^{page 11 note 2} See below, p. 15.

^{page 11 note 3} Diog. Laert. 5. 37.

^{page 11 note 4} Ib. 5. 58.

^{page 11 note 5} Ib. 60.

^{page 11 note 6} The actual development, and the part taken by the first Ptolemy and Demetrius, are not very clear. The question was opened by Ritschl, F., Opusc. 1, 5, 15Google Scholar, and has been debated often.

^{page 11 note 7} Diels, , Sitzungsber. d. pr. Ak. (1893), p. 106Google Scholar; Wellmann, , PW s.v. ‘Erasistratos’ (1907)Google Scholar, col. 334; Jaeger, W., Herm. xlviii, 1913, 62Google Scholar; Diohles von Karystos, 1938, p. 221 ff.

^{page 11 note 8} Hermippus in Suet. fr. 1 R., Satyrus often in Athenaeus.

^{page 11 note 9} Leo, F., Gr. Röm. Biog. 118Google Scholar.

^{page 12 note 1} There seems to have been a third stage in which the name lost its connexion even with Alexandria, and did not mean more than ‘grammarian’ or ‘literary critic’. This may be the explanation of the title Peripateticus Tubursicensis of the grammarian Nonius. I should have made that point in my article ‘Peripatos’ in PW, offprint 1936, col. 6; conclusion for the mainhistory of the Peripatos, ib., col. 37.

^{page 12 note 2} Rohde, E., Gr. Roman (1876), p. 99Google Scholar, n. 3.

^{page 12 note 3} Athen. xi. 477 c, Schneider ad fr. 109.

^{page 12 note 4} Knaack, PW, s.v. ‘Aratus’, col. 392 (but fanciful on Callimachus); Wilamowitz, , Hellenist. D. ii. 275Google Scholar.

^{page 12 note 5} Oxy. Pap. no. 1362 (vol. xi. 1915)Google Scholar; fr. 8 Pf., from the Aetia.

^{page 12 note 6} I quote the biographies after the edition of Maass, E. in Commentariorum in Aratum Reliquiae (1898)Google Scholar. The same scholar discovered Theon as the author of the biography mentioned in the text: Anal. Eratosthenica, 1883, pp. 38 ff. It is, however, uncertain whether this Theon is the fourth-century mathematician of Alexandria as Maass, following one Renaissance MS., maintained, or the grammarian under Augustus who specialized in commentaries on Hellenistic poets (K. Ziegler, PW, s.v. ‘Theon 15’, col. 2079).

^{page 12 note 7} The retranslation given by Maass, E., Aratea, p. 243Google Scholar, is convincing, and even Carystius for the name Gecraustius looks right, but cf. Wilamowitz, , Hellenist. D. ii. 275Google Scholar, n. 2.

^{page 13 note 1} The translation does not even make a complete sentence; on the rendering of propter cf. n. 3.

^{page 13 note 2} Below, p. 20. Praxiphanes T 5a = Callim.fr. 100 g Schn.

^{page 13 note 3} propter Praxiphanem Mytilenum is not a translation of ⋯ν τοȋς Пρòς П., as Susemihl pointed out (Alex. Lit. i. 287, n. 10); the translator throughout renders πρός or εἰς in titles by the rather inapposite preposition apud. Nor is adsistens ei ab infantia in any way similar to ὡς πρεσβυτέρου. If the words mean anything the original text might have had κατ⋯ Пραξιϕάνην Μυτ.; that would account at the same time for Μυτιληναȋον which does not form part of the Callimachean title. Beloch, , Gr. Gesch. iv. 2², 1927, 591Google Scholar, and Herter, PW, Supp. vol. v, 1931, s.v. ‘Kallim.’, col. 388, although right in rejecting Rohde's fantasia, were a little sanguine in declaring the Latin passage to square with the Vita Achillis.

^{page 13 note 4} Wilamowitz's, explanation (Hellenist. D. 212Google Scholar, n. 1) that this is merely inferred from Callimachus' praise of Aratus in Epig. 27, and that the man who makes compliments was taken to be younger than the man complimented is not exactly convincing. But he rightly infers that this fact also discredits the same statement purporting to come from Пρòς П.

^{page 13 note 5} Γένος Ἀράτου, § 2, p. 326 M. γηραɩῷ δ⋯ τῷ Κυρηναίῳ ⋯πεβάλετο, παρ' οὖ καì ⋯πɩγράμματος ἠξιώθη. This of course refers to Epig. 27, and here the reference is justified.

^{page 13 note 6} Theon, 3, p. 150 M. ⋯ Καλλίμαχος συνεγγίζων αὺτῷ κατ⋯ τοὺς χρόνους. Then there follows again a sensible reference to Epig. 27.

^{page 13 note 7} A particularly involved specimen is Couat's account, Alex. Poetry 50.

^{page 14 note 1} This has often been tried. Beloch even sought to refute Rohde's, hypothesis by reference to one of the other unestablished versions, Gr. Gesch. iv. 2², 1927, p. 591Google Scholar (but iv. 1², 1925, p. 488, he talked of Aratus as a pupil of Praxiphanes). Another cul-de-sac was to emend one of the divergent testimonies; so several times since Ritschl, , Die alexand. Bill. 1838, p. 88Google Scholar (Opusc. 1. 72) and Clinton, , Fast. ii. 7Google Scholar.

^{page 14 note 2} Rohde's hypothesis is to be found not only in older handbooks like Susemihl, , Alex. Lit. 1891, i. 287Google Scholar, n. 384, Knaack, PW, 1896, s.v. ‘Aratus’, col. 392, Lübker's, Reallex. 1914Google Scholar, s.v. ‘Praxiphanes’, but also recentlyin Christ-Schmid, , Gr. Lit. ii ⁶, 1920, pp. 80, 126, 163Google Scholar, in popular editions like the Loeb Callim. and Aratus 1921, pp. 2, 362, and was revived again in such recent Wilamonographs as Cahen's, E.Callimaque, 1929, pp. 26Google Scholar ff., 347 f., and A. Rostagni's articles to which I must return later. Against Cahen and Rostagni see Herter, H., Bursian's Jahresb. cclv, 1937, p. 84Google Scholar f.

^{page 14 note 3} In fact, on our evidence we can well believe what Theugenes of Ikos says to Callimachus at the banquet mentioned before: τρισμάκαρ, ἦ παύρων ⋯λβιός ⋯σσι μέτα, ναυτιλίης εἰ ν⋯ιν ἓχεις βίον (fr. 8. 33 Pf.).

^{page 14 note 4} So, after Bentley, Hecker, A., Comment. Callim. 1842, p. 68Google Scholar; Dilthey, K., De Callim. Cydippa, 1863, p. 18Google Scholar; and, after Rohde, Susemihl, , Alex. Lit. i, 1891, p. 145Google Scholar; Knaack, , PW, s.v. ‘Aratus’, 1896Google Scholar, col. 392; Crönert, W.Kolotes u. Men. 1906, p. 74Google Scholar, n. 355 a. Recently propounded again by Cahen and Rostagni, cf. above, n. 2.

^{page 14 note 5} Preller, L., De Praxiphane Peripatetico, 1842, pp. 7Google Scholar and 18 (= Ausgew. Aufsätze, pp. 97, 105). Both meanings of πρός are so usual in titles as to make examples unnecessary. Schmidt, F. (Die Pinakes d. Kallim. 1922, p. 102)Google Scholar, however, was too ready to assume the meaning against in certain titles without proof.

^{page 14 note 6} Schneider, , Callimachea, 1870, p. 351Google Scholar; Wilhelm, A., Öster. Jahresh. viii, 1905, p. 4Google Scholar; Wilamowitz, , Hellenist. D. i, 1924, p. 212Google Scholar; Beloch, , Gr. Gesch. iv. 2², 1927, p. 587Google Scholar, and others. H. Herter, PW, Supp. v, 1931, s.v. ‘Kallim.’ col. 403 was rather undecided but inclined towards the party, later he changed his mind, as he well might, and cautiously argued for the second party, Bursian's Jahresb. cclv, 1937, p. 110Google Scholar.

^{page 14 note 7} A.Rostagni, Studi It., N.S. ii, 1922 ‘Aristotele e Aristotelismo nella storia dell' estetica’;Google ScholarRiv. Fil., N.S. i, 1923Google Scholar, and ii, 1924 ‘Filodemo contro l'estetica classica’; iv, 1926, and v, 1927 ‘II dialogo Aristotelico Περί ποιητ⋯ν’; vi, 1928 ‘Nuovo Callimaco’; xi, 1933 ‘I nuovi frammenti di commento agli Αἴτια e la polemica letteraria di Callimaco’; xii, 1934 ‘Nuovi frammenti Callimachei nel contesto degli Αἴτιαφ and ‘Le nuove Διηγήσεις e l'ordinamento dei carmi di Call.’; there are also the valuable introductions to his editions of Aristotle's, Poetics, 1927Google Scholar, and Horace's, Ars Poetica, 1930Google Scholar.

^{page 15 note 1} First in his Poeti Alessandrini, 1916, p. 240, n. 18; cf. Riv. Fil. v, 1927, p. 172Google Scholar; vi, 1928, p. 21; ‘Arte Poet, di Orazio,’ 1930, p. xiii; Riv. Fil. xi, 1933, pp. 196Google Scholar ff.

^{page 15 note 2} Arist. Poet, chaps. 8, 18, 23 f. on the Cyclic Epic; Duris 76, FGH, F 83 Jac. (= Antimachus T1 Wyss) and Callim. ib. on Antimachus, also Epig. 28 on the Cyclic Epic and the wellworn fr. 74 b Schn.Λύδη καί παχù γράμμα καί οὐ τορόν. Whether τò μέγα βιβλίον ἴσον εἶναι τῷ μεγάλῳ κακῷ fr. 359 Schn. was used in any particular context is unknown. Cf. also below, p. 17, n. 3. From the one certain mention in Aristotle, of Antimachus, (Rhet. 3. 6)Google ScholarWyss, B. concludes that Aristotle was a modicus et cautus amator of the poet (Antim. Col. Frag. 1936, p. xlii)Google Scholar.

^{page 15 note 3} Pap. Soc. It. no. 1219, publ. by Norsa, and Vitelli, , Bulletin de la Société archéol. d'Alexandrie, xxviii, 1933, p. 123Google Scholar, also in Pap. Soc. It. (PSI), xi, 1935, p. 139Google Scholar.

^{page 15 note 4} Oxy. Pap. 2079 (vol. xvii, 1927); cf. Mr. Lobel's edition of 1935 in Herm. lxx. 32 f.

^{page 15 note 5} Controversy was rife in connexion with the end of the Hymn to Apollo and its possible reference to Apollonius.

^{page 15 note 6} So Pfeiffer, R., Herm. lxx, 1935, p. 340Google Scholarf.; Rostagni's, mind, however, was set on Apollonius: Riv. Fil. vi, 1928, pp. 5, 36 ff.Google Scholar; xi, 1933, p. 194. On Apollonius' name in the Florentine papyrus cf.Herter, , Burs. Jahresb. cclv, 1937, p. IIIGoogle Scholar.

^{page 16 note 1} Vitelli, , Bull. Soc. Alex, xxviii, 1933, p. 130Google Scholar: ‘C'erano sembrate bubne le ragione del Rostagni, (Riv. Fil. 1928, 21)Google Scholar per intendere che lo scritto Callimacheo Πρòς Πραξιϕάνην fosse “diretto a” Prassifane non “contro” P.; ma il nuovo testo ci consiglia di intender “contro” (cosl intende ora anche il Pfeiffer) ritornando perciò alla interpretazione del Preller (Ausgew. Aufs. p. 97 e 105).’ Pohlenz, , Herm. Ixviii, 1933, p. 319Google Scholar: ‘Dass die Schrift Πρòς Πραξιϕάνην gegen einen Gegner governgerichtet war, muss jetzt als sicher gelten.’

^{page 16 note 2} Antim. T 14 Wyss (Asclep. A.P. ix. 63); T15 W. (Posid, . A.P. xii. 168)Google Scholar. Therefore I cannot agree with Dr. P. Maas's opinion, expressed PRIMI i, 1937, 159, that the names are without foundation. We can, after all, check the list at some important points.

^{page 16 note 3} In the article of 1933 referred to p. 14, n. 7.

^{page 16 note 4} On Rostagni's supplement μεμϕομένο[υ]ς instead of μεμϕομένο[ι]ςsee Vitelli PSI, xi. 143, n. 8 ‘nella piccola lacuna è possibile ι, non è possible; ci. also Herter, , Burs. Jahresb. cclii, 1937, p. 109Google Scholar with references. Rostagni thought Gegner that the datives in the list of names were governed by an expression like περì τούτων πολùς ⋯ λόγος παρ⋯ …, thus turning the list of Callimachus' opponents into as many sources for the Long Epic.

^{page 17 note 1} The usual terms are fiaxpos (in the comparisons Aet. Oxy. 2079. 10 and 15; Dr. Pfeiffer considers it possible also in 13) and μ⋯κος Aet. 6. τάμοι δφ ἄπο μ⋯κος ⋯οιδῇ and, perhaps, Iamb. Oxy. 1011, fol. vi ν, 1. 6; cf. notes 2 and 3.

^{page 17 note 2} The strict unity of an ἄεισμα διηνεκές excludes variability of metre and literary style. In the last Iambus of his collection the poet defends himself against τοùς καταμεμϕομένους αὐτòν ⋯πì τῇ πολυειδείᾳ (Dieg. 9. 32). Dr. Pfeiffer points out to me that this also may have been connected with the idea of μ⋯κος, since the word perhaps occurs in the Iambus (Pap. Oxy. 1011, fol. vi ν, 1.6, ed. Lobel, , Herm. lxix, 1934, p. 176)Google Scholar.

^{page 17 note 3} Therefore the well-known fragment τò μέγα βιβλίον ἴσον εἶναι τῷ μεγάλῳ κακῷ(359 Schn.) might well belong to the discussion on the Long Epic despite current opinion to the contrary. Cf. μέγας in the comparison with the μέγας ῥόος Hym. Apol. 108, and the μεγάλη γυνή Aet. Oxy. Oxy. 2079. 12.

^{page 17 note 4} Maass, E., Aratea, 228Google Scholar, stated a difference in the use of λεπτόν he thought that it was applied to the quality of Aratus' style in χαίρετε λεπταì ῥήσιες (Epig. 27), but had a quantitative meaning in the title of Aratus' (and Virgil's) collections of small poems ⋯ν τοις Κατ⋯ λεπτόν, for which he cites the parallel ⋯ν τοȋς Κατ⋯ βραχὺ ⋯πομνήμασιν. The vagueness with which Callimachus uses his terms should warn us against pressing that distinction.

^{page 17 note 5} Reitzenstein, E., Festsch. f. R. Reitzenstein, 1931, p. 47Google Scholar.

^{page 18 note 1} Arist, . Poet. 8Google Scholar. 1451² 19 ff. (in Butcher's translation). This idea of unity recurs throughout the work. Aristotle introduced it, readymade as it were, from his speculative philosophy, and finds that it is attained in epic poetry by Homer though not by his followers (cf. 23. 1459² 33). In chap. 23 we have the same idea, but there he condemns the external unity of time in the Cyclic epics and praises Homer for concentrating on one aspect of the story only, and for using the technical device of ‘episodes’ to deal with other elements of the myth. Tragedy is a still stricter unity because it excludes, or at any rate should exclude, the πολύμυθον (18. 1456^a II) This is also one reason why tragedy is ‘better’ than epic; it fulfils its end within a narrower compass and attains a higher degree of unity (chap. 26).

^{page 18 note 2} l.c. 8. 1451^a χρ⋯ … τ⋯ μέρη συνεστάναι τ⋯ν πραγμάτων οὕτωςὥστεμετατιθεμένου τινòς μέρους ἤ ⋯ϕαιρουμένου διαϕέρεσθαι καì κικεȋσθαι τò⋯λον ⋯ γ⋯ρ προσòν ἥ μ⋯ προσòν μηδ⋯ν ποιεȋ ⋯πίδηλον, οὐδ⋯ν μόριον τοû ⋯λου ⋯στίν.

^{page 18 note 3} Tragedy: I.c. chap. 7, esp. 1450^b 40 μέγεθος ὑπάρχειν μ⋯ τò τυχόν (… οὕτε πάμμικρον… οὕτε παμμέγεθες); Epic: 23. 1459^a 35; 24. I459^b 20 ff.; Epic and Tragedy compared 26. 1462^a 18 ff.

^{page 18 note 4} I.c. 7.I450^b 35 μήθ' ⋯πόθεν ἔτυχεν ἅρχεσθαι μήθφ ⋯που ἕτυχε τελευτ⋯ν:. ^b39 ταûτα τεταγμένα δεȋ ἔχειν: ^b40 τò γ⋯ρ καλòν ⋯ν μεγέθει καì τάξει ⋯στίνκτλ.

^{page 18 note 5} εὐσύνοπτον 7. 1451^a 3; εὐμνηόνευτον ^a6.

^{page 18 note 6} Chaps. 5, 6, etc.

^{page 18 note 7} Rhet. 3. 6. 1408^a 10 τό δ⋯ πρέπον ἔξει ⋯λέξις ⋯⋯ν ᾗ παθητική τε καί ἠθικ⋯ καì τοȋς ὑποκειμένοις πράγμασιν. τ⋯ δ⋯ ⋯νάλογόν ⋯στιν ⋯⋯ν μήτε περί εὐ⋯γκων βδήλως λέγηται μήτε περί εὐτελ⋯ν σεμν⋯ς, μηδφ ⋯πì τῷ εὐτελεȋ ⋯πῇ κόσμος εἰ δ⋯ μ⋯ κωμῳδία ϕαίνεται κτλ.

^{page 18 note 8} Poet. 24. 1460^a 8 αὐτòν γ⋯ρ δεȋ τòν ποιητ⋯ν ⋯λάχιστα λέγειν οὐ γάρ ⋯στι κατ⋯ ταûτα μιμητής

^{page 18 note 9} Ib.^a10 οἱ μ⋯ν οὖν ἅλλοι αὐτοì μ⋯ν δι' ⋯λου ⋯γωίζονται, μιμοûνται δφ ⋯λίγα καì ⋯λιγάκις.

^{page 19 note 1} Arist, . Poet. 4Google Scholar. 1449^a 7, text uncertain.

^{page 19 note 2} Cf. Poet. chaps. 23 f., esp. 24. 1459^b 22 where Aristotle confronts the old epics (τ⋯ ⋯ρχαȋα) with other possible ones which conform to his standard. These ideas have left important traces in contemporary and later writers like Apollonius Rhodius and Virgil.

^{page 19 note 3} I should contend the same even if it could be proved that the doctrine of the ‘characters of style’ was worked out in the Peripatos and had become known early enough to make possible a connexion of Callimachus' λεπτότης with the Peripatetic γένος ἰσχνόν; cf. Herter, H., Bursian's Jahresb. cclv, 1937, p. 214Google Scholar.

^{page 19 note 4} The non-Aristotelian character of the Hellenistic combination of poetry and criticism has been recently emphasized by Pfeiffer, R., Arch.J. Kulturgesch. xxviii, 1938, p. 192Google Scholar.

^{page 21 note 1} All the arguments put forward by Wilhelm (ref. T 1) in favour of an exact date have proved of no avail. In particular, no argument based on Agathostratus is valid since the inscription adduced by Wilhelm does not contain Agathostratus' name, cf. Roussel, , CBH 1911, p. 443Google Scholar and Tarn, , Antig. Gon. 1913, p. 469Google Scholar. This fact also disposes of Beloch's date for Praxiphanes in Gr. Gesch. iv. i². 1925, p. 598Google Scholar.

^{page 21 note 2} The existence of two similar works on poetics has sometimes been doubted, e.g. recently by Rostagni, A., Riv. FiL., N.S. iv, 1926, p. 465, n. 2Google Scholar, on Пεπ⋯ ποη[μ⋯]των: ‘si non è falsa lettera, è probabilmente un equivoco’. This inference seems rash; Пεπ⋯ πδημ⋯των: may have been an acroamatic work. Later Hellenistic works bear the same title, e.g. Philodemus's book in which it is quoted; the title seems to appear here first. Aristotle also wrote two main works on literary criticism, a dialogue Пεπ⋯ ποητ⋯ν and an acroamatic work Пεπ⋯ ποητκ⋯ς, our Poetics.

^{page 21 note 3} Cf. also T 2 b.

^{page 24 note 1} Cf. particularly the introduction to his edition of the Poetics.

^{page 24 note 2} I am thinking of the ἒν κα⋯ ⋯λον, the καθ⋯λον and καθ᾽ ἒκαδτον, the οίον ἂν γἂνοτο and τ⋯ γεν⋯μενα, and, most important, the κατ⋯ τ⋯ ε⋯κ⋯δ ἤ τ⋯ ⋯ναγκαῖον.

^{page 24 note 3} Arist, . Poetics 7Google Scholar. 1450^b 36 μ⋯γεθος generally; 1451^a 1 ft., esp. ^a10, for tragedy; chap. 24, esp. 1459^b 28, on epic; cf. above, p. 18, n. 3, Praxiphanes T 6 μ⋯κοδ. On Callimachus' use of μ⋯κοδ see above, p. 17.

^{page 24 note 4} Criticism seems to have been directed particularly against ‘dualistic’ doctrines which had a Platonic ring, such as the immortality of the soul, the νο⋯ς θὑπαθεν ε⋯δὡν, and the primacy of the ‘theoretical’ life. I have tried to give an account of this type of heterodoxy in my article ‘Peripatos’, col. 19 ff. A different, and perhaps more important, type is represented by non-Aristotelian conclusions drawn from Aristotelian premisses.

^{page 24 note 5} Ar. Poet, chaps. 23 f.; 26. 1462^b 4.

^{page 24 note 6} With the evidence at our disposal we are at present unable to tell how Theophrastus or Praxiphanes solved the problem. It remains to be seen whether the so-called ‘Peripatetic history’ was really conceived by genuine Peri-patetics like Theophrastus and Praxiphanes, or whether it is merely an adaptation made by outsiders like Duris. Prof. Ullman (ref. F 10) warns us against the name ‘Peripatetic tragic history’ because Aristotle did not favour that application of poetry to history (l.c. 33, n. 49). His article has raised the problem afresh. Unparticularly fortunately he goes to the opposite extreme by attributing to Isocratean influence the whole of ‘tragic history’. That is not convincing, and Duris' proem alone should have precluded any such attempt. The problem is really more complicated since we have to deal with Isocratean as well as Peripatetic influences. Also, Prof. Ull-man talks as if we knew Aristotle's ideas about possible applications of poetical principles to history, and holds that ‘it was a betrayal of Aristotle's creed to apply his pronouncements about tragedy to history’ (37). That surely is an over-simplification. Paradoxically, despite Poet. chap. 9, we do not know Aristotle's ‘creed’ in this matter. Aristotle never gave precepts for what history should be, and, as at other points where he indicated problems without giving solutions, there originated an intense discussion in his school, to which he himself may or may not have contributed his own ideas. We are unlikely to make much headway before we have re-captured the trend of that discussion, and, in particular, Theophrastus' system of literary criticism-a task to which Prof. Rostagni has brought many stimulating problems in his article in Stud. It., N.S. ii, 1922Google Scholar.

^{page 25 note 1} In a different historical scheme he appears, along with Aristotle, as the culmination of the pre-Alexandrian literary scholarship which is based on the interpretation of Homer (T 9). There may be as much justification in this view as in the one given in the text.

^{page 25 note 2} Dilthey, K., DeCall.Cyd. 18Google Scholar, and Immisch, O., Festsch. Gomperz, 1902, p. 273Google Scholar, assigned to Пο⋯ς Пπαξφ⋯νην the fragment of Callimachus' critiwherecism of Plato. This is not at all unlikely but cannot at present be proved. The fragment occurs in Proclus Plat. Tim. i, p. 90. 25 Diehl: μ⋯την οὒν φληναφ⋯ον Καλλμαχος κα⋯ Δο⋯ κπ⋯νεν ποητ⋯ς. Callicaptured machus was concerned with Plato's interest in Antimachus, cf. FGH, 76, F 83 Jac., Antim. T1 Wyss, , Rostagni, Riv.Fil., N.S. v, 1927, pp. 166 ff.Google Scholar, and Duris was probably referred to by him. The fragment certainly suits our context well.

^{page 25 note 3} Our criticism above, p. 13 n. 4, concerned the chronological conclusions, not the actual fragment.

^{page 26 note 1} Callim. Epig. 27 Wil. (A.P. 9.507) ‘EEgr;σιóδον τó τ’ ἂεισμα καì ó τπóπος οὐ τòν ⋯οιδ⋯ν ἒσξατον, ⋯λλ’ ⋯κν⋯ω μ⋯ τò μελιξπóτατον τ⋯ν ⋯π⋯ων ó Σαλε⋯ς ⋯πεμ⋯ξατο χαιπετε λεπται π⋯σιες, ‘Απ⋯τον σὑμβολον ὑγπνπνíης. Fr. an. 388 Schn. (Marcus ap. Front, ad M. I. 4, p. 11 N. = I, p. 94 Haines) ποιμ⋯νι μ⋯λα ν⋯μοντι παπ’ íξνιον Oacute;ξ⋯ος íππον ‘σιóδω, Μονσ⋯ων ⋯σμòς óτ’ ⋯ντíασε, cf. Aet. fr. 9. 84 Pf. (Oxy. 1011)… τῷ Μονσαι πολλ⋯ νὐμοντι βοτ⋯ σὐν μὑθονς ⋯β⋯λοντο íξν[ι]ον óξ⋯ος íππον (cf. Hes. Theog. 22 f.).

^{page 26 note 2} Eratosth. against Aristotle's political opinaions: Strabo I, p. 66; Tarn, W. W., Alexander and the Unity of Mankind, 1933. p. 7Google Scholar. Cf. Wilamowitz's, convincing argument, Briefwechsel m. Mommsen, p. 250Google Scholar, against Mommsen's, remark in The Provinces of the Rom. Emp. ii. 241Google Scholar, n. 2 = R.G. v. 562) that the Ptolemies professed Aristotelian politics and that Eratosthenes was opposed to them.