaster ground. To side of mouth, on right check, line of dark tesserae, probably usual dark orange colour. Face extremely expressive, with slanting eyebrows indicating sorrow.

o 10 20 30 Cm. LLL cL Fic. 10. — The Virgin and Longinus, cat. no. 7. Diagram showing the condition of the fragment before its last restoration.

The costume. Headdress, maphorion, same modelling of folds as reclining Mary from Nativity (cat. no. 3). Form of the uplifted hands seen outlined through folds of paila covering them. The last restoration, during which the original part of the Virgin's head and body was cut away and partly reset before the work was abandoned, must have been madé ca. 1940-50, before the mosaic fragments were transferred to the Grottoes. Cat. no. 8. Sk. Peor, from Peter preaching lo the Romans. Fragment in Room IV, Vatican Grottoes, Rome (PL, XVI). Measurements : Height of fragment: 0.57 m, width: 0.68 m. Restorations. Fragment completely restored. In resetting arm, tesserae of pink and brown marble, probably from original, utilized. τάχ

PER JONAS NORDHAGEN

Sourcis for the provenience and the original state A) Until this century, the fragment in its original place in the Grottoes bore the following inscription: IMAGO B. PETRI PREDICANTIS ROMANI EX SACELLO IOANNIS VII PAPAE PAULUS V PONT. MAX. (). B) The fragment is mentioned many times in Grimaldi's manuscripts, each time with the additional τος ‘mark that it had been taken from the scene where the Apostle is preaching to the Romans ('). C) Torrigio mentions the presence of the fragment in the Grottoes in 1635 and gives the text of the inscription (see A); at that time the figure of Peter still had his keys, and on the background were the letters RO MA 6). ‘The fragment was kept in the Grottoes until 1924, when it was moved to the Museo Petriano, It was returned to the Grottoes after their rearrangement in 1940-50. ‘The main restoration mast have been made before 1877. The keys are still seen in the print of the fragment in Dionisio's work from 1828 (PI. XXII a), but the text testifies to the utter decay of the mosaic surface (). When Mintz describes the fragment in 1877, the keys are not mentioned, and only a small area (the arm) seemed to Mintz to be original (). Further, from Garrucci's 1877 drawing one can also see that by that date the fragment had acquired its present, heavily restored appearance (). Mintz identified the fragment as a part of the scene where St. Peter preaches to the Romans, in the second zone of the north wall mosaic cycle (7. Grimaldi's drawing (Pl. XIX a) shows Peter standing with his right hand lifted in a gesture of speechand his left hand holding the keys. To both sides of him are groups of kneeling people, and on either side of his head, the letters RO MA appear. Lost FRAGMENTS Cat. no. 9. Salome at the Crib, from the Nativity, Fragment formerly in the Vatican Grottoes, Rome.

‘The fragment is known through a watercolour drawing from about 1630, preserved as folio 12 in the Cod. Barb. lat. 4410 in the Vatican Library (PL. XXI δ). It was first published by Garrucci (). The leaf, which measures 0.26 X 0.22 m, bears the following text: Ex ruinis antiquissimi Sacelli Joannis Papas VII in Veteri Basilica MDCVIII in sacris cryptis Vat. Description. Salome, terned profile right, kneels to left of high rectangular Crib, one arm — that with withered hand — supported by other, outstretched before her. She wears same type of dress as in fragment of Washing of Child (Cat. no. 4): bonnet with red and white stripes and long sleeveless chiton. In the watercolour, chiton rose colour, yellow-brown belt. Arms faintly shaded in light brown along underside. Out nes of figure rendered as rows of black dots. On Crib — dark grey sides and red top — lies Child. Swaddling cloths white, face pink. His nimbus red colour imitating gold, with inner contour white, outer blue. Over edge of Crib, ox and ass look at Child, ox light yellow-brown, ass dark grey. Above Crib is star, formed by greyish blue disc with stripes of gilding, three broad rays emerge from it. Background yellow, with red outlining cubes, imitating golden ground. Other sources concerning the fragment A) Grimaldi reports that a fragment from the Nativity scene on the east wall, representing « the miracle where a woman holds her withered hand towards the Crib», was brought to the Vatican Grottoes (). B) Torrigio describes the fragment among the other precious remains in the Grottoes and states that it was accompanied by the inscription: EX RUINIS ANTIQUISSIMI SACELLI IOHANNIS VII. IN VETERI BASILICA MDCIX (»).

(1) (2) (3) (4) (5)

Torrigio has copied this inscription, op.cit., p.82. (6) MS. Florence, Bibl. Nar, II-III-173, fol. roo. (7) Torrigio, op. cit., p. 82. (8) Dionisio, op.cit.,p. 45. Οἷς pl. XVIII, 3 forprint. — (9) Mantz, L'Oratoire, p. 149. (ro)

14

Garrucci, Storia, IV, p. 282, 2. Mintz, L'Oratoire, p. 149. Garrucei, Storia, IV, p. 99 and pl. 280, 3. MS Florence, Bibl. Naz, 111-173, fol. 17. Torrigio, op. cit, p. 57.

THE MOSAICS OF JOHN VII (705-707 A.D.) ‘The fragment must have disappeared already in the seventeenth century, for it is not mentioned in the otherwise very thorough itinerary of Ciampini's De sacris Aedeficiis from 1693, nor in later descriptions of the Grottoes. Garrucci, who was the first to utilize the watercolour drawing for research on the Oratory mosaics, identified its subject as the small scene of Salome at the Crib, which belongs immediately to the left of the reclining Virgin in the Nativity (1); the scene is reproduced correctly in the pen drawings of Grimaldi’s manuscripts (Pl. XVIII), but strangely misinterpreted in the pencil drawing from the ‘Album’ (PI. XX a) (+). A comparison between the watercolour drawing and the existing fragment of the Washing of the Child (PI. VII δ) demonstrates the high quality and reliability of the drawing, cf. the rendering of such details as Salome's chiton, bonnet and curved back. Cat. no. 10. The Raising of Lazarus, fragment formerly in the Bibliotheca Angelica, Rome.

Grimaldi writes that this scene, which, in the east wall mosaics, was located in the left comer of the panel containing the Entry into Jerusalem (PI. XX 0), was taken out as a fragment and kept at the Bibliotheca Angelico () Cat. no. 11.

The Last Supper, upper, fragment fragment formerly in the Bibliotheca Angelica, Rome. This fragment, which was located in the upper right corner of the panel with the Entry into Jerusalem (PL XX δ), Grimaldi writes was kept in the ‘triclinium’ of the Bibliotheca Angelica (). Griineisen writes about the fragments which were brought to the Bibliotheca Angelica: ‘Our own and Müntz' search, both in the old monastery of St. Augustine in Rome and in the Bibliotheca Angelica, for these valuable monuments unfortunately has shown no results’ () Cat. no. 12. Christ on the Cross, fragment formerly in the Bibliotheca Angelica, Rome. In Angelo Rocca's book on the Sacred Cross, from 1609, the fragment is said to be in the Bibliotheca Angelica (ἡ and is reproduced in a woodcut (Pl. XXI e). ‘The woodcut shows the crucified Christ, with His body covered by a long colobium and His head, surrounded by the nimbus with cross, inclined to the left. Christ is shown as living, with open eyes. Above the head of Christ is a fabula with the inscription INRI. This inscription is also mentioned in Rocca's text, where he further gives the information that the colour of the colobium was purple [. Fmacwxwr Nor FROM Joux VII's Mosarcs Attached to the wall at the entrance of the circular confessio corridor in the Vatican Grottoes is a mosaic fragment which has caused some confusion, since earlier authors included it among the remains of John VII's Oratory mosaics. The fragment (Pl. XXII δ), which is extensively restored, measures 0.75 m in height, 1.18 m in width and represents the enthroned Virgin with the Child between two standing figures. It bears the inscription: VETUSTISSIMA HAEC B. VIRGINIS MUSIVA IMAGO ERAT AD CIBORIUM IOANNIS PAPA VII; HIC ANNO 1631 EXPOSITA. Müntz believed that the fragment was identical with the small mosaic which, according to Grimaldi, existed below the mosaic with the Christ cycle on the east wall of the Oratory, and which contained a Ma-

(1) C£, note 8, p. 142.

(5) de Grüneisen, op. cit, p. 295, mote 1

(2) Cf. the authors forthcoming study on Gri(6) A. Rocca, De particula ex pretioso et vivificatio maldís drawings of the Oratory and its mosaics. Hgmo sacratissimae crucis, etc, Rome 1609, p. 43. (3) MS Florence, Bibl. Naz, lLIIL173, fol. τοῦ. ὀ(7) Ibid. loc, cit. (4) Ibid., loc. cit. 143

hi

TECHNIQUE (Observations based on the original, unrestored parts of the mosaic fragments). 1) The plaster

The type of plaster used in the mosaics of John VII is called stucco romano by modern mosaic makers (), It consists of a mixture of lime and powdered marble, extremely fine in consistency, and yellowish white in colour. It has a smooth surface and can be scratched with a fingernail (?). 2) The tesserae i) Material ‘Two different types of material are used for tesserae in the mosaics of John VII, marble and glass (smalto).

Marble with coarse crystals is used in white garments such as pallia, in background details, in the outlines of the halos, in ornaments and jewellery such as pearls, and in patterns on fabric. Marble with fine crystals is used predominantly in the flesh areas of the figures.

Occasion-

ally, marble of this type is also used for certain dress colours and in parts of the background. -

‘The stone is firmly bound by the plaster, so that very few of these tesserae have fallen out.

But in the course of time the material has undergone other changes: it has to some extent lost its colours, This process of fading has gone very far in the Maria Regina fragment (Cat. no. 1), where the face is now a general grey colour, and where only by close examination one can distinguish the differences in surface structure which indicate that originally there was modelling with many different colours. In the other fragments the colours of the stone are better preserved. Glass is the material most extensively used in the mosaics, where it is found in most of the costumes and in the background.

It contains few bubble-holes and has an even, shiny surface.

The type of glass often occurring in the Ravenna mosaics, with a dull, scabrous surface, is not found in the John VIL mosaics. The distribution of colour pigment in the glass is uniform, with a tendency to stripes only in the tesserae of a certain dark rust colour, where the mixture of red and black is uneven. In the gold tesserae, the gold leaf is attached to greenish, transparent glass, ‘The leaves have darkened and have fine cracks in the surface. Between one fourth and one fifth of the gold tesserae have lost their leaves. (x) Information from the Scuola del Mosaico, Vatican City. (2) This plaster is distinguishable from that of the

restorations, which is often of a granular texture and has a brown or greyish colour.

145

PER JONAS NORDHAGEN

ii) The colours The following colours are used in the Oratory mosaics: Marble:

White, coarse crystals White, fine crystals

Grey with darker spots, fine crystals Greyish purple, dark, fine crystals Pink » > Yellow-ochre » > Cream » > Glass, opaque: Black White : 1) Pure white, 2) Pearl grey Blue : 1) Blue-black of indeterminable tone, 2) Ultramarine, one dark and one light value, 3) Cobalt, 4) Greyish blue, different values, 5) Azure-blue, different values, 6) Lavender, different light values. Green : x) Grass-green, many values from light to dark, 2) Green-ochre, 3) Olive-green, with one variety of tone towards yellow, 4) Sea-green, different light values, 5) Applegreen, 6) Emerald-green, different light values, 7) Green khaki, 8) Azure-green, 9) Icegreen, light. Red : 1) Vermilion, 2) Rust-red, 3) Orange, dark, 4) lacquer-red. Brown: 1) Burnt Sienna, 2) Burnt umber, Purple: 1) Caput mortuum. in values from light greyish to nearly black. Glass, transparent :

Light bottle-green with values towards colourless. Glass, gold: , The gold is of three types: one red, one lighter orange, one light lime-green, iii) Size of the tesserae

The following measurements are taken from the fragment of the Adoration of the Magi, S. Maria in Cosmedin, Rome (Cat. no. s). Pl. X. Tesserae in the hair and flesh areas of all figures (both glass and marble): Normal size from

3 x 3 mm. to 5 x 5 mm.

The dark orange tesserae down to 2 x 2,5 mm. Tunic and palla of the

Virgin: 7 x 7 mm. to 8 x 10 mm. The dark contour lines of the dress are of somewhat larger tesserae, from 7 x 8 to 10 x 12 mm. Golden pallinm of the Christ Child : From 6 x 6 to 8 x 8 mm. White pallia of Joseph and the angel (marble tesserae): From 7 x 7 to 10 x 15 mm. The throne and its cushion : As in the Virgin’s garments, Dress of the first Magus: As in the Virgin's garments. Gold tesserae (in the nimbus of the Virgin): Irregular sizes, from 4 x 5 mm to 10 x ro mm. In the representations of pearls some very large tesserae are used. In the Maria Regina fragment in Florence (Cat. no. 1), the marble tesserae depicting pearls in the diadem and on the

dress measure up to 15 x 20 mm.

146

THE MOSAICS OF JOHN VII (705-707 AD.)

Characteristic for the mosaics of John VII is the great difference in size between the tesserae of the flesh areas and those used in the rendering of garments and other objects. The tesserae of the first category measure, as the list above shows, from a quarter to half the size of those of the second. In this particular aspect the mosaics of John VII differ from the other early Christian and early medieval mosaics in Italy which display greater uniformity in the size of tesserae (see below pp. 155 f.). Another special feature of the John VII mosaics is the use of particularly large tesserae for the contours and the dark modelling lines of the garments. iv) The setting of the tesserae ‘The tesserae are generally pressed so deeply into the setting bed that the plaster often lies flush with the mosaic surface. The tesserae lie more deeply embedded here than they do in other Italian mosaics of the early medieval period. Another technical peculiarity is that the tesserae are placed with very large interstices, so that each tessera stands nearly isolated in the white plaster. In extreme cases these intervals measure up to about 5 mm. The plaster itself is often used as a positive element to represent the white cornea of the eyes, supplemented by only a limited number of white tesserae (1). This sparseness of tesserae seems to be a common trait in mosaic art of thè seventh-eighth century (p. 155).

3. The artist at work Some indication of how the artist made the mosaics can be gathered from cases where the work was not completed as intended. Unfinished details occur in many of the fragments, above all in the hands and faces. The faults may be due to ordinary negligence or to the fact that the plaster hardened before the work could be finished.

Among the orbiculi on the shoulders of Maria Regina, only the one to the right is fully

completed, and shows an outer band of pearls with each pearl, between borders of dark ultramarine tesserae, made up of one central tessera surrounded by a circle of azure-blue (Pl. I). ‘The pearl band on the other shoulder is incomplete, the central white piece of each pearl stands surrounded by an area of unset plaster ~ reserved for the azure-blue tesserae which were never put in, Evidently, in the rendering of pearls, the outer border and the central piece were placed first, and the circle of azure-blue tesserae was then filled in as the final part of the operation. Outlines of hands and fingers are lacking in the fragment of the Virgin at Orte (Cat. no. 3,

pl. VII a) and in the Adoration fragment in S. Maria in Cosmedin (Cat. no. 5, pl. X). In the

latter, the outlines have been omitted in the angel's left hand (holding the staff), in the left hand of Mary, in the Child's hands and feet, and finally in the hand of the King holding the box with gifts. In all of these examples only the marble tesserae of the flesh have been set, forming thin rows which are isolated from each other by the broad zones of plaster that were reserved for the dark glass tesserae of the finger outlines. We may therefore assume that the flesh-coloured tesserae of the hands and feet were set first, and after them the tesserae of the outlines (*). In order to work with the sureness this procedure indicates, the musearii must have had the whole com-

(1) In the mosaic of the emperor Alexander (886- — (2) An example of missing hand contours is found 913) in Hagia Sophia at Istanbul, the iris of the eyes in the mosaic above the central door of the narthex, is left unset, with the colour rendered solely by the Hagia Sophia, in both of the Emperor's hands; cf. painted plaster;cf. Underwood and Hawkins, op. cit., Whittemore, op. cit., I, pl. XIV. On contours left P. 194 and pl. 7. ‘unset on purpose, cf. Underwood and Hawkins, p. τοῦ.

147

THE MOSAICS OF JOHN VII (705-707 A.D.) to fall out. Probably in the seventeenth century, the-lost tesserae were replaced by new ones,

also of marble, but of a darker colour and of somewhat larger dimensions (PI. ILI). Thus, rows of dark tesserae run in different directions through the face and throat areas of the fragment. These rows look like strong lines at a distance (PI. I), and have been mistaken for the borders of restored zones (). Complete restorations of older date, affecting large areas, are generally not very difficult to detect, since in most cases they are made with little thought of imitating the original parts. Typical examples are the whole figure of Peter (Cat. no. 8, pl. XVI 5) and the body and dress of John VII (Cat. no. 2, pl. VI a), where a close setting of the tesserae, completely different from the untouched areas, is found. The restorations are harder to point out when they are deliberately made to blend with the original style and technique of the mosaic. In the hands of Maria Regina the flesh areas are set with small marble tesserae like those in the untouched parts, and the cubes are placed in the plaster at changing inclinations (PI. I). However, the tesserae protrude high above the setting bed, while in the true technique of the John VII mosaics, even the smallest tesserae of the flesh areas are deeply embedded. The hands were restored, possibly with cubes from the original mosaic, in the seventeenth century (together with the cracks in the face of Maria Regina) when the fragment was installed over the altar in S. Marco. It is a very early example of style imitation in mosaic restoration. A more recent example of the same kind of restoration is found in Pope John's face (Pl. VI a). The face at first glance seems original, while the body, as stated above, immediately betrays its later date, The brow shows typical original traits: the tesserae set irregularly and marble cubes used in the flesh. The lower part of the face, however, in spite of the deep-set tesserae and their sparse distribution, is a restoration; the material in the flesh zones here is not marble, but glass, as shown by the bubble holes. The work nevertheless displays an excellent knowledge of the ancient mosaic technique and must have been done recently, probably as late as about 1940-50 (p. 129).

5. The use of colours A. Special effects

i) Transparent glass In nearly all the fragments of John VII's mosaics, transparent glass is used in certain outlines of the flesh areas, especially in faces. In Maria Regina (Cat. no. 1), glass of this type is employed for the shade line which rans immediately under the eyebrow on the right side of the nose, and which continues all the way down to the tip of the nose (PI. III). Furthermore, tesserae of the same material constitute the lower row in the double line between the right side of the face and the throat, Finally, a horizontal row of transparent tesserae divides the flesh area of the throat from the ultramarine outline of the collar. In the original part of the Pope's portrait (Cat. no. 2), transparent tesserae run in a row between the flesh area of the face and the darker hair area, following the undulating line of the brow (PI. V). In the Madonna at Orte (Cat. no. 3), rows of the same kind of tesserae are found both above and below the black outline of the chin (PI. VII a). The Adoration fragment (Cat. no. 5) has transparent glass used in the outlines of chins in all faces (Pls, XI-XII). In the head of Christ from the Entry into Jerusalem (Cat. no. 6) a similar line goes from the brow down (1) Kitzinger, loc. cit.

149

PER JONAS NORDHAGEN

to the mouth, dividing the flesh area from the dark glass areas of the hair and beard (PI. XV). A few transparent tesserae are also found in the vertical lines flanking the mouth.

In all these instances the transparent glass is apparently used to create neutral shadows which

will harmonize with the subtle modelling of the marble colours in the flesh parts. It also creates soft transitions between the weaker pigments of the natural stone and the much stronger ones of the coloured glass. In some examples, tesserae of gold glass, turned with the gold leaf in and the transparent side out, are used among the purely transparent ones (1). In the Washing of the Child (Cat. no. 4) transparent glass traces the outline of Salome's dress, here again in connection with natural stone areas: her chilon is set with tesserae of light yellow marble, with contours and folds in light green, transparent glass (PI. VII 5). In the angel's Pallium. of white marble tesserae (Cat. no. 5), some of the contours are also of this type of glass. (On transparent glass used in the rendering of hair, see below p. 153). ii) Dark orange glass Dark orange glass has a special function in the faces of the mosaics, In all the fragments

it is used to render the colour of the lips, but it is also applied to other parts of the physi-

ognomy, as colour spots on the chin, on the point of the nose and on the cheek. Most commonly the orange glass is applied in clusters of extremely small tesserae. In some cases the cubes are arranged in short rows from the corner of the eye out to the cheek (Pls. I, XV). The function of these dark orange spots or rows must have been that of concentrated warm shadows, meant to accent the relief in the faces.

iii) Gold Gold is used for the background (cf. the unrestored parts of cat. no. 6, p. 138), for halos (the only exception is the blue halo of the angel in the Adoration fragment, cat. no. 5), and, finally, in garments, jewellery and objects of different kinds. In this last, decorative function, gold is used lavishly for the seat and back of the {throne in the Adoration fragment (Colour plate), and is the dominating colour of the tub in the Washing of the Child scene (Cat. no. 4). ΤῈ is furthermore employed in large quantity in the imperial dress of Maria Regina (Cat. no. 1), as setting or frame for the jewels and pearls and as pendants for the large ornamenta of the collar. Goldis also used in the smaller emblems on the shoulders, breast and head of the Virgin's palla in the other fragments. In one case, in the Orte fragment (Cat. no. 3), gold even lines the folds of the Virgin's maphorion. Small gold emblems also decorate the Magus’ garment in the Adoration, Gold is the material for the clavi on Christ’s tunic in the Entry into Jerusalem (Cat. no. 6), and it is also the dominating colour of the Christ Child’s pallium in the Adoration. In the last instance (as in similar representations in other mosaics) gold is used together with tesserae of light yellow-green opaque glass, a material of a colour intensity close to that of the gold itself. The yellowish tesserae outline the folds and the contours of the pallium, which still retains the effects of a pure golden fabric (*). Gold tesserae are used

(2) Barlier examples of the use of transparent {glass in the modelling of faces: a) The apostles in the cupola mosaic of the Orthodox Baptistery, Ravenna (fifth century). δ) The prophets-in the second decoration zone of S. Apollinare Nuovo; Ravenna (early

sixth century), In both cases the transparent cubes seem to be used to create neutral shades. (2) A modelling with yellow-green in golden robes is seen in the frabea costumes of the Holy Virgins in S. Apollinare Nuovo; Ravenna (middle sixth cen150

THE MOSAICS OF JOHN VII (γ05:707 A.D.) also for the contents of the King’s box of gifts, and they are mixed with coloured tesserae of many different hues for the angel’s wings in the Adoration fragment (Colour pl). ‘Nowhere in the preserved fragments do we find silver, which is common in East Christian mosaics and is also found in Italian mosaics from the sixth century onwards (1). In spite of the scarcity of the fragment material, there is reliable evidence to the effect that silver was not employed in the John VII mosaics: if this material had been used, it would inevitably have been preserved in the fragments with representations of Christ (Cat. nos. 4, 5 and 7), since silver was used in Christ's halo more than in any other connection, either in the disc or in the cross (?). In the preserved figures of Christ from the mosaics of John VII, gold is used in the halo disc and white glass in the cross. B) Other coloured tesserae and their use

i) Background Part of a landscape background is preserved in the Washing of the Child (Cat. no. 4). Here some components of the Nativity scene's rocky slope can be studied (Pl. VILI): 1) An area set with yellow marble tesserae to the left of the Child's head. 2) The rock on which

Salome is

sitting, made of tesserae of gray, spotted marble, 3) Green zones to both sides of Salome (the

blue around her head is a restoration), set with opaque glass tesserae in different hues of green, x) and 2) are in one colour without any shading, 3) has a wide variety of different tones and values (p. 133). Unfortunately, it is not possible to know what function these parts held in the picture, whether they represented the sides or the top of the cliff etc. ii) Water The Washing of the Child (Cat. no, 4) contains the only preserved representation’ of water

from the Oratory mosaics. Water is depicted in the tub in which the Child stands and is also shown running in a trickle from the jug at the far left (PI. VIT 3). The water is rendered by a chaotic juxtaposition of all kinds of blue and green glass tesserae, mainly of light values (p. 133). There is no articulation with wave lines. mixture of colours.

The shimmering

tury), and in one of the donor figures in St. Demetrius, Thessaloniki (middle seventh century), cf. Wilpert, op. cit, III, pl. 78, 2, and Grabar, Peinture byz., op. cit, illustration p. 50. In the Orthodox Baptistery in Ravenna (fifth century) yellow-green glass is used next to gold in the columns which divide the ‘thrones and the altars in the lower zone of the cupola, with an effect as if gold is used throughout the columns, cf. Wilpert, III, pl. 81-82. In S. Maria Maggiore in Rome (fifth century), a zone of yellow-green is used to form the transition from the blue and dark green planes of the background to the inserted golden stripes, ef, colour pletes in A. Grabar and C, Nordenfalle, Le Haut Moyen Age, ed. Skira, Geneva 1957, PP.(1)38-39. In the early sixth century mosaics in S, Apollinare Nuovo, Ravenna, Christ's halos, the halos of the prophets in the window zono, and many details

effect is created solely through the

of the costumes and objects are set with silver. In Rome only a few examples of the use of silver are mown: a) the halo of the Christ Lamb in the lower zone of the apse mosaicin SS. Cosma e Damiano (sixth century); δ) the halo of the Lamb in the apse mosaic of S. Prassede (ninth century), obviously copied after that of SS. Cosma e Damiano. On the use of silver as a characteristic of the Byzantine mosaics, cf. Mintz, Mosaïque, pp. 41 f. On the uso of silver in general, see H. P. L’Orange and P. J. Nordhagen, Mosaik von der Ánihe zum Mitelaltr, Munich τοῦο, pp. 65 f. (2) In Christ's halos in S. Apollinare Nuovo (early sixth century), silver is used in the disc.while ‘the cross is of gold. In later Byzantine works, howfever, the disc is always of gold with the cross im ; this is the caso a) in the vestibule mosaic in Hagia Sophia af Istenbul (late ninth century), cf. 15r

PER JONAS NORDHAGEN

iii) Objects

In representations of jewellery a simple form of material imitation occurs, e. g. in the τὰς bies » and « emeralds » on the throne of the Adoration fragment (Cat. no. 5), formed by dark orange and azure-blue tesserae respectively (Colour pl). The imitation is especially elaborate in the jewels of the Maria Regina figure (Cat. no. 1), where the « diamonds » of the diadems, the collar, and the shoulder orbiculi are framed with gold as in the real setting of precious stones, In the orbiculi and on the throne we even see reproduced the small golden prongs which secure the diamonds, Objects of more complex colour structure are preserved only in limited number. In the Washing of the Child, the tub is set with vertical rows of gold and glass in many different colours. The front of the cushion on the throne in the Adoration fragment is set in different green tones of glass, but is uniform in value, which gives the side of the cushion an even, flat form. The lower part, however, is shaded in green mixed with ultramarine and blue-black tesserae, as is the

part where the cushion is weighed down by the sitting figure. This shading is intended to create 8 plastic effect.

iv) The costumes

In the fragments parts of seven figures are preserved. The examples of costumes are: the imperial dress of Maria Regina (Cat. no. 1), two representations of the Virgin in palle (Cat. nos. 3 and 5), Christ's tunic and pallium (Cat. no. 6), Salome's chifon (Cat. no. 4) and the garments of Joseph and the angel (Cat. no. 5). The costumes of Salome, Joseph and the angel are all rendered by simple colour combinations, the dress colour being one dominant, light tone, with the contours and fold lines set in thin rows of darker tesserae. On the angel’s left shoulder, a few short rows of very light grey and blue glass tesserae are inserted in the white marble area, which gives the pallium a faint shading. The other garments are more complex in their colour structure. In the costume of Maria Regina the main colour is purple, with a range of values from very light to almost black. For the darkest shades, black and brown-black are also used. Of the two examples of the Virgin's falla, only that in the Adoration fragment offers a certain variety in colours (Colour pl.). Here, the blue is in different hues, each with light and dark values, While in the imperial costume of Maria Regina and in Mary's palla, there is one main colour with a restricted range of tones — purple and blue respectively — a somewhat richer juxtaposition of colours occurs in the pallium of Christ from the Entry into Jerusalem. Here two hues of blue are used together with purple to create the effect of purple blue. This is the only example of optical colour blending in these mosaics, although the technique is often found in earlier Christian mosaics, where strongly contrasting colours of tesserae blend in the eye of the spectator and create a new effect when seen at a distance, One example is the purple in Christ’s pallium in the

mosaics in the small apses of S. Costanza, Rome (fourth century); this purple was created through a mixture of red, violet, blue and green tesserae (1). This ‘pointilliste’ principle seems gradually Grabar, Peinture byz, illustration p. 88; δὴ in all Christ figures in Daphni (eleventh century), cf. Grabar, pp. 114, 116-117; c) Cappella Palatina, Palermo (twelfth century), Grabar, pp. 130-131; d) Deesis in Kariye Camii (fourteenth century), cf. D. Talbot

Rice and M. Hirmer, The Art of Byzantium, London 1959, pl XXXI; e) Pantokrator in the cupola of Fethiye Camii (fourteenth century), cf. Talbot Rice, pl. XXXII. (1) Wilpert, op. cit, ΠῚ, pl. s. 152

THE MOSAICS OF JOHN VII (705-707 A.D.)

to have been abandoned in the Early Middle Ages, especially in Byzantine mosaics. As we have already seen, the red purple used in the mosaics of John VII was a ‘ready-made’ caput mortuum which had variations only in value. v) The flesh All fiesh parts in the mosaic are set with marble tesserae, while the stronger lines and contours both in faces and hands are in darker coloured glass of the same tessera size as the marble, Pink marble is generally used for the shadows (occasionally purple-grey marble is used for dark shadow); for the lighter parts yellow is often used, while the highlights are set with thin lines or small clusters of white tesserae. There are no abrupt colour contrasts, but a gradual transition from one not very strong colour to the other. Only the white stripes and patches stand out with some force. The general effect is that of softly shaded areas where the white lights pick out the points of anatomy that are in highest relief, The orange glass tesserae which add a stronger glow to certain parts of the physiognomy (p. 150), were probably less prominent originally, when the marble colours were of greater intensity. vi) Hair

‘The hair is rendered by a combination of colours not unlike the optical colour blending of earlier mosaics. The contours of the locks are in black or dark purple glass, while the hair colour is created by a juxtaposition of green, yellow and transparent green glass tesserae. In the hair of Joseph and of the angel in the Adoration fragment, some grey marble tesserae were also employed. In the dark hair and beard of Christ from the Entry into Jerusalem, purple is mixed

with almost black glass tesserae, and some cubes of very dark transparent green can also be found.

153

II COMPARATIVE STUDIES

Mosaic art in Italy from the seventh In Roman mosaic art of the seventh schools (). The chief monument of the first is where strongly linear tendencies are manifest

to the ninth century century we can distinguish two clearly different the apse mosaic of S. Agnese fuori le mura (625-638), (), The face of St. Agnes has pale, unarticulated

flesh areas set with coarse tesserae of marble, The features are strongly outlined in dark glass tesserae, and extremely small orange glass tesserae are concentrated in large patches on the cheeks, In the light part of the garment, white marble tesserae are extensively used, The mosaic. surface is uneven, with the tesserae sparsely distributed, and the white plaster bed, which is

visible in the intervals, contributes to the general shimmering effect of the mosaic, ‘The apse mosaic of 8, Stefano Rotondo (642-649) is closely related in style to the S. Agnese mosaics, with its pale, simply drawn faces and extreme linearity. The technique of the mosaic is also related to the S. Agnese school, in the coarse, irregular tessera work of the faces and garments (), and in the two colours of marble (pink and red) of the flesh areas. In both S. Agnese and 8, Stefano Rotondo the tesserae of the faces seem to be slightly smaller than those used in the garments and background (*) (1) The picture Kitzinger gives in his Rémische Malerei, op. cit., of the stylistic changes which took place in this period, agrees at many points with the evidence of the mosaic techaique. Important for the knowledge of mosaic art in Rome in the Early Midle Ages aro also G. Matthiae, Tradizione e reazione noi mosaici romani dei sec. V1 e vi, « Proporzioni ν, III, 1950, pp. 10 ff, and P. Baldass, The Mosaic of the Triumphal Arch of San Lorenzo fuori le mura, « Gazette des Beaux-Arts s, XLIX, 1957, pp. 1 f . (2) For illustration cf. R. Koemstedt, Vormittlaltriche Malerei, Augsburg 1929, pl. 123. (3) Cf. Matthiae, Tradizione, fig. 7. (4) The mosaic of S. Lorenzo fuori le mura, from the time of Pelagius II (578-590), has been commonly regarded as a forerunner of the S. Agnese school; ef, the survey of acories im Kitzinger, op. cit, pp. § ff. However, as P. Baldass has shown in a recent article (cf. note x), only a small part of that mosaic, namely the figure of the Papal donor and that of St. Lawrence, is from Pelagius own ime; the rest is the product of a series of restorations. The results of Baldass’ stylistic analysis corro-

lates closely to the present author's conclusions, which are based on the technical evidence: the faces of both Pelagius and St. Lawrence are executed with large glass cubes, in a style and technique close to that of 8. Teodoro in Rome (middle of the sixth century, cf. G. Matthiae, SS. Cosma e Damiano e S. Teodoro, Rome 1948, pp. 86 41); whereas the face of St. Hyppolite, also in glass, with its somewhat heightened colourism, may be contemporary with the S. Venanzio mosaics or is from a recurrence, in the eighth-ninth cent, of the S. Venanzio style. The St. Peter — Christ — St. Paul — St. Stephanus group is, on the other hand, of much later manufacture and shows heads composed of very small, rather pale cubes which are put together in a coarse technique reminiscent of that of later medieval mosaics of Rome, e. g. apses of S. Francesca Romana (ca. 1160) and S. Maria in Trastevere (ca. 1140). The particular technique displayed by the Pelagian part of the mosaic seems to demonstrate that this "work belonged to a school different from that which produced the apse mosaics in S. Agnese,

154

THE MOSAICS OF JOHN VII (755-707 A.D.)

To a totally different school belong the mosaics of the Cappella di S. Venanzio (ca. 640649) in the Lateran Baptistery (!). Here the heads are strongly plastic, although the garments especially the white chlamydes — have linear traits. The flesh of the faces, and the hair and beards are all set with tesserae of the same large size as those employed in the garments, and the material is exclusively glass (). We do not find here those rather pale faces with cheek patches or the careful delineation of features with sharp contours. Instead, a juxtaposition of tesserae in strong, often complimentary colours creates an effect of plasticity. This ‘impressionistic’ colour modelling is brought to its peak in some of the faces on the triumphal arch, where the form nearly dissolves in the absence of outlines (). The tesserae of heads and costumes are sparsely set, as in the mosaic of S. Agnese and in other monuments of the S. Agnese group (ἢ). The mosaic fragment with the standing figure of St. Sebastian, in the church of S. Pietro in Vincoli, belongs to the tradition of the S. Venanzio mosaics and shows the degeneration of this particular style towards the end of the seventh century (5). Glass is still the only material used, and the colours are brilliant, with a reminiscence of ‘impressionistic’ modelling in the strong, freely distributed red accents of the face. But the figure is more stylized, with harder outlines

and a stiffer pose.

Almost contemporary with the St. Sebastian fragment is the mosaic of Bishop Reparatus on the left wall of the choir of S. Apollinare in Classe at Ravenna, a work executed ca. 673-679 (ἢ). In the panel commemorating the presentation of privileges to Reparatus, little more than the heads seem to be original (. Marble tesserae are used in great number in these heads, and they are coarsely set in a simplified technique (). ‘The Oratory mosaics of John VII (705-707) stand out conspicuously among the works discussed above, Aesthetically and technically they contrast markedly with the S. Venanzio mosaics and their tradition; instead of glass in uniform tessera size for all parts of the picture, John VIT's

mosaics have small cubes of marble in the flesh areas, and glass in the garments and backgrounds

only. And while the S. Venanzio mosaics create the effect of plasticity in the faces through ‘impressionistic’ modelling and contrast in colour, the faces of John VII's mosaics are built up through changes in values, through combining white highlights with gradual shading by means of cubes of subdued colours. (1) Kitzinger, op. cit., pp. 12 fi; Baldass, op. cit., pp. 9 1. (2) For illustration, cf. L'Orange and Nordhagen, op. cit., pls. 82-83. (3) To the left, especially in the faces of St. Paulinianus, St. Telius and St, Asterius; to the right, St. Antiochianus and St. Gaianus. (4) The hypothesis, put forward by Matthiae (SS. Cosma e Damiano, p. 62), that the angels on the triumphal arch of SS. Cosma e Damiano belong to the last decades of the seventh century, does not stand a test based on the technical factors; the angels have none of the sparse tessera setting so typical of the seventh and eighth centuries. On the contrary, fhey are set with densely placed and regularly arranged tesserae in a technique characteristic of the early sixth century and identical with that of the apse. From a stylistic point of view, Kitzinger has shown the reason for the strong «Hellenistic » ap-

pearance of the angels on the triumphal arch; cf. his Byzantine Art in the Period between Justinian and Iconoclasm, «Berichte zum XI. Intern. Byz. Kongress», Munich 1958, p. 47. (5) For illustration cf. S. Bettini, Früchristliche Malerei, Vienna 1942, pl. 99. This fragment is generally dated about 680, cf. Kitzinger, Rom. Mal, p.21. (6) This is the only one of the side panels that can be dated on firm historical grounds, î. e. from the theme which is represented. For a summary of discussions concerning the date of the two panels, cf. M. Mazzotti, La basilica di Sant'Apollinare in Classe, Vatican City 1954, p. 176, n. 26. (7) The restorations of the panel are marked in C. Ricci, Tavole storiche dei mosaici di Ravenna, Rome 1935, VII, pl. LXVIII. (8) C. Ricci attests to the use of marble in panels of the presbytery (Marmi ravennati erratici, « Ausonia», IV, 2, p. 248, n. 2). 155

PER JONAS NORDHAGEN

In their extensive use of marble, the John VII fragments are more closely related to works of the S. Agnese school of the seventh century. Marble was also, as shown above, used in the Ravenna mosaics at the end of the same century, not long before John VI own time. But even these works ~ also belonging, apparently, to a tradition different from the S. Venanzio school — do not entirely resemble the Oratory mosaics in style and technique. There is a great difference between the way marble was handled in S. Agnese fuori le mura and the way it was used in the Oratory mosaics. In the seventh century works, the, flesh areas were set with rather large, coarse marble tesserae in a limited range of colours, resulting in generally pale flesh zones accented only by the large red patches of tesserae, None of these works displays a variety of marble colours comparable to that of the Oratory fragments, or a technique employing tesserae of such minute size, Nor did they show such refined combinations of materials, as for instance the use of transparent glass zs a neutral element between the marble and the stronger coloured glasses. These differences in workmanship place the mosaics of John VIIin marked contrast to the examples of the earlier ‘marble tradition’ in Italian mosaic art. After the brief reign of John VII, there is a gap of nearly a hundred years before we again encounter preserved mosaics in Italy. Because of this lacuna, little is known about the immediate impact of the Oratory’s style and technique upon the mosaic art of subsequent decades. Fragments from as late as ca. 800 testify that the coarser form of the marble tradition did survive. In, for example, the preserved heads from the mosaic of the Lateran triclinium of Leo ΠῚ (795-816), large marble tesserae are used in the flesh, and are chaotically placed in the setting bed. Their colours seem to be only two, white and yellow (1). However, another work attributed to the time of Leo III, the mosaics in SS. Nereo e Achilleo, reveals that the ‘glass tradition’ also persisted. In the figures on the triumphal arch, glass is employed in the flesh as well as in the garments, with a uniform, large tessera size throughout. In the first half of the ninth century, the mosaics made for Paschal I (817-824) in S. Prassede, S. Maria in Domnica and S. Cecilia in Trastevere, and for Gregory IV (827-844) in S. Marco are all set entirely with glass. The style, especially in the garments, is extremely linear, with broad, band-like outlines (); the faces, nevertheless, show strong reminiscenses of colour modelling: blue or bluish green tesserae are often used in the outlines, where they

contrast with the vivid orange or red highlights (*). The tesserae are generally irregular in

form, but are set closer than in earlier works, so that the white plaster intervals are almost eliminated. There can be little doubt that the pure glass mosaics of the late eighth and early ninth century both technically and aesthetically represent a continuation of the seventh century S. Venanzio school. The style has changed considerably, but the principles of colour modelling (especially in the faces) with strongly coloured glass tesserae has survived. These characteristics, moreover, can be traced far back in the history of Italian mosaics. ‘Impressionistic’ modelling with strongly coloured glass tesserae of large format is seen in the original mosaic heads in S. Lorenzo fuori le mura (^j, and in S. Teodoro, and SS. Cosma e Damiano (also sixth century) (+), in the S. Maria Maggiore mosaics in Rome and in the cupola mosaic of the Orthodox Baptistery (2) For illustration cf. G. Galassi, op. cit, I, (8) Ilustration in Grabar, Le Haut Moyen Age, fig. 104, p. 217. In his text, Galassi asserts that pp. 42-43. Baldass, op. cit, p. ro, calls this the ‘poinunlike the mosaics of John VII these heads are rep- — tillistic’ style. resentative of a Roman school, ef. op. cit., pp. 215 f. (4) Cf. n. 4, p. 154. (2) For a description of this style, cf. Kitzinger, (5) Matthiae SS. Cosma e Damiano, pls. 4-9, Rim. Mal. pp. 33 f 2528, 37-38.

156

THE MOSAICS OF JOHN VII (705-707 AD.)

(S. Giovanni in Fonte) at Ravenna (both fifth century) (1), and, finally, in the few preserved heads in S. Pudenziana, Rome, from the late fourth century (). There are good reasons, therefore, for believing that this particular school represents an indigenous line in the mosaic art of Italy, with roots in late Roman mosaic art (). It then remains for us to locate the source of the marble tradition, which formed the other main trend in the preserved material of Italian early medieval mosaics, and to which Pope John's mosaics belong. If, as we have proposed, the ‘impressionistic’ modelling with glass probably

represented a native tradition in Italy, the marble tradition may then have resulted from foreign, i. e. Byzantine,

influence.

In fact, the

technical

similarities

which

exist between

the

mosaics of John VIT and Eastern, Byzantine works (similarities to which we shall return in the next chapter), support this hypothesis and even suggest that the Pope had a group of Byzantine mosaicists working for him in the Oratory. Two traditions seem to have existed side by side in the mosaic art of Italy in the Early Middle Ages, one native and one Byzantine, the latter nourished by the Byzantine political domination of parts of the peninsula during this period (*). In Ravenna, the native glass tradition reigns uninterrupted until about 500 A. D.; in the

Orthodox Baptistery and the so-called ‘Mausoleum of Galla Placidia’ glass is used for the flesh as well as for the other parts of the figures. In the Baptistery, however, there is an occasional use of marble for ‘stone imitation, e. g. in the ‘marble screens’ between the thrones and altars in the lowest zone of the cupola decoration. Then, in the mosaics from the time of Theodoric the Great (493-526), typical Byzantine

features such as the general use of gold backgrounds and the varied application of silver appear

in conjunction with an occasional use of marble in the flesh areas. Thus marble, in two to three colours, are found in the faces, hands and feet of all the Miracle scenes of the S. Apollinare Nuovo Christological cycle, as well as in a few of the Passion scenes (Gethsemane and the Betrayal), while the greater part of the Passion scenes have glass (*). In the mosaics of S. Apollinare in Classe, marble is again used in the flesh areas of the figures of bishops in the apse. However, in the almost contemporary mosaics in S. Vitale, in the panels with Emperor Justinian and his consort Theodora (about 547 A. D.), glass was preferred for rendering the flesh. In Ravenna,

accordingly, the Byzantine influence, registered from Theodoric's time, seems to have resulted in a mixed technique, where both glass and marble were used for the flesh, without any pre-

ference being given to either of the two materials. In Rome, the preserved mosaics show that the glass tradition persisted throughout the

sixth century (ἢ); no influence of the marble technique is discernable here before the S. Agnese and S. Stefano Rotondo decorations towards the middle of the seventh century. It seems evident

(ἢ S. Maria Maggiore: cf. illustration in Grabar, 1888, especially pp. 257 f. (on the popes of Greek deLe Haut Moyen Age, pp. 37-39. The Orthodox scent), pp. 277 f (on the Greek colony in Rome), Baptistery: Wilpert, op. cit, III, pl. 79. pp. 280 ff (on the penetration of Byzantine culture (2) P. Maratoft, La pittura bizantina, Rome 1927, into Italy). The adherence of John VII's mosaics to the Byzantine tradition has been stressed by Gapl XL (6) In the third century Silvanus mosaic from lassi, op. cit, ΤΊ, p. 203; it has also been stated Ostia at the Lateran Museum (ill, in L'Orango and by V. N. Lazarev in his History of Byzantine PainNordhagen, op. cit., pl. 38) glass in cubes of uniform fing (in Russ), Moscow 1048, I, PP. 53 f, p. 268, size is used throughout; the same technique is found 7. in the third century Christ Sol in the Mausoleum M (5) Observations made at the exhibition of the under St, Peter's (ill, ibd., pl. 30). Christ cycle panels in the Museo Nazionale, Ravenna, (4) Ct. C. Diehl, Études sur l'administration by- in 1958 santino dans l'exarchat de Ravenne (568-751), Paris (6) CE n 4 pe 154 157

THE MOSAICS OF JOHN VII (705-707 A.D.)

Beard: Purple-black glass set together with dark greyish purple marble tesserae. "Hands: Marble tesserae in greyish purple and pink, the pink concentrated as light patches on knuckles. Dark glass contours in red and black. Mappa: White marble tesserae with folds and outlines in dark glass. Sceptre: Staff set with one row of white marble tesserae, one row of gold and silver tesserae; cross arm οἱ same materials. Background (the white ‘wall'): White marble with horizontal lines of greyish purple marble. Along right border of panel, white field is shaded with a zone of light marble tesserae with red spots. The ground: Mostly greyish purple marble tesserae. Above ‘wall’: light blue background, glass tesserae. Costume: Toga set with rows of gold tesserae alternating with rows of yellow-green tesserae. At figure's right elbow is part of dalmatica, in purple and red glass tesserae. Shoes: Yellow-green and green glass. St. Demetrius Face: Flesh of marble tesserae, mainlyin pink and yellow, some white highlight stripes and small shadow areas in darker greyish pink. Halo: Gold tesserae, White marble outline, round which follow. some circles of dark blue glass tesserae, then light blue background. Costume: All white parts in marble, also orbiculus on shoulder on tunic. Fold lines of chlamys in bluegreen glass. Big purple emblema has greyish purple tesserae of marble in lighter parts, dark purple glass in darker folds. Square ornaments with a central dot: on white dress area material is gold and silver mixed, on purple emblema it is white marble. Fibula set with gold tesserae. Part of tunic to left of chlamys made of alternating rows of gold and light purple tesserae. Shoes: White marble, black glass. The donor holding a book Face (1): Same materials as other two figures. Paenula : White marble tesserae, folds in red-brown glass, stronger shade lines in dark brown glass. Book: Front gold, side with pages outlined in alternating rows of white and dark natural stone. Inscription below panel: Letters black glass tesserae on white background of marble tesserae. White mosaic frame of panel also marble. Size of tesserae: Flesh areas one third to one fourth of those of rest of mosaic.

The analysis of the materials establish their close similarity to those of the John VII mosaics. In both mosaics we find a comprehensive and richly varied use of many-coloured marble in the garments and flesh, and a high degree of refinement in the combination of marble and coloured glass (). In addition, the faces are modelled in a manner which closely resembles that of the Oratory mosaics: the greyish purple and pink marble tesserae create the shaded parts of the flesh, while the highlights of white stone, concentrated in single lines, sharply illuminate the form, producing an effect of plasticity. 2) Mosaic fragment in the church of Hagios Nicolaos. Istanbul (PI. XXVI). Published by H. Zidkov (). The fragment measures ca. 0.39 x 0.31 m. Its provenience is uncertain (). Restorations. Description. Face beardless, features and bluish tunic can be

An outer zone along modern frame restored with mosaic tesserae, Fragment shows head and shoulders of figure which Zidkov identifies as an angel (ἢ). youthful, thin band runs through Iuxuriant curls of hair. Parts of white pallium seen. A staff, baculum, seen against left shoulder. Head turned in three quarter

(1) Illustrated in A. Grabar and M. Chatzidakis, Byzantine Mosaics in Greece, New York 1959, pl. IX. (2) Special marble types used in both H. Demetrios and John's Oratory: 1) Greyish purple, for deep shadows in complexion (fragm. of John VII, cat. no. 2); 2)White with spots ~ used for backgrounds (Washing of the Child, cat. no. 4).

(3) H. Zidkov, Bin Wandmosaikfragment aus Konstantinopel, « Byzantinische Zeitschrift a, XXIXXXX, 1929]30, pp. 6or ff. (4) According to the tradition on the site, the fragment came from Gul Camii (St. Theodosia), but Zidkov, p. 602, regards this with doubt. (5) Zidkov, op. cit, p. 602. 159

PER JONAS NORDHAGEN

profile towards left and somewhat raised. Technique, colours and materials. Plaster white, of fine consistency. Tesserae, sparsely set, especially in hair and costume, are deeply embedded, so that plaster in interstices lies flush with mosaic surface. Tesserae of very small size (about one quarter to half size of tesserae in costume and halo) used in flesh areas of face and throat. Darker lines of costume have tesserae slightly bigger than those of lighter coloured areas. Halo: Gold tesserae with outline of two rows of green glass tesserae, As in mosaics of John VII, gold tesserae of halo placed in somewhat irregular circles around head. Gold leaf attached to green, transparent smalto. Many leaves fallen off. Gold background (restored). Originally another type of background must have existed here, since a cluster of original glass tesserae of different colours is preserved outside the halo ta left of head. Head: Face and throat set with marble tesserae, now so covered with dirt that one can hardly discern the modelling with different colours. Highlight spot over left eye and on chin. Line of highlight runs along ridge of nose. Eyes: Central tessera of black glass (pupil) surrounded by one circle of brown glass tesserae and one of light marble. At right, in both eyes, two white glass tesserae mark highlight of eyeball Eyebrows, eyelashes, contour of nosetip, line between lips all Set with black glass. Green glass lines between brows and lashes, along whole length of nose, between flesh areas and hair, as contour of check from ear to chin. Mouth has red glass tesserae in lips, brown shadow line of glass running immediately below lips. Hair: Dark brown and light brown glass tesserae arranged in clearly defined curls. Band through hair one row of white glass tesserae and one of green. Costume: On angel's left shoulder pallium set with light (white) tesserae of marble, slightly smaller than glass tesserae of costume. Outlines of Pallium : long line of large tesserae leading from car down along throat is of red transparent glass, other outlines of opaque red. In pallium two different marble colours, whose exact character cannot be determined because of dust layer. Tunic, on angel’s right shoulder and at his breast, red glass contour tesserae; colours of tunic in glass tesserae: vertical rows of dark blue tesserae alternate with rows of light turquoise-blue and rows of silver. Silver leaf, like gold, attached to green transparent glass. Baculum: One single row of brownish red tesserae.

Stylistically, Zidkov found the fragment closely related to the mosaics of John VII (1). V. N. Lazarev has since stated that this work provides strong evidence of a Byzantine background for the Oratory mosaics (), Our observations on technique and material confirm these conclusions, The mosaics of John VII and the fragment in Istanbul have the following technical features in common: A) The deeply inserted tesserae which are set leaving wide interstices. B) The use of marble tesserae of extremely small size and of different colours in the flesh areas (). C) The arrangement of these tesserae in curving rows separated by broad plaster interstices (for the surface effect, compare the throat area of the Istanbul fragment with the similar area in the fragment of the Virgin at Orte, from the mosaic of John VII, PI. VIL a). D) The use of extra large tesserae for the dress folds and contours. These features create a surface structure which is almost identical in the two mosaics. In the choice of materials there is one conspicuous difference: the silver used in the tunic of the Istanbul angel is a material highly characteristic of Byzantine mosaics, and is not found in John VII's mosaics (see p. 151). However, given the many traits, stylistical and technical, which connect the fragment from Hagios Nicolaos with the Oratory mosaics, the absence of silver in

the latter cannot diminish the impression that both spring from the same tradition. The Istanbul fragment, being the closest extant parallel to the mosaics of John VII, must date from the same period, probably some time in the late seventh or early eighth century (ἢ. (1) Ibid., (2) Ct n. (3) Since size of the

p. 604. 4, p. 157. the mosaic today is under glass, the tesserae could not be measured. They

seem, those (4) venth 160

however, to correspond fairly accurately to of the Oratory mosaics, Zidkov, op. cit, p. 607, is in favour of a secentury date.

THE MOSAICS OF JOHN VII (705-707 A.D.)

3) Mosaics of the bema and apse in the church of the Dormition, Nicaea. The following observations are based on the publication of T. Schmit, Die Koimesiskirche von Nikaia, Berlin 1927, and are concerned mainly with the heads, since these are the only sections of the mosaics that Schmit reproduces in detail in colour (colour plates made from watercolours). By minute comparison of the plates in black with the corresponding colour plates it may be possible to supplement Schmit's own text, especially as regards the technique and the use of materials. Natural stone will always be distinguishable in detail photographs because of its smooth surface, whereas the glass is betrayed by the characteristic bubble holes. By following these

clues, a more thorough description of the tessera work in the mosaic can be given (1).

The angels of the bema, represented by AYNAMIC (PL XXIII); cf. Schmit, pls. XVI and XVII: AIL along face oval, inside outline, runs a broad peripheric shade zone, to left formed by three rows of tesserae, to right — shaded part of face - by four to five rows. Outer rows of these shadows apparently of green glass, of relatively light hues on left side of face, but with darker components on right. Central parts of face, the eyes and nose, also inscribed in broad, green shade lines; nose has two rows of tesserae on left side, two to three on right, all in glass. In this symmetrical system of shading, lighter areas are distributed thus: on brow a concentration of pink natural stone tesserae from which white lines marking horizontal wrinlos of brow stand out in contrast. These white lines probably also set with natural stone. Light zones of white and pink natural stone tesserae also on cheeks, under eyes, to both sides of mouth and on chin. A highlight stripe in pink down ridge of nose. Probable that light green tessera rows immediately above eyelashes on both eyes, and under eyeball to left, are of natural stone. Centre of left cheek an oval shadow, probably οἱ red glass (cheek spot). Schmit mentions eyebrows of EZOYCIE made of colourless, transparent glass, giving dark effect when seen at distance (Schmit, p. 26). On throat, short rows of white and pink tesserae, evidently of natural stone, mark lighted areas which are surrounded by broad curved shadows of darkegreen and brownish glass. These curved shadows give throat highly stylized shape, like series of ‘double chins’. Special technical features: In both eyes of AYNAMIC the highlight spot to right of pupil set with single, unusually large tessera of pink or white natural stone. Extra large pieces of natural stone used singly as highlight patches seen in all other heads. In KYPIOTITEC, a huge oval tessera at point of nose (Schmit, p. 25), in corner of both eyes similar pieces. In ARXE, outsized tessera at point of nose, same in EEOYCIE. Same technique used in faces of Virgin and Child in apse: Virgin in corner of both eyes; Child has such tesserae on point of nose and in corner of one eye. Size of the tesserae (Schmit, p. 27): in garments, ca. 8 x 9 to 11 x 12 mm; hair and flesh from 2 x 4 to 4X6 mm; gold smallo 4 X 5, 5 X 5 and 5X6mm. Setting of the tesserae: There are intervals between the tesserae, but not so large as those in John VII's mosaics. Tessera disposed in even, regular rows with strictly parallel curves in contours and shadow areas. Roundness of angels’ faces stressed by this regularity. Technically, close correspondence exists between the mosaics of the church of the Dormition and the John VII mosaics, above all, in the use of different coloured natural stone tesserae and transparent glass in the face modelling, but also in the distinction made between the size of tesseTae employed in the faces and those in the garments. This technical evidence further substan(1) Schmit makes the following observations on the material used in the tesserae: a) in describing the apse mosaics (p. 22), he mentions tesserae of «eine feste aber nicht glasharte Masse », used for the rays of light in the Hetimasia, According to Schmit, ‘these tesserae had become very dirty. Their material therefore was probably marble, to which dust and soot will easily cling because of the coarse grained surface. The same kind of material had also been

‘used for the pearls on the throne, which originally had been white but now were completely black (Schmit, P. 22); b) in the description of the boma mosaics (Schmit, p. 27) is found the statement: «Nicht von Glas sind nur die weissen Würfels, Although Schmit suggests mother-of-pearl, since the tesserae were too soft to be of glass, this material probably also was marble. Schmit further mentions that this enigmatic material was used in the garments. 16r

PER JONAS NORDHAGEN

tiates the theory that John VII's mosaics developed from a Byzantine tradition. But stylistically there are important differences in the distribution of light and shade on the faces in the two mosaics. In the church of the Dormition, highlights are prominent on the faces, e. g. the white touches on the brow, along the nose and on the chin of the angels. These highlights, however, do not stand out sharply as thin lines or dots as in the John VII mosaics. On the contrary, they tend to be arranged in larger, regularly formed light zones, which are balanced against evenly shaped and strictly symmetrical shadow areas, And instead of a naturalistic reproduction of anatomical form (such as that of John VII's angel of the Adoration (PI. XIII) with its superlative rendering of the throat muscles) the bema angels are composed of more synthetic, abstract, form, as in the ' double chins’ described above. This more advanced stage of abstraction would seem to indicate that the mosaics of the church of the Dormition, i. e. the parts that include the angels of the bema and the Virgin of the apse, were made, as P. A. Underwood suggests (1), in post-Iconoclastic times. The stylistic affinities with works of the later ninth century are obvious (); however, in the technique many characteristics of the mosaics of John VII prevail. 4) Mosaics in the church of Hagia Sophia, Istanbul. A) The Virgin of the apse and the Archangel of the bema, probably 866-67 (). It can be established with certainty that a quantity of different marble colours were used in the figures, large cubes together with glass tesserae in the garments, and extremely small cubes in areas of the flesh. In the colossal faces ()), the modelling of light and shade is somewhat different from that seen in the Nicaea mosaics; the shadows are asymmetrically distributed in patches of soft hues, not, as in the Nicaea heads, arranged in broad, regular bands. This greater 'painterly’ freedom in the choir mosaics of Hagia Sophia seems to indicate that they are slightly earlier than the angels and the Virgin at Nicaea, B) ~ a) Figures of saints on the north tympanum of the nave ©). b) Panel over door in S. vestibule; the Virgin between Justinian and Constantine (9). Both these mosaics belong to a slightly later phase of the post-Iconoclastic redecoration of Hagia Sophia, and can probably be dated to the last quarter of the ninth century (ἢ. (1) Underwood gives a brilliant solution to the problem of the different periods of work in the mosaics of the apse and bema, in The Evidence of Restorations in the Sanctuary Mosaics of the Church of the Dormition at Nicaea, ¢ Dumbarton Oaks Papers», 13, 1959, pp. 235 ff. His conclusion is that the angels of the bema and the Virgin of the apse are contemporary, all dating from the time immediately after the end of Iconoclasm, i. δι, from the middle or the second half of the ninth century. (2) Compare especially the heads of AvNawic (Schmit, Koimesishivche, pls. XVI and XVII) and of John Chrysostom from the north tympanum in Hagia Sophia (reproduced in colour on the cover of «Archaeology », 4, 1951, 2). (S) As the commemorative inscription on the apsidal arch in Hagia Sophia mentions rulers in the plural, the mosaics in the choir seem to have been

made during the combined reigns of either Theodora and Micael INI (843-55) or of Michael ΠῚ and Basil I (866-67). Preference should be given to the latter date, as it scems highly probable that the Virgin in the apse is the one alluded to as newly installed in the sermon of 867 by Photius; cf. C. A. Mango, The Mosaics of St. Sophia at Istanbul, Washington, D. C., 1962, pp. 94 f. For arguments in favour of the earlier date, cf. A. Grabar, L'iconoclasme bysantin, dossier archéologique, Paris 1957, pp. 189 ft. (4) Talbot Rice, op. cit, pls. 88-89. (6) For reproductions of these mosaics, cf. Mango, op. cit, pls. 62, 70 and 72. (6) T. Whittemore, op. cit., Il: The Mosaics of the Southern Vestibule, Oxford 1936. (7) The arrangement of the hierarchy on the tympana follows tho new principles for church decorations which are heard of in the second half of the 162

THE MOSAICS OF JOHN VII (705-707 A.D.) On the technique of the Virgin panel in the vestibule, we have the following information (1):

‘The faces and hands are formed of small marble and glass tesserae. The marble, in delicate shades, is used for the lighter parts of the flesh only, while shadows and semi-shadows are set with glass of brown, olive-gréen and greyish blue. Large, pear-shaped marble tesserae are set as extra-strong highlights on the point of the nose in all figures, as in the Nicaea mosaics (p. 161) (*). Transparent green glass is used on the throat of Justinian, outlining the larynx, and

as semi-shadows on the emperor's hands (*), Although marble still dominates in the flesh areas, glass is more widely used than in any of the Byzantine mosaics we have studied so far. In the heads, the regularity of the tessera work transforms features into abstract patterns.

Shadows in the frontal face of the Virgin are rendered as broad bands along the oval outline of the face (‘); these bands also circumscribe the eyes and the nose. In the Nicaea mosaics, which must be almost contemporary, the semi-shadows in the cheek were still rendered as irregular patches and accents, while here they are formed by an orderly series of short, hatched lines which extend from the broad band of shading along the outline of the face. In the balfturned faces of Justinian and Constantine, the lights and shades have now become stylized wave lines, Together with the use of stronger colours for the flesh (olive-green, greyish blue), this simplification of anatomical form heralds a new phase in Byzantine mosaic art. These tendencies were to reach their purest expression in the great mosaics of the eleventh and twelfth centuries. The style and technique employed for the saints on the north tympanum of Hagia Sophia

seem to correspond closely to those of the vestibule panel. In the figure of John Chrysostom (*), the typical bands of shading, in olive-green and greyish green, are seen in the face, where

they impose upon the features a hieratic rigidity. To this stage of the evolution also belong the cupola mosaics of Hagia Sophia at Thessaloniki, whose colours and materials are not yet published (*). From the photographs available, a certain amount of information can be gathered. The usual marble is apparently used both in the flesh and in parts of the garments. In some of the faces, one extra large white tessera forms the light on the tip of the nose. Broad shadows, similar to those we have seen in the face of the Virgin in the Hagia Sophia vestibule, encircle the area of eyes and nose. C) - a) Panel with Emperor Leo VI (886-912) before Christ, over central door of narthex (). b) Portrait figure of Emperor Alexander (886-913) in the north gallery (9). In the Leo panel, the tesserae are not set so closely or so regularly as in the vestibule panel, but a similar stage of stylization can be observed. Strong yellow-green glass tesserae are used

ninth contury; the gallery of saints, however, cannot be from before 878 A. D, since it includes Ignatius the Younger; see S. der Nersessian, Le décor des églises du IX* sidle, «Actes VI. Congr. Intern. Bt. By. +, Paris ror, pp. 315-320, and Mango, op. cit, p. 57. The vestibule pancl, vaguely datedby Whittemore to the end of the tenth century (op. cit, II, . 31), belongs, for stylistic reasons, to the late ninth century: its stage of stylization is closely related to that of the Nicaea mosaics, which must have originated some time in the second half of the ninth century, cf. note 1, p. 162. (2) Whitiemore, op. cit, IT, pp. 42 8. (2) On the use of these largo highlight cubes in

the mosaics of Hagia Sophia, cí Underwood and Hawkins, op. cit. p. 195, n. 29. (3) Whittemore, op. cit., II, pls. XIII, XVI-XVII. (4) Ibid, pl. X. (5) For a reproduction in colour, cf. note 2, p. 162. (6) Grabar (L'iconoclasme, p. 195) is of the opinion that the mosaics in both cupola (Ascension) and apse (Virgin) are from the end of the ninth century. ‘The Virgin of the apse, however, with its thin lines of light and shade in face and hands seems to be the work of a later period (twelfth century?) (7) Published in Whittemore, op. cit, I, pp. 14 ft. For the date, see Mango, op. cit., p. 96. (8) Published in Underwood and Hawkins, op. cit.

163

PER JONAS NORDHAGEN

in the shadows of the face, while the softer hues are rendered with marble, As in the mosaics of John VII, transparent green glass is used for soft contours of the flesh areas in the Virgin medallion to Christ's right. Transparent glass (turned gold cubes ?) forms the outline of the Virgin’s brow and chin, and the shade lines on her throat. The same type of glass is used in a large number of cubes in the emperor's beard (1). While marble continued to be employed for the flesh zones in all later Byzantine mosaics (?), it seems that the transparent glass, which made the different materials blend more harmoniously (pp. 149), was not used after the ninth century. It does not seem to have been employed in the figure of the Emperor Alexander, whose modelling in most other respects closely follows that of Leo. (1) Whittemore, op. cit, I, pls. XV, XXI. (2) For instance, cf. Whittemore, op. cit, IV,

on the Deesis panel in Hagia Sophia (ca. 1300); description of colour and material, pp. 38 ff.

164

IV ConcLusION

In this study of the artistic afüliations of John VII's Oratory mosaics, the point of departure for discussion has been the mosaic technique, or to be more precise, the technique involved in rendering the flesh of figures, especially that of faces. A detailed analysis of John VIT's mosaics has shown that the mosaicists chose mainly marble for the flesh and that they used it in a wide variety of subdued colours and in tesserae smaller than those otherwise employed in the mosaic. We have established that, in these respects, the mosaics of John VII stood out conspicuously among Italian monuments of the Early Middle Ages. The technique used in one group of Italian works seemed significantly different from that of John VIT's mosaics, namely in those where large tesserae of strongly coloured glass were used for the flesh, and where the total colour effect was obtained by vigorous contrasts.

Another

school

showed qualities more related to those of the Oratory mosaics: a predominance of marble in the flesh areas, but with coarser tesserae in a more limited range of colour. Only a few mosaics of the early sixth century in Ravenna presented a variety of marble tints comparable to that of the mosaics of John VII. When we extended our comparative studies to include Byzantine mosaics of the Early Middle Ages, we found in them a series of perfect counterparts to those qualities which are characteristic of the Oratory mosaics. The technique of depicting flesh by using a rich assortment of marble colours in small tesserae is seen in numerous Byzantine mosaics, from the earliest of the monuments examined by the author, the St. Demetrius panel in Thessaloniki (seventh century), down to the post-Iconoclastic works; in fact, this technique seems to have been peculiar to Byzantine mosaic art throughout its existence.

Further, in Eastern monuments it was possible to

observe

workmanship analogous to that which distinguishes the Oratory mosaics, for example such details as the highly regular, though wide, spacing of the minute marble tesserae, and their deep setting in the cementing medium. On the basis of these similarities, we have formulated the hypothesis that the mosaics in the Oratory of John VII were executed by Byzantine artists. In anelabo-

ration of the same hypothesis we have suggested that the school that used strong glass colours

for the flesh may represent the native, non-Byzantine mosaic tradition in Italy. Concerning the above suggested division of early medieval mosaics into two schools, one Western (Italian) and one Byzantine, we might add a few more remarks. In the author's opinion, it would be wrong to explain the two schools merely as two different traditions of workmanship, and to regard their distinguishing features purely from a technical point of view. It proved practical, for description and comparison, to contrast the two schools strictly in terms of technique, while in fact far broader implications are involved. It cannot have been mere preference

for a material that led one group to choose glass of strong and contrasting colours, while another used marble with more subdued tints; the difference in media seems to spring from a divergence 165

PER JONAS NORDHAGEN

of aesthetic principle. The school using glass renders the human face through ‘impressionistic’ juxtapositions of intense colour points; the marble school, however, tends to seek subtler effects — softer transitions and finer gradations, Thus, while the first, in a stale continuation of the illusionism of ancient mosaic art, employs a technique that was meant to reflect the changing surface of things, the second seems to have developed with the purpose of grasping more permanent aspects of form — above all the large, immovable planes of the Holy visages.

166

P. J. Nordhagen, The mosaics of John VII

The Adoration (Cat. no. 5). S. Maria in Cosmedin, Rome, Colourplate, Enciclopedia Universale dell'Arte.

P. J. Nordhagen, The mosaics of jchn VII

Maria Regina (Cat. no. 1). S. Marco. Florence. Phot. Alinari

PLATE I

P. J. Nordhagen, The mosaics of John VIT

Maria Pagine (Cat. no. 1). The head and its diadem. Phot. P. J. Nordhager.

FLATE Ii

E. J. Nordhagen, The mosaics of John VII

Maria Regina (Cat. no. 1). The face. Phot. P. J. Nordhagen.

PLATE III

P. J. Nordhagen, The mosaics of John VII

Maric Regina (Cat. no. i). The jewes at the neck. Pho:. P. J. Nordhagen.

PLATE IV

P. J. Nordhagen, The moscics of John VII

PLATE V

Pope John VII (Cat. no. 2). Vatican Grottoes. The heed before the last restoration, Phot. Alinari.

4) Pope John VII (Cat. no. 2). Present state. Phot. Rev. Fabbrica di S. Pietro in Vaticano. b) Pope John VII (Cat. no. 2) The fragment as depicted in Grimaldi, MS Barb. lat. 2733, fol. 93. Phot. Biblioteca Apostol. Vaticana.

P. J. Nordhagen, The mosaics of John VII PLATE VI

4) The Virgin (Cat. no. 3). Duomo, Orte. Phot. Gabinetto Fotografico

Stato. 5) The Washing of ti

(Cat. no. 4). Vatican Grottoes, Phot. Reverenda

P. J. Nordhagen, Tie mosaics of John VII PLATE VII

P. J. Nordtagen, The mosaics of Tom VIT

PLATE VOI

The Washing ot the Chik: (Cat. no. 4). Christ, detail. Phot. P. J. Nordhagen.

P. J. Nordhagen, The mosaics of John VII

PLATE IX

The Wasning of the Child (Caf. no. 4). Salome, détail. Pto:. P. J, Nordhagen.

P. J. Nordhager, The mosaics of John ἘΠῚ

FLATE X

The Adoration (Cat. no. 2). S. Maria in Cosmedin, Rome. Phot. Archivio Fotografico delle Gallerie e dei Musei Vaticani.

P. J. Nordhagen, The mosaics of John VII

The Adoration (Cat. no. 5). Joseph, detail. Phot. P. J. Nordhagen.

PLATE XI

P. J. Nordhagen, Tie mosaics of John VII

The Adoration (Cat. no. 5). The Virgin detail. Phot. P. J. Nordhagen.

PLATE XII

F. J. Nordhagen, The mesaics oj John VII

The Adocetion (Cat. no. 5). The angel, detail. Phot. P. J. Nordhzgen.

PLATE ΧΠῚ

P. J. Nordhagen, The mosaics of John VII

PLATE XIV

Christ (Cat. no. 6). Vatican Grotfoes, Phot. Reverenda Fabbrica di S, Pietro,

P. J. Nordhagen, The mosaics of John VII

Christ (Cat. rc. 6). The head, de-ail, Phot, P. J. Nordhazer.

BLATE XV

P. J. Nordhager, The mosaics f John VII

PLATE XVI

a) The Virgin and Longinus (Cat. no. 7). The Virgi 1, Setail, Phot. Anderson. δ) St. Peter (Cat. no. 8). Vatican Grot-ces. Phot. P. J. Nordhagen. ¢) The Virg= and Longinus (Cat. ro. 7). cula del Mosaico, Vatican, State before the last restoration. Phot. Andersor.

J. Nordhagen, The mosaiss of John VII

PLATE XVII

‘The Virgin and Longinus (Cat. no. 7). Longinus, detail, Phot. P. J. Nordtagen.

PLATE XVIII

P. J. Nordhagen, The mosaics of John VIL

'oouo104 'opeuorzex, vooporta “Iota “ELI-ITI-II "ZON “INE “SOURIOLT SK 'tprvaro ut For-For ‘19; Suwenq ΠΛ tof jo osea 'op£o Ist) SUL

PLATE XIX

P. J. Nordhagen, The mosaics of John VII

cona qoasody ONE 3oud “ona *

ραν ‘eumary ow 6b “oy Supe TEA wiof po Aaore10 “ILA WHOL pue use PMO (2 “soUEIOLL

P. J. Nordhagen, The mosaics of John VII

PLATE XX

a) The Nativity, in the Christ cycle, Oratory of John VII. Drawing fol. 45 in the «Albums, Arch. Bas.S, Pietro. Phot. Biblioteca Apostol, Vaticana. δ) The Raising of Lazarus, the Entry into Jerusalem and the Last Supper, in the Christ cycle, Oratory of John VII. Drawing fol. 42 in the « Albums, Arch. Das, S. Pietro, Phot. Biblioteca Apostol. Vaticana.

P. J. Nordhagen, The mosaics of John VII

PLATE XXI

5 a) The Virgin and Longinus (Cat. no. 7). The fragment as depicted fol. zr in MS Cod. Barb. lat, 440. Paot Biblioteca Apostol. Vaticana. 3) Salome at the Crib, fragment do-merly in the Vatican Grottoes. Watercolour fol. 12 in MS Cod. Barb. lat. 4410. Pho=. Biblioteca Aposto Vaticana. e) The Virgin and Longinus (Cat. no. 7). The fragment as it appeared at the beginning of tbe nineteerth century. From Dionisio, d) The Virgin and Longinus (Cat. no. 7). The fragment after 1840. e) The crucified Christ, fragrent formerly in the Biblioteca Angelica, Ro:

P. J. Nordhagen, The mosaies of John VII

PLATE XXI

LIS SIYE RERUM

4) The fragments with St. Peter (Cat. no. 8) and John VII (Cat. no. 2) at the beginning of the nineteenth sentucy. From Dionisio, 1848. 3) The Virgin, mosaic panel from the Ciborium of the Sacred Sudzrium in the Old St. Peter's. Vatican Grottoes. Phot. Anderson.

P. J. Nordhagen, Tie mosaics of John VII

PLATE XXIII

Head of the angel Dynamis, from the cst mosaics in the church of the Dormition, Nicaea. From Schmit, Koimesishirche.

P. J. Nordhagen, The mosates of John VII

PLATE XXIV

St. Demetrius between donors, mosaic. Church cf Hagios Demetrios, Thessaloaiki. Phot. P. J. Nordhagen.

P. J. Nordhagen, The mosaics of John VII

PLATE XXV

SMa nese ia re

St. Demetrius between donors. Head cf donor, detail. Phot. Hj. Torp.

P. J. Nordhagen, The mosaics of John VII

Angel, mosaic fragment. Church of Kagios Nicolaos, Istanbul.

PLATE XXVI

Phct. F. J. Nordhagen.

Notes on the dates of some early Giant Bibles KNUT

BERG

E. B. Garrison's important studies on the development of Central Italian initial styles and on Umbro-Roman painting during the late eleventh and the twelfth centuries will long remain the fundamental work for the study of Central Italian art of this period (1). Realizing the prime importance which must be given to dated monuments, he has carefully scrutinized all the information available concerning the dating of individual monuments from the period with which he deals (). In this work he has been able to dispel many erroneous traditions concerning the dating of certain monuments and at the same time has brought

forth valuable new material for

the dating of others. Some of the conclusions at which Garrison has arrived may, however, be

open to question.

I

The S. Cecilia Bible in Rome, Bibl. Vaticana, Barb. lat. 587, a key monument for the study of the development of Roman Romanesque painting, has traditionally on the basis of certain internal evidence been dated to 1097 or shortly before, and was believed to have been made in the Roman church of S. Cecilia (3). In his study of the manuscript Garrison has clearly demonstrated that the pages on which the information concerning the date and the origin occur are inserts and do not belong to the original manuscript (). Between the Canon Tables and the beginning of the Gospel of Matthew, five leaves, fols. 304-

308, have been inserted, insert I (5). The inserted leaves contain St. Isidore’s Allegoriae sacrae (1) E. B. Garrison, Studies in the History of Mediaeval Italian Painting, I-IV, Florence 1953-62. (Henceforth referred to as «Garrison, Studies0). (2) E. B. Garrison, Notes on the History of Certain Twelfth-Century Manuscripts of Importance for the History of Painting I, «La Bibliofilia», LIV, 1952, PP. 1-34 (henceforth referred to as «Garrison, Notes I5) and II, Garrison, Studies, I, pp. 10-17. (3) H. M. Bannister, Monumenti vaticani di paleografia musicale latina, Leizig 1913, p. 139 no. 145; H. Quentin, Memoire sur l'établissement du teste de la Vulgate, « Collectanea. biblica latinas, VI, Rome 1922, pp. 364 fL; P. Toesca, Storia dell'arte italiana, Il mediosvo, Turin 1927, p. 1053; idem, Miniature romane dei secoli XI ὁ XII, « Rivista del R. Istituto d'archeologia e storia dell'arte», 1, 1929, pp. 74-76: A. Boeckler; Abendlandische Miniaturen, Borlin-Leip-

E

167

zig 1930, pp. 688; G. Ladner, Die italienische Malerei im ii. Jahrhundert, « Jahrbuch der kunsthi-

storischen Sammlungenin Wien », N.F. V, 1931, pp. 518; N. Gabrielli, La Bibbia atlantica della Biblioteca Beriana di Genova, « Academie e Biblioteche d'Itaia », VI, 1932-33, pp. 46f.; L. Coletti, I Primitivi, I, Novara 1941, p. XV; F. Hermanin, L'arte in Roma dal sec. VIII al XIV, Bologna 1945, P. 393. (4) Garrison, Notes I, pp. 11-30. (5) As it is of no importance for the dating of the Bible, we leave aside the first eight folios with Jerome's letter to Pope Damasus and the beginning of Genesis. These are, as has been pointed out by earlier. writers, a later replacement, probably from the beginning of the thirteenth century, and made as a copy of the original. This part contains a full-page Genesisillustration and an illustration of S. Cecilia be-

KNUT BERG

scripturae, Ex Veteri Testamenti and Ex Novo Testamente (), with the rubricated titles Libellus Typicam gerens Figuram Sci Hysidori Veteris Testamenti (fol. 304), which open with an initial D (PI. ΠῚ e), and Item Libellus Typicae Sci Hysidori in Evangelio (fol. 305). These are followed by some shorter texts taken from St. Isidore's De etymologia:

De clericis, De monachis (®), De ceteris

fidelibus (), (lib. 7, cap. 12-14), De ecclesia et synagoga, De religione et fide (lib. 8, cap. 1-2), and De canonibus conciliorum (lib. 6, cap. 16) (+). These texts are written continuously and by the same scribe, ending in the middle of the left column on fol. 308. In the margin on fol. 307, outside the beginning of De monachis, is written in minium: ANN.DALXCVILIND.VALJUL. framed by a simple wavy line (PL IV o). In the right column on fol. 308 is a short text, Aileranus, Carmen in Eusebii canones (), filling about one third of the column. The script of this text is contemporary with that of the preceding one, but written by a different scribe. On fol. 308v is a long list concerning the deposition of relics and the dedications of altars in the church of S. Cecilia in Rome. It begins with the consecration of the church by Pope Paschal Τ in 822 and the deposition of the relics of S. Cécilia and other saints under the altar of the church. Thereafter follow notices concerning the deposition of relics in the altar of the Saviour on the azst of January, 1055; in the altar of the Virgin on the 2sth of May, τοι; in the altar of St. Andrew on the goth of August, 1073; in the altar of St. John the Baptist on the 3rd of January, 1072; in the altar of S. Cecilia on the roth of September, 1073; and the dedication of a new altar to S. Cecilia on the 3rd of June, 1080.

The list ends with a notice

concerning a silver statue of S. Cecilia made in the'same year. Immediately below is written in minium capitals: CHRISTI SACERDOTI VENIAM CONCEDE JOHANNI (ἢ. The whole list is written continuously and in a script very close to that of the main text of the insert, but by a different hand. As a continuation of this list, a notice dated the 24th of February, 1098, concerning the deposition of relics in the altar of 8. Marmenia, has been added by another hand and this again is followed by an undated notice concerning the altar of St, James the Apostle by yet another hand (). These notices show that the insert was written in the church of S. Cecilia in Rome. The original list, which is clearly a retrospective record of the events it records, must have been drawn up after 1080, the latest date recorded, but before 1098 when the notice about the altar

of S. Marmenia was added. A more precise dating of the insert is probably given by the date 1097 written in the margin on fol. 307. There is no apparent motivation for the recording of this date, but since it is most likely written by the scribe who wrote the main text of the insert, and since it fits well into the date limits for the addition of the list about the altars on fol. 308v, there is good reason to accept the date 1007 as the year in which the insert was written.

Between Epistola Sci Petri II and Epistola Sci Johannis I, a single leaf, fol. 361, has been

inserted, insert II. It contains prologues and arguments to the Acts of the Apostles and to each

tween SS. Tiburzio and Valeriano (Ladner, op. ci figs. 20, 21). This last illustration thus proves nothing beyond the fact that the Bible at that time was in tho possession of S. Cecilia. (1) Migne, P. L., LXXXIII, cols. 97-130. (2) Wrongly rubricated as by St. Jerome. (3) Correctly rubricated as by St. Isidore. (4) Migne, P. L. LXXXII, cols. 2438, 290-206. (5) E. Dümler, Kanon Evangeliorum, « Anzeiger

für Kunde der deutschen Vorzeit», N. F. XXVI, 1879, cols. 84ff.; D. de Bruyne, Préface de la Bible Latine, Namur 1920, p. 185. (6) Garrison, Notes I, fig. 4. (7) A copy of this inscription with minor variations must have been in the archives of S. Cecilia. ‘This has been reprinted by J. Laderchi, Caecilia Virg. et Mart. Acta ef Transtyborina Basilica, IL, Rome 1723, PP. 10-15. 168

KNUT BERG

serts; they can also be found in the very body of the Bible itself. In the margin on fol. 283, where Book I of the Maccabees begins is written St. Jerome's prologue to this Book, and on fol. 370v, on the page preceding insert III, after the prologues and arguments to the Pauline Epistles, is written Pope Damasus’ poem De S. Paolo Apostolo ; a text often used as a prologue to the Pauline Epistles (). The script of both these additions is contemporary with and of the same characteristic type as that of the inserts. Furthermore, short inscriptions can be found in the margins of fols. 26v, 118 and 192, all framed by the same wavy line as the date in the margin of fol. 307, and written in the same type of script as the inserts. On fol. 301 is written in rather large capitals in minium: JOHANNES PRESBYTER VIVAT IN CHRISTO (9), and below, there is a long inscription, entirely scratched out except for the frame, which again is of the same type as that on fol. 307. Throughout the manuscript, innumerable smaller or larger corrections have been made in the text of the Bible. At times the original text has been erased and replaced by a new one, at times by a word written in between the lines or in the margin. -Most of these corrections have been written in the characteristic minuscola romanesca. The condition of the parchment after the larger erasures makes it difficult to judge the script, but in some cases, as those on fols, 215v, 216, 218, 372, 373, 381v, and 389v, we meet a script very similar to that of the inserts. The script of these notices and corrections in the body of the Bible itself shows that they must be about contemporary with and made in the same scriptorium as the inserts. In a later study of the S. Cecilia Bible, where he deals with its initials, Garrison reports a few of these additions in the body of the Bible (*). He recognizes their script to be of the same type, and possibly by the same hand, as that written by the principal scribe of the inserts. Garrison believes the Bible to have been made in S. Cecilia in the decade rrrs-25 (see below), and in his opinion these additions and corrections indicate only that the scribe of the inserts must still have been alive at that time. In this connection he also proposes the theory that the Bible was planned from the beginning to receive the inserts. If this were the case it certainly would have been possible to have arranged for the inserts to be incorporated into the Bible in a better

and more harmonious way; and would it not, especially concerning inserts II and III, have been much more natural to have incorporated their text into the Bible proper ? The evidence discussed above shows that if the inserts were to be considered fragments of one or several other and earlier manuscripts, it would, as we have attempted to demonstrate, be necessary to find plausible explanations for a number of strange coincidences. If, on the other hand, one assumes the inserts to have been made for the Bible, there are no problems involved. In addition to this, what one might call external evidence, there is also what one might

call

intemal evidence which indicates that the inserts were made for the Bible. The text of insert II bears every sign of having been arranged for the purpose of being inserted into the S. Cecilia Bible. The Canonical Epistles of the Bible have no individual prologues and arguments; these are provided by insert IZ, The text of the insert also includes a prologue and an argument to the Acts of the Apostles, although the Acts already has a prologue in the

text proper, but this is different from that of the insert. In many other Bibles we find these prologues and arguments included in the text, but it would be very unlikely to find the prologues and arguments to both the Acts and the Canonical Epistles collected in one place. I have

not found any examples of such procedure in Italian Bibles from the eleventh and twelfth cen-

turies. And it would be a strange coincidence if they were all written on the same leaf. (1) A. Ferrua, Epigrammata Damasiana, Rome 1942, p. 816.

(2) Garrison, Notes, fig. 1. (3) Garrison, Studies, I, p. 58, n. 3. 170

NOTES ON SOME EARLY GIANT BIBLES

Also the index rerum of insert III can best be explained as having been made for the S. Cecilia Bible, The same motivation must have existed for the addition of the prologues that have been written into the body of the Bible as for those that were written in insert ZI.

Also the poem

by

Aileranus in insert I, which is a very rare text, is used as a prologue to the Gospels (), These added prologues show that the clergy of S. Cecilia desired commentaries on the scriptures to be

added to the Bible. This explains the reason behind the addition of the excerpts from St. Isidore in insert I. Since these too, are commentaries on the scriptures they can be regarded as intended for the same purposes as the added prologues, though they have never been used as such in the text proper of any known Bibles from the period. As the excerpts from St. Isidore concern both the Old and the New Testament, they have been placed between them. There is therefore no difference between the nature of the texts that have been written into

the Bible proper and those that have been added on the inserted leaves. It seems, thus, very likely that all the additions were planned together and made at approximately the same time, On the basis on this evidence, both the external and the internal, it seems safe to conclude that the inserts were made for the Bible, and that their date, 1097, gives a terminus ante quem for its dating. The style of its early-geometrical initials and its illustrations, as well as its script, shows that the Bible cannot have been made long before the inserts, The Bible can therefore best be dated to shortly before 1097. During the latter part of the eleventh century, both the script and the initial style of the scriptorium of S. Cecilia are well documented. In addition to the inserts of the Bible, three illuminated manuscripts have been preserved which by original inscriptions can be shown to have been made in S. Cecilia: a St. Remigus, Commentary on fhe Pauline Epistles, dated x067, in Oxford, Bodleian Library, Add. D. 104 (*); a Gradual, dated 1071, previously in the Phillipps Collection, MS. 16064, later in the possession of the London bookseller,

W. H. Robinson, Ltd. (*);

and an Evangeliary in Florence, Bibl. Laurenziana, Plut. 17.27. (‘). The last manuscript is w dated, but it can on stylistic grounds be shown to be a few years younger than the S. Cecilia Bible (see below).

All these manuscripts are written in the same characteristic script as the inserts

of the S. Cecilia Bible, a script that as we have seen, differs markedly from that of the Bible proper. They are also all decorated with the same type of stylistically closely related vegetal initials, which again is markedly different from the initials of the Bible (PL. ΠῚ d). ‘The Bible has been written by several scribes who all use the same type of script. It is difficult to imagine that the scriptorium of S. Cecilia maintained two teams of scribes each writ-

ing in its own type of script, and that the one team would be employed to make additions, notes, and corrections in a manuscript written shortly before by the other team.

In the same way it is

difficult to believe that this scriptorium had artists who worked in entirely different initial styles. It is, therefore, fairly certain that the Bible cannot have been written in the scriptorium of S. Cecilia.

Both Toesca (5) and Garrison (*) have pointed out the intimate stylistic relationship be-

(1) De Bruyne,

op. cit., p. 185.

tions

of three

initials; Garrison,

Studies,

I, p. 58,

(2) Garrison, Studies, ΤΙ, p. 81, fig. 8x Ὁ. 4. (3) J. Hourlier and M. Huglo, Un important (4) Toesca, Miniature romane, p. 76; Garrison, Moin du chant « views-romain », « Revue Grego- — Notes I, p. 23%. rienne » XXXI, 1952, pp. 26-37. W. H. Robin(5) Toesca, loc. cit. son, Lid, Catalogue 83, Rare Books and Manu(6) Garrison, Notes I, pp. ast; and Studies, III, soripts, London 1953, pp. 59-63, with reproduc- pp. 17ff. i7

KNUT BERG

tween the illustrations in the main body of the Bible (PI. II a, ὃ, c) and the portraits of the Evangelists in the Evangeliary, Bibl. Laurenziana, Plut. 17. 27, mentioned above (PI. ΠῚ a, D). This relationship is so close that it cannot be explained merely as the result of one work having influenced the other. In the technique of the execution, in the colours used, and in the handwriting of the execution — seen particularly in the rendering of the facial features -- such a complete correspondance exists between the two works that there can be no doubt they were made in the same workshop. Toesca thought the illustrations in the Bible to be the earlier of the two works, while in Garrison’s opinion the illustrations of the Bible ‘can best be explained as derived from those of the Evangelary, the illustrators best understood as having received instruction from the illustrator of the Evangelary’ (). Garrison seems to have arrived at his opinion mainly because he finds the illustrations of the Bible to be of a lesser quality than those of the Evangeliary (ἢ). Both works are among the finest of the preserved Roman illuminated manuscripts from this period. While the three illustrations in the Evangeliary are by the same hand, several hands must have participated in illustrating the Bible. Garrison distinguishes between at least five different artists (9. Even if most of the illustrations in the Bible are of an inferior quality compared to those in the Evangeliary, the difference is not so noticeable when one considers the illustrations of highest quality in, the Bible. ΤΕ we compare the illustrations in the two manuscripts, we find in both a style where the play of folds in the drapery has been transformed into an ornamental system of abstract geometrical forms. In the illustrations of the Bible the ornamental play of the folds is conceived as a surface pattern. The surface is covered by a web of lines that in free-flowing rhythmic combi nations make up the design of the drapery, and there is only a superficial correspondance between the drapery and the body underneath. In the Evangeliary illustrations,the interplay of geometrical forms is more firmly organized in order to create a better correspondance between drapery and body. An attempt can also be felt at a more plastic rendering of the figure. Also in other respects can be seen how the illustrations in the Evangeliary constitute a refinement and maturation of tendencies apparent in the Bible. When seen in connection with the general development of Roman painting during the late eleventh and early twelfth centuries, as it can be followed in works like the frescoes in the lower church of S. Clemente, those in Castel S, Elia andin S. Pietro in Tuscania (4), it becomes evident that the illustrations in the Bible are older than those in the Evangeliary. There is no question of the Bible illustrations in their style copying or imitating that of those in the Evan-

geliary. On the contrary, a clear stylistic development can be traced from the Bible to the Evangeliary. Though it is impossible to ascertain, it seems highly probable that the artist who lustrated the Evangeliary participated in the decoration of the Bible at an earlier periodin his career.

(1) Garrison, Studies, III, p. 17. (2) On the basis of this close relationship between the illustrations in the main body of the Bible and those in the Evangeliary, Garrison attributes the Bible to S. Cecilia. He dates the Evangeliary to 1110-20 and the Bible to 1115-25. The marked difference between the script of tho Evangeliary and that of the Bible is due, he assumes, to a change in the type of script used in that scriptorium during

‘the interval between the making of the Evangeliary and the Bible.

(3) Garrison, Studies, III, p. 19. (4) The frescoes in S. Clemente, reproduced, Garrison, Studies, ΤΙ, figs, 189-192, those in Castel S. Elia, idem, ΠῚ, figs. 39, those in S. Pietro in Tuscania, idem, II, figs, 23, 24 and C. A. Isermoyre, Die mittelateriichen Malereien. der Kirche S. Pietro in Tuscania; « Rómisches Jahrbuch für Kunstgoschichte ν, IL, 1938, figs. 255-258, 263, 266, 268. 172

NOTES ON SOME EARLY GIANT BIBLES

The Bible and the Evangeliary thus offer an interesting example of two manuscripts produced in different scriptoria, but illustrated by either the same artist or by artists belonging to the same workshop, Any explanation of how this was brought about must largely be guesswork, but it does not seem unreasonable to assume that the priest Johannes who is mentioned in inscriptions in both works may have been responsible for selecting the same artists. On fol. 301 in the main body of the Bible is the following inscription: Johannes presbiter vivat in Christo. The inscription is an addition, probably made when the manuscript came to S. Cecilia. Tt does not tell anything concerning his connection with the Bible, but we may suppose that his name was inscribed in the Bible because he had donated it to the church, On fol. 308v in insert I is written: Christi sacerdoti veniam concede Johanni. As this inscription first of all must be seen in relation to the added list about the dedications of altars in the church of S. Cecilia, it does not permit us to draw any conclusions concerning his connection with the making of the insert. The Johannes of this inscription is called sacerdos, and it is questionable if he can be the same Johannes who is called presbiter in the other inscription, In the aforementioned list, the notice about the statue of S. Cecilia relates that it was made: Temporibus domni. Johannis archisacerdos, Johannis presbiter, Nicolai presbiter, Benedicti presbiter, Franconis presbiter, Petri diaconi, Nycolai diaconi, Silvestri subdiaconi, Johannis subdiaconi, Gregori subdiaconi, Romanii clerici, Petri clerici. This is obviously a list of the clergy of S. Cecilia at that time, i.e., 1080, and they are probably listed according to rank and seniority. The Johannes sacerdos mentioned in the inscription is very likely to be identified with the Johannes archisac-

erdos, who was then the head of the church, The collophon of the Evangeliary, fol. 150, Accipito parvum munus Caecilia sanctum, quod tibi Johannes devote presbiter offert…, clearly tells that the manuscript was donated to the church by a Johannes presbyter. It does not say that he wrote it, but this is a possibility which cannot be excluded. The Gradual dated 1071; previously in the Phillips Collection, has a collophon which tells that it was written by a Johannes presbiter (1), and there is a strong possibility that he is the same one mentioned in the other inscriptions.

Even though Johannes was one of the most common names of the period, as is amply de-

monstrated by the list of the clergy in S. Cecilia, it does not seem unreasonable to suppose that

the Johannes presbyter who may have acquired and donated the Bible to the church, and the Johannes presbyter who shortly afterwards commissioned and perhaps wrote the Evangeliary are one and the same. He may also possibly be identical with the Johannes presbyter who in 1072 wrote the Gradual. The most likely person with whom he might be identified among those mentioned in the list of the clergy of the chuirch, is the senior priest. At the time the Bible was bought and the Evangeliary written, he may very well have been the head of the scriptorium and in charge of the church library. If this were the case, it is easy to imagine that Johannes, who clearly intended the Evangeliary to be a luxury manuscript by Roman standards, commissionedan artist from the workshop in which the Bible had been decorated to make the portraits of the Evangelists. The reason for this may be that although the scriptorium of S. Cecilia had artists well trained in making initials, it had none who were capable of making illustrations of a quality suitable to the standard he intended for the Evangeliary. The initials were therefore made in S. Cecilia, while the illustrations admire in the Bible.

were

made

by one of the artists

(1) Hic tempore domni Alexamdri papas scriptus neumatusque est por Johannes presbiterum indic.

whose

work Johannes

-had come

to

VIII antiphonarius iste. Hourlier and Huglo, op. cit, p. 33. 173

KNUT BERG I

Another important illustrated Giant Bible for which it is possible to establish a reliable terminus ante quem is the one in Munich, Staatsbibl, Clm. 13001. This Bible has inscriptions placed in the margin at the beginning of every Book of the Bible, saying that it was a gift to the Monastery of St. Aurelius in Hirsau by Henry IV. The inscriptions read: HemRIcus ΠῚ REX DEDIT HUNC LIBRUM S. AURELIO, or abbreviated, Hemricus IMI Rex D.H.L.S.A. (Pl. I). The script of the inscriptions is of a much finer quality than that of the Bible, The inscriptions are probably by a German scribe, and made at the time when the mianuscript was given to Hirsau. Swarzenski assumed that the gift was made to Hirsau around 1075 in connection with the refoundation of the monastery and Henry's confirmation in that year of its possessions and freedom (1). This date has been accepted by both Toesca (ἢ and Ladner ('). Garrison has, however, argued against this early dating of the Bible, He considers that the gift could have been made at any time during Henry’s reign, as he assumes the relationship between Henry and the monastery to have been excellent throughout his reign, In support of this assumption he states that Hirsau sided with the Emperor in his struggles with the Pope over Lay Investiture.

Henry

abdicated in 1103 and died in 1106, In Garrison's opinion these inscriptions only give a firm terminus ante quem of 1105/06 for the making of the Bible (ἡ. He was therefore free to date the Bible to around 1100, the period to which in his opinion it stylistically belongs.

Such a late dating of the Bible is certainly impossible. As Henry in all the inscriptions is called King, the gift must have been made before he was crowned Emperor in 1084. On the

basis of historical consideration there is, however, every reason to accept the date, 1075, pro-

posed by Swarzenski as the most plausible moment for the gift to have been made. The history of the relationship between Hirsau and Henry IV shows that the only period

in which it would be natural to assume Henry to have made a gift to Hirsau was around 1075.

In that year Henry, as said, confirmed the possessions and the freedom of the monastery, and it is likely that the Abbct of Hirsau, Wilhelm, in order to obtain that confirmation courted the friendship of the King. With the excommunication of the King in 1076, the struggle between the King and the Pope broke out, and Hirsau did not, as Garrison says, side with the King in this war. On the contrary, Abbot Wilhelm was one of the leaders of the Papal party in Germany, and Hirsau soon became a center of opposition to the King. As early as 1077, we hear that the anti-king Rudolf visited Hirsau, and the Papal legate stayed there for a long time in 1077-78 (). Hirsau became the center for the reform movement in Germany, and we never hear anything about a reconciliation between the King and the monastery. It is therefore highly unlikely that Henry after his wars with the Pope had begun, would have made a gift to Hirsau (*). (1) G. Swarzenski, Die Regensburger Buchmalerei, Leipzig 1901, pp. 175f. (2) Toesca, Miniature romane, p. 81. In 1927 Toesca (Storia, p. 1054) wrongly placed the reign of Henry IV as from 1111 to 1133, which was repeated

Reformbewegung in Deutschland, « Studi Gregoriana v, II, 1947, p. 13. (6) As the Bible did not come to the Staatsbibl. in Munich from Hirsau, but from the Stadtbibl. in Ratisbon, Swarzenski (op. cit, p. 176) tentatively

(5) A. Brackmann, Gregor VII and die Hirchliche

im Ratisbon and loft it behind there when he mov-

by Hermanin in 1945 (op. cit, p. 393). (3) Ladner, op. cit, pp. 58% (@) Garrison, Studies, 1, p. 12.

suggested that thetheBible never that arrivedAbbotat Wilhelm Hirsan, and he mentioned possibility received the gift while he stayed at St. Emmeram

174

KNUT BERG

‘This inscription consists of three notices, the first (line 1-3), records the death of the Abbess Gerlenda on the roth of May 1078; the second (line 4-14), records the dedication of three altars in the crypt of a church dedicated to S. Ponziano by an unknown bishop on the 19th of June 1078; and the third (line 14-19 col. b 1-4), lists the dedication of three altars in the church itself by Bishop Rudolfo of Spoleto on the 28th of January, 108 (?). Garrison has convincingly demonstrated that the convent in question must be that of S. Ponziano outside Spoleto (1). Unfortunately, very little has been published about the history of this convent (*). The inscription mentions the altars que subfer sunt and que sursum sunt. This would indicate that the church in question had a crypt which was dedicated first. The church of the Monastery of S. Ponziano outside Spoleto has a crypt (). Garrison considers the whole inscription to have been written by the same scribe and at the same time. Moreover, he finds that the scribe who wrote the inscription is the same as the one who wrote the Bible itself, and that it was probably written together with the text of the Bible. In his opinion the inscription must be a retrospective record of the events it mentions, According to Garrison, the last mentioned date in the inscription, 108 (9), therefore furnishes a terminus post quem for the dating of the Bible (4).

‘A closer comparison of the script of this inscription with that of the text of the Bible will, however, clearly demonstrate that they cannot have been written by the same hand. Small differences can be observed in the ductus of nearly every letter. The differences are most pronounced in the letters a, 1, and g. But it is not only between the script of the inscription as a whole and that of the Bible proper that differences can be noticed. Differences may also be discerned between each of the three notices of which the inscription consists. First of all there is a slight difference in the colour of the ink, and also in the character of the script, which can best be seen by comparing the second and third notices, Furthermore, there are differences in the wording and in the spell ing (gue in notice two, guae in notice three), inconsistencies that would be unlikely to occur if the whole inscription had been written at the same time and by same hand. These facts show that the various notices of the inscription were written at different periods. ‘They are thus likely to be contemporary records of the events they describe. Furthermore the difference between the script of the Bible proper and that of the notices shows that they must be later additions to the Bible. Neither do the notices present any evidence for assuming that the Bible itself was produced in the Convent of 8, Ponziano, although this is a possibility which cannot be excluded. As the script of the notices and that of the Bible proper must be about contemporary, not many years can have passed between the writing of the Bible and the addition of these notices. On the basis of the notices this Bible can therefore be dated to shortly before 1078, to the bottom of the page, the end of the lines have been cut away and from five to eight letters are missing in each line. We cannot know that the ensis at the beginning of the thirteenth line is the last part of the word Veronensis. (1) Garrison, Studies, I, pp. τοί. (2) Kehr, Ital. Pont. IV, p. 16; G. A. Rota, Spoleto ὁ il suo Territorio, Spoleto 1920, p. 127.

(3) On the architecture of the church, cf. G. Martelli, L'abbaziale di S. Felice di Giano e un gruppo di chiese vomaniche intorno a Spoleto, « Palladio +, N. S., VII, 1957, p. 82. The inscription is also an important document for the dating of this romanesque church, which, however, has been greatly modified in more recent times. (4) Garrison, Studies, I, pp. x6.

176

PLATE I

K. Berg, Notes on some early Giant Bibles

= fmt mute onere peru TER Gromo, nee i, aprumsim rapentuperammen Seaream: deserunmqu mtricl msae fa? not, mobecte

PTUS

Nru sign. alam Dae gandieboc txprewanceinches nto. elucsmar xorer pere τελεία fuel. i Nehnramem era fana. RU Perea ag genere ame Re Battasb orevroriquemfugaus cercare ceelfincu udine cor qu polline facerdooum lure Pcerloraleteu, Isuurmanclaucor cano diit abengerir ges: simmiftenoflo: ponb-confiice banane

portum gh snfinaftro bal cumeape

En E TE romano: ccbelia epi ieroni pleno attodero Maran o ino: dem cc on mitra: Ext geni uzbrum chulcico formone, conferspaim. aclleonum fhiumtrafum! Libnmiung, robe quombebper decachi-

m

rae n cai meo Feuso:ar Argumde cim (UN EO Tuum cinoneWn mLennl au itabneve ferre» Senelu ufc vacan e plane "Sum vichetochrpbcevire-epstyporam, unes ιν.. Dm

ἰβεταιίεκιρότι

è chatcleonambngua er

men peneomit. rachiteaure spes opere: Deme am oclo uror: quarterobo fecero a. ἡ ceenform volar fusfebas Sram cperieebiracieim alee plume: cei adorabar dine?”

fua: echece mua e Ormapre r fear ra manfiaFilef

ne. ped me eclerenegrnibel gore /

mee

docu

‘erabfhnere-aboms

È ananasor ou

München. StaatsbibL, Clm. 13001, fol. 84%.

PLATE Il

K. Berg, Notes on some early Giant Bibles

|

va

Pais@

a

gumar- uacnf «fincm

(agora imaginari: opum

&angrioy

‘que fübal |

rr pexaonc ali

enr breurc deforaza. cormaxi derer: ur pira a

amp apra led loribufyedderenaf que: qancxplica.

anffé.bifton Libri guara a regum ec fedi parabpo

- menon dica |o

Achag uero dereliceodo y doli

famane ecfine quib; ena filoffuof infacnficiit 14m. concremandof obrulerar coluif fe;con flax.

Rome Biblioteca Vaticana, Barb, lat. 587. a) fel 173. δ) fol. 358". c) fol 264. d) fol. 304. e) fol. 186,

K. Berg, Notes on some early Giant Bibles

PLATE IIT

RATER

Florence, Biblioteca Laurenziana, Plut 17.27. a) fol. 6. δ) fol. 49. c) Rome, Biblioteca Vaticana, Barb. lat. 587, fol. 304. d) Florence, Biblioteca Laurenziana, Plut. 17.27, fol. 77°.

PLATE IV

K. Berg, Notes on some early Giant Bibles

a Serene qui to Ce uem Brani.aio

ie

Ee ee, ;

p

ΡΣ

ij

pie eni

ἢ mpl Der vod :

i

Vo gat