¹ Kanzler, R., Avori della Bibl. Vaticana, Roma, 1903Google Scholar, Tav. I. (front and reft profile views, arm from the outer side), merely mentioned in the text, No. 1.

² Memorie romane di antichità e belle arti, i., sez. 2, p. 8 ff.; in the estate of Monte Calvo on the Rieti road, at the 33rd milestone from Rome, among the ruins of a ‘nobilissima villa,’ together with several sculptures in marble.

³ F. 152: Mask of ivory and fragment of arm, representing Pallas, found in an excavation made about 1825 in the Sabina and sold for 50 scudi to the Government by the dealer Capranesi.

⁴ Archivio del Camerhngato, iv. 2570 (1836): Note of the antiquities purchased by the Chancellery of the Holy Roman Church, and now in the rooms called Zelada in the Holy Vatican Palace, by formal delivery to Mons. the Prefect of the Holy Apostolic Palaces:—

Objects from Capranesi.

154 A head and a fragment of an arm, in ivory, of exquisite workmanship.

⁵ The order to transfer the ‘Etruscan’ antiquities of the Library to the new Museum, preserved in the archives, f. 238, is dated 23 Jan. 1837. It seems, however, that only a selection of the vases and better specimens was made.

⁶ In 1843 they were not exhibited, for Braun's Guide (‘Ruinen und Museen’), issued in that year, does not mention them, although it lays stress on the small ivories always preserved in the same case (p. 836)

⁷ Vol. i. 1913, p. 258 sq.

⁸ Pauly-Wissowa, Real-Encyclop. ii. col.2366. The strangeness of the fact is the more remarkable, if we remember that the cases of the Museo Profano were always open to students and, for example, Furtwängler knew the collection of Roman phalerae in onyx, which is kept in the same case (cf. Antike Gemmen, iii. p. 366, 1).

⁹ Total height, mm. 142; greatest breadth, ca. 125; height of face, 125; breadth at the eyes, 117; nose, height from wings to root, 48; eyes, cas. 30; mouth, 37; chin, from the lower lip, 37.

¹⁰ Height in the middle of the forehead, 14 mm.

¹¹ In an archaic acrolith of the Vatican Museum the eyebrows are in lead (cf. Helbig, Führer³, n. 400). In another colossal figure of Roman date in the Vienna Museum the surface of the cavities shows traces of bronze patina (cf. Oenterr. Jahresh. 1908, p. 175). For eyebrows worked separately in bronze technique cf. Furtwängler, , Intermezzi, p. 4Google Scholar(Olympia, iv. Bronzen, p. 14).

¹² The pieces must have been finished together on a small vice, according to the technical processes expounded by Quatremère de Quincy, , Jupiter Olympien, p. 418Google Scholar, sqq. In this sense we must understand the συνρυθμίζειν of written tradition; cf. Pauly-Wissowa, ii. 2364.

¹³ Published here by kind permission of the Director, Prof. Roberto Paribeni. It is a left foot (length, ca. mm. 120, about half natural size) and seems to be a woman's. With it was found the little child's band, which suggests that we are concerned with a group of Aphrodite with Eros. The styliza-tion of the fingers, as well as the scrolls which ornament the upper leather of the closed sandal, recalls fairly closely the Camillus of the Palazzo dei Conservatori, or the Atticizing style, perhaps of the Pasitelic, perhaps of the Hadrianic, period.—For the literary sources on sculpture in ebony, cf. Pauly-Wissowa, v. 1994 (M. C. P. Schmidt).

¹⁴ The superiority of ivory to marble as plastic material, in regard to hardness and compactness, has been acutely observed by Quatremère de Quincy, l.c. pp. 394, 395.

¹⁵ Helbig³, no. 70.

¹⁶ Heibig³, no. 1820; Hanser, , Oesterr. Jahrpsh. 1913, pp. 57sqq.Google Scholar The affinity with the Parthenon sculptures asserted by Klein is merely generic. The drapery of the chiton especially shows notable differences from the fine stuffs of the pediment figures. Comparison with the Weber head inclines us to accept the date proposed by Amelung and Kalkmaan.

¹⁷ Brunn-Bruckmann, , Denkmäler, Taf. 362Google Scholar; Winter, Kunstgesch. in Bilder, 2nd ed., reproduced without the restorations. For comparison I have used the cast in the Gabinetto archeologico of the University of Rome.

¹⁸ Brunn-Bruckmann, Taf. 194; Smith, , Marbles of the Parthemm, Pl. 36, p. 54Google Scholar. It is known that it was chiefly on this example that Matz based his observations in order to prove the date of the Suppliant; cf. Klein, , Gesch. d. Gr. Kunst, ii. p. 119Google Scholar.

¹⁹ Heibig³, 1336.

²⁰ Oesterr. Jahresh. 1908, pp. 169 sqq.

²¹ L.c.; Heibig³, 1367.

²² Oesterr. Jahresh. l.c. 175 seq.

²³ Ibid. pp. 196 ff., Taf. 11.

²⁴ Cf. Furtwängler, , Meisterwerke, pp. 1 ff.Google Scholar Admitted on the whole by Amelung in the Index to Helbig³, ii. p. 530. It is to be observed that he places among the personal works the head of Aphrodite (‘Sappho’ type), which in view of the elongated oval of the face, a peculiarity constant in all the known copies, is rather to be associated with the works of the ‘School,’ such as the ‘Ceres’ of the Rotunda of the Vatican, and similar works. The similarity which Arndt would, for this. reason, find in it with the Parthenos (Brunn-Bruckmann, on Taf. 576) is in plain contra diction with all that he notes about the copies of the latter. See below, note 39.

²⁵ The Athena of Brescia and especially the Bologna head. Amelung, Jahresh. l.c. and Schrader, ibid.; Winter, l.c. We shall see in its place what account is to be taken of the view of Preyss, , Jahrb. des Inst. 1912, pp. 102 ff.Google Scholar, who persists in the hypothesis of Furtwängler. The Albani herm, Brunn-Br. 632, is quite akin to the Bologna head as well as to that of the Heraion, and the opinion of Arndt (ad loc.), that we may have to do with the Amazon of Pheidias, does not deserve discussion after all that has been written about the already known examples

²⁶ The Demeter, Kekule v. Stradowitz, Winckelmann-Progr. No. 57. For the Kore Albani which has already been excluded from the Pheidian cycle by Furtwängler, M. W. p. 100, cf. Helbig³, 1922.

²⁷ ii. p. 125.

²⁸ Ibid. 1094.

²⁹ l.c. Alinari 34, 922–24. The ‘industrial’ character of this double example is seen above all in the clumsy coupling, which is especially disagreeable in the profile, where the faces have very slight relief. The Wörlitz copy (Einzelverkauf, 381, facing, bending forward, a very bad photograph), so far as I can judge from photographs, shows the face differently stylized, in a more elongated oval.

³⁰ Helbig³, n. 1922.

³¹ We may say that on this depends the revision of almost all the results of the study of ancient sculpture. It is necessary to this end to study again, using purely experimental criteria, all the Roman material, seeking to re-group it and to distinguish the marble workshops, with minute and exact knowledge of the technique. The typology of supports and puntelli will be of the first importance (up till now we have nothing but the deplorable article by Maviglia, Ada, Röm. Mitt. 1912, pp. 1 ff.Google Scholar, in which the study of supports is reduced to a hermeneutic exercise on S. Reinach's Répertoire). The well-known memoir by Furtwängler (Abhandl. d. K. Bayer. Alead.) is only a ‘prolegomenon,’ put together generally out of his personal opinions. Thus the alleged ‘mechanicity’ of the copies from the time of Pasiteles onwards, which he would deduce from the passage of Pliny (where 21.4. probably refers to the ‘proplasmata’ of the original works) and from some cases of points of measurement found on ancient statues (of these, too, we are not told whether they are taken from originals or copies) is belied by the numerous variations which we find even in a single type. No weight can be allowed to the rule established by Reinach, Salomon (C. R. de' Acad. des Inser. 1900, p. 535 ff.Google Scholar; cp. Pollak, L., Oesterr. Jahresh. 1901, p. 145)Google Scholar, because if two identical ancient copies exist, their identity is not enough to prove that the original was a bronze, and so much the less that the copy was made on the basis of a cast; instead of which we meet with the fact that when we can prove that the original was a bronze, the copies do not differ less than those taken from sculptures in other material. It is an excellent plan in Furtwängler to compare the copies with Roman works of fixed date (op. cit. p. 26). We have, for example, a series of statues which range from the age of Augustus (e.g. the so-called Germanicus of the Louvre) down to the third century of the Empire; to the latter period belong the two colossal statues of the Palazzo Odescalchi (Matz-Duhn, Ant. Bildwerke in Rom, Nos. 1028, 1036. Cf. P.B.S.R. viii. pp. 75 f. The head of the older man does not seem to me to be modern, as has been asserted; the type is very like the portraits of Maximinus; for the other see Einzelverkauf, Nos. 2058–60); the ‘hunter’ and the group of ‘Mars and Venus’ in the Salone of the Museo Capitolino, of identical technique, the headdresses and masculine types about 250–270 after Christ. In these statues we have portrait heads, of which the date is ascertainable, placed on Greek ideal bodies, from which much light may be shed on the copies and variations of classical works. Mariani's latest observations (Atti e memorie dell' Accademia di S. Luca, 1913–11, pp. 35 ff.) make greater confusion of our ideas, perpetrating an utter travesty of the conception of ‘copy’ as the tradition of the motives of Greek art. At any rate, since we are working to-day on copies as on originals, it is well to insist on this preliminary inquisition. Only when we have acquired some notion, even though only approximate, of what really remains to us, shall we be able to attempt to restore what we have lost.

³² Helbig,³ ii. p. 395.

³³ For the copies of the Parthenos cf. Athen. Mitt. 1908, p. 128 (Pagenstecher).

³⁴ This fact weakens considerably the assertion of Schreiber (cf. Arch. Zeitung, 1883, p. 306) that the provenance is of no importance in deciding the value of copies.

³⁵ The isolated heads which are connected with the tradition of the Parthenos have to be judged in a different way, because the loss of the figures considerably restricts the range of comparison. Cf. Oesterr. Jahresh. 1901, p. 146, n. 2 (L. Poliak), and Monuments Piot, vii. pp. 167 ff. (Michon). In fact it has not been proved that they all belong to true ‘copies’ of the Parthenos simply because they have a helmet more or less like hers. The identity cannot be confirmed except in regard to complete figures. With the first group of copies must be connected above all the Kaufmann head (Denkmäler d. Inslit. i. Taf. 3, p. 1, Fränkel), together with the Riccardi head (Amelung, Antiken in Florenz, No. 204); while the Jacobsen head (Pollak, l.c.; much better are the reproductions from the cast in Studniczka, Kalamis, Taf. 12. 4; 14, 4; cf. Anielung, , Oesterr. Jahresh. l.c. p. 175Google Scholar: too badly preserved to allow of a critical opinion), the Sacchetti head (Mon. Piot, l.c.), and the Dresden fragment (Arch. Anz. 1898, p. 53, Fig. 24, Hermann; cf. Jahrb. 1899, p. 143) are different in character and value, and represent the other tradition. An intermediate type is provided by the Cologne head (Bonner Studien, Taf. iv., the best reproduction, from a cast), in which the modelling of the forehead and the eyes, the only portion preserved, is rather that of a Hellenistic sculpture. The others (Acropolis, Pollak, l.c. Fig. 173; Verona; Aquileia) need not even be mentioned, as we need not discuss the head published by Pollak (ibid. Fig. 172) and at that time in the market.

³⁶ Cf. Kabbadias, Ἐθνικὸν Μουσεον, No. 65, Kastriotris Catal. No. 129. For comparison I have used the cast in the Gabinetto of the University of Rome and Alinari's photographs 24215–16 n.

³⁷ Bibliography: Pagenstecher, l.c.; Alinari, 24219.

³⁸ Brunn-Bruckmann, 512 (“Minerve au Collier.”)

³⁹ Loewy, , Scoltura Greca, pp. 43 ff. Figs. 82–90.Google Scholar Helbig,³ n. 1304. Anderson, 2039–40. Only the nose is re-worked.

⁴⁰ Brunn-Bruckmann, 511. Jahrb. d. Inst. 1912, Beilage 3, Abb. 9, (d). Anderson, 17102.

⁴¹ Cf. Schreiber, l.c. p. 19.

⁴² Jones, Stuart, Cotal. of the Capitoline Mus. i. p. 293, Pl. 72Google Scholar: ‘polished but not worked over’ (the date ca. 470 B.c. proposed must be regarded merely as a slip). The re marks which are made in the text of the Einzelverkauf, ii. p. 35, are devoid of critical value. There, in addition to the alleged reworking of the anterior side, it is noted that the head is a poor repetition (‘in schlechter Wiederholung’) of the Apollo of the Tiber, and that the attribution to Pheidias made by Petersen was based on the resemblance to the Athena Ludovisi. Furtwängler's determination of characteristics of the style of Calamis in a pretended original (Meisterw. p. 381) is pure phantasy.

⁴³ Cf. Amelung, Helbig³, ii. p. 125 I cannot understand why Loewy, (Scoltura Greca, p. 47)Google Scholar finds it so difficult to establish the originality of Pheidias, and limits his work, at lea3t for the Parthenon, to the summing up and completing of the cycle of archaic art. If we had nothing more of Pheidian style left to us than the fragment of ivory which is the subject of this paper, yet the soul of the new epoch, in contrast to the types immediately preceding it, would be revealed to us; as, for instance, in Duccio's panels and Giotto's frescoes are revealed the beginnings of great Italian painting as distinct from the Byzantine manner. The type of the Pheidian Athena is an innovation of this kind, in virtue of the divine ideal which dawns in its countenance, holding the same rank in Greek art as the Madonna of Santa Maria Novella or the Maestà of Duccio holds with regard to older images. The traditionalism of the motive, which seems to carry such weight in the judgment of the Austrian scholar, is common to the great Greek innovators as it is to the Italian ‘primitives,’ and does not limit, but rather enhances, the creative rôle which they play in the development of the type.

⁴⁴ Furtwängler, , Masterpieces, p. 60Google Scholar; Ruesch, , Guida, p. 36Google Scholar, Fig. 10. Mariani's view, that the herm may be derived from a statue without aegis, does not seem to me to be acceptable, since the attribute would have been hardly visible owing to its small size.

⁴⁵ Jones, Stuart, Catal. p. 240, 54Google Scholar; the form of the face is more triangular and shorter than in the Naples copy.

⁴⁶ Jahrb. d. Inst. 1912, pp. 88 ff.

⁴⁷ Masterpieces, p. 13.

⁴⁸ Furtwängler, l.c. p. 90. The relations which Mahler would see (cf. Klein, , Gesch. d. griech. Kunst, i. p. 408)Google Scholar with archaic works are absolutely imaginary.

⁴⁹ Cf. Jahrb. l.c. pp. 102 ff.

⁵⁰ After Amelung's acute observations (Oesterr. Jahresh. 1908, pp. 194 f.), which confirmed the doubts advanced by Gardner and Reisch, and after Schrader's most felicitous comparison (ibid. 1911, pp. 61 f.), it does not seem to be possible to reopen the question as long as our ‘Pheidian’ material remains what it is. The last author's conclusions were accepted by Winter (Kunstgesch. in Bildern, 2nd ed.). I think by this I have shown how little substance there is in the comparisons adduced by Preyss.

⁵¹ The attribution to the young Pheidias of the relief of Eleusis (bibliography in Helbig³, 1922) is without secure foundation, and the comparison with the figures of the Parthenon frieze shows generic affinities combined with strong stylistic differences. The scheme of the composition is somewhat primitive in comparison with the sculptures mentioned. It lacks, in fact, that continuity of action which we find expressed with so much mastery by the vase-painters as early as 450, and which consists in placing the middle figure to the front, with the head turned to one of the sides, so as to give harmony, variety and unity to the composition. I do not know whether Curtius's attribution of the Cassel Apollo is to be regarded as an exception to the general tendency, in so far as the dating of this sculpture was based on an inaccurate idea of the complete copy. (Cf. below, note 55.)

⁵² Cf. Klein, l.c. p. 409.

⁵³ Cf. Amelung in Helbig³, Nos. 912–973.

⁵⁴ For the chronology of Hegias see Klein, l.c. p. 375.

⁵⁵ The copies of the head that we possess vary greatly in the style of the face. From the more archaic type of the Barracco example (cf. Furtwängler, , Intermezzi, p. 6Google Scholar; Klein, l.c. p. 404) we pass by degrees to the more elegant and rounded copies of Cassel (Arch. Anz. 1914, p. 7, Fig. 3) and of Florence (Brunn-Bruckmann, 304) and to the others (quite unrecognizable) at Athens (Nat. Mus. No. 47, Alinari 24293) and in the Naples Museum (Ruesch, , Guida, p. 44, Fig. 13)Google Scholar. The short form of the face of the Florentine copy, which makes the eyes seem large and gives them a more powerful expression is singularly out of keeping with the type of all the others; the modelling is softened down to the point of disappearing altogether. The projection of the cheekbone in the profile view, the stiffness of the transitions, belong to the Roman academic style of the Augustan or Hadrianic age. The copy at Athens is distinguished by the roundness of the cheeks and chin, by the smoothness of the transitions, especially in the mouth and eye. I cannot compare the examples in the Jacobsen (Curtius, l.c.) and Ince Blundell Collections (Furtwängler, , Statuenkopien, p. 565 (41))Google Scholar. Obviously I cannot accept the opinion of Curtius (cf. M. Bieber, l.c. p. 7) as to the classification of the copies, seeing that, if the archaic characteristics of the Barracco example find their genuineness confirmed by original monuments, no one can suppose that it was the copyist who introduced them. The head in the Palazzo Vecchio seemed to Curtius to be the most beautiful; so be it, but this is only a subjective aesthetic impression. Fräulein Bieber would infer that the craftsman worked with his compasses on the original, which, to speak frankly, is a strange way of reasoning, the more so that supposing it to have been executed in the Augustan aga, or whenever the Neo-Attics worked, this mechanical method is more problematical than ever. In the profile there is close analogy between the Harmodios (Brunn-Bruckmann 328. 2, 1893) and the Apollo (Barracco example) in regard to the line from eye to chin. In the hero the lips, protruded by rapid breathing, mitigate the exaggerated expression caused by the projection of the chin. Very close is the modelling of the neck and of the cheek, which in the Apollo is broader; the position and height of the ear are identical. As to the affinities generally recognized between the head of Harmodios and other sculptures, these, except in the boy from the Acropolis (Dickins, , Catal. of the Acrop. Mus. p. 264, No. 698Google Scholar; Schrader, , Auswahl archaischer Marmorskulpt. Taf. XVI. –XVII. pp. 53 ff.Google Scholar) are sufficiently generic, and their only relation is chronological. So we may see an earlier phase of development in the Ludovisi head (Klein, l.c. p. 380; Helbig³, n. 1288), in the Vatican acrolith (see above, note 11) and in the ‘Boudeuse’ of the Acropolis (Catal. Nos. 241–4; Klein, i. p. 407). Contemporary, in all probability, is the Charioteer of Delphi (cf. Klein, p. 407); of which, for the rest, the Attie origin has been much disputed, and the attribution to Pythagoras proposed by Mahler and most recently maintained by Von Duhn, (Ausonia, viii. 1913 (1914), pp. 37 ff.Google Scholar) is very disputable.

⁵⁶ Arch. Anz. 1914, pp. 6–10, Figs. 1, 2, 4.

⁵⁷ If we study the Barraeco copy in relation to kindred archaic sculptures, these peculiarities, which produced in Furtwängler such enigmatic sensations of mysticism, come into clearer light as stylistic characteristics.

⁵⁸ Especially perceptible in the paintings by Brygos and Hieron.

⁵⁹ So in the polychrome cup by Euphronios (Hartwig, , Meisterschalen, Taf. 51, p, 484Google Scholar, cf. ibid. p. 490; ca. 480–70), in that of Hera in the Munich Museum (Furtwängler-Reichhold, , Griech. Vasenmalerei, Taf. 61, ii. p. 25Google Scholar, ca. 470–60).

⁶⁰ Griech. Vasenmal. Taf. 108, vol. ii. pp. 245 ff. Especially remarkable is the same manner of placing the cheekbone too high. This does not clash with the date (460) established by Hauser, in relation with the Olympian pediments, with a comparison which is of great generic value, in relation more to the scheme of composition and the motifs, than to the singularity of the types.

⁶¹ Cf. J.H.S. 1914, pp. 179 ff. The technique (drawing of the eyebrows, nuances in the painting of the eye) is certainly so much superior to that of the ordinary vase painter that the author must have had a true artistic training.

⁶² Furtwängler, Intermezzi, Taf. 2.

⁶³ Intermezzi, p. 11.

⁶⁴ The resemblance in style to the ‘Perseus’ (cf. Klein, l.c. p. 403) is quite generic; and that which has been asserted between it and the boy of Stephanos (ibid, p 406) is absurd. The ‘Iacchos’ of the Braccio Nuovo and the Athena Albani with the wolf-skin are, in the form of the face, fairly close to the ‘modernized’ copies of the Cassel Apollo; but we have no sure criterion to enable us to decide whether we have in them free copies or later phases of style.

⁶⁵ Cf. the Minoan Cretan statuette of the ‘Serpent Goddess’ carved in ivory with gold ornamentation, Amer. Journ. Arch. xix. 1915, Pll. XII.–XIV.; and the Diver from Cnossns, B.S.A. vii. Pll. II., III.

⁶⁶ Egyptian excavations are especially rich in small carvings in ivory. The stele of the sculptor lritesen (cf. Perrot et Chipiez, Hist. de l'Art, i. p. 839), where it is mentioned together with ebony, gives ground for believing that it was worked in pieces larger than are known hitherto. What has been said above as to Syrian ivory of prehistoric times (cf. Poulsen, F., Orient, u. frühgriech. Kunst, p. 37)Google Scholar does not detract much from the general value of the fact, the more so that we are completely in the dark as to the differences of quality which doubtless existed, as they exist now, and are recognized in commerce, between African and Indian ivory.

⁶⁷ Herod, i. 383; cf. Koldewey, , Die Tempel v. Babylon u. Borsippa, p. 39Google Scholar. The assertion of the priests that the statue was solid and weighed 800 talents (about 16 tons) is fabulous. For Egypt, a statue in stone (green basalt?) of the twentieth or twenty-first dynasty, recently acquired by the British Museum (Egyptian Sculptures in the British Museum, Pl 42) gives us au example of a head plated with chased gold.

⁶⁸ Paus. x. 4. 1.

⁶⁹ Cf. Koldewey, l.c. pp. 47 ff.: remains of polylithic statues, life-size, found near the Temple of Marduk, Esagila. N. 2, Fig. 78: several portions of eyes: corneas of shell or white stone, irises of two or three different materials. N. 3, Fig. 79: fragm ents of hair, beards, and eyebrows in lapis-lazuli, ninth to thirteenth century B.C.

⁷⁰ Cf. especially Amelung in Helbig³, ii. p. 159. The polychrome head of Athena is worked at the back, as is clear from a drawing in Amelung's possession, in a way analogous to the acroliths of Athens there noted.

⁷¹ Paus. ix. 4. 1. Overbeck, Schriftquellen, n 655.

⁷² In the experiments made with so great technical skill by Quatremère de Quincy, the head of life-size is composed of five pieces (l.c. p. 412, Pl. 29). It is noticeable that the type of the Vatican fragment has been spontaneously repeated in the most recent attempts at chryselephantine sculpture (reproduction of a Munich head made by Fr. Stuck, Klein, l.c. p. 411; bust of the Republic in the Musée du Luxembourg). The dimensions of our fragments recur with a certain frequency in Roman copies of sacred images. I note, for instance, that they are constant in examples of the Aphrodite in which Furtwängler wished to recognize the chryselephantine statue of Alcamenes called ἐν Κήποις (cf. Collignon, , Sculpt. Grecque, ii. p. 118)Google Scholar, and they occur in the Florentine copy (Milani, , Mus. Archeol. p. 329Google Scholar, No. 180) of the Artemis which Studniczka attributed to Menaichmos and Soidas. Including the beginning of the neck, the figure is 98 cm. high, so that the complete statue may be calculated at 120–125 cm., or a little more than two-thirds natural size. (The copy at Naples is a good deal smaller.) I do not think it is necessary to take much account of the tradition, preserved in the texts, of the softening of ivory (cf. Blümner, l.c.; for the negative result of modern attempts see Jacob in Daremberg et Saglio, ii. 1, p. 447). The method indicated by Plutarch, by means of a decoction or beer made of barley, seems to me absurd, no less than that of boiling for six hours with mandragora, of which Dioscorides speaks. Pausanias' text alone enables us to explain this fable, which is derived from the ignorance of technique in the popular mind, which supposed that it was possible to fuse elephants' tusks, as also oxen's horns κέρατα καὶ βοῶν καὶ ἐλεφἁντων . . .ὑπὸ πυρὸς κ.τ.λ.). Everything then is based on the uncertain reading of a passage of Seneca (Epist. 90), in which ‘polire,’ attributed to Democritus, may with probability be referred, in a generic way, to every other manipulation of the surface in regard to colouring or preservation.

⁷³ In regard to polychromy, the best object for comparison is the Kaufmann head (see above, note 35). It is to be noted that in the acroliths, in which the inserted eyes remain, there is always an accentuation of the chromatic separation of the cornea from the marble, in order to get the effect which Pheidias, for instance, had obtained with white stone (Plato, , Hippias maj. 290 B)Google Scholar. In the archaic Vatican acrolith (see above, note 11) the cornea is of opaline quartzite; in the Athens head, of bone (Stais, , Guide illustré, i. No. 177, p. 24)Google Scholar.

⁷⁴ ii. 13. The calculation, which is nothing more than approximate, is not affected by the alleged bronze sphinx, the existence of which has been maintained on the basis of a rather doubtful reading of a passage in Pliny. (Cf. Schreiber, l.c., pp. 66 seq.)

⁷⁵ The price of ivory is determined by a ratio between the weight and the volume, as is usual with precious organic or mineral substances (tortoiseshell, amber, pearl, precious stones). The thicker parts of the tusk have greater value, which is yet more increased in older specimens. In antiquity the value was certainly increased by the difficulty of transport and the multiplicity of middlemen.

⁷⁶ Cf. Overbeck, Schriftquellen, Nos. 331, 350, 407, 479, 515, 539, 634, 645 ff., 692, 755, 819, 847, 848, 853, 855, 932.

⁷⁷ Paus. i. 40, 4.

⁷⁸ Vita Apo'lod. v. 20. Blümner, l.c., thinks that he may refer to small figures; more probably, I think, to imitations in cheap material.

⁷⁸ Cf. Jahrb. d. Inst. 1913, p. 353 ff.

⁸⁰ vii. 27, 2.

⁸¹ The supposition of Klein, accepted by Furtwängler, , M. W. p. 57Google Scholar, that the statue of Pellene was regarded as earlier than the other Athenas by Pheidias, as being archaic and therefore not Pheidian, is purely arbitrary. Pausanias, as always, reports what he heard on the spot, and the traditions were not based on stylistic comparisons. We may have our doubts about the detail of its precedence in regard to the others, but we have no reason to suppose that the information was a complete invention. We get no additional light from the xoanon reproduced on the coins of Roman date (Loewy, , Studi ital. di filol. v. p. 33Google Scholar; Klein, , Gr. Kunstgesch. ii. p. 38Google Scholar; cf. Pausanias, , ed. Hitzig-Blümner, ii. 2, pp. 673 ff.)Google Scholar. It requires a great deal of good will to suppose that in Greece it was possible to confuse it with a work of Pheidias or of his school, while everything suggests that it was the more ancient and venerable image. If we had the coins of Athens as our only source for discovering the Parthenos, the case would b not very dissimilar. Equally arbitrary is Furtwängler's denial that the Athena of Plataea was by Pheidias, because if so it must have been executed about forty years after the dedication of the tithe of the spoil of the battle. The details furnished by the Periegete about technique make his report peculiarly worthy of attention; and, on the other hand, what do we know of the causes—and they are many—which may have delayed the execution of the image? The date 465 proposed by Loewy (l.c.) has no foundation.

⁸² Paus. vi. 26, 3; Pliny, , H.N. 35, 54Google Scholar. The identification with the Athena Medici proposed by Frickenhaus, (Jahrb. 1913, pp. 341 ff.)Google Scholar is not founded on any secure data. The colossal dimensions (about one-third of the Parthenos) make it very difficult to recognise in it the chryselephantine statue of a secondary temple.

⁸³ The question is in no degree answered by Dinsmoor, (Amer. Journ. Arch. 1913, pp. 70 ff.)Google Scholar, as de Ridder supposes (Rev. Ét. Gr. 1915, pp. 191 ff.). The dates of the Parthenon works give no certain support to the opinion of Furtwängler as opposed to that of Puchstein. Frickenhaus's résumé (l.c. pp. 342–352) merely shows the impossibility of extracting anything certain from the documents which are in our hands, and makes it clear that the dates which we have for the architectural work possess but secondary value for the plastic decoration.

⁸⁴ See Jordan, , Topogr. d. Stadt Rom, ii. pp. 572Google Scholar, 576 (dei eburnei LXXUII or signa eburnea deorum LXVII); cf. Michaelis, , Die archäolog. Entdeckungen des neunzehnten Jahrhunderts, p. 2Google Scholar.