Lucian and the Authorship of "De Saltatione" Anderson, Graham Greek, Roman and Byzantine Studies; Fall 1977; 18, 3; Periodicals Archive Online pg. 275 Lucian and the Authorship of De Saltatione Graham Anderson THERE has never been any single cogent argument for denying the authenticity of De Saltatione, generally ascribed to Lucian. But the treatise is clearly eccentric in a number of ways. It is an encomium of pantomimic dance, a subject Lucian scarcely men­ tions in his canonically accepted works; it deals sympathetically with the stage performer, whom he usually despises; and its author has presented his material in a form grotesque by any standards: he digresses into a long catalogue of pantomime themes, and his dis­ cussion is at times completely disorganised.1 I wish to show that these and other eccentricities need not suggest that the author is an anony­ mous expert in dancing but are in fact typical of Lucian's methods when he sets out to improvise an erudite treatise on something he knows very little about. (To avoid circularity I shall refer to the author of De Salt. hereafter as 'the author', as opposed to 'Lucian'). There should be no problem about the motives which would have encouraged Lucian to write such a piece. He wrote the pair Imagines and Pro Imaginibus as compliments to Panthea, Lucius Verus' mistress prior to the latter's marriage with Lucilla in A.D. 164;2 the co-emperor was also interested in fidicinas et tibicines et histriones scurrasque 1 This led R. Helm to condemn the work (Lucian und Menipp [Leipzig 1906] 370), with the dubious argument that Lucian couid never cast himself as a Stoic (cf Salt. 2); but Helm later recanted, RE 13 (1927) 1759f.]. Bieler had already advanced detailed arguments against Lucian's authorship on linguistic grounds (Ueber die Echtheit des Lucianischen Schrift de Saltatione [Progr. Wilhelmshaven 1894]); these were easily refuted by D. S. Robertson (The Authorship and Date of Lucian's de Saltatione, in Essays and Studies presented to William Ridge­ way [Cambridge 1913] 180-85). M.Kokolakis, IIAATQN 11 (1959) 3-11, usefully expands the historical background to Robertson's article. For the technical background to pantomime, see H. Bier, De Saltu Pantomimorum (Diss. Bonn 1920), with the supplements by L. Robert, Hermes 65 (1930) 106-22, and Kokolakis llff. The parallels between Salt. and Libanius'inr£p TWV Op)('7CTWV are discussed by Helm, Lucian und Menipp 365ff, and]. Mesk, WS 30 (1908) 58-74; they are incancl usive for the relationship between Salt. and Aristides' lost KaT' 0PXTJCTWV (see n.6 infra), and are of course no help towards determining the authenticity of Salt. itself. 2 For the chronology see J. Schwartz, Biographie de Lucien de Samosate (Brussels 1965) 17. 275 276 LUCIAN AND THE AUTHORSHIP OF DE SALTATIONE mimarios (SHA Verus 8.11), and Lucian could have been expected to supply appropriate diversion in the same way.3 H. Richard thought the comments so exaggerated that the author of De Salt. must be parody­ ing the technique of encomiasts;4 but we should note that Lucian for his part could flatter with assurance, as in Nigrinus, Pro Imaginibus and Harmonides; and an encomium of pantomime would have been no challenge to a writer who could extract compliments even out of a grammatical solecism (Pro Lapsu 15f). Such a writer might of course be privately amused at his own volte-face, but that is not a matter for his audience; Lucian is an opportunist of this kind in his use of material, and it would be no surprise if he were to praise a dancer as XEtpoccxpoc in front of the enthusiast Verus without expecting the emperor to know that he criticises the use of the word elsewhere (Salt. 69: Rhetorum praec. 17, cf Lexiph. 14). Bompaire doubted Lucian's authorship because the interest in dancing is so obviously 'un-Hellenic' and therefore alien to Lucian's temperament and good taste.s But any encomiast has to take his patron's enthusiasms into account,6 and the author of De Salt. makes it perfectly clear in the introductory dialogue (§2) that he himself would be regarded as a very unlikely advocate for pantomime-as a man 7TCXtSEtff CVVTPcxpOC KCX~ rfotAoccxptff \, t, I Tex /LETptCX W/Ltll7JKWC. Lucian himself shows little interest in dancing of any kind in his canonically accepted works: he feels at liberty to mention the KOpSCX~ or the 7TVpptX1J, since both of them are 'sanctioned' by references in classic passages of Plato or Aristophanes; but for the contemporary 0PX1JcTOStSaCKcxAoc (De Merc.cond. 27)1 he has the utmost contempt. But he has no committed interest in historiography either, and yet he 3 Robertson, op.cit. (supra n.1), developing the view of W. Schmid, BursianJahresb. 108 (1901) 254. , Ueber die Lykinosdialoge des Lukian (Progr. Hamburg 1886) 36f. 5 Lucien ecrivain, imitation et creation (Paris 1958) 356f. 6 As Lucian himself admits elsewhere (Merc.cond. 35). For Lucian and Aelius Aristides see A. Boulanger, RevPhil 47 (1923) 150f. This sophist enjoyed considerable prestige with Marcus Aurelius (Philostr. VS 582f). It would certainly not be unreasonable for someone wishing to fraternize with Lucius Verus to compliment the co-emperor's frivolous tastes and challenge a particularly ponderous rival at the same time. Cf Kokolakis, op.cit. (supra n.1) 9. 7 KopSag: lear. 27, Ar. Nub. 540; 1TVPPLX'J: D.Deor. 8, PI. Legg. 8158. For the oPX'JcToS,McKMOC see Kokolakis, op.cit. (supra n.1) 11. Harmon (LCL V ppA02f) assumes that the Pseudologist must have been a pantomime actor (Pseudolog. 25), which Kokolakis denies, ibid. 47. Neither has suggested that his performance of the blinding of the Cyclops may itself be a panto­ mime (Pseudolog. 27). It could certainly be an amateur's improvisation in the style. GRAHAM ANDERSON 277 could improvise half a treatise of armchair advice when the Parthian War made the subject topica1.8 The author of De Salt. emphasises that he himself is a new convert to dancing, and an unlikely one at that (2f); and the treatment he adopts coincides in three broad respects wirh Lucian's inreresrs. De Salt. includes a review KaT' E8VT) ro explain how the dance arrived in Greece; Lucian applies the same technique to philosophy (Fugitivi 6-9) and religion (Iupp. Trag. 42). De Salt. also focuses attention on the solo performer, who figures frequently in Lucian's similes and anecdotes from the stage.9 And above all the author of De Salt. has found an entirely new angle from which to attack tragedy: instead of condemning a pageant of ranting actors, he has the opportunity to prove that even a single performer in a dumb­ show can do better! The last point alone would have been sufficient incentive to Lucian to go out of his way in this manner. But would a writer like Lucian with no real interest in dancing have been able to write the work? The author falls into Lucian's usual habit of taking shortcuts whenever there is any specialised research to be done: the aforementioned Pro Lapsu is also a classic example of how Lucian could twist a handful of hackneyed quotations and exempla into an impressive treatise on the most recondite trivia. Most of the material can be accounted for in school reading, and the rest can safely be inferred to be random information pressed into service, rather than the fruits of conscientious antiquarianism.Io The author of De Salt. proceeds in exactly the same way-a flourish of Homer and Hesiod is enough to prove that dancing has a respectable antiquity: Hesiod's Eros (Theog. 120: Salt. 7); the Muses dancing (Theog. 4: Salt. 23); Meriones the dancer (II. 16.617f: Salt. 8); dancers on the shield of Achilles, itself a 'textbook' of rhetorical ecphrasis (ll. 18.593-606: Salt. 13); Odysseus watching the danCing Phaeacians (Od. 8.256ff: Salt. ibid.); and dance as cXfLVfLWV (ll. 13.636ff: Salt. 23) or as the art of peace (II. 13.730f: Salt. ibid.). The author has made the most of all this, but his real standby, as he himself betrays in passing (34), is the intermittent discussion of danc­ ing in Plato's Laws (Books 6-7). While this is not a work which Lucian 8 See esp. G. Avenarius, Lukians Schrift zur Geschichtsschreibung (Meisenheim am Glan 1956) 178, who rightly concludes that Lucian has patched together a large number of secondhand platitudes about historiography in the second half of the treatise. 9 e.g. Adv.lnd. 8-10; Pisco 36; Apol. 5; Gall. 26; lear. 17; Neky. 16. 10 See my Lucian: Theme and Variation in the Second Sophistic (lvlnemosyne Supp!' 41, 1976) 13off. 278 LUCIAN AND THE AUTHORSHIP OF DE SALTA TIONE himself uses often,u the whole context of the discussion of dancing would be an important text for Plato's views on paideia and one which an author of Lucian's limitations might be expected to be able to draw on with no great effort: even Maximus of Tyre, who represents the lowest and most mechanical level of sophistic culture, resorts to the same nexus of passages for his discussion of dancing and music (see Hobein's parallels to Or. 37.4). The Laws would provide the author of De Salt. with the 1TVPPLXYJ and EILlu.AEia (814E-815A); he would have to find only the Kopoa~ and CLKLVIILC elsewhere, and the latter is mentioned together with the EILIL~A~La in a simple classification of dances associated with satyric drama, tragedy and comedy (Salt. 22). The author would then only have to know that the lmoPXYJILa is the characteristic dance of choral lyric (Salt. 16); the context suggests that he has included the OPILOC in his treatment of Spartan dancing, and he most probably recollected it from a source which dealt with Spartan training (12), like the YVILvo1TaLoLa which follows it. From Timaeus 40c he could have extracted the xopEia eXCTEpwv, which provides the beginnings of dancing in De Salt. 7. plato would also supply him with numerous references to the Corybants (e.g. Symp. 215E, Crito 54D, Euthyd. 277D). Harmon (ad Salt. 8) notes that he does not make the usual confusion of these dancers with the Curetes: it would be a reasonable guess that this is because he has extracted the latter direct from Hesiod once more (fr.23 Merkelbach-West, cf Theog. 477ff). And once we subtract this material directly available from a sophist's standard reading, we are left with only isolated ingredients. There is some bizarre paradox­ ography: Indians dance in honour of the sun; and Ethiopians do a war dance (Salt. 17, 18). We cannot safely speculate on the source in either case, but there is nothing to suggest that it was a treatise on dancing; we should bear in mind that Lucian himself certainly knew some exotic Indica,12 and Ethiopia was an equally magnetic locale for sophistic writers.13 The information in De Salt. is minimal here-as much, in fact, as Lucian knows about Indian and Ethiopian philoso­ phies in Fugitivi 7. Nor is the author of De Salt. better informed about 11 A few scattered reminiscences are colJected in W. H. Tackaberry, Lucian's Relation to Plato and the Post-Aristotelian Philosophers (Toronto Studies, philo/, seT. 9, 1930) 62-85. 12 For his use of Megasthenes and Ctesias see A. Stengel, De Luciani veris historiis (Berlin 1911), passim. 13 As suggested by the parallel in Heliodorus' bizarre novel (9.19), cited by Harmon ad loco GRAHAM ANDERSON 279 the Salii, and their perfunctory appearance was presumably prompted only by a Roman patron. The author may not even have needed to go as far as a handbook for the information that the 7TVPPLXYJ took its name from Pyrrhus,I4 while the mention of dancing in the mysteries is only a means of introducing a respectably 'Attic' institution, also available in Leges 815. As for Dionysus in India (Salt. 22), again we should note that Lucian himself wrote a prolalia on the subject (Bacchus 1-4). Beside this the equally absurd picture of Priapus teaching the young Ares the war dance will come as no surprise (Salt. 21); we should notice that Lucian in turn makes much of the tradition of an equally unlikely warrior, Pan (Bis accusatus 9-11; Dial. Deor. 22.3); and the Priapus tradition, otherwise unattested, is not likely to have been culled from a technical manual on dancing itself.1 5 The supposed inscription in honour of dancers in Thessaly (Salt. 14) suggests a 'Lucianic' technique: Lucian was given to improbable inscriptions in exotic places,I6 and he liked to try his hand at small stretches in non-Attic dialectY The greatest absurdity is the euhemerising of Proteus as a pantomimic dancer (Salt. 19), but that would again be no surprise if the author is Lucian, who makes the god draw a learned analogy between his own versatile person and an octopus (Dial.Mar. 4)! Apart from these sophistic flourishes there is very little left of the 'history' of dancing. The author pretends that there is a development to present-day perfection (34), just as Philostratus pretends that there is a direct connection between the first and second sophistic, despite the gap of so many centuries. IS It is also instructive to compare the sheer bulk of information with which the genuine antiquarian Athenaeus is able to inundate his reader (628c-631E): he sets out to supply the very kind o~ information which the author of De Salt. makes such a point of avoiding: the literature 7TEp't EVPYJfLaTW)) looms large in this impressive collection of names and founders of dances. I9 14 For the real explanation see K. Larre, De saltationiblls Graeeorllm (Giessen 1913, repro Berlin 1967) 28. 15 For the explanation of this unique but perfectly probable Priapic tradition see G. Kaibel, GottNachr 1901, 488ff. It is not necessary to assume (contra Roscher, Lex. 2972) that the author took the tradition from a collector of Bithynian cult-lore about the god; he might easily have known the tradition that Priapus is a son of Dionysus (Roscher 2971). 16 Cf Dips. 6; Ver.Hist. 1.7,2.28; Seytha 2. 17 e.g. Vit.Allet. 3-6, 13f; De Domo 20; Bompaire, op.cit. (sllpra n.5) 633. 18 Cf G. W. Bowersock, Greek Sophists in the Roman Empire (Oxford 1969) 9. 19 For these authorities see Larre, op.cit. (sllpra n.14) 1-17. 280 LUCIAN AND THE AUTHORSHIP OF DE SALTATIONE Again, however, Athenaeus has little more actual evidence at his disposal and concentrates on what he can supply about the dances of tragedy, comedy and satyr drama, or Spartan athletic dancing, from literary interests as much as from technical treatises (630c). Sophists who derive their information from this kind of source find their stocks of information running short before very long. It is therefore not surprising that the author of De Salt. should break off his 'history' of dancing to attack those who have written at length on the forms and history of pantomime, 7ToAvfLa{}{ac 7"av7"'Y]v E7Tlo€tgtv ~yoVfL€VOt 7Tapegetv: this is mere cf>tAonfLlav a7TetpoKaAov ... dlptfLaBfj ... aKatpOV (33). Having made these criticisms of the' experts' he proceeds to emulate them himself: each of these objections can be applied to his own account of the themes which the dancer has to memorise, spread over the next twenty-four paragraphs. This curious preamble confirms that the writer is an amateur, tilting at the professionals before he begins his own display: one notes that Lucian is equally ponderous when setting himself up as an adviser on historiography (Hist.conscr. 34-36). Since the author of De Salt. has been a convert to dancing, he has to overwhelm the audience with his command of the subject. All this amounts to is a list of (mainly tragic) themes (37-61), which are, we should note, a special interest of Lucian's. The author of De Salt. may not have had either time or means for adequate research into dancing; or he may not have been prepared to put the effort into a work which he might not be able to use again. He avoids naming dancers, possibly because he was neither sufficiently interested nor well enough versed in the subject to know them. Lucian for his part has the same limitation when naming actors ;20 and there are at least two apparent misunderstandings in the De Salt. account. 21 The author is safest when he is posing as a professor of repertoire. But how are his dancer's themes related to Lucian's stock in his canonically accepted works? The list of myths in De Salt. corresponds well with those used by Lucian: it is based chiefly on Homer, Hesiod and tragedy (61) with some unacknowledged additions from Herodo­ tus. Bompaire hastily dismisses them as "souvent peu connues" (supra n.l, p.582) and therefore un-Lucianic, though he admits else­ where (195 and n.4) that Lucian does in fact use recondite myths from 20 Bompaire, op.cit. (supra n.5) 437. 21 Kokolakis. op.cit. (supra n.l) 20f. GRAHAM ANDERSON 281 time to time. It is easy to lose sight of the fact that if Lucian were the author, he would know the myth of Boreas and Oreithuia (Salt. 40) from its prominent appearance in Plato, Phdr. 229c (in fact it also appears in an undoubtedly genuine work at Philopseudes 3); and he would also know of rhe rwo Erores (Salt. 38, cf 7; Hes. Theog. 116f) from Plato, Symp. 178B. Dido and Aeneas (46) might well be of slight significance to a second-century Greek, but if Lucian could mention Augustus in Pro Lapsu 18 as a concession to his Roman patron, he could also produce the most hackneyed of Roman traditions as a compli­ ment to Verus. This leaves only a small proportion of myths still outside Lucian's own usual repertoire (Nisus-Scylla, Salt. 41 ; Hypsipyle and Archemorus, 44; Medea's dream, 53). Harmon notes that the last of these is not in fact a normal part of the Medea legend, but duly compares Apollonius Rhodius, Arg. 3.616-82. It is not necessary to assume that there was a pantomime theme based on this, though the possibility is certainly not excluded; but it would be a passage which a sophist might be expected to know for other reasons as an exercise in CVYKptctC. Lucian does not use all the rest of the material which occurs in this section of De Salt., but he does exploit a cross-section of each 'block' of mythology sufficiently large to suggest that he would have been aware of the remainder.22 The author of De Salt. introduces an unusual element into the story of Polycrates, probably by confus­ ing the tyrant's brother with his daughter (54): Harmon insists that Lucian knew his Herodotus too well to make such a mistake (LCL V p.261), but the version of Polycrates' death at Charon 14 is not quite accurate,23 and in the same dialogue Lucian transposes Tellus of Athens and Cleobis and Biton in Solon's list of happy men (10). This is clearly the kind of material for which Lucian trusted to memory. The selection of myths from outside Greece reflects limitations very similar to Lucian's own: the Aetolian cycle (50) culminates in Heracles on Oeta (Peregrinus 21; Hermot. 7; Dial. Deor. 13); Italy (55) is repre­ sented by Phaethon and the poplars (De Electro; Dial. Deor. 25.2); the west (56) by Heracles' labours. ending with the oxen of Geryon (Heracles 2); Phoenicia (58) by Attis and Stratonice (Syr.D. 15, 17ff); 22 e.g. Salt. 49: Lucian is not interested in Pasiphae but mentions both Daedalus (lear. 2) and Talus (Phi/ops. 19); of the Argonauts he selects only Orpheus and Heracles (Fugit. 2) as well as the talking keel (Gall. 2). 23 S. Walz, Die gesehiehtliehen Kenntnisse Lukians (Diss. Tiibingen 1921) 46. 282 LUCIAN AND THE AUTHORSHIP OF DE SALTATIONE and Egypt (58) by Osiris (ibid. 7), Epaphus (cf Dial. Mar. 7.2) and the metamorphosis of the gods (De Sacrificiis 14). Lucian insists on the mystical character of Egyptian mythology (Deorum Concilium 11, cf Dial.Mar. 7.2); in De Salt. the dancer is to use more symbolism in presenting it (-rpac (Pseudolog. 15). 26 Bompaire. op.cit. (supra n.5) 583 and n.2. GRAHAM ANDERSON 283 barbarian misinterpreting a civilized specracle27 (Anach. Iff, wrestlers; ibid. 23, actors). Another barbarian imagines that a dancer with five roles must also have five souls (66); Lucian gives the cock a series of comic transmigrations (Gallus 19f). The author of De Salt. applies the classic simile of the cuttlefish to the dancer (67); Lucian applies an elaborate example to Proteus (Dial. Mar. 4.3). The dancer can repre­ sent the three parts of the Platonic soul, and the dancer's silence represents the silence of Pythagoras (70). Caster28 objected that symbolic interpretation would be "tout a fait etrangere a l' esprit de Lucien." In fact this would be in favour gf Lucian's authorship rather than against it: the author is careful to quote such a theory at second­ hand (ijKOVca 8E TLVOC Kat 7TEpLTTOT€pOV TL VEaVLWOfLEVOV). But all is fair in the a8ogov: in Muscae Encomium Lucian makes ridiculous connec­ tions between the same two philosophers and the fly-its reviving 'proves' the immortality of the soul (7), and Muia was associated with Pythagoras (ll)! The author of De Salt. enlarges on blood and dust in wrestling (71) as Lucian does in Anacharsis 2f; he combines Heracles and Aphrodite (73), as Lucina combines Heracles and a cinaedus (Adv.Ind. 23) or Heracles and Omphale (Hist.conscr. 10); and the dancer's physique is held up to the Polyclitan canon, as is Peregrinus' in De Morte Peregrini 9. The final anecdote (83f) is about a man who overacted the madness of Ajax. This very theme is the basis in Lucian's acknowledged works for a number of favourite anecdotal motifs: the mad performer CHist.conscr. 1) who forgets his proper role and runs amok (Piscator 36), and whose misconducr nearly causes a grave injury to his supporting actor Odysseus (cf Pseudologista 27). Hirzel (Der Dialog II 285) connects De Salt. as a whole with Anacharsis since both are conversations with a Cynic and the approach is Peri­ patetic; but this is entirely superficial. In Anacharsis the 'Cynic'­ Anacharsis himself-goes away at the end unconvinced ;29 the didactic element is much less prominent and is cleverly used for humorous ends; and in De Salt. there is none of the "freie BI lithe und Kraft des Lebens" which Hirzel saw in Anacharsis. A more fruitful comparison can be made between De Salt. and Quomodo historia conscribenda sit: dancer and historian alike must present their events clearly (62: 27 The story itself need not be the author's; cf Philostr. VA 5.8: the Gadeirans thInk that Nero won his Olympic victory by literally conquering the Olympians! 28 Lucien et la pensee reiigieuse de son temps (Paris 1937) 43 n.97. 29 Bompaire, op.cit. (supra n.5) 356 n.l; 678. 284 LUCIAN AND THE AUTHORSHIP OF DE SALTATIONE Hist.conscr. 43); accommodate to a succession of different situations (65-67: Hist. 49); combine mental and practical disciplines (69: Hist. 37); create a vivid realism in descriptions (63: Hist. 51); and maintain correct proportion (75: Hist. 16, 23). Both may combine pleasure with profit (71 : Hist. 9) and the effect of eye and ear (78: Hist. 46f). Of course such headings are applied to vastly different circumstances, and the writer of De Salt. leaves out no absurdity: the dancer is shown as having solecisms (80), K(xKO~TJ>..t(X (82), and knowledge of past, present and future (36) as readily as the bogus historians. It is quite clear that 'Lucianic' treatises on medicine or lyre-playing would be as easy to produce from the same set of headings as these two; and they would look just as uncomfortably alike. Scholars analysing De Salt. have tried to explain it only in terms of textbook schemes. There is general agreement on the break at ch.35:3o Bompaire (supra n.5, p.28! n.4) sees the work as two distinct and self-contained encomia (of dance, 7-34; of dancers, 35-84); others use the formula which Norden applied to Horace's Ars Poetica: an dcaywy~ under the headings Ars and Artifex.31 But this classification is not entirely satisfactory even for the Ars Poetica itseIf,32 and it is still less so when applied to De Salt.; once the labels are attached, what happens to the catalogue of myths? As dancer's repertoire it is Ars, but the writer has introduced it by a tenuous connexion with memory (36), so that it also falls under the heading of Artifex. The writer indeed has already announced that he is going to speak about the dancer (35) but does not reach any conventional discussion of his faults or merits until 62; 37-61 is a plain catalogue quite distinct from the material on either side and has to be seen as a single unit. The general layout is then familiar: 1- 4 erato is shocked to hear about Lycinus' enthusi- Introduction I asm for dancing, 5- 6 but agrees to hear his defence. II 7-25 'Historical' instances of dance. Catalogue Ia 26-32 Its role compared with that of tragedy and Ib comedy. (33-34) A defence of the writer's omissions. 30 Th. Sinko, Eos 14 (1908) 134; RE (supra n.l) 1759; L. Miiller, Eos 32 (1930) 569-71. 31 Hermes 40 (1905) 481-528. 32 For criricism of Norden and summary of later developmems in the search for a 'scheme' for the Ars, see C. O. Brink, Horace on Poetry, Prolegomena to the Literary Epistles (Cambridge 1963) 20-40. (35-36) 37-61 37-48 49-59 62-84 85 GRAHAM ANDERSON The cultural value of dancing. The dancer's repertoire, by regions of Greece, and foreign parts. Other aspects of the dancer: the merits of a good one, and the faults of the rest. Concluding frame: Crato's 'conversion'. 285 lIa lIb (IlIa III b) Epilogue The arrangement just described is in accord with the schemes of many of Lucian's canonical works. I have described its operations in detail elsewhere.33 Here it will be sufficient to note that dialogues as diverse as Navigium or Prometheus start out with a two-panel intro­ duction followed by three more extensive panels dealing with the main theme; so does Timon, before Lucian tags on an ending which mimics the closing scenes of an Old Comedy.34 In De Salt. also there are three marked divisions for the main subject; and the author has also introduced a characteristically 'Lucianic' element of contrast between the middle section and the other two panels. In De Salt., however, the third section is particularly confused. The writer has set out to discover the dancer's qualities and faults; but he announces the topic rather belatedly, as he himself admits (74): 'EfUAw DE TfDTJ Kat f ~ -I:, , .... \ , ('..... \ 1" ,,, " ,,, V7TOOELsat COt TlfJ /\OYlfJ, 07TOLOV XPTJ Etvat TOV aptCTOv 0PXTJCTTJV EV TE tPvxil Kat CWILaTt. KaLTOt TijC /LEV tPvxijc 7TPOEt7ToV T