The first Athenian prose writer

BY F. JACOBY I In a paper on Pherekydes Wilamowitz 1) has re-affirmed and for the first time tried to prove his well-known theory that Pherekydes is to be regarded as a collective name for early Ionian prose writing in the same way as Homer’s name once covered the whole Ionian epos, and Hippokrates became in the course of time the author of all Ionian medical books and treatises 2)— for we have to make this distinction, though Wilamowitz did not 8). The paper in its second part, which is a sort of appendix, is rich in results for the text, and also for the explanation of a great number 1) Sb. Berlin 1926 = KI. Schr. V 2, p. 127 ff. Attention should be paid to the Appendix of the editor L. Malten (p. 194). rag 5p6al ACOBY*X. + 2) Gr. Lit., 1912, p. 55f. ‘der name hat genau so viel wert wie Homer und Hippokrates’, and the impatient remark in Platon II, 1919, p. 238 n. 1: ‘ich betrachte auch jetzt den namen genau so wie den Homers; das kann man ignorieren, aber nicht widerlegen’. I hope that this investigation, although restricted mainly to the biographical tradition, will serve to refute this thesis which in FGrHist. 1, 1923, p. 386 I passed over in silence, because I considered it arbitrary and wrong. After its recent detailed discussion: by Wilamowitz W. Schmid.Gr. L. I 1929, p. 711 n. 4 declared it to be ‘wenig wahrscheinlich’, while Geffcken Gr. L. I A, 1926, p. 102 n. 106 called it ‘durchaus richtig’. The latter failed to understand Wilamowitz when he tried to prove by general considerations about the spirit and the style of the work that there existed ‘one Pherekydes only’, viz. the theologian of Syros, to whom also the fragments of the Genealogiai belong—an assertion quite contrary to the opinion of Wilamowitz. My defence of the tradition (for such it is) was approved by W. Schmid Lc. and Momigliano (‘Per l’età di Ferecide Ateniese’ Riv. di Fil. N.S. 10, 1932, p. 346 ff.). About Laqueur ‚R.E. XIX 2, 1938, col. 1991 ff., who also pleads ‘die geschichtlichkeit des Pherekydes (scil. of Athens) im sinne der antiken tradition’ see below p. 44; and for the particular problem of the Lerian, in whom Laqueur with many others does not believe, see p. 48 ff. 3) Cp. below p. 46f.

of fragments as printed in my edition 4). But I am doubtful whether in its first and main part Wilamowitz did succeed in establishing his thesis, as against my brief—and as I now recognize perhaps too brief—introductory notice 5), in which I upheld the general opinion that, apart from the Hellenistic author Pherekydes of Leros, we have to acknowledge two clearly distinguishable writers of this name, viz. the theologian or philosopher from the island of Syros, who belongs to the second half of the sixth century B.C. 6), and the Athenian genealogist, author of a book on 4) FGrHist. 1, 1923, no. 3. 5) Op. cit. p. 386. 6) Diels-Kranz Die Fragmente der Vorsokratiker® I, 1934, p. 43 ff. In view of Diog. Laert. 1, 119 who ultimately derives from Kallimachos’ Pinakes — o@itera. 5% TOU Zueiou t6 te fifliov 5 ovvéyoarwpev od à dexn *Zàe piv xaì Xeóvos Heav dei xtA.» — and of what we know of the history of the book-title there is not much sense in arguing about ‘the correct title’ (Kern Rel. d. Gr. I, 1926, p. 297f.; von Fritz R.E. XIX 2, 1938, col. 2028, 29 ff. is vague). But the contradiction between Damaskios De princ. 124b (Vorsokr. 7 [71] A 8) & &v év xévte puyois ôimenuévov scohhùv &AAnv yeveëv ovornvar Oev tiv Ilevréuvyxov xalovuévnv, tavtov 5% tows sireiv mwevré- -xoouov and the title in the Suda s.v. Wegextbdnco—éoutt 5% Ava & ovvéyeawpe “Estapvyoc frou Ocoxegacia i Oeoyovia — can hardly be explained away as H. Gomperz W. Sf. 47, 1929, p. 18 ff. tries to do. What the Suda has is a title and two variants which are taken from the contents and are “rather superfluous as they are followed by one of the explanations, frequent in the Suda and mostly quite correct: got. d& Oeokoyia Ev BuBAtoius Öéxa Exovoa OesHv yéveorv xai ÖLudoxdc. This explanation is well in accord with the earlier statements of the contents made by Damaskios xoÂÂñv GAAnv yeveùv ovorijvar Oev, and Theopompos (FGrHist. 115 F 71) xeei pboewe. xai Oewv. Gomperz’ insertion of yevéoews before Oe@v is unnecessary and probably wrong. Kuester’s transposition in the Suda ñ QeoAoyia- gor ö& Oeoyovia is also wrong, and I was wrong in referring the whole explanation to the Hisforiai of the Athenian genealogist (FGrHist. 3 T 1). It is merely the number of books which crept in from the next article about the Athenian Pherekydes. The quotations do not help much because they mostly occur in ‘theological’ writers, who naturally think of the Syrian and simply quote. D. gnoiv, Eieyev efc. Only in a few cases they add the distinctive epithet 6 Zeros: Achilles Vorsokr. 7 B 1a; Porphyrios ib. B 6. Thé supplement in Philodem. De pietate (ib. B 13) is not quite certain (cf. n. 10). In Clement . FGrHist. 3 F |174 the epithet 6 Zveros is due to confusion (see below p. 39ff.), and it is very doubtful whether 6 saAatéc in Plutarch ( Vorsokr. 7 B 13a = FGrHist. 3 F 177) refers to the Syrian (see n. 137). But the two older grammarians who record words from Pherekydes (Vorsokr. 7 B 9—11)}

Greek mythical history, whose time we shall have to define later on?). As the problem is somewhat important not for Pherekydes alone but for the history of early prose writing generally, a new investigation seems indicated: I could have wished that somebody else - had undertaken the task. For in view of the friendliness with which Wilamowitz judged my attempt of an edition of the remains of the Athenian Pherekydes, even when differing on fundamental questions, my rather violent criticism of his treatment may seem churlish. I comfort myself with the principle which was the lodestar of Wilamowitz’ own life-work: amicus Plato, magis amica veritas. . Starting as we must from the ancient evidence about the several bearers of the name Pherekydes we meet with a fact which we cannot push aside as unconcernedly as Wilamowitz does. The evidence is quite clear: it ascribes one and only one work to each ‘writer; to the Syrian a treatise, probably not a very lengthy one, - which according to what we know of its contents we should describe as theological or (there is no real difference) philosophical; to the Athenian a rather bulky work, divided by the ancient editors into ten books, which from the likewise later titles—Teveadoyiar or "Iotogiau—we have to regard as historical in the ancient sense of the word 8). Eratosthenes regarded these two writers as two distinct quote accurately: (1) Herodian. II. pov. AsE. p. 7, 4 says td tod Zvetouv, and the use of the epithet shows that he knew another Pherekydes. Actually we read in Choiroboskos (FGrHist. 3 F 162) maga Pegexbönı tat "AOnvaior, and ‘Choeroboskus ist ja Herodian’; it is sheer prejudice that Wilamowitz p. 131 calls the epithet t@r “AOnvaicor in question. (2) Apollonios Dyskolos De pron. p. 65, 15 Schn. has Depexdôns év tht Oeokoytar xai Erı Anuóxeurog, and in spite of theschronological sequence tv negoi Annöxeırov, Depexbönv, “Exaratov, he probably means the same author on p. 93, 1. About Pegexbônc év tHe Oeoyoviar in Schol. Apoll. Rh. 2, 1209/15a see n. 8 and 14. 7) See p. 28 ff. 8) He is called '‘AOnvatos and yeveakóvos by Eratosthenes and Dionysios of Halikarnassos; 'AOnvatos by Strabo and hisforiarum scriptor by Eusebios (the passages see p. 26 f.). His activity is called yeveaAoysiv by the Scholiast on Pindar (FGrHist. 3 F 171; pu0oxowiïv of the Syrian Origenes Vorsokr. 7 B 4). The quotations furnish more information about him than about the Syrian, because they mostly derive from Hellenistic grammarians who had the Alexandrian edition before them and were accustomed to quote accu-

persons 9), and the quotations of the works bear him out by adding either the respective title of the work 10) or distinguishing their authors by an epithet such as Zúetoc, BeoAöyog, “AOnvaios, yeveaAöyoc. More important perhaps that even the most superficial comparison of the remains of these two books seems to show a striking difference as to their subject-matter, their treatment of it, their style, and the attitude of mind of their authors 11). Of course, rately. To be sure, abbreviations to a simple Depexúönc or Degexbôns Ev tht @ are frequent in the sources preserved to us, and sometimes they happen before our own eyes (see e.g. FGrHist. 3 F 18; 25; 41; 69; 90). But the distinctive epithet ‘AOnvatos (and no other) occurs almost regularly in Philodemos (FGrHist. 3 F 35c; 91; 130; 165; supplied in-F 166; the majority of the quotations derives from Apollodoros) who perhaps cited. the Syrian too (n. 6, 10); in the isolated quotations by Dénys (3 F 156), Hyginus (3 F 90d, from Eratosthenes?), the Scholia on Homer (3 F 169), and Choiroboskos (3 F 162) from Herodianos. The title of the book is preserved in three cases only, but always in the form év tye a (Y, 1) tev iotogt@v; they are Didymos (3 F 2), Scholia on Euripides (3 F 35a from Apollodoros), Athenaios (3 F 18a). The one exception ®egexböng Ev Osoyoviaı in the Scholia on Apollonios (3 F 54) refers to a certain part of the work in which the author had dealt in detail with the fight against Typhon and his companions (s. n. 14). 9) Diog. Laert. 1, 119— FGrHist. 241 F 10. It still seems probable to. me that the passage derives from the Xeovoyeagia:. But as Eratosthenes. repeatedly consulted the "Iotogiaı in other works, as e.g.in the Katasterismoi (see 3 F 16c; 59; 99), it is conceivable that he rejected the astrologer of Andron (p. 24 f.) in another context. . 10) See n. 6; 8. It is unfortunate that the passage of Philodemos De piet. 41a p. 19 Go (cf. Philippson Herm. 55, 1920, p. 271) is defective. The supplement ®eee[xdöns] 5’6 [Zó]etos is not certain; for Diels’ argument that the Lerian Pherekydes is called Athenian by Apollodoros, and consequently by Philodemos, is wrong and it is not quite impossible to refer the passage to the Lerian. About the confusion Mm Christian writers of the Athenian and the Syrian or Lerian see n. 35; p. 52f. 11) I do not attach much importance to real or alleged contradictions in details, as e.g. between Vorsokr. 7 B 1—3 and FGrHist. 3 F 16. But I am afraid I must sharply contradict Wilamowitz who—in order to efface the borderline between the two works and incidentally to make it credible how ‘der Syrier die ionische mythographie erben konnte’ (Lc. p. 133)— declares: ‘seitdem der fetzen des papyrusbuches <Vorsokr. 7 B 2> gefunden ist, habe ich immer geglaubt und gelehrt, dass ein solches buch, das die gôtter sogar redend einführte, eigentlich unter die mythographen sehr viel besser passte als unter die doxographen’. Even apart from the wellknown fact that Hesiodos in the Theogony introduced speeches of the gods, and

there remain doubts as to the authorship of some single fragments: the contents of a theology and a genealogy occasionally over- Jap 12); the scholars who used one of the two works belong to leaving aside the improper use of the term doxography, this passage seems to show that Wilamowitz failed to recognize what the ancients called theology. The compromise of Von Fritz (dc. col. 2025 no. 4) who under the influence of Wilamowitz calls the Syrian ‘mythograph und kosmologe’ is equally unsatisfactory. The position of Pherekydes as a philosopher, or a predecessor of the real philosophers is unambiguously determined by Aristotle Metaph. N 4 p. 1091b 8 of ye uepryuévor adtT@v xai t@v un pvOxo Östavra Aéyew, olov Degexbdnc xaì Eregoi tives (cp. Diels Sb. Berlin 1897, p. 148 n. 1) and the remains of the treatise show quite clearly why Eratosthenes distinguished the Syrian Pherekydes as a Oeokóyos from the Athenian yeveadAdéyos. As late an author as Philon of Byblos (Vorsokr. — 7 B 4) uses the correct expression é@e0AGynoev as distinct from the simple Aéyerv in the introduction of quotations from the genealogist; and it does not matter that a Christian controversialist ({b.) replaces BeoAoyeiv by uv- 8Soxotëiv. I need not enter upon details. But it is worth mentioning that H. O. Schroeder (Herm. 74, 1939, p. 108 ff.) at last interpreted correctly the quotations by Origenes: pnotv ‘actually refers to Celsus, not to Pherekydes, as was universally assumed (Diels; Wilamowitz Gil. d. Hell. I p. 266 n. 3; Kranz Herm. 69, 1934, p. 114). We thus get rid of the far-reaching conclusions which were drawn from these quotations for the work of Pherekydes. Schroeder’s triumphant conclusion ‘umsonst heisst dieser Pherekydes nicht historicus ebenso wie der Athener’ is, of course, wrong (cf. n. 35). The Alexandrian Ilivaxes naturally listed the Oeokóyos among the œiiéooæor and the ysveaAöyos among the iotroorxot. 12) The question whether the ‘Iotogiac began with a theogony is of no great importance for our purpose. A rather extensive theogony is certain for Akusilaos and for the later compilations of mythic history, it is improbable for Hekataios, and Hellanikos too in his five mythographical works seems to have dealt with the human pedigrees only. There was no rule. As to Pherekydes’ Hisforiai Wilamowitz p. 133 ff. contested that they contained a theogony. I admit his criticism of my arrangement which assumed that a theogony filled the first two books. But there is a not inconsiderable number of fragments unfortunately mostly without a distinctive epithet for the author or the title and number of the book (3 F 44ff.), for which it is not easy to find a place in the human pedigrees, and which, nevertheless, nobody gave (or could give) to the Syrian. They deal with Hekate, Daeira, Kyklopes, Daktyloi, Kabiroi, Thriai, Typhos, and similar deities or half-deities. Wilamowitz, who assigns to the story of Herakles e.g. the eagle of Prometheus, offspring of Typhos and Echidna (3 F 7 from the second book)—an attribution which is hardly favoured by Akusilaos 2 F 13—paid attention only to the fragments which can with some probability be connected with a heroic myth. But there are others which cannot. Mnemosyne XIII «

x widely different circles, and therefore naturally did not always add a distinguishing epithet to the name of the author whom they quoted, because there was no danger of misapprehension; lastly and chiefly we seldom have the. original and accurate quotations of the Hellenistic authors from the Genealogiai of the Athenian, but only the abbreviations in the Lexica and Scholia which in the course of time were moré and more epitomized and cut down 13). Nevertheless, if we eliminate the eight dubious quotations which certainly do not derive from the theologian of Syros there are very few among the 172 fragments of the Genealogiai in regard to which a serious doubt arises as to whether they come from this work or the Theology 14). Even the use and (or) misuse of the name It is a task for the much needed special edition to consider those fragments and the various possibilities for their position in the Genealogiai. If the work did not begin with a short theogonic introduction, monsters like Echidna or the eagle of Prometheus may have found their place in digressions in the. history of the heroes and the gods at the head of the several pedigrees, as in at least two books of Hellanikos, the Asopis and the Atlantis. This last supposition has much to recommend it: one is reminded of Akusilaos who evidently opened with the marriage of Okeanos and Thetys. In Pherekydes Triptotemos sprung from the marriage of Okeanos with Ge, and the majority of the fragments from book I can be connected with Zeus. 13) See n. 6; 8. 14) This is the case with 3 F 109 in Schol. Apoll. Rh. 1, 643/8e Degextbing dE now, St S@eov elye mapà tod ‘Epuod 6 AiBaktöns vd tv woziv abroù wot® pev év “Atbov, wort 5% év rois Dorèe Tv ynv téso1g elvar. Diels (Vorsokr. 7 B 8) rather hesitatingly and without a cogent reason listed it among the fragments of the Syrian; Rohde (Psyche? I, p. 167 n. 1) assigned it ‘without any doubt’ to the genealogist. I believe the latter to be right, because the Scholia on Apollonios quote the Athenian, and only the Athenian, frequently (see on FGrHist. 3 F 106; Wilamowitz p. 133, 1). In spite of the quotation ®Wegexbinsg Ev Oeoyovia (cf. n. 8) Wilamowitz did not convince Diels-Kranz or me that the Typhos story derives from the Pentemychos. For 3 F 46, which Wilamowitz is ‘inclined to give to the Syrian’ see n. 36. Concerning the eight Dubia 3 F 173—180 one has only to look at the authors who quote them in order to understand why most of them are really dubious, i.e. not belonging to an author of the sixth or fifth century at all, with the exception of F 177 in Plutarch. De fac. in orb. lun. 24, where we hesitate between the theologian and the genealogist. The quotation Pegexböng 6 sadatédcg does not help much, but the Syrian may have the better claim (FGrHist. I p. 429, 40 ff.; Wilamowitz p. 133, 1; Vorsokr.> 7 B 13a). F 175 and the second quotation in F 179 may belong to the genealogist; if they do they are genuine. In F 174 from Clement Depexóöns

Pherekydes in late authors hardly ever makes the question of distinction between the two authors more intricate. In these circumstances we are surely justified in asking for very strong reasons if anybody maintains that both works had ever been regarded as being written by the same author; that ‘the scholars of the library (in Alexandria) were the first to recognize that the (genealogical) book could not have been written by the well-known Syrian’; that consequently ‘they distinguished between two homonymous authors’; and that ‘they inferred from the contents of the genealogical book that its author must have been an Athenian’ 15). I shall not enter here upon the second question of Wilamowitz about the scheme of the genealogical work, and the distribution of the subject-matter on the ten books; nor shall I examine at this early phase the surprising phenomenon of a pre-Herodotean writer in Athens, who composed a voluminous work in the Ionic dialect 16); and where it can be avoided, I shall not deal with the text and the interpretation of single fragments 17). Our only task is the examination of the biographical, or rather literary problem which in any case must be the starting-point of any special investigation whether of the fragments or of the works to which they belong, or of the authors who wrote them. This examination must be very full. For in spite of its circumstantial character the paper of Wila- mowitz did not exhaust the tradition, and he admitted a number of factual errors, some of them fatal. Moreover his main thesis ‘we cannot but imagine that the work of ten books, inscribed with the name of Pherekydes, came to Alexandria, where however nothing was known about the personality of this Pherekydes or about the 6 Zveuos is chronologically impossible; the fragment most probably belongs to the Lerian (see p. 51 ff.), as perhaps also F 173 does (p. 54). F 176; 178; 179 derive from the AûrtóyOoves, a late forgery (see p.55 ff.). In F 180 the change of Kedrns into epexdôns is more than doubtful. In all other fragments, whenever a doubt arises, the greater probability is in favour of the genealogist who is utilized much more frequently than the Syrian even by ‘theological’ writers with the exception of the Christians: Philodemos in De pietate quotes the Athenian at least six times, the Syrian once, if at all (cf. n. 10). 15) Wilamowitz p. 132. 16) See p. 34f. 17) See p. 42 ff.

existence of an Athenian writer. Pherekydes at all’ 18) begs the question; and the inference built upon this petitio principii, viz. that it was Eratosthenes who invented the Athenian because the said book apparently was not written by the well-known philosopher of Syros, is a vicious circle. It is this theory which entangled a problem not so very difficult in itself, if only we interpret without parti pris the testimonies handed down to us. I shall therefore begin -with stating the simple fact that tradition knows of five bearers of the name Pherekydes, and I shall discuss them in due order 19). Il 1) Depexdônc Baßvos Eúetos 2°), the ‘theologian’, author of the so-called IHlevr&uvxos or ‘Exréuvxoc”'), who was usually regarded ‚as the earliest 18) L.c. p. prose writer *"). 132. With His existence the specious arguments on has which never been Wilamowitz bases his assertion I shall deal in the course of this investigation. 19) The name is formed as correctly as @Wegexedtns, Pegexins, and others (Bechtel Die historischen Personennamen 1917, p. 444f.) and seems to have been used for human beings only. In Athens we find Deeexóöns Kespaindev on a tombstone of the fourth century (P.A. 14197; the genealogist is lacking), a fact of some importance for our no. 3. The son of Teiresias (Peisandros Schol. Eurip. Phoin. 834) owes his existence. to a rather questionable conjecture of Ed. Schwartz whe changed a ®segoexéoedns into Depexdône (see Robert Heldensage 1920, p. 128 n. 1; Kloesel R.E. XIX 2, 1938, col. 1991 no. 1; 2056). 20) Diog. Laert. 1, 116ff.; Suda s.v. Begexbing Baßvos; Damaskios De princ. 124 b = Vorsokr. 7 [71] A 1-2; 8. 21) Cf. n. 6. 22) Plin. N.H. 7, 205 (see p. 23); Apuleius Flor. 15 p. 22, 8 Helm Pherecydes Syro ex insula oriundus, qui primus versuum nexu repudiato conscribere ausus est passis verbis, soluto locuto, libera oratione; Suda s.v. Beoexbins BéBvoc- TwE@tOV dt goyyeaphv éEeveyxerv neloı Adyou tives totogovov, étéopwv tovrto eis Ké&dpov tov MıAnorov pegévrov (cf. n. 29); the same s.v. *Exataiog "Hynoavögov- noros d& ioropiav mets EEhveyxev, ovyyeaphv 5: Depexböns- ta yao “Axovotddov voBeberar (cf. FGrHist. I p. 376, 7ff.); Porphyrios (FGrHist. 260 F 21) in Suda s.v. Pegsxhöng 'AOnvaios (cf. n. 120). In the light of these testimonies we have to explain Diog. Laert. 1, 116 tovtov noi Osérouxos (FGrHist. 115 F 71) nowrov mepi poems xai Oev yoawar. From Pliny and Theopompos we infer that the claim of the Syrian Pherekydes to having been the first prose writer goes back to the fourth

doubted, and we know the first words of his book from a bibliographical statement evidently taken from the Ilivaxes of Kallimachos °°). The inference drawn from the statement as to the time of the author has been amply corroborated by the well‘known papyrus which gives us an idea of the style and the authors manner of presenting his subject ?*). The Suda (as often) has two dates, placing them side by side without (as usual) noticing their discrepancy: yéyove 5& xat& tov Avidv: Baoriiéa *Alvadtrnvy (617 or 605—560 B.C.) de OUYXQOVELV totic T cogoic, nai tetéyGar negi tiv pe éAvuriéôa (600/597 B.C.). The first date is the simple synchronism, entered as such into a chronicle, and does not make difficulties, for in Hellenistic chronology oi éxté gopoí (like t& IIzooıxd) is a fixed date, which means Ol. 48, 4 — 585/4 B.C., about the middle of Alyattes’ reign 25). In the second century B.C. In fact, the Pentemychos was a real book, listed in the catalogues of the Alexandrian library, while the books of all his competitors, viz. Kadmos (Strabo 1,2, 6, following Dikaiarchos and Poseidonios), Pseudo-Akusilaos, and the pre-Homeric Pherekydes, author of the Adréx8oves (our no. 5) are later forgeries. Whether or no the claim of the Syrian is justified is hard to tell, as the publication of his book cannot be dated exactly. But Thales left nothing in writing, at least nothing in prose (Diog. Laert. 1, 23); the prose book quoted by Galenos (Vorsokr. 11 [1]B 2) is a late forgery which did not influence the chronology of the Hellenistic scholars. Anaximander’s t@v decoxdévtmv abta@u xeqodarmdns ÉxOeorc, Hu mov mweerézuxev xai "Axok6ôwooc 6 “AOnvatoc (Diog. Laert. 2, 2) has been overlooked by the earlier historians of literature and style; for he is never named as a competitor of Pherekydes. In any case it was and is a matter of opinion — whether the contemperary of the Seven Wise Men was not earlier than thephilosopher of Miletos, who according to Apollodoros (FGrHist. 244 F 29) was born in 610/9 B.C. and published his book (if he published it at all} at the end of his life, again according to Apollodoros, in 547/6 B.C. 23) Diog. Laert 1, 119= Vorsokr. A 1; B 1. The inscription containing the author’s name, which must have preceded the first words of the treatise itself, has unfortunately dropped out. But we may safely assume that the name of Pherekydes’ father, and perhaps his native place was known from his own introduction. . 24) Grenfell-Hunt Greek Papyri Ser. I {no. 11 = Vorsokr. 7 B 2. Cf. Diels Sb. Berlin, 1897, p. 144 ff. 25) Ph. U. 16, 1902, p. 179 n. 9; 182f. The epochal year 585/4 B.C. is based on the eclipse of the sun mentioned by Thales, and it remained the Jloruit of that philosopher in the chronology of Apollodoros, while the other meri of that group were dated more accurately each by himself. It replaced

date, which can definitely be traced back to Apollodoros 26), the number of the Olympiad is slightly corrupt. Wilamowitz, who upheld it as the date of the publication of Pherekydes’ book 27), i.e. for his floruit, failed to recognize the cogency of the parallel passages, particularly Diog. Laert. 1, 121 yéyove ÔÈ xate tiv mevrnxoothv xal évérnv Ókvuniáöa (544/1 B.C.) who derives. from the same source as the Suda. A floruit in Ol. 59requires OI. 49 for the birth. We must therefore either change the ME of the Suda into u9 (as Rohde did), or the évérnv of Diogenes into aéurrnv. I still favour the former alteration (1) because it is easier from the earlier epochal year of Demetrios of Phaleron (FGrHist. 228 F 1) in the archonship of Damasias 582/1 B.C. xa6’ Sv oi ésrtà oopoi éxAñOnoav. Demetrios was preceded by Andron év Toinoöı, who gave the simple synchronism ovvexoövıoav 5& oi éExta ovyxarakeyévres oogpoi tr Oadci (Clem. which Pherekydes appoints as. the executor of his testament Thales oùv x Al. Strom. 1, 129). It is possible and even probable that Andron mentioned the Lydian king at whose court they met; but we must not call this an epochal year. It is further uncertain whether Andron and Demetrios placed Pherekydes among the Seven Wise Men. Apollodoros does not seem to have done so, nor did Alexander Polyhistor, for he calls him a disciple of Pittakos (FGrHist. 273 F 85). Diogenes Laertios, although: he treats Pherekydes in the same book as the Seven, does not expressly count among them either Pherekydes or Epimenides. Aristotle év y Ilfeoi somruxñc (F 75 Rose) even speaks of literary hostility between Pherekydes and Thales. The only witness for the former being one of the group is the forged letter in toig dAkoıg oogois (Diog. Laert. 1, 122). It is therefore uncertain who introduced Pherekydes into that circle, though it must have happened in the fourth century at the latest, because he appears among the seventeen competitors of Hermippos (Diog. Laert. 1, 42), but is missing among the ten names given by Dikaiarchos (ib. 1, 41). It probably was an early writer, for we can observe that the legend which connects Pherekydes with Pythagoras, thus incidentally separating him from the Seven Wise Men, made steady progress. Usually one believes Ion of Chios to have been the first representative of that legend or tradition (see H. Gomperz W. St. 47, p. 14; v. Fritz /.c. col. 2027, 52 ff.). 26) Ph. U. 16, p. 210 ff.; FGrHist. II D p. 805, 36 ff. No progress is made by von Fritz Le. col. 2026 f. 27) Lc. p. 128 f. (followed by v. Fritz Zc.). He assumes that Pherekydes was dated in Ol. 45 in order to prove that the first prose book was ‘older than Kadmos, or a book of Thales, or perhaps the Laws of Solon’. I do not believe an elaborate refutation to be necessary. Apart from the parallel passages quoted above which decide the question the Suda here does not date the floruit of Pherekydes or his book, but his birth.

the palaeographical point of ‘view; (2) because Euseb. Chron. a: Abr. 1473/6 = Ol. 59, 1/428), who ultimately derives from the same source, confirms the. floruit in Ol. 59; (3) because a testimony of Plin. N.H. 7, 205—-prosam orationem condere Pherecydes Syrius instituit Cyri regis aetate, historiam Cadmus Milesius *)—which derives from another source, viz. a book Megi ebenudtwv, requires a date, which places the publication of the first prose book in the reign of Kyros 560-529/8 B.C.: Ol. 59 fulfils this requirement, Ol. 45 does not. Neither date rests on documentary tradition, but the second probably is nearer to the truth. We cannot say for certain whether it was Apollodoros himself who gave the reign of Kyros 30) where the earlier date had Alyattes, or whether the name of the Persian king is taken from an intermediate chronicle which followed the dating of Apollodoros and therefore entered Pherekydes in the middle of Kyros’ reign. But it seems fairly certain that the year 544/3 B.C. for the akme of Pherekydes is based on his relation to Thales; for the latter me@toc xeoi ptoews SieAdxyOn, Go tivec (Diog. Laert. 1, 24) whereas the former motos sweet mpicems Fyoawe (Theopompos in n. 22). The interval of forty years suits the method of Apollodoros, who prob28) See also Cicero Tusc. 1, 38 Pherecydes Syrius primum dixit animos esse hominum sempiternos, antiquus sane: fuit enim meo regnante gentili (578—535 B.C.); cp. n. 35. 29) The relation between Pherekydes and Kadmos is twofold: either they are competitors for the first prose book (Suda s.v. Depexúöns BáBvos; see n. 20), or a distinction is made between the first prose writer (Pherekydes) and the first historian (Kadmos). In the first case both competitors are legendary and pre-historic persons, viz. Degexbing 'AOnvatos esoßürsoog tov Zvpiov, Sv A6yos ta ’Oepéws ovvayayetv (our no. 5) and Kadmos dc WEWtTOG naTd tivac ovyyoaghyv Eyeawpe zutakoydönv, prxoari vewtegos Ovopéos (Suda s.v. Kédpoc Wavdiovoc). In the second case they are writers of the sixth century: the Syrian Pherekydes with a floruit in 544/3 B.C. and the Kadmos belonging to the group of ot tac ioropias émyerofjjoavres ovyyedperv ae’ aùrots (scil. vois "EAAnoıv), AÉyo ÔÈ rods weei Kéôuov te tov Miioiov xaì tov “Ageyetov "Axovoikaov xai peta rodrov et tives &ŒAkor Aéyovrar ye- | véc8at, who Beayd the Ilspowv éxi Thv 'EAAdÒa oteateiac TOL YOdver seobAaßov (loseph. c. Ap. 1, 13). It does not matter that this Kadmos too owes his existence to a later forgery. 30) As he did for Xenophanes FGrHist. 244 F ‘68.

ably under the year 544/3 digressed on the literary history of the sixth century in the same way as he did for the fifth century under the years 456/5, 444/3, and 432/1 B.C. There are some further facts from the life of the Syrian Pherekydes, either real or alleged, which we may safely leave aside here 31). 2) Degexbing Eúetoc, the astrologer: “Avigwv 8’ 6 ‘’Eœpéoioc œnoi Ôvo yevovévar Peoextiacg Zvpiovs, tov HEV aoreoAdyov, tov © 5& Oeokáyov vidv Baßvoc, Ot xai IIvOayóeav oxoidoar- “Egatoo8évng 5’ Eva uóvov, xai Eregov 'AOnvatov yevealéyov 37). It is universally acknowledged that the. author of the Tripous invented the earlier Syrian on the basis of two dates: these are firstly the lines in Homer et mov axoverc, Od. o 403 f. vnoéc tig Svein xixÂAñoxetar, “Ootuvying xaOdrepOev, 601 teoscai ñAioro, and secondly a tradition of Syros (let us hope that it is a tradition) which connected the heliotropion in some way with Pherekydes 33). 31) For his relations to Pythagoras and the Seven Wise Men see n. 25. The former could be made to agree with the floruit in 544/3 B.C. (cf. FGrHist. II D p. 805, 36 ff.), while the latter had to be discarded. Theopompos (FGrHist. 115 F 71) knew of a sojourn of the Syrian Pherekydes in Sparta, and he is doubtlessly meant in Plutarch Pelop. 21, 3 (the passage is missing in Vorsokratiker) <@v 5° tjoregov DegexGdnv te viv coqdv Ind Tv Aaxe. Sawoviev dvaigeBévra, xai Thv doegdàv adrod Hard TL Aöyıoy dard tOv BaGUEwV peoveovuévnv. The same story is told by Sosibios about Epimenides (Diog. Laert. 1, 115; Pausan. 3, 11, 11; cf. the proverb td "Esttpeviderov déoua’ éxt tv dnroBérwv Diogen. Prov. 8, 28). The place of his death and his tomb were also disputed: we hear of Delos (Aristoxenos), Samos (Aristotie), and the Magnesian territory where Ephesos performed his cult. It is obvious that the authors of the (fifth and) fourth centuries had much to tell about a personality which later on became somewhat shadowy. 32) Diog. Laert. 1, 119. The Teitsrovc belongs to the fourth century and seems to have been known to Theopompos (FGrHist. 115 F 70). The Suda has no Life of this pre-Homeric Pherekydes, whom Eratosthenes had eliminated once and for all. 33) Diog. Laert. 1, 119 oérberar 5% tod Evolov té te PrPAfov..., c@ietar 5& xai td NAoTeönıov Ev Zeo tHe voor; Schol. Hom. Od. o 404 éEvOa paoì slvar jAiov oocfjkarov, du’ où onperodvtat tas Tod HAtov teonds. Wilamowitz p. 128 believes that Homer’s lines refer to a real fAtotoémow, partly on the strength of a fourth century inscription from the Cretan Itanos (Syll.? 1264 with the note of Hiller). I am rather doubtful because the poet does not seem to have had any local knowledge of Syros (see Z.G. XII 5,2 p. XXXI); he does not seem to say that there is a nAuLorgömıov on the island Syrie,

The inference of Andron that the man who used the heliotropion was not the theologian but an astronomer living before Homer, is not quite stupid—the Heptamychos certainly does not give the impression of its author having made scientific observations—, but of course it is wrong: it would have been correct for, Andron to deny the connexion of the theologian with the heliotropion; but the horror vacui prevented this simple expedient; the heliotropion had to have an ‘inventor’. Nevertheless, the inference is important in so far as it shows that even in the fourth century the theologian could not be dated earlier than the sixth century because as yet scholars did not like to assume pre-historic prose books. Later on, in early Hellenistic, and even more in Roman times, one became less discreet; the man who invented the author of the AûtógOovec (our no. 5) had no such inhibitions. This invention of a pre-historic Athenian genealogist was rejected by Porphyrios, the pre-historic Syrian of Andron by Eratosthenes, who as far as we Can see opposed only the splitting up of the Syrian into two persons, a theologian and an astronomer. 3) Degexdting 'AOnuatoc yeveakóyoc. This is the writer to whom most of the fragments belong and whom Eratosthenes acknowbut that the island itself is the point where from Ortygia one can observe the turning of the sun. I absolutely distrust the account of Politis in a letter to Hiller von Gaertringen written in 1911 about the discovery of the ävrgov tov PDeperböovs on the N.-E. side of the island. The re-examination by Peek in 1933 (A.M. 59, 1934, p. 64ff.) has yielded no facts from which such a use of the two small grottoes could be inferred. True, Peek does not mention the Zeliofropion, and I am afraid that (like Politis and Hiller) he mixed it up with the dwelling-place of Pherekydes during his last illness. The archaic inscription on the rock in a distance of about 200 metres from the grottoes, in which the lively imagination of Hiller found ‘den letzten todesseufzer’ of the philosopher, Peek read as the parallel to the oige of the Thera inscriptions, and the various graffitti found beside it may favour his explanation. However that may be, I do not even believe in Peek’s alternative suggestion that Gvteov tov Yılocögyov, as the peasants are said to call one of the grottoes, means the dwelling place of ‘the hermit’. I suspect that the peasants heard of ‘the philosopher’ from local scholars (Politis himself?). We demand stronger arguments than the mere existence of two osfjAata for making us believe that the memory of Pherekydes was preserved on the island ‘durch die jahrtausende’.

ledged besides the Syrian. On Eratosthenes depend directly or indirectly, supplementing the short quotation of his opinion in Diogenes Laertios as well as each other, Strabo 10, 5, 8 Zdeoc.... & he Degex’ing 6 Baáfvoc Tv: vedteQoc 8° gotiv 6 “AOnvatoc éxeivov; Dion. Hal. A. R. 1, 13, 1 &viea rev dexalov ovyyoagéwv ....Degextdnv tov “AOnvaiov **), yeveadéyov obözvög Sedtegov; and Euseb. Chron. Ol. 81,1 (456/5 B. B.) Ferecydes secundus historiarum scriptor agnoscitur 35). The same writer is » 34) tov “AOnvaiov Cas tv 'AOnvaiov Mss. The testimony is terminologically interesting and factually important: (1) As to the terminology, ovyyeaq~etcs means ‘historian’. It is used here in the same narrower sense of the word as e.g. in De Thuc. 5 and De imit. 2, 3 p. 207, 5 U.-R. We do not know whether in the Pinakes, where the historians had a section to themselves, they were listed as Zyvyyoageis or ‘Iorogwxot (for the evidence see F. Schmidt ‘Die Pinakes des Kallimachos’ Jacoby’s Klass.-philol. Stud. 1, 1922, p. 49 ff.). In Hellenistic times the term iorog.oyed@os (iotogixéc) is common, and used almost exclusively in the biographical source(s) of the Suda. Confusions of the general term ‘prose writer’ with the special term ‘historian’ are not lacking. (2) As to the matter, the words of Dionysios show that the Genealogiai of the Athenian were regarded as authoritative for the West. Pherekydes was the predecessor of Hellanikos (particularly for the Phoronis and the Hiereiai) and of Antiochos of Syracuse whose relations to him may (mutatis mutandis) be compared to the relations between Herodotos and Hekataios. The quotation further proves that the Genealogiai was accessible in any good library. It does not matter much whether Dionysios looked up the book himself or took the quotation from a source, which then was not Alexander Polyhistor (as Wilamowitz p. 131 n. 3 believes). About Polyhistor as the source for Italy and the West generally see Commentary on FGrHist. 273. | 35) The note is correctly affixed to Ol. 81,1 in L only; P shifted it to Ol. 81,2; B and the Armenian to 81,3. The Armenian Eusebios has preserved a second entry under the same year: Die Pythagorianer fielen in jene zeiten. It has never been doubted that the first note does refer to the genealogist; for secundus refers back to the entry under Ol. 59 (544/1 B.C.) Deperböns tovogiuxdc éyvweiteto IIvBayóeov Ötödoxakos (the last two words are missing in Hieronymus, but attested for Eusebios by Synkell. p. 451, 17; Dionys. Telm. p. 28, and the Arabic tradition; see Ph. U. 16 p. 211, 4). Both notes most probably derive from Porphyry who wrote about the Syrian (FGrHist. 260 F 21; see below p. 55 ff.; De antro Nymph. 31 = Vorsokr. 7 B 6 and Proklos ib. B 3). There seems to be some confusion here, though it may be due to excessive abbreviation. The date and the connexion with Pythagoras prove that the first entry refers to the theologian of Syros; the designation historicus shows that the second entry refers to the Athenian. If confusion there is, we cannot decide (and it is of no great impor-

meant by Cicero De or. 51-53 when he picks out three names of early historians from the same source from which Dionysios of Halikarnassos De Thuc. 5 took his long (though far from comcomplete) list of degatot ovyyeagetc: ‘age vero’ inquit Antonius ‘qualis oratoris et quanti hominis in dicendo putas esse historiam scribere’ .... ‘atqui ne nostros contemnas’ inquit Antonius ‘Graeci quoque ipsi sic initio scriptitarunt ut noster Cato, ut Pictor, ut Piso. erat enim historia nihil aliud nisi annalium confectio ... hanc similitudinem scribendi multi secuti sunt, qui sine ullis ornamentis monumenta solum temporum, hominum,' locorum, gestarumque rerum reliquerunt. itaque qualis apud Graecos Pherecydes, Hellanicus, Acusilaus fuit aliique permulti, talis noster Cato et Pictor et Piso®®). The Athenian ovyygawpeúc, yeveaAöyoc, or historian tance) whether it is due to a misinterpretation of ovyyeagets (see n. 34), or because later writers (not the Christians alone) were more interested in the Syrian and therefore understood him, whenever a Pherekydes was quoted. Nor is it possible to decide whether Porphyry himself is at fault: he had to deal with the alleged second historian of Athens (our no. 5). In any case, if Porphyry or Eusebios made a mistake they do not stand alone. Clem. Al. Strom 5, 8, 44 (FGrHist. 3 F 174) quotes ‘the Syrian Pherekydes’ for Dareios’ expedition against the Scythians, which is manifestly wrong (see p. 51 f.). A typical case is found in Tzetzes (FGrHist 3 F 136) who gives xarà Depexbinv tov Zúeov (sic) iotogıxöv, while his sources, the Scholia on Homer and Euripides, simply quote Depexdônc (viz. the genealogist). Perhaps the same confusion must be acknowledged in Ps.-Lucian Macrob. 23, who has “xaì Depsxböng 6 ZbprLog éuoiwoc óydonxovra mévre (scil. EEnoev Eryn), alongside. of Hellanikos who reached the same age, among the ovyyoageis who follow the philosophers. There is little doubt that the ‘historian’ Pherekydes is meant, and then we have to infer that the source of the Makrobioi (ultimately Apollodoros?) not only treated the two historians Pherekydes and Hellanikos under the same year—both their floruits are assigned to 456/5 B.C. — but made their lives coincide entirely. Wilamowitz (to put it crudely) rejected the whole of the Hellenistic tradition in favour of this late confusion of the Syrian with the Athenian. | 36) Cf. also Ioseph. c. Ap. 1, 13 (text in n. 29) who is interested only in ‘the dates of the earliest Greek historians. The Cicero passage is proof sufficient against the: assertion of Wilamowitz p. 131 ‘dieses werk (scil. the Genealogiai of the Athenian writer) ist von der antiken stilkritik ganz unbeachtet geblieben; es fehlt in der langen namenreihe der vorthukydideischen historiker bei Dionys. Hal. Thuk. 5’. It does not matter whether Dionysios forgot Pherekydes, or whether the name dropped out in the Mss. tradition. I cannot deal here with the list the preservation of which is notoriously bad. But in the face of the context of Cicero and Iosephus, who both speak about earliest historical writing, nobody can take seriously the

is called ‘the later’ by Strabo, ‘the second’ by Eusebios, and he is a round ninety years younger than his namesake, the 6sok6yoc of Syros.. . The exact date given in the Chronicle, of course, has an approx’- mate value only. We need not trouble whether it has been established by Eratosthenes, or by his successor Apollodoros; in any. case, 456/5 B.C. was one of the great epochal years in Apollodoros’ Xoovıxd, under which he grouped together a variety of poets, philosophers, and historians, most of whom he was not able to date more accurately because’ he had no documentary evidence. It is quite possible that he dated the several persons of such a group with due caution and that he even discussed doubts as to their real time 37); but mistakes were inevitable. It is sufficient to point to the fact that 456/5 B. C. is far too early for the floruit of Hellanikos and we cannot tell with certainty how Apollodoros arrived at this date 38). That applies even more to Pherekydes, of whose life it seems that no details were known. Modern scholars believe next words of Wilamowitz: ‘und wenn Cicero De or. II 53 eine ähnliche aufzählung mit Pherekydes beginnt, so hat der Syrier den nächsten anspruch, der ja bei Eusebios ‘historicus heisst’. This is the desperate pleading of a bad cause. No reputable author has ever called the Syrian philosopher a historian (cf. n. 6); whether Eusebios did is doubtful (cf. n. 35); and if he, or even Porphyry, did it would not matter. To complete the case against the attempt of Wilamowitz to minimize the evidence for the Genealogiai: it is not correct that ‘the grammarians too paid little attention to the ten “books of the Athenian’. Herodianos quoted the Syrian and the Athenian, and carefully distinguished between them (see n. 6); Apollonios Dyskolos. perhaps quoted the Syrian only, and Pollux (FGrHist 3 F 154) only the Athenian. But if F 46 belongs to the Syrian (as Wilamowitz believes), the case of Pollux would be the same as that of Herodianos. The Athenian is further quoted for rare words by Athenaios (F 13a) and Hesychios (F 25b; 164); in the Lexica of Kyrillos and Photios (F 49; 90a); in the Etymologica (F 76; 163); and by the authorities of Stephanos of Byzantion (F 5; 147; 160; 161; 172). It is not surprising that the number of quotations from the pure grammarians is less tham that of the quotations in mythographical. works and scholia for both of whom the Genealogiai was one of the most important reference books. It is the same with Hekataios, Akusilaos, and Hellanikos. For the Ionic dialect the grammarians had at their disponal a sufficiency of great literary writers, poets as well as prosaists. They rarely drew upon the early historians—which is a pity. The local historians in particular would probably have yielded much more if worked through thoroughly. 37) See for example FGrHist 244 F 7; 32. 38) R.E. VIII col. 107 ff.; FGrHist I p. 430 f.; II D p. 722. x

it natural that one ascertained the several dates, as far as possible, from chronological allusions in the works themselves; but it is not at all certain that even an Apollodoros used throughout or even primarily this method, which alone could have furnished him with certain, if not complete results. For example, the Life of Hellanikos abounds in synchronisms with Macedonian kings, with Athenian tragic poets, with his chief predecessor Hekataios; but it does not contain a single date from the works, although at least the Priestesses of Hera and the Affhis could have furnished such documentary dates. When we connect with the floruit of 456/5 B.C. the 85 years which Hellanikos is said to have reached we arrive for his life at the long period from 495/4 (496/5) to 411/0 (412/1) B.C.; and we state at once that the computation is contradicted by the fact that the Afthis included the year 407/6 B.C., and probably the whole Peloponnesian War 39). This makes it appear dubious whether Apollodoros has made an attempt at all to determine the floruit of Pherekydes from the ‘Iotogiar, which by their very nature furnished few, if any dates serviceable for this purpose. Nor can we correct Apollodoros’ date as to a certain degree we can his date of Hellanikos; for there are four fragments only which might bear on the question, viz. the pedigrees of the Philaidai (F 2), of Daidalos (F 146), of the Asklepiadai (F 59), and the expedition . of Dareios against the Scythians (F 174). Of these four two must be eliminated at once: in F 59 the imposing sequence of authors — *Eoatoo8évyns nai Degexting xai 'Asrokkódwgos xai “Agetog 6 Tagcete — in the Vita of Ps. Soranus, written not before the first century A.D., neither proves that Pherekydes gave the whole pedigree down to the great Hippokrates, nor that he gave it for his sake at all 40). F 174 from Clement, quoting the Syrian (not the 39) FGrHist 4 F 171/2. With this question I shall deal extensively in my English edition of the Atthidographers (The Ancient Historians of Athens, to no. 1). . Introduction 40) See FGrHist I p. 409, 38 ff.; If D p. 750, 24 ff., and recently L. Edelstein R.E. Suppl. VI, 1935, col. 1295, 27ff. who in a careful investigation traces back the so-called Soranus to Philon of Byblos and Hermippos’ of Berytos book Tegi tav da wardciav oeuvnOévrov évôólov aview@v iatowv. To get rid of F 59 by assigning it to the Lerian Pherekydes, as Lipsius does (Quaest. Logogr. 1886, p. 17ff.), is a subterfuge. But neither can I take seriously the argumentation of Wilamowitz p. 136: ‘von den koischen

Athenian) Pherekydes, must in my opinion almost certainly be assigned to the Lerian 41). Anyhow F 174, of dubious authorship as it is, is less important for the problem of the time of Pherekydes than F 2, to which not enough, if any attention has been paid 42). It refers approximately to the same period as F 174; for the pedigree ends with Miltiades 6c &@rxioe KXegedvnoov, and it never went farther down; otherwise Didymos would have copied it to the end, as he wishes to prove that the historian Thukydides was an axéyovoc tev svdoxupwrat@mv oteatnyav, A€yw 5é tHv megi MiArtıdönv xai Kiuwva **). We need not deal here with the several difficulties in the tradition of this pedigree; and one may even ask whether dc œuuoe Xeepoövncov means Miltiades I or Miltiades Asklepiaden hatte man schwerlich veranlassung einen stammbaum zu geben bevor ein koischer Asklepiade zu allgemeiner beriihmtheit gelangt war; der Asklepios von Kos ist ja erst durch die koischen ärzte berühmt geworden’. There can be no doubt that any genealogical work had to mention Asklepios, and his descendants as well, because they played a part in the Iliad and in the Nostoi. In fact, his birth occurred in the first, and his death in the eighth book of Pherekydes (F 3; 35, cf. n. 48). Wilamowitz himself apparently had not much confidence in his argument, for he declares that the whole passage probably was a later addition (‘nachtrag’). But if it was, it loses at once its value for the dating of Pherekydes. I leave aside the computations of v. Gutschmid (Kl. Schr. IV p. 299) who on the strength of F 59 and some dates for the activity of Hippokrates dated the death of Pherekydes in c. 423 B.C. Nor is there any need to refute the use which he made of F 111. If Pherekydes and Antimachos (F 58 Wyss) agree about ~ Herakles’ share in the adventure of the Argonauts, the natural inference is that the learned poet followed the genealogist, and Wyss in his commentary simply refers to FGrHist I p. 419, 24 ff. This furnishes us with a terminus ante for Pherekydes which is not very helpful, as Antimachos did not flourish before the turn of the fifth to the fourth century. His principal “works, which interested Plato, may well have been published in the first u decades of the fourth century B.C. 41) See p. 48ff. | 42) It may be sufficient to refer to Momigliano (Lc. p. 348) who in dating the Athenian Pherekydes uses only F 174 which he maintains (for a quite insufficient reason) ‘deve esse posteriore alla spedizione Scitica dt Dario, ma anche all’inizio delle guerre Persiane, quando appunto tali anedotti si dovettero diffondere nel mondo Greco, richiamati dall’interesse per le cose Persiane’. 43) It must remain open to discussion | (but see on FGrHist 3 F 60) what Hellanikos said év tH éxryoapouévnr "Acwnidı (FGrHist 4 F 22). Didymos quoted him after Pherekydes, but Marcellinus did not copy the passage.

Il *), whom xartakauwdgevov ra refñyuata éxi Xegoovicov azrootéAAovor terjest où Terovoteatibias *°). But there is one certain inference to be drawn: the fact that Kimon is not mentioned gives us a definite terminus ante. We may safely assume that the Athenian genealogist carried the pedigree of the great clan down to his own time, or near to it; and if he does not mention the victor of Eion and Eurymedon, who in 476/5 B.C. brought back to Athens the remains of Theseus, he must have written his Genealogy before ‚that, year 46). We cannot determine the final year more exactly firstly because of the uncertainty about the last Miltiades mentioned in the pedigree, and secondly because the Persian war is not mentioned. The simple reason for the second fact may be that the work was written before 490 B.C.. There may be (though I am not inclined to believe it) the concurrent reason that for the contemporaries the battle of Marathon was a victory not so much, if at all of Miltiades (who was regarded with a certain degree of distrust in some quarters) but of the polemarch Kallimachos, who claimed it in the additional epigram inscribed on the statue of Hermes dedicated by him before the battle 47). But perhaps we had better be cautious, and content ourselves with a statement that cannot be doubted: viz. that Pherekydes not only carried down the pedigree to-at least the second half of the sixth century, but that by making additions to the individual names he set forth the titles to glory of the Philaid clan. The settlement of the Philaidai in Athens confirmed the claim of Athens to Salamis, acquired not very long ago; to the Philaidai the Athenians owed the creation _ 44) We have to put the question, because the Mss. of the pedigree are not free from corruption, though personally I agree with Wade-Gery (who believes Miltiades II to have been the adopted son of Miltiades I) that for Pherekydes the oixıorng can only be the first Miltiades. 45) Hdt. 6, 39, 1. It is evident that a new light is thrown on this piece of tradition too by the fragment of an archons’ list in the Agora (nv. I 4120; Meritt Hesperia 8, 1939, p. 59). 46) The Athenian genealogist was interested in Theseus and treated his story in detail (3 F 148-153). It is unfortunate that none of the fragments has the number of a book. I do nat doubt that for his narrative he used the Theseis which had been published not long before: the Scholion on Pindar Ol. 3, 50 (= 3 F 71) states their agreement in a fact from the Herakles story. 47) See Hesperia 14, 1945, p. 158, 4.

(or at least the enlargement) of their greatest national festival; and lastly the Philaidai extended the sphere of Athenian influence in a direction vital for the interests of her life and her commerce. It does not matter that the first two claims were not uncontested, and that about the third there were widely different opinions in Athens. The pedigree in itself and standing as it does in a myfhic history of Greece justifies the inference that Pherekydes stood in close relations to the Philaidai 48). If it is permitted for a moment to indulge in fantasies, we might suppose that he went to the Chersonnese with Miltiades II and returned with him to Athens in 493 B.C., or otherwise that he himself (a supposition not wholly impossible chronologically) or his father before him, belonged to the settlers who (after the battle of Pallene?) went to the Chersonnese with Miltiades I. At any rate Pherekydes wrote his ‘historical’ work in the time of Miltiades II. For F 146, fortunately also ‘48) One almost automatically compares with the pedigree of the Philaids the passage in Hdt. 6, 131 where he carries the pedigree of the Alkmeonids down to Perikles. But the divergences are greater than the resemblance: -the mind of Pherekydes is directed towards the point of departure of a pedigree, in which Telamon, Aias, Philaias occur, which is part of the great Aiakos pedigree, and at the head of which Zeus himself is standing. Herodotos with the words où 5% ’AAxusovidaı foav piv xai ta dvsxadev Aauxçgoi év trou 'AOfvnor (6, 125, 1; cf. about the Philaids 6, 35, 1) lightly passes over the pre-history because he is primarily interested in the time when the immediate ancestors of his contemporary Perikles fought against the tyrants and the Persians, and because in this special case he wishes to refute some recent and definite attacks on the Athenian patriotism of th2 Alkmeonids. The point of view expressed in the preface éy@ 5é xepi tobtwv oûx Egyonar éogéwv «th. (1,5,3) dominates his whole work even when he reaches back occasionally into epic times with which it is the task of genealogy to deal. For Pherekydes as for all genealogists just this touching upon facts of contemporary history is the exception, and we had better treat these passages as such. We are unable to decide how often Pherekydes admitted contemporary allusions. The Daidalos fragment used above in the text is another example; and the pedigree of the Asklepiads may be a third, though this is far from certain (see p. 29). Even more uncertain is Apollodoros FGrHist 244 F 162 who evidently excerpts an Ionian author: viòv ÔÈ xaralımeiv (scil. tov SedtEQov Odpvew) "Avtioyov, ÉE où xaì Tis EA* *Ov<yatodc> Ev "AOjvats vevéoBar tovs 'A<vroytôas? >. Here Hellanikos has the same or even a better claim than Pherekydes, whom Wilamowitz (GL. d. Hell. Il p. 16 n. 5) believes to be the author quoted by Apollodoros. These exceptions may each have a special reason, but we are seldom able to find out what it was.

PROSE WRITER 33 a verbatim quotation,.and even less noticed than F 2, provides us with a definite ferminus post: Mntiovı ôè tat 'EoeyOéoc xai "Igıvönı yivetar AatÖakoc, aq’ où 6 Ôônuos naAcitar Aadadridar -Aönvnoı. There really can be no doubt that the Mistoriai were published after 508/7 and before at the very latest 476/5 B.C. 49). We shall find a corroboration of this early date in the use of the Ionic dialect and in the mental attitude of Pherekydes 50). For the moment it is sufficient to state that when Apollodoros made Pherekydes and Hellanikos contemporaries by fixing both their floruits in 456/5 B.C. he dated the latter far too early and the former far too late. We return from our own inferences from the fragments to the biographic tradition. It is an undisputed fact that at least from Eratosthenes onward the Athenian genealogist was universally distinguished from the theologian of Syros, and the conscious distinction of the two writers may have been attested still earlier if (as seems to be the case) the enumeration of ‘ancient historians’ used by Cicero and Dionysios ultimately derives from Theophrastos, who then knew also the ‘historian’ Pherekydes, as without any doubt he knew the theologian. On the other hand we do not know of anybody who identified the theologian and the genealogist, and again judging from the tradition preserved to us, I believe that we are justified in stating that nobody did. Against the theory of Wilamowitz that before Eratosthenes the Pentemychos and the 49) This result agrees fairly well with the date which Iosephus (n. 29; 36) has for the early Greek historians generally and the Suda for the Lerian Pherekydes: yeyovds ned Odiyou tic oe éAvuuziäôos (480/79 B.C.). But we cannot say whether the source of the Suda (whichever it was) actually meant the Athenian genealogist, and if it did how it arrived at its date. It is quite possible that yeyova@c does not refer to the floruit but to the birth of this Pherekydes, and in the latter case the date perhaps would not be at variance with the Apollodorean floruit 456/5 B.C. Wilamowitz p. 143 is content with dating the work ‘on account of its form and its contents before Herodotos, apart from later additions’. Laqueur (R.E. XIX 2, 1938 col. 1991, 36 ff.) gives ‘the first half of the fifth century’ and (col. 2019, 10) ‘the time of Kimon’, and seems inclined to believe in a documentary reason for the floruit in 456/5 B.C.. W. Schmid Gr. Lit. II, 1934, p. 707 does not date the Athenian at all. About v. Gutschmid, who put Pherekydes’ life between c. 508 and c. 423 B.C. see n..40. 50) FGrHist 1 p. 386, 31 f.; Momigliano l.c. p. 351. See below p. 34 f.; 62. Mnemosyne XIII

Historiai were regarded as having been written by the well-known theologian of Syros, and that it was Eratosthenes who first attributed the latter work, though it was written in the Ionic dialect, to an Athenian author invented by him for this purpose, it may be sufficient to state that such an unqualified assertion on the part of Eratosthenes is quite inconceivable; for that it was unqualified we learn from him and from the witnesses dependent on him: Wilamowitz’ assumption contradicts all we know about Eratosthenes’ method e.g. from his history of geography, not to mention the method of Kallimachos in the Ilivaxecs, and the opinions as well as the prejudices of ancient scholars generally. Even if in a work of ten books Athenian mythical history had been treated extensively, „this fact (which is by no means so self-evident as Wilamowitz seems to believe 51)) would never have been a sufficient reason for any ancient scholar for making the author an Athenian if the 51) The evidence of the fragments makes this first assumption rather doubtful. The facts (never clearly stated) are these: the pedigree of an Athenian clan occurred in the first book (F 2), but evidently in connexion with the Zeus-Aiakos stemma (cf. n. 48). The other fragments pointing to Athens (cf. n. 52) are without the number of a book. They refer to the goddess Daeira (F 45), Triptolemos (F 53), Daidalos (F 146), Theseus (F 148-153), and Kodros (F 154). We may perhaps add the Thriai F 49 (though they are not exclusively Attic), the rape of Oreithyia F 145, (although the locality remains vague), and F 102; 142; 155, probably alt referring to the Ionian migration. Taken together these fragments may confirm the tradition that the author of the book was an Athenian, but . they are certainly not sufficient to support the thesis that Eratosthenes invented an Athenian author for it. We do not know how much space in the ten books was required for Athenian mythic history, but we should not overestimate it; it probably was relatively less than in the Bibliotheca. It should be noticed that according to F 2 the Panathenaia were established under the archonship of Hippokleides 556/5 B.C., not in the reign of king Erichthonios, as the Atthides have it. This (though an incidental remark) makes for an early writer, and it explains why we do not meet Pherekydes’ name in our ample evidence for the institution of this festival. Nor is he ever quoted for Kekrops or any other of the pre-Thesean kings, apart from Erechtheus who appears in the pedigree of Daidalos (F 146). It is improbable that he gave the whole series of pre-Trojan kings which was first arranged by Hellanikos. We infer from the evidence of the fragments that Pherekydes did not know much more about early Athenian history than Herodotos who may have read the Genealogiai (cf. W. Schmid op. cit. ‘p. 554 n. 9). In the formation of the Athenian archaeology he represents together with Herodotos .the first, or (as we most probably must date Hekataios before him) the second stage.

: THE FIRST 35 dialect. was at variance with the assumption of Athenian origin of the work. The dialect was the first and decisive point noticed by literary historians and critics of style: Dionysios of Halikarnassos distinguishes two groups among his dexaioı ovyyoageic, viz. ot re thv [dda srooeAönevor Öudkextov thy Ev toicg TöTE YXoÉvoic udAtota GvOodoav xai oi Tv degatav 'ArTO(Öa pixeds Tuva Éxovoav ötapopäs mag Tv “ArOida. The difference assumed between ’lés and dexaia ’Arôis may have been ever so slight, noticed it was. And so far as we are able to judge from the few verbatim fragments, which have passed through many hands and the linguistic form of which is sometimes more or less obliterated, there is no noticeable difference as to the dialect between Pherekydes and e.g. Hellanikos. 52). If Eratosthenes called the genealogist an Athenian (and we know that he did) the only possible assumption (in strict opposition to the theory of Wilamowitz) is that the Historiai came to Alexandreia as the work not of a simple DBegexúôns, or for that matter of the Syrian Pherekydes, but of Pherekydes the Athenian. It is quite conceivable that an Alexandrian scholar would have doubted the inscription, just because the book was written in the Ionic dialect; but then he would have noted his doubts in the usual manner ei vyvñoiov tod ovyyoaæpéæoc TÒò fiBAiov, or in a similar form 53). We may even go farther: the doubt may have been justified either fully or partly. An author who in his preface called himself ’A@nvatog need not have been a native Athenian: Herodotos called himself @oúgtoc, though he was born in Halikar52) We can also compare Hekataios and Akusilaos, not the artificial language of Herodotos. It may be noticed that the fragments show no hyper-ionisms. But considering the many stages through which they came down to us nobody will rely on ®8tav (F 1), ToAvidea (F 8), taig trois (F 18), and the like. Wilamowitz (Aristophanes Lysistrate 1927, p. 32 n. 2; s. also Kl. Schr. V 2 p. 143) suggests that the Athenian passages may have been written in the Athenian dialect. This is a consequence of his regarding the Genealogiai as a ‘compilation’. If he wishes to imply that the passages which contain Athenian subject matter are all later additions | cannot follow him at all (see n. 51). But it is in my opinion a very likely suggestion of Wade-Gery, that Denys’ source by dexata *“ArOics indicated the voluminous work of Pherekydes of Athens; for there is no doubt, that this source registered this writer (s. n. 36). 53) For the formulas used by Kalimachos and his successors cf. F. Schmidt Die Pinakes des Kallimachos 1922, p. 21 ff.; 79; 91 ff:

nassos 54). And even a forgery of the fourth century cannot be excluded forthwith. The question must be considered, though we have no reason to suspect a forgery. Also as to the nationality of Pherekydes, nothing favours the assumption that he was an adopted Athenian, while something may be said for Athenian origin 55). In any case the Lerian Pherekydes must be kept out of the discussion, for at least Biography proves him to have been a much later writer 56). As to the use of the dialect, if the genealogist was a native Ionian and an adopted Athenian it was as natural for him to write in the Ionic dialect as it was for Herodotos; if he was an Athenian by birth he wrote in this dialect because in the first decad(s) of the fifth century Attic could not be used for a work of literature 57). Akusilaos at about the same time did not write in the Argive dialect, though local historians of Argos in the fourth century did 58); nor did Hellanikos write Aeolian in the second half of the fifth century; and all medical books of the schools of Kos and Knidos are written in Ionic. These are well-known facts. Not one of the negative arguments with which Wilamowitz in passing tried to support his a priori thesis is in any way convincing, and positive proofs he did not, and could not furnish. If he insists that ‘nothing was known of the existence of an Athenian writer Pherekydes’ we may well ask what was known for example of Dionysios of Miletos or Akusilaos of Argos, to mention only two distinguished writers of the first half of the fifth century, apart from what they said about themselves in their prefaces. Incidentally 54) R.E. Suppl. II col. 205 ff. 55) See n. 51. 56) See p. 48 ff. | 57) See Lipsius Quaest. Logogr. 1886, p. 19; Jacoby FGrHist I p. 386, 29 ff.; Momigliano Lc. p. 350. In answer to the repeated request of Wilamowitz for another example of an Athenian writing in the Ionic dialect (Kl. Schr. V 2 p. 144; Aristophanes Lysistrate p. 32 n. 2; in Gesch. d. gr. Sprache 1928 p. 16 ff. and p. 24 he burkes the problem) we may state that the request is not justified for reasons of method so long as we know of one Athenian prose writer only in the first half of the fifth century. There would be much more cause for demanding a parallel if an Athenian of that time had written a voluminous work in Attic. Before Thukydides the language of literature was Ionic. 58) Derkylos FGrHist no. 305.

we miss the former in the long enumeration of the ‘ancient historians’ in Dionysios’ Treatise upon Thukydides, just as we miss the name | of Pherekydes; and for Dionysios we have no other witness, while we know from Cicero that Pherekydes, mentioned by Dionysios as an ‘ancient historian’ also in the Roman Antiquities, has been in the source from which Dionysios got his list. What was known even about Herodotos or Thukydides apart from the facts to be found'in their works? It was Wilamowitz who in an authoritative paper showed up and explained the miserable condition of the biographies of the great prose writers of the fifth century 59). It was Wilamowitz who in the paper on Pherekydes 60) rightly said ‘the name of an author is guaranteed only when it occurred in the text at the opening of a book’, but failed to prove that this was not the case here. For an unbiassed scholar the very fact that Eratosthenes spoke without reserve about the ‘Athenian genealogist’ is proof sufficient that he was listed as such in the Pinakes, and that makes it at least very probable that the name and the native country occurred in the introductory words of the Historiai. But, says Wilamowitz 61), ‘the person of the Athenian is so completely forgotten that he has not even a father’. Again for an unbiassed person this point is proof only for the well-known fact that the authors of the Pinakes strictly adhered to the statements which they found in the Mss. Kallimachos sometimes (and particularly in questions of authenticity) may have remarked on these facts 62), but idle speculations he left to biographers who were less conscientious, or to modern scholars who like to ‘guess at the unattainable’. Of course, the lack of a father’s name is never a reason for doubting the existence of an author, or the authenticity of his work. Alkmaion and probably Akusilaos in their prefaces called themselves by their own names, their fathers’ and their native coun59) ‘Die Thukydideslegende’ Herm. 12, 1877, p. 326 ff. For the foundations of the biography of Herodotos see R.E. Suppl. II 50) p. 132. 1913, col. 213 ff. 61) p. 131. 62) The attribution of Hekataios’ ’Acin to the author of Evoewxn (RE. VII col. 2672, 66 ff.) is a case in point. It seems probable that Eratosthenes (Strabo 1, 1, 11) simply followed Kallimachos (cp. p. 47).

tries 63); Hekataios, Herodotos, and Thukydides did not 64), and the same seems to hold good for Hellanikos, as the Suda gives. two or three father’s names for him 65). Obviously, there was no fixed rule, though it seems. that in the course of the fifth century, when writers began to envisage a pan-Hellenic public, the father’s - name began to be dropped though even then there seems to have been no fixed rule 66). Later local research found the father’s name for Herodotos, just as the periegetai of Athens tried to establish the kinship of Thukydides to the house of Kimon. If they tried at all to do the same for less famous writers like Pherekydes they either found nothing or were unable to identify their man 67). Finally and lastly: I totally fail to understand how the fact (let us assume that it is a fact) of the work not being uniform or well arranged can be made to bear on the question of the author’s personality 68). 63) Vorsokr. 24 [14] B 1 ’Adxpaiov Kootwvintns téde Eieke IIeugí0ov vióc; FGrHist I p. 375. 64) FGrHist 1 F 1 ‘Exatatog Mumoroc de pvGcitar; Hdt. 1, 1 ‘Heoddédtov Ooveiov iotogins anmödeEıs Hôe. Thuk. 1, 1, 1 Oovxvötôns ’AOnvatos Evvéyeawe tv wölznov. 65) Suda s.v. "EAAdvıxog” MutiAnvatoc, iotopixés* vidc ‘Avôoouévoug, ot dé “Aqutopévouc, oi Öè Zxäuwvoc, où Su@vupov Eoxev vióv. The first two are variants of the same name, but the third looks like an inference from the name of his son, Zxœuwv “EAAavixov, author of a book on Lesbos (cf. R.E. VIII col. 106; III A col. 437). 66) See the case of Charon of Lampsakos FGrHist 262 T 1. 67) About the name of Pherekydes in Athens see n. 19. 68) This is in fact the chief argument of Wilamowitz (l.c. p. 132 ff.). Again his reasons are far from convincing, and: again he himself does not seem to put much confidence in them, since he makes great play with later ‘additions’. In order to show up ‘contradictions’ in the Genealogiai (which, if they really existed, can be easily explained) he makes a surprising use of the torogia: of the so-called Mythographus Homericus, not only those bearing the name of Pherekydes, but even of anonymous ones, though he himself appreciated it that—following the famous investigation of Eduard Schwartz (Jahrb. Suppl. 12, 1881, p. 405 ff.)—I treated. these quotations with the utmost reserve. The consequence is that Laqueur (n. 77) draws his picture of Pherekydes almost wholly from them. Wilamowitz further finds a ‘disproportion’ in the fact that ‘Herakles and the Argonauts together fill almost the fifth of the whole work’. But the same two myths fill more than a quarter of the Bibliotheca. I leave aside minor points. Of course, I do not regard as final my own reconstruction which relied mainly on the constancy of a literary genos.

I believe we may safely assume ‘that the work did not write itself 69). The ten books under the name of Pherekydes must have had an author, even if (and Wilamowitz does not think this ‘impossible’ they contained stories of different origin and different value, and told them without a real connexion as Wilamowitz asserts (for insufficient reasons as I venture to say) they did 70). The problem 69) One sometimes has the impression that Wilamowitz believed it did, to judge for example from his brief characterization of Pherekydes’ style (l.c. p. 141; but cp. below n. 73) “Wir finden bei Pherekydes einen stil, der keinem individuellen menschen zusteht, sondern der märchenstil der pvGodoyin ist...dieser naive märchenstil ist in allem der bunten fülle und redseligkeit des epos entgegengesetzt, ein schöner beweis für die stilsicherheit hellenischer rede. Ebendarum ist er ebensowenig persönliches eigentum eines erzählers wie der epische’ (the italics are mine). In my opinion the extensive use of the fairy-tale style and, for that matter, the great number of fairy tales, is typical of this writer. Hekataios and Hellanikos do not seem to have used that style much; perhaps Akusilaos did, as the Kaineus story seems to imply (I do not pretend to understand the doubt of Wilamowitz Griech. Verskunst p. 43 n. 1). Actually in questions of style it does not help much to speak of ‘the’ Genealogy: there are merely individual genealogists who have certain traits in common as e.g. the arrangement of the subject-matter by pedigrees, and its restriction to the contents of epic poetry. A sentence like this ‘die genealogen bezwecken 6t8aoxaAia hier (scil. by Pherekydes) wird pwuyaywyia angestrebt’ does not seem to make much sense considering the state of our knowledge of the single authors. If we had more we should perhaps be able to distinguish more accurately between the scientifically-minded Milesian, the historically-minded Hellanikos, and authors like the more local writers Pherekydes and Akusilaos, between whom a certain resemblance seems to have existed, though as to the subject-matter the Argive writer was perhaps more interested in the theogony than the Athenian. In any case, wuyaywyi« and di6acxaAta do not exclude each other. I leave aside the characterizations of Pherekydes by Aly (Volksmärchen, Sage und Novelle 1921, p. 236f.) and Laqueur (l.c. col. 2019, 58 ff.; cf. n. 77), both more or less fantastic in my opinion. For the . former the apparent simplicity of Pherekydes is not simplicity at all but ‘wissenschaft am untauglichen objekt’, and the style in the more detailed narratives is what in Germany one calls ‘ein papierner stil’, which ‘can derive only from books, not from the fresh fountain of oral tradition’, because Pherekydes ‘does not know’ what Aly mysteriously terms ‘the logos’. For Laqueur he is a teacher of his people who did not care so much for collecting the mythographic stuff but who ‘durch form und inhalt wie er die mythen erzählte hat weltanschaulich wirken wollen, genau so wie Hesiodos oder Aischylos und Pindar’. 70) L.c. p. 140. Here I for my part should be grateful for a parallel from the fifth, or even the fourth century. What Wilamowitz describes are Yrouviuata and ‘Iotogia: of Hellenistic times.

concerning the primary nature of a literary work and its subsequent fortunes in the course of tradition through the centuries is quite different from the problem concerhing the author. The former problem must be divided into two subquestions which are closely connected, viz. the question concerning the literary achievement of _ Pherekydes and the question concerning the history of the text. As to the first sub-question it would, of course, not matter at all, if in the work ‘plain and unadorned passages’ 71) alternated with fairy-tales narrated in detail, and Pherekydes obviously was rich in fairy-tales. Such a disparity would not constitute a difference from Hellanikos or even Herodotos, let alone the dexaioı ovyyeaqetc of Dionysios, for whom the narration of u%0o1 among the plain facts is a characteristic feature. As far as we can see such circumstantial stories were in many cases, if not always, part of a pedigree, which in its main part consisted in a mere series of names: they were attached to one or several names in the pedigree, thus forming what technically one may well call a digression 72). For Herodotos, who did not admit fairy-tales in the strict sense of the word, it is sufficient to call to mind the position of the novellistic tale of Gyges, which in the first section of his Lydiaka interrupts ‚the rather dry sequence of the Lydian kings with its short and mostly chronicle-like entries. Otherwise the comparison between the historian and the genealogist is, of course, not a good one. But let us suppose for a moment that we had only 180 partly brief and ' 71) ‘Ganz diirre stiicke’. 72) Examples of this kind of arrangement are easily available, not only. from the late Bibliotheca. For Pherekydes we may refer to F 1 and F 60 which provides the context from which the former is taken; to the great Perseus story F 10-12, the opening of which has been preserved in paraphrase: “Axeictoc yauei Evovôtenv tiv Aaxedaipovos’ tev 52 yiverar Aaván ; to the now brief additions in the pedigree of the Philaids (F 2); to a secondary pedigree like F 20 or F 146. Other examples are furnished in my commentary. We may compare with them Akusilaos FGrHist 2 F 22 and Hellanikos ib. 4 F 19b (both verbatim and from papyri), or the Kodros story 4 F 125. It would be a suitable task for a disciple of Dornseiff to follow up the development of this stylistic phenomenon, which is not restricted to Greek literature and not to Genealogies. Mutatis mutandis it is also found-in ethnographers and periegetai. For the latter we have a late example in Pausanias, as we have the Bibliotheca for the Genealogies, and here K. Robert made the fundamental distinction of Bewejpare and Aöyoı.

rarely verbatim fragments of Herodotos’ nine books: it would be easy with the objections raised by Wilamowitz against Pherekydes to delete the literary personality of the author, and to maintain that. a work including passages of so different.a character and style cannot have been ‘uniform’ or ‘well arranged’, thus arriving at the result that ‘at the utmost we could imagine a man who arranged the various pieces’, and that even this modest supposition ‘is contradicted by the lack of an accomplished design’ 73). Or does anybody think that he could recognize and restore the design of Herodotos even in a rough outline from 180 fragments selected at random, and with the help of the remains of Ktesias, Theopompos and Diodoros? I am of opinion that it would perhaps be easier to restore the arrangement of Pherekydes’ work because apart from the 41 fragments with the number of a book attached to them we have a late descendant of genealogical writing in the Bibliotheca, and from the. fifth century at least the titles of Hellanikos’ genealogical works, which treated the great pedigrees in special books, nine in all. So far as we can judge genealogical works had a certain scheme of arrangement: they gave the pedigree, which is naturally dry because it consisted almost exclusively of proper names, and they interrupted this sequence of names by digressions, short remarks, and circumstantial tales; they consisted, one may say, of ovéuuata and u%0o1 The objections of Wilamowitz rest on a misunderstanding of this method of arranging the subjectmatter: because he did not recognize this method he inferred that the work was without order and without a leading idea, and consequently that it had no real ‘author’. It is hardly possible to refute 73) L.c. p. 142 ‘höchstens könnte man an einen ordner des ganzen denken, der vorhandene erzählungen verschiedener art zu einem ganzen vereinigte; aber dem widerspricht der mangel eines durchgefiihrten planes’. The following sentence (in which the italics are mine) I must confess I do not understand at all: ‘natiirlich hat doch ein mann einmal die sammlung veranstaltet und unter den namen Pherekydes gestellt, den vorher manche, sicherlich nicht alle, erzählungen getragen hatten; da die erzählung ganz unpersönlich war, verlangte niemand nach dem verfasser’. These are the ‘almost inevitable consequences’ of a hasty a priori theory: Wilamowitz has to find an explanation for an assumption that cannot be explained, namely that somebody at some time made a collection of myths which had neither style nor order, and inscribed it with the name of the theologianphilosopher Pherekydes of Syros.

properly a vicious circle based on an a priori assumption. But I cannot admit Wilamowitz’ final statement 74) that his conception ‘is better in accord with the general development of early prose writing and the fate of early books than the tradition (for a tradition it is, not an inference of Eratosthenes) about the Athenian Pherekydes, author of a comprehensive genealogical work of about the same size as the special books of Hellanikos. The Alexandrian scholars published this work in an edition of ten books 75), evidently because they valued the contents highly. In fact, it was used extensively and permanently as an authoritative source for mythic history not only by grammarians, commentators, theologians, but demonstrably also by Kallimachos 76), Apollonios, and probably other learned poets. I cannot deal here with the second sub-question, viz. the fate of the ancient book between the time of its publication by the author.and the Hellenistic edition, that is the first period of its textgeschichte. This would mean a new investigation of the fragments, which would probably alter my text in many details, and moreover a detailed review of Laqueur’s conclusions, which for their greater part appear fantastic to me 77). But I think I can 74) L.c. p. 144. 75) Even this editior gets a bad note from Wilamowitz: ‘diese sammlung ist nach Alexandreia gekommen und dort höchst ungeschickt in zehn biicher geteilt, ganz ohne riicksicht auf den zusammenhang, also nur auf gleichen umfang’. This opinion is selfcontradictory and can be explained only from the subconscious tendency to discredit everything connected with Pherekydes the genealogist. If the work was such as Wilamowitz believes it to have been, a big chunk of prose without a leading idea or a clear plan, the Alexandrian scholars could not be expected to re-write it. If it was not—and I am sure from the fragments with the number of a book that it was not—there was no reason for them to treat it in another manner than other bulky books. One has often remarked on the ‘clumsy’ division of Herodotos’ work into nine books which apportions a uèv— de sentence to different books. 76) See e.g. on FGrHist 3 F 92. For the use which earlier poets (and historians) had made of the Genealogiai see n. 82. 77) L.c. col. 1991-2025. Laqueur, making use of the compiled iotogia (see n. 68) applied to Pherekydes the same method as to Polybios and Thukydides. It consists in the distinction of a ‘grundquelle’ and a ‘nebenquelle’, or a ‘first draft’ and ‘later additions, which, where we are able to control them, in length surpass the first draft’ (col. 2019, 18 ff.), though not a word was altered in the latter (col. 1995, 11 ff.). The first draft in

FIRST ATHENIAN PROSE WRITER 43 predict the results for the history of the text, modest though they are compared with the claims of Laqueur. It will appear that the text of the Historiai came to Alexandria in one or more copies which (probably and apart from every-day corruptions) were interspersed with additions, variants, and perhaps insertions of whole stories. We may imagine that the copy (or copies) bought by the Egyptian king had formerly belonged to a tragic poet or a genealogist, or a mythographer, who had used it for his work; let us say (only in order to give an idea of the existing possibilities) that the copy of Asklepiades of Tragilos was among them, or that it was his copy. This writer who ‘tragoediarum argumenta non narravit, sed historiam fabulosam, qualis post tragicorum curas longe alia reddita erat atque qualem ex epicis lyricisque carminibus narraverat’ 78) had, at least according to our current texts, quoted Pherekydes 79). In any case, we have to assume that in pre-Hellenistic times too ample use had been made perhaps particularly by the tragic poets of the book which contained much material for his opinion was non-Attic, the additions, made considerably later, were Attic (col. 2016, 5 ff.). Between Pherekydes and mere compilers, like Diodoros and Nikolaos of Damaskos, there is this difference, which places Pherekydes at the side of Thukydides (col. 2019, 18 ff.): he is the representative of a ‘weltanschauung’, or rather of two (see n. 69); and as he is the author of the first draft as well as of the ennlarged edition we have to acknowledge ‘two periods in his literary activity’ (col. 2008, 18 ff.) in the interval of which Pherekydes changed his ‘weltanschauung’ under the influence ‘of the Attic world, which surrounded him’; ‘auf eine einfache formel gebracht lautet die weltanschauung des friihen, des ionischen wer kes: ohne die götter, ja auch gegen die götter; die des späten, des att schen: mit den gGttern’ (col. 2020, 65 ff.). 78) See Wilamowitz Analecta Euripidea 1875, p. 181 n. 3, cf. FGrHist I p. 484. If we exclude the tragic poets, the same characterization applies to the genealogists who tà ‘Houddov pethAAcéav eis aebdv Adyov xai de tira ëEúveyxav, as EdunAös te xai ‘Axovoilaoc of iotogroyedwpor (Clem. Al. Strom. 6, 26, 7 = FGrHist 2 T 5). Granted individual differences (cf. n. 69) it applies to Hekataios and Pherekydes as well. 79) Schol. Pind. Pyth. 2, 40b — FGrHist 12 F 3: ’AoxAnmıaöns Öë év Y Toayoidouuévov ottw yeéper” ,,xoootorogodor ÖÌ Zvior Gti xai uavein Ô TElov, ds Peoexböng, xai Tv émi rod teoxov 5é xöAacıv adtae Tape yxEyerohxacw” xt. But the quotation des Depexúòns probably is a marginal note, as Asklepiades usually quotes his authorities anonymously. Drachman’s transposition of the quotation, approved of by Wilamowitz (KI. Schr. V 2 p. 57 n. 1) seems wrong to me.

the mythic history of Athens as well as for other towns 80). Otherwise the unwieldly work would hardly have been preserved through two centuries. For it is hard to believe that Kallimachos was the first to recognize its value; nor should I take seriously the assertion that the tragic poets did not read prose books. As things are now, after Wilamowitz wholly denied the existence of the Athenian writer, while on the other hand Laqueur drew a full picture of him, describing his method and even the development of his weltanschauung 81), a special edition has become necessary which must give more attention than I can to the use that has been made of Pherekydes in the fifth and the fourth centuries. Such an edition, it is to be hoped, will arrive at more and more certain results 82). It requires a very full mythographical 80) But see n. 51. 81) See n. 77. 82) The investigation must be conducted separately for each fragment or for each myth, and it is never simple. Nor will the results be equally, if at all, certain. The cursory survey of W. Schmid (Gr. L. I p. 713; see also Laqueur /.c. col. 2025, 8 ff.) mainly groups together the remarks scattered through my commentary, where I did not treat the question systematically. The chief points to be considered are: (1) the relations existing between the lyric and particularly the tragic poets on the one side and Pherekydes on the other. The formula of Schmid ‘die grossen attischen tragiker berithren sich so gut wie garnicht mit Pherekydes’ now seems to me to be too negative. At least the scholiast on Eurip. Or. 1654 (3 F 63) says that Euripides used Pherekydes. (2) It is certain that about 400 B.C. Antimachos consulted him (see 3 F 86; 111; and cf. n. 40), and in the third century B.C. Kallimachos and Apollonios. As to the latter the challenge of Wilamowitz Hell. Dicht. UI p. 229 n. 1 does not seem justified to me. But the point must be re-examined. (3) Of historians and mythographers Hekataios seems to be excluded, if only for reasons of chronology, and unfortunately we cannot prove that Herodotos made use of the Genealogiai, though Pherekydes is said (erroneously?) to have been the first who mentioned the river Eridanos in the Phaethon story (3 F 74). Hippias also seems to have known him (3 F 90), and Hellanikos certainly did, as further Herodoros and Asklepiades of Tragilos (cf. n. 79). He is often quoted by later historians and antiquarians—Eratosthenes in the Katasterismoi, Suidas in the Thessalika, Ar(i)aithos in the Arkadika, Istros in the Eliaka, Lysimachos, and others. The Genealogiai clearly was an indispensable book of reference also for the commentators: the Scholia on Homer, the tragic poets and Apollonios are full of quotations. But it may be sufficient to refer to Didymos (F 2; 68; 159), and for Pherekydes’ authority generally to citations like otve “Ouneoc otte “Hatodoc ottve Degexbdyc Aéyovorv (Schol. Apoll. Rh. 1, 45—F 110), or ot ôè srhetovs Ov Eotı xaì Degertôns xai "EMävixoc xat Ho6édweoc (Plutarch T hes. 19

commentary which must try to determine the place of Pherekydes in the development of Greek mythography, his relations to epic (and lyric) poetry on the one hand, and to tragic poetry on the other; lastly the peculiar qualities of his work in comparison with the other genealogists of the fifth century from Hekataios down to Hellanikos and Herodoros. The edition will be so much the better the broader and firmer it lays its foundations and the less it starts from pre-conceived opinions, no matter whether they concern the questions of authorship, or the arrangement of the subjectmatter, or the linguistic form. The new edition must take into account all we know about Ionian prose books, and about the tradition of the individual books, some of which were re-edited in the Hellenistic time. Here, too, the result as far as it concerns the history of literature can be predicted: the anonymous prose narratives of the sixth century with which Wilamowitz in the course of time worked more and more 83) will have to disappear from Greek literary history. From the first there is this difference between epic poetry and prose writing: epic poetry, inspired or dictated by the Muse, is anonymous on the whole 84), even in the sixth century and even if the public knows the names of some of its poets; prose books, from the very first, appeal to a greater or smaller public as the works of individual men who have something special to say and who introduce and legitimate themselves in the preface or introductory sentence. The preface at the same : 83) Ar. u. Ath. ii p. 20; Griech. Lif. 1912, p. 55; Griech. Verskunst, 1921, p. 43 n. +; Gl. d. Hell. II, 1932, p. 94. It is a great triumph for forgers like Amelesagoras, Bion, or Antiochos that Wilamowitz infers from them (for actually they do not say so; Wilamowitz got his facts wrong) that ‘the Milesian chronicle’. was published under the name of Kadmos, and ‘the Ephesian chronicle’, under the name of the epic poet Kreophylos (see R. E. X col. 1473 no. 6; XI col. 1710 no. 2). There is only one book of this character, the Korinthiaka of Eumelos, and here the reason is obvious: Corinth had produced neither a chronicle nor a local history. Moreover the prose paraphrase seems to have been late; it does not prove anything for the hypothesis of Wilamowitz. 84) In epic poetry we find authors’ names first for theological poems, and they are pseudepigrapha. The prototype is neither the Syrian Pherekydes nor the lyric poets, but obviously Hesiod under whose influence the poet of the Delian hymn to Apollo introduces himself to his public Berlin 1933, p. 700).

time serves as the title of. the book by giving the name of the author and his native place. It authoritatively begins with a téôe Agyeı, and thus impressively opposes the new wisdom to the ‘EAAjvev. Aöyoı. For titles in the literal sense of the word these books had not , and this fact explains that we sometime have more than one title for the same book from the sixth or fifth century 85). The introductory words of Hekataios and Alkmaion are preserved; we infer them for Anaximandros, Anaximenes, the Syrian Pherekydes (where at least the first words of the treatise itself are preserved from the Pinakes) and Akusilaos; for there has never been a doubt about the genuineness of Anaximandros’ tv deeoxóvraov abta@. xeqadkarbing ExBeorc, or of the book in which Anaximenes xéxontar AéEer ‘I&ôr arckqe xai dscegitte@:, or of Pherekydes’ ‘Exréuv0c. Of course, forgeries of a name or a book, or forged quotations were possible; Amelesagoras of Athens (or Eleusis), the enlarged edition(?) of Akusilaos, the pre-Homeric Pherekydes of Athens (our no. 5), the prose book of Thales are cases in point. Some of these forgeries are rather early, and Hellenistic scholars recognized them as such. But anonymous — prose books (if they ever existed) are rare exceptions, confined (or so it seems) to certain spheres of literature, if we may ¢all it literature, when the author had not destined his memorandums for public use 86). And there are cases ofa published book losing the author’s name and being attracted by a greater one: the Corpus Hippocraticum furnishes some examples, but it shows at the same ‘ time that this was a slow and later development, and that in the fourth century the real authors were still known 87). This case 85) See n. 6, and as a further example lon’s Triagmos. j 86) Certain technical treatises, e.g. about the arrangement of the stage, which begin to appear in the fifth century and soon become fairly numerous, have genuine authors’ names; they were published. 87) It is well known that the problem of the beginning and the growth of the Corpus is much disputed. See Edelstein R. E. Suppl. vi, 1935, col. 1307 ff.; 1331 ff.; Jaeger Diokles von Karystos 1938, p. 4 f.; 228 ff.; W. H. S. Jones Proceed. Brit. Ac. 31 (1945), whose suggestion is rather attractive that the Corpus is a ‘school library’ and ‘that the bulk of the writings in the Corpus were associated with the name of Hippokrates, not because they were thought to be written by him, but because they originally belonged to him or to his school’.

is wholly different from the history of what we may call the Corpus Homericum, or the Corpus Hesiodeum, and neither these nor the Corpus Hippocraticum furnish a parallel to the alleged Corpus Pherecydeum of Wilamowitz. There was no such Corpus; there were two books belonging to different categories of literature; and the theory that the one theological book written by Pherekydes son of Babys from the island of Syros attracted one ‘historical book’ written by an Athenian genealogist has nothing to recommend it. In order to bring to a conclusion the long discussion of the theory of Wilamowitz about the Athenian Pherekydes and the relations between him and the theologian of Syros, we can state simply and confidently that there were no relations at all. It might be conceivable that the two homonymous authors had been confused in pre-Hellenistic times, but there is no reason at all for such an assumption. Tradition shows that there was no confusion: it gives us two authors and two books. And tradition is amply confirmed by the remains of the two books. If Eratosthenes (and probably Kallimachos before him) without reserve and without being contradicted called the author of the historical work an Athenian, we may safely assume that he was justified in doing : so by the introductory words of the work to which Kallimachos in the Pinakes gave the title Historiai or Genealogiai. There is one argument, and one only, which could be advanced against our believing in the. tradition, viz. that there is no Life of the Athenian genealogist in the Suda 88). Wilamowitz was right not to stress this argument; for it is a very weak one. It is a wellknown fact that the Suda gives Bio. for a great number of writers of whom we have no fragments at all, and that on the other hand it has no ßiog for quite a number even of distinguished writers. whom: we know from their fragments. Though there is no real need for it, in a case in which the existence of a writer whose Life is lacking in the Suda is denied, it is useful if the gap can 88) Momigliano rather begs the question in beginning with the assertion that the three authors assumed by him—the Syrian, the Athenian, and the Lerian—are distinguished in the Suda. True, the Suda has a Life of an Athenian Pherekydes, but is not the author of the Historiai (see p. 55 ff.).

be explained. The Life of the fourth Pherekydes furnishes the explanation.. 4) Degexdinc Aéeros iotogixds °°). Eratosthenes (to put it neutrally) does not mention this Pherekydes, and that is perhaps the reason why the views of modern scholars differ so widely, though mostly they get rid of him by denying his separate existence and iden- | tifying him with the Athenian genealogist 9°). This is understandable because the two components of the short Life in the Suda (the date of this Pherekydes and. the catalogue of his books) obviously contradict each other. As to the date yeyovòc od SALyou thc oe éAvuréôoc (480/77 B.C.), which is somewhat vague, it seems obvious that it is the same, or rather a variant 91) of the floruit 456/5 B.C. of the ‘second’ Pherekydes, the Athenian ‘historian’ who was ‘younger than the Syrian’ 92). This sort of variant 89) Suda s.v. Degexting Aépros: totoguxds* yeyov@cs med GAtyou Tijs oe Ökvusuddos | (480/79 B.C.). Heeì Aggov:- leet "Ipıyevsiag‘ Negi Tóv Auovdcov éoptdv* xai da. 90) The hypothesis advanced by G. J. Vossius De Histor. Graecis 1623, p. 24 ff. ed. West. ‘ut Lerius diceretur quia esset in Lero insula natus, Atheniensis vero creditus sit quia apud hos viveret’ has been accepted by C. Mueller F. H. G. I p. XXXV; v. Gutschmid Al. Schr. IV p. 299 ff.; Jaceby Ph. U. 16, 1902, p. 214f.; Christ-Schmid Gr. Lit. 1 1912, p. 454f.; Tresp Fragm. d. griech. Kultschriftsteller 1914, p. 53 and others. It still represents the general opinion (Diels-Kranz Vorsokr® I, 1934, no. 7 [71] B 13; Index Nominum p. 547; Laqueur R. E. XIX 2, 1938, col. 2020, 6 ff.). The Lerian has not even a lemma in the À. E. I gave the correct facts briefly in FGrHist.: (1) that the titles of his works assign the Lerian to the Hellenistic period; (2) that there is no reason to doubt his existence; (3) that Eratosthenes had no cause for mentioning him. This thesis has been accepted by W. Schmid and Momigliano (n. 2). The objections of Wiiamowitz Lc. p. 130f. ap. n. 98) rest on the date in the Suda, combined with the wrong assumption that the use of the Ionian dialect is attested for the Lerian. His arbitrary solution—that we have to acknowledge an Ionian and therefore pre-Alexandrian book on Leros, which bore the name of Pherekydes and was either ignored by Eratosthenes as unauthentic, or left to’one of the two name-sakes—is wrong for this reason alone. Actually it leaves us with the main question how the book came to be regarded as written by Pherekydes. Wilamowitz’ answer ‘it had been given this author’s name because it seemed suitable for the contents and the language’ is no answer. 91) See n. 49. 92) See above p. 26.

is not rare, and this is not surprising, because it was impossible in most cases to date accurately a fifth century prose writer: it is not only for Hellanikos that the Suda gives two dates—the quite definite statement of Apollodoros for his floruit in 456/5 B.C. which probably. underlies the synchronism with Euripides and Sophokles, and the rather vague statement which dates him from his name and his relation to Hekataios with yeyovdc xat& và IHeoocıx& fj mixed wzedoGev °°). Even for Herodotos, whose floruit in Alexandrian chronology is 444/3 B.C., Ktesias has the same vague date xarà EéeEnv yeyov@c totic xedvorc °*). In most cases the rough synchronism with the IIegouá is the earlier dating which Apollodoros replaced by the more precise (though not always correct) attribution of the floruit to a definite year. On the other hand it is even more obvious that a list of books which begins with the three titles IIeoì A&oov, Iegi ’Ipıyevsiac, Meei ta@v Atovúoov éogta@v (the rest, perhaps many more books, being suppressed by xai &Àka, words which often are conveniently forgotten by modern scholars) cannot refer to a writer of the beginning of the fifth century. It stands to reason that it is the list of books of a Hellenistic writer 95). Taken together with the dif. 93) FGrHist 4 T 1 with Rohde’s and my corrections. See ib. I p. 430f.; il D p. 722, 16ff.; RE. VIII col. 107 ff. 94) Diodor. 2, 32, 2; R.E. Suppl. II col. 229, 19 ff.; XI col. 2040, 57 ff. 95) Rohde Ki. Schr. 1 (1878) p. 168; Jacoby FGRHist I p. 386, 42 f. As the simple statement of an obvious fact did not sufficiently impress scholars, it must be proved, though it is a tedious task. As to Ilepi Azoov the Hellenistic form of the title shows at least that the book was no ancient chronicle, no Aseimv “Qoou Ilspi ‘Tepryevetas is, to judge again from the title, one of the many special works written in Hellenistic times about single mythic persons. The ‘epic poet’ Simonides of Karystos or Eretria (Suda s.v.; the name is missing in the R. E.) wrote Ileoi ‘Iqryeveiac & in prose as the title is preceded by n eis AdAôa oúvoôos tev “Axarmv, teipeteov, PipAia B. Heeì tav Awovóoov éopgr@v also is one of the almost innumerable special books about ‘festivals of individual gods or cities. It is nota ‘buch über die opfer des Dionysos’ (as Tresp Kulfschriftsteller p. 7 has it): éogtai and Gvoiar are different things; for example Philochoros wrote a book Ilegi éoegtóöv and another Ilegi Gvoiwwv. In the titles of these books Ilegi éogtwv is often connected with Tegi unvov, and Ilsoi Guoiwv with Ilepi tega@v. A book Ilegi tav Atovécoov éoeta@v is not confined to Athens (against Tresp p. 27). There is only one question which may be put and which in the absence of fragments we are unable to answer, viz. whether it dealt with cults only, Mnemosyne XIII .

ference in their respective native places, this makes impossible at once the identification of the Lerian writer with the genealogist — of Athens. We cannot simply disregard a list of 3 + x books; we cannot regard their titles as denoting parts of the Genealogiai; and lastly we cannot take seriously any attempt at dating or otherwise determining the Lerian which takes into account only one or two of the books ascribed to him 96). There is no escape; we have to acknowledge the existence of a later writer of the same name whom the biographic source of the Suda listed as a ‘historian’, either rightly on account of the book on Leros and probably the subject-matter of his other books, or erroneously because the confusion with the genealogist, to which he owes his early date, has affected not only the date but the description also. We on our part should prefer to call the author.of such books a grammarian or antiquary 97); the ancient librarians who had no special term for this sort of scientific writing used the two terms historian and grammarian for Hellenistic antiquaries either indiscriminately or a potiori. . The book IIeoì Aggov, not Aseiwv Goo, favours the assumption or, as we had better say, confirms our belief in the tradition. It is of no great importance whether literary history (represented by books like Ileei évôéEov dvôe®v, Heoi Su@vöuov somt@v xai ovyyesææpéwv) inferred the native place of the Lerian from the book about the island, or whether he gave his name and home country in the title of one or more of his books—the latter being perhaps the natural thing for a Hellenistic author. Nor need we ask, whether _Agevosg in the title of a Hellenistic book means the citizen of an | independent Leros, or a man belonging to the Milesian ‘deme’, . or to the community of the Muñoror ys Aggov which we find in or whether it belongs to the domain of musical history like Dikaiarchos” Heeì Atovuoiaxév dyovov (cf. now Wehrli Dikaiarchos 1944, p. 67). The question of Tresp p. 53 is naive: the Tederat of Stesimbrotos (FGrHist 107 F 12) are quite different, and the Eis “Apywvoc davaßeoıs of Hellanikos (FGrHist 4 F 56) does certainly not belong to the ‘special books about cults’. 96) This objection holds good for the attempts of v. Gutschmid (Kl. Schr. IV p. 300 f.); Wilamowitz Lc.p. 130 f; and others. 97) Momigliano l.c. p. 346 calls him ‘antiquario, mitografo, e storico locale’.

the Athenian tribute lists 98). We cannot decide this question, and if we could it would be of interest only for the political state of the island during the Hellenistic period. Finally no importance at all can be attached to the fact that Eratosthenes did not mention the ‘historian’ of Leros. Of course, one can infer (and it has been done): that the Lerian was later than Eratosthenes. That may be so. But even if Eratosthenes knew a colleague of that name he really had no reason for mentioning him. He was concerned merely with the rejection of Andron’s pre-Trojan astrologer 99), and there was no need to mention even the Athenian genealogist. But the latter was an old and famous author, and a ‘there were no other men of this name except’ nicely rounded off his polemics. It would have been ridiculous to enumerate contemporaneous bearers of the name. As to the work of the Lerian Pherekydes we know for certain only what the Suda tells us. Nevertheless one should be cautious 98) I really cannot discuss here the argumentation of Wilamowitz Lc. p- 130, which has impressed Altheim Griech. Goetter im alten Rom 1930, p. 119f. and Laqueur Le. col. 2020, 6 ff. Wilamowitz argues: ‘nun hat es aber Lerier nur in der frühzeit gegeben, da Phokylides und Demodokos mit feinen versen einander schraubten. Bald darauf ist Leros milesisch geworden und immer geblieben; selbst die tributlisten des attischen reiches kennen nur MtAjouo éx Aggov. Dann ward es ein milesischer demos, und nach dem konnten sich die Milesier so wenig nennen wie ein Athener nach Eleusis oder ein Rémer nach Arpinum’. Taken strictly the argument would bring the Lerian into a time impossibly early for a prose writer. For Phokylides, and consequently Demodokos, belong to the first part of the sixth century B.C. (see Schmid Gr. Lit. I p. 299; 374), and Leros was annexed by Miletos or ‘colonized’ (Anaximenes FGrHist 72 F 26, unfortunately without a date) at the latest some time, perhaps a long time, before 494 B.C. (see Wilamowitz Kl. Schr. V 1 p. 427; Büchner R.E. XII 1925, col. 2097, 53 ff; Hiller von Gaertringen R.E. XV 1931, col. 1590, 49 ff.). But Wilamowitz does not take his own argument seriously; it is for him only one more. reason for rejecting the Lerian Pherekydes as the author of a book on Leros which he believes to have been written in the fifth century (cf. n. 90). We are again left with the question who wrote a book on Leros, if there were no Lerians. This and all other difficulties are mostly self-created; it is surely not the tradition which creates them. in 99) Von Gutschmid in his lectures on Greek Historiography, published 1893 (Kl. Schr. IV p. 299), protested against the improper use which Vossius and his followers made of Eratosthenes in trying to prove from his words that there could not have been a third writer of the name of Pherekydes. In vain; see e.g. Laqueur l.c. col. 1992, 6ff.

of making the casual remark ‘there is no trace of them’ 100). There is at least one quotation for which the authorship of the Lerian must be considered very seriously. Clement 101) quotes at length a passage from the ‘Syrian Pherekydes’ which narrates a warning given by the Scythian king to Dareios when he crossed the Danube. It is universally acknowledged that the author. of this story cannot be the Syrian, if only for reasons of chronology; but neither does it fit the subject-matter of the Pentemychos 192). The Syrian would be possible only if he played an active part in the story warning © Dareios or interpreting for him the puzzle of the message of the Scythian king; i.e. if we had before us a story similar to those told by Herodotos about Sandanis 103), Thales, Bias, Hekataios, or—in Herodotos’ narrative of Dareios’ expedition against the Scythians 104)_— of Gobryes who corresponds to the two advisers in the story told by Pherekydes, viz. Orontopatas the chiliarch and the quite unknown Xiphodres. This is not the case: Clement does not tell an anecdote about a Pherekydes who might have been the philosopher and one of the Seven Wise Men, but the quotes a passage from Pherekydes, a real ‘fragment’ from a book which inter alia narrated the Scythian expedition. Since J. G. Vossius scholars therefore have frequently corrected 4 Eúetoc into à Agguos, and from a purely palaeographical point of view the alteration seems to be easy enough. But the methodical objection of Gutschmid 195) is deserving of consideration: in a certain group of witnesses we have to state confusion between the Syrian and the Athenian Pherekydes, never between the Syrian and the Lerian. This confusion (I will add) sometimes led to the inconsiderate addition of a 6 Zúetoc, where the source had a simple Pherekydes 196). We should have to suppose that the Athenian Pherekydes, who perhaps was a follower of Miltiades and in any case 100) F.H.G. I p. XXXVI; v. Gutschmid Le. p. 300; Wilamowitz Lc. p. 130 and many others. 101) Strom. 5, 8, 44 = FGrHist 3 F 174. 102) I do not understand: Laqueur /.c. col. 2024, 36 ff. 103) Hdt. 1, 71. For the text see Herm. 51, 1916, p. 477. 104) Hdt. 4, 131-132. 105) L.c. p. 299. 106) See n. 14; 35.

took a particular interest in him 107), did not only mention the Scythian expedition, but gave a full account of it. In the face of the pedigree of the Philaidai in F 2 which I treated above I do not feel happy about this supposition 108). There may be another way of dealing with the fragment of Clement, provided we leave aside the epithet 6 Xderog as an inconsiderate and erroneous addition, as (I submit) we are entitled to do. Now, Momigliano 109) and De Sanctis110) have again compared the two versions of the Dareios story extant in Herodotos and Pherekydes. They arrived at a result which can hardly be doubted and (as far as I know) has not been doubted: the two different versions are independent of each other; they derive from what De Sanctis calls a ‘tradizione originaria’ and takes to be a Greek invention. One might well call it a folk-tale, and it is quite probable that the story is as old as the Scythian war itself, and that the tradition of it was at first purely ral. The story may further have been written down as early or even earlier than the Herodotean version, but—and this is the mistake in logic made hitherto—there is not the least need to assume that it was written down for the first time by Pherekydes, -and that it entered literature under his name. If the story is impossible for the Syrian, and if it is hard to find a place for it in the Genealogiai of the Athenian, it is easy enough to find a way out of the difficulty: we have to assume that the Pherekydes quoted by Clement and erroneously called ‘the Syrian’ took it from a fifth century author, a contemporary of Herodotos or an earlier writer, —Dionysios of Miletos, Hellanikos, Charon or whoever it was, 107) See p. 30f. 108) Wilamowitz Lc. p. 136 is more confident. He makes the following short but surprising statement: ‘daraus den Lerier zu machen ist vergewaltigung des textes; ihm die geschichte zuzuschreiben kaum etwas besseres; denn von seiner benutzung fehlt jede spur, und dies ist ja auch nur eine ausrede, um das zitat loszuwerden, das man auf den Athener beziehen muss, einerlei ob es fiir einen genealogen passt’ (my italics). If Clement quotes the Syrian and Wilamowitz understands the Athenian is not that also an ‘excuse in order to get rid of the quotation’? Momigliano p. 348 on the strength of the same F 174 (it is his only reason) makes the Athenian (sic!) contemporary with the Persian wars. 109) L.c. p. 346 ff. 110) In Memor. Pârvan p. 110f. (my copy is not dated, and I cannot find out the year).

‚who wrote either Persika or Horoi, and narrated in this work the expedition of Dareios against the Scythians. We cannot even guess at the name of the early writer who first introduced the story into history, nor is it of great importance. There are two points which really matter. The first we infer from the author’s name given by Clement: the story was taken over by Pherekydes of Leros; who is the only possible claimant, if we have to exclude his two earlier namesakes; and there is not the least reason for refusing to believe that a book about Leros mentioned, or even dealt circumstantially with Persian history. The second point is that there are many examples for the relation of sources which I assume here: Oral tradition(s) entered literature in the work of Herodotos, and in a different form in that of an Ionian author of the fifth century who in his turn was exploited by a later writer. I shall content myself with one rather illuminating example, viz. the story of Battos and Pheretime as told by Herodotos and Menekles of Barka, a writer of the second century B.C., the source of whose version is as unknown to us as the source of Clement’s Pherekydes 111). Menekles is followed in his turn by Polyainos and Aelianus, neither of whom quotes his authority. The same Aelianus 112) tells the story of the peAcayoiéec of Leros, again without an author’s name, but here with one of his silly anonymous quotations: xai Ars A aitia tcaci te oi Thv vioov oixotvteg Tv Aéoov, xai Evsorı uaGetv AAAax6Bev. The ultimate source obviously is a book on Leros 113). This was most probably the only special book about the island which inter alia told about the sacred guinea fowls kept in the sanctuary of lokallis-Parthenos 114) in Leros. And then perhaps we may even regard as a third point the enigmatical enumeration of praesaga animalia in the so-called Aetios, who quotes Simonides and Pherekydes 115). But I will not lose myself in more or less plausible conjectures, nor shall I repeat the learned enquiry of Wilamowitz about the Parthenos of Leros and Patmos, or the 111) FGrHist 270 F 5-6 with the commentary on these fragments. 112) N.A. 4, 42; cf. 5, 27. 113) It is astonishing how Wilamowitz l.c. p. 131 n. 1 evades the inference. 114) About this goddess see Wilamowitz l.c. p. 84; 149, 2. 115) FGrHist 3 F 173. 130f.; Gi. d. Heil. I p.

combinations of Altheim 116) in order to determine the contents of a second book from the pen of the Lerian, viz. his treatise Hegi *Ipuyeveiac. It is sufficient to have made out here a case (a rather strong one, or so it appears to me) for traces of at least one book of the later writer Pherekydes of Leros, a book used in Hellenistic times by one of Aelianus’ sources and later by Clement, who confused him (as others did) with the better known theologian of Syros. To return to the biographic tradition: it seems fairly clear what the matter is with the several articles in the Suda on real or alleged writers of the name of Pherekydes. We find a confusion, not rare in this lexicon 117), of at least two men,—the Athenian generalogist, who has the same claim to the description iotogıxös as Hekataios, Akusilaos, Hellanikos and perhaps a better one than the Lerian. In the conflation the date (and perhaps the literary genos) belongs to the Athenian, the native place and the list of books to the Lerian. The primary reason for the confusion may be found in the fact that the Suda listed another earlier Athenian (our no. 5) whose Adréy8oveg allegedly also had ten books like the Genealogiai of the later Athenian. Therefore the first confusion may have been between the two Athenian writers. This, of course, we cannot prove; but confusion there certainly was, as we find the ten books also in the Life of the Syrian Pherekydes 118); and in any case the consequence is that the Suda gives us a Life for the forger of the AùróyBovec but dropped the real author of the Genealogiai 119). 5) Degexdbing “AOnvaiog meszofiüregos tot Zveiov has a comparatively full Vita in the Suda. °°). Though he is dated rel116) Op. cit. p. 101 ff. 117) FGrHist I p. 386, 38 ff. and the three Poseidonioi 87 T 1; 279 T 1. See also v. Gutschmid Lc. p. 301 ff.; Rohde KI. Schr. I p. 168; Momigliano l.c. p. 349 n. 1. | | 118) We have to delete only év BiBAfiou t (Wilamowitz Lc. p. 129), not the whole sentence &orı 5% — Ötaôoyds (as I proposed on 3 T 1) which is a correct statement of the contents of the Heptamychos of the Syrian (cf. n. 6). 119) Cf. n. 88. 120) Peperböns 'AOnvatos® seeoPitegos tov Xvetov, Ov Adyos ta Oppéws ovvayayetv. Eyeawpsev Abdtöxdovas— Eorı 5% repi tis ‘Artus aoxarodoyiac—

atively only, the time is determined more accurately by another addition öv Aöyos ta “Og~éws ovvayayeiv. This statement is made in order to date him very early, perhaps at the time when according to an Athenian tradition Eumolpos t& tot satedc Movoaiov rouoeig ÉEéOnxev, or when Musaios wrote down the hymns of Orpheus öiiya éxavoeOa@oac !?'!). The author of this tradition did not wish to create a rival to one of the historical writers: he simply added one more name to the circle of pupils of Orpheus, into which, besides Musaios and Eumolpos, Thamyris, Linos, ‘Kadmos, and others had been introduced. The invention was convenient if the author was merely concerned with giving an early date to his. Pherekydes; but it was less innocuous than the pre-Homeric Syrian invented by Andron; for it was made in order to introduce a forgery and to guarantee the authority of it. That this was so is proved by the wording of Porphyrios’ polemics: Tog@iguoc 5é tov TTEOTEEOV oùÖéva Séxetar, GAA’ Exetvov uÓvov Tyetrar Goxnyov ovyyeagiic. The list of writings, against which the contradiction is directed, gives Atdtézx8oveg and Iopaivéoeis in alphabetical sequence 122). That is surprising, because sound pinacographic tradition arranges the writings of an author by genera of literature, and always enumerates first his poetical productions. It is equally ' surprising that the Athenian is dated by comparison not with his compatriot, the genealogist, but with the theologian of Syros, who Ev BiBAforg 1° Ilaparvéoers 61° éxwv. Iloepúguos (FGrHist 260 F 21) ôë row agotégov (scil. tov Lugiov) oúdéva woeoPitegov Séxetar, GAN’ Exetvov uóvov Hycita: doxnyöv ovyyeapijs. 121) Marm. Par. A 15; P. Berol. 44 — Kern Orph. Fragm. F 49. Von Gutschmid Lec. p. 300 f. overlooked the second date (for the consequences: see n. 124) and Wilamowitz p. 129, who referred it to the Athenian genealogist, called it ‘eine fiir uns ganz unkontrollierbare angabe’. The disregard of the biographic tradition is spreading: see e.g. M. P. Nilsson Gesch. d. griech. Rel. 1 1941, p. 646 ‘bei Suidas ist zu lesen, dass der athenische logograph Pherekydes, der im 5. jahrhundert lebte, die orphischen schriften gesammelt hat’. Kern /.c. T 228 still records the wild conjecture Pepexedeng év toig Aûürouókors for Degexböng Ev toig AdröxBocı in F 176 which (Iam sorry to say) I also mentioned in my app. crit. 122) Laqueur Lc. col. 1991, 56ff. rightly referred to the Ilagaıuv&ceıs Isidor. Origg. 1, 39, 11 according to whom Moses used the hexameter longe ante Pherecyden et Homerum.

was generally regarded as the first prose writer 123). These irregularities all point in the same direction: what was invented was less the author than the work. It is merely to guarantee the genuineness of the AûróyBovec which are described as dealing with Attic archaeology that the forger had to create an author and give him a personality by fixing his time, enumerating his books, and bringing him into connexion with Orpheus 124). Porphyry was as justified in rejecting this writer, as Eratosthenes was in rejecting the astrologer, though we cannot decide with certainty (nor is it of great importance) whether or no he followed his own judgement, basing himself as on an established fact on the general belief that the Syrian was the first prose writer. It is more important to ask whether the AûrtóyBoves, which is repeatedly quoted though only by rather late writers 125), did or did not exist, i.e. whether we have to deal with a full-fledged forgery, or merely with fraudu125) See n. 22. 124) We infer that in this case the ten books are not wrongly repeated from the Life of the genealogist as they were in the case of the Syrian: (n. 118), but that they are part of the invention: they corroborate my opinion that the forger of the AúróyOoveg had in mind the Historiai of the genealogist. The tradition is clear and simple. Gutschmid, who overlooked the alleged collection of the Orphic poems by the pre-Homeric Pherekydes _(n. 121), made an awful muddle of the tradition by assuming ‘a later Pherekydes of Alexandrian times, but evidently an Athenian’, assigning to him two of the books of the Lerian, viz. Ileoi I@uyevetas and Hegi tev Atovicov éopr@v; the Ilagawéoerg he gave to ‘the double of the Syrian’ invented by Andron, who actually was an astrologer, and whom not even Andron seems to have regarded as a writer. Lipsius (Quaest. logogr. 1886, p. 19) ‘saw no reason why the author of the Aufochfhones should not have been the Lerian’, though the Suda has a separate Life of him. Wilamowitz p. 129f. and Laqueur col. 2024, 39 ff. are quite at sea as regards the alleged author of the Aufochthones. 125) The title of the book is given only in the Etym. Gen. (Magn.) s.v. Bonòeombv (FGrHist 3 F 176). The Scholia on Aristeides p. 313, 20; 320, 1 (3 F 178—179) quote ®egexböng xai ner’ éxetvov "Avrioxos and Deoexdbing xaì “Avtioyosc. The second quotation recurs in the Ef. M. s.v. IIlaAAdödLov with a simple Aéyer Deeexúöns, but the contents leave no doubt that the Autochthones were meant. From the contents it is further probable that this Antiochos is the ’Avtioyoo év évétor tov ‘“Totoet@v quoted by Clem. Al. Protr. 3, 45, 1 for the tomb of Kekrops. About the quotation Degextöns 6 xaXkarés in Plutarch De fac. 24 (3 F 177) see n. 137.

lent quotations. In the former case we have parallels in the books forged on the names of very early writers like the Eleusinian Amelesagoras 126), Pseudo-Akusilaos127), Pseudo-Kadmos128), Diktys, and Dares; in the latter case we have to compare the forged quo‘tations from the poetical Affhis of Hegesinus 129) and the masses of invented authorities in Ptolemaios Chennos, Ps.~Plutarch and others 130). The manner in which the Scholia on Aristeides quote this Pherekydes together with a certain Antiochos (Depexdône xai per’ adtov "Avrioxog *') might favour the first possibility, and it seems tempting to regard this Antiochos as the inventor of the old Athenian Pherekydes. But it is not possible to speak with assurance for the following reasons: (1) both ‘Avtíoyoc Ev T@r Evaroı t@v Totoetwv and Degexbing Ev Trots AùtróyBoorv are once quoted.each by himself, the former by Clement, the latter in the Etymologicum; and it is just these quotations which alone give us the titles of their respective books; (2) the name Antiochos. is common, and lacking a distinctive epithet we cannot identify this Antiochos with any of the bearers of the name otherwise known to "us 132); (3) and chiefly: the matter may be more complicated than 126) The Ancient Historians of Afhens no. 9 (not yet published). 127) FGrHist 2 T 1; 7. 128) Bion of Prokonnesos FGrHist 14 T 2: 489 T 6; cf. The Anc. Hist. of Athens no. 11. 129) Pausan 9, 29, 1-2; cf. The Anc. Hist. of Athens no. 10. 130) Mnemos. III 8, 1940, p. 73 ff. 131) Cf. n. 125. 132) It is no longer necessary to prove that it cannot have been the Sicilian historian Antiochos of Syracuse who cited the theologian of Syros or the genealogist of Athens: the contents of the fragments exclude the fifth century B.C. It is pure chance that Clement quotes the ninth book of the ‘Iorogia: and that the XuceAuxe of the Syracusian were divided into nine books. But it would not be pure chance if the forger wrote not more than nine books (a hypothesis which Clement’s quotation does not prove). For in this case we should have to compare the Ilavtodaxai iotogia: Ev BıßAioıs 6, Œtiva éExryoaq~er Modoac ‘Téôr Stadéxtwr of Kephalion from the time of Hadrian (FGrHist 93) and assume:that the forger (if he really existed) took Herodotos as his model. Of later bearers of the name Antiochos the_sophist from Aigai from the second century A.D. (R. E. I col. 2494 no. 65) would perhaps be possible in point of time. The author of IIeeì Oev, most probably Antiochos of Askalon who took part in the battle of Tigranokerta in 73 B.C. as a member of the cohors of Lucullus (Plutarch. Lucullus 28,8; cf. v. Arnim R.E. I. col. 2493 no. 62), seems too early.

it appears in the simple alternative just set forth. Ptolemaios Chennos, who hardly wrote before the last quarter of the first century A.D. 133), quotes for the obviously invented story of a beauty competition between Thetis and Medeia one ’A0nvööweog 6 ’Eeetoveds Ev n ‘Yzouvnuétov, who on his part (or so Chennos says) cited ’Avtiozoc Ev B tev Kare sókuv pvOuxe@v '°*). A similar quotation of Antiochos and Pherekydes by Ptolemaios (or another impostor) would explain why later writers sometimes quote these two authors together (Pherekydes in these quotations being the earlier of the two) and sometimes each by himself. And the same assumption would fix the time of the forgery (or of the forged citations) which fits the contents of it excellently. The discussion in F 178 about Dionysos, Isis, Osiris and other Egyptian gods reminds us of the theological sources which Plutarch used in De /side. But the standpoint of ‘Antiochos’ is exactly the reverse from Plutarch’s: Antiochos protests against the theories which seek the origin of Greek gods in Egypt, and he decides the old issue whether the Athenians had founded Sais, or conversely the Saites had colonized Athens 135) in favour of Athens by introducing ‘the Attic autochthons Ogygos 133) The Suda s.v. 'Enagpoööıros dates him in a synchronism with Epaphroditos zwi Néewvos xai uéxer -NéeBa (54—98 A.D.), and s.v. IIto- Aeuatos “AAcEavigets probably more correctly éxt te Toataevod xai ’Adpıavoü (98—138 A.D.). The first testimony cannot be used for proving that he lived in Rome: Av is a purely chronological term which does not allow of supplying év "Popyne On the other hand it is not improbable that he did live in Rome, at least for some time, because he dedicated his book to a Roman lady Tertulla. If he saw himself the temple of Peace built by Vespasianus (Phot. Bibl. 190 p. 149b 32ff.)— and again it is not improbable that he did — he was in Rome after 75 A.D. 134) Phot. Lc. p. 150a 37 ff. = FGrHist 29, where I collected citations of this kind (see also ib. I p. 386, 20 ff), but did not distinguish the possibilities clearly enough. C. Mueller’s confident identification of the author of the Mutixd with the author of the ‘Iorogia:r(F.H.G. IV p. 300 f.; 639) is correct only in so far as both books were invented by the same Ptolemaios. We cannot deside, nor is it of any importance whether or no he identified his two homonymous authors. 135) Proklos in Plat. Tim. 21 E (= FGrHist 124 F 51). The priority of Athens was supported in the fourth century by Kallisthenes and Phanodemos, the priority of Egypt by Pseudo-Theopompos (= Anaximenes FGrHist 72 F 20) and Hekataios of Abdera (ibid. 264 F 25 ch. 28-29). The political bias is apparent in the first three writers.

and Thebe’ 136) as settlers in Egypt, bringers of the Isis mysteries, name-givers of Egyptian gods, and founders of Thebes in Egypt. It is for these assertions that the alleged Antiochos cites the evidence of an Athenian author from the time of Orpheus. Wary scholars will perhaps be content with dating Pseudo-Pherekydes and (Pseudo-)Antiochos somewhere between Eratosthenes and Porphyry, because these two fermini are quite safe: Porphyry expressly denies the existence of the very old Athenian, and in this case we can and even must infer from the silence of Eratosthenes that he did not know the Orphic Pherekydes. If he had known him, he could not simply have distinguished the Syrian theologian and the Athenian genealogist as the earlier. and the later Pherekydes. Personally I do not doubt that the forgery of the book (or of the author) is much nearer to the later date, and not much later (if-at all) than ca. A.D. 100. It is possible that Plutarch already knew it 137). 136) The curious couple points to later times. For this author the difference between Athens and Thebes has become obsolete. Ogygos and Thebe are both Greeks, and this is what matters to him. Gutschmid (lc. p. 300 f.) probably was right in assuming that the author (or for that matter Chennos who invented him) tried to prove with bold etymologies that the Attic ‘autochthones’ were the fountain-head of all theology. But he was, of course, wrong in the conclusions which he drew from his assumption: bold etymologies are no privilege of Chennos or ‘Antiochos’, and to assign to him e.g. F 49 about the Thriai is simply nonsense. 137) In the quotation in Plutarch De facie 24 p. 938b (== FGrHist 3 F 177; VorsokrS 7 [71] B 13.a) the following points are remarkable: (1) the distinctive epithet Depexdône Ò awadatdc, which is not only unique (see n. 6; 8) but not really distinctive; (2) the identification of Athena with Selene, which seems to be late (Gruppe Griech. Myth. p. 1291 n. 3 quotes too many witnesses and does not pay attention enough to their time and value); (3) the statement that ñ ‘'AOnvá T@t ‘AyiÂker véxtagds te éuBoooïtas évéotake pi weoctepéevewr Toopnv, which seems to be an invention like that of Ptolemaios Chennos about the beauty competition between Thetis and Medeia. In any case, it is typical of Ptolemaios’ method, for it evidently connects the rich tradition about the rearing of young Achilleus by Thetis, who uses ambrosia for an ointment (Apoll. Rhod. 4, 869 ff.; Bibl. 3, 171), with the equally rich tradition about the food he got in Chiron’s cave, a tradition, which for example explains the change of his name from Aıyboov to ’AxAkeds from the fact that he ta xeiAn pacroic où xpooñveyxe. I placed the story amoug the Dubia and gave it hesitatingly to the theologian restricting it however to the divine food (FGrHist I p. 429, 19; 430, 18), while Wilamowitz (Zc.

We can sum up now the results of our investigation. An unbiassed examination of the biographical evidence as well as of the fragments fully corroborates the statement of Eratosthenes. This ought not to surprise us. For if the scanty information which we got from Diogenes. does not deceive us (and there really is no reason for believing that it does) the decisive point was not problematic to Eratosthenes. It does not look as if he passed judgement in a disputed and involved question. Apart from rejecting Andron’s pre-Trojan astrologer of Syros, which was not much more than a matter of interpreting the Homeric evidence, he simply stated a fact which followed automatically on this rejection: there were only two writers of the name Pherekydes, one from Syros, the well-known theologian, the other the genealogist of Athens. It is almost selfevident that in this as in othier cases he based his statement of facts on the manuscripts stored in the great Alexandrian library and catalogued by Kallimachos, who was still living when Eratosthenes became librarian 138). I therefore submit that Eratosthenes’ statement rests on, and must be regarded as documentary evidence, and I further submit that the remains of the two works, the authors of which were listed respectively as a philosopher and a historian, fully confirm the distinction mady by the cataloguer and Eratosthenes. Nor is there any reason for doubting the biographic facts which also derived from the Pinakes and therefore ultimately from the opening words of the two works. In these the older Pherekydes called himself Syrian and son of Babys; the younger perhaps only Athenian, though (as we have no biography for him in the Suda) we had better speak cautiously. What more we are (or believe ourselves) able to state about the Athenian is not evidence in the sirict sense of the word but inferences from the remains of his work. We learn from ihem that the Athenian Pherekydes wrote the work which the librarians entered as Historiai or Genealogiai not p. 133 n. 1) dictates ‘F 177 gehört sicher dem Syrier’ and impressed Kranz. -I now prefer to derive it from the forged Autochthones or to regard it as one of the forged quotations from’ the alleged book. 138) See Wilamowitz N. Jahrb. 1914 = Kl. Schr. I p. 412 f.; Herter R.E. Suppl. V, 1931, col. 389, 63 ff.

before 508 B.C. 139). He probably was younger than Hekataios of Miletos and may well have known ‘his historical work 140), he may even have got the impulse for his own writing from him; and perhaps he was contemporary (though he may have preceded him by some years) with Akusilaos, whose Argive nationality has never been doubted although he too wrote in the Ionic dialect. Of course, no testimony does, and no fragment can tell us whether Pherekydes was born in Athens or (to express it more cautiously) whether he was Athenian by blood, perhaps one of those who went to the Chersonnese with the second Miltiades and returned with him, or for that matter the son of one of the original colonists. In this case the fact that he came into personal touch with Ionic culture may explain how an Athenian came to write a prose book on the mythic history of Greece, just as the journeys of Solon may have made him familiar with Ionian elegy and iambic poetry. But I do not attach much importance to such considerations: books travelled quickly even before the fifth century, and the existence of an Argive genealogist is a sufficient parallel to an Athenian one. That both wrote Ionic really ought not to surprise us, for again the parallel with Akusilaos applies. Nor is it surprising (though it is most interesting and has its true parallel in the poems of Solon) in how different a spirit (compared with the Milesian Hekataios) these first ‘historians’ from Greece proper dealt with their subject-matter: there is no trace in their remains of the scientific spirit which provoked, and the rationalism which pervaded the ‘historical’ work of the Milesian 141), while Pherekydes had an even stronger liking 139) See p. 29 ff. 140) Hekataios published the Periodos most probably before 500 B.C., though Thorikos as wöAıs (FGrHist 1 F 126 as compared with Pherekydes: 3 F 146) is perhaps not sufficient proof for the assumption that he visited Attika before the reform of Kleisthenes. The Genealogiai are almost certainly later than the Periodos, but we are not able to determine accurately the time of their publication (R.E. VII col. 2671, 48; 2741, 48ff.). The reasons for the eighties of the fifth century (FGrHist I p. 319, 7; 326, 40 ff.) are very weak. A date between 510 and 491 B.C. is more credible, but the argumentation of Prakken (Transact. Am. Phil. Ass. 71, 1940, p. 467 ff.) is futile. . : 141) See FGrHist 1 p. 386, 31; W. Schrhid Gr. Lit. I 1929, p. 711; Momigliano l.c. p. 351. I deliberately refrain here from trying to characterize at length the literary personality of the first Athenian prose writer

than Akusilaos for what we must term fairy-tales, which apparently were preserved by oral tradition. This explanation, simple as it is, of a fact which may be surprising in itself (though properly understood it actually is not so surprising as some scholars try to make out) I much prefer to flights of imagination, such as a Lerian writer leaving his native island for Athens when the Milesians ‘colonized’ the island, or (for the former date seems much too early 142) ) after the suppression of the Ionian revolt and being granted citizenship perhaps as a reward for introducing Athens into the domain of heroic myth, in which she had hitherto played a rather unimportant part: for the Theseis, which had been published not so long before, never formed part of the Epic cycle, and did not influence Pan-Hellenic genealogy in any perceptible degree until the fifth century, and then through tragedy. The foundations for such fancies, quite apart from their openly contradicting our evidence which clearly distinguishes the Athenian and the Lerian, would be extremely weak. Grants of citizenship in the time of Kleisthenes and Themistokles are as rare, if not rarer than they were in ‘Lycurgean’ Sparta 143); and it does not look as if they became more frequent in the time of Perikles, who in 451/0 B.C. had attached severer conditions to entries into the citizen’s list. In the forties Herodotos was given a monetary reward by the state (such as Pindaros received in the sevenresting content with referring to Gutschmid Lc. p. 303 ff. and W. Schmid op. cit. p. 710f. The latter perhaps overestimates the ‘historical’ character as inferred from the elimination of the cosmogony: ‘darin liegt ein abriicken von philosophischer spekulation und eine annäherung an die ionische logographie, die sich mit sammlung und ordnung der “geschichtlichen” überlieferung begnügte; gerade in dieser beschränkung und sachlichkeit mag man etwas von der eigenart attischer wissenschaftlichkeit, eine vorbereitung auf Thukydides erkennen’. But he certainly does more justice to the figure of Pherekydes than Aly (Volksmärchen etc. 1921, p. 227) does: ‘Pherekydes ist das langweiligste stoffsammelbuch, das irregeleitete wissenschaftlichkeit je produziert hat’ and so on. Aly cannot have read the fragments. 142) Cf. n. 98. 143) The Peisistratids seem to have applied rather generously the law of Solon (Plutarch Solon 24, 4) who regulated the question prudently but not in an illiberal spirit, and Kleisthenes does not seem to have cancelled their grants. I agree in the main point with Wade-Gery Cl. Qu. 27, 1933, p. 25f. I do not know of any individual case between Kleisthenes and Perikles. For Sikinnos, the Persian (or so Plutarch Them. 12, 4 has it) slave

ties, whom the Athenians made a proxenos) ™*) and a xAñooc in Thurioi. Perhaps the two gifts are connected, and it may well be that they also made him their proxenos there; but they did not grant him the citizenship of Athens. Of the three more or less tangible writers, viz. the Syrian, the Athenian, and the Lerian, the Athenian probably was the best known, at least in wider circles. His genealogical work seems to have been abundantly used in Athens. We infer this from the probable fact that it came to Alexandria in a condition which showed traces of such use, though they are not very clear to us. In Alexandria a new edition was made which again was abundantly used by scholars, and here the details are better recognizable. We may safely leave aside the two other bearers of the name, as they owe their existence to conjectures and forgery. The first undoubtedly is the case with the pre-Homeric Syrian, who is but a double of the sixth century philosopher; but the pre-Homeric Athenian, author of the Autochthones (if a book under this title existed) is a plain forgery of probably late origin. of Themistokles, was made a citizen of Thespiai through his influence (Hdt. 8, 75, 1); and if there is anything real in the story of slaves fighting at Marathon (which I rather doubt in spite of Notopoulos A.J.Ph. 62, 1941, p. 352 ff:), they were liberated not enfranchised (Pausan. 7, 15, 7; cf. 1, 29, 7; 32, 3; 10, 20, 2). In Sparta we have at least the case of Tisamenos the Iamid (Hdt. 9, 33). . 144) Wilamowitz Pindaros 1922, p. 272 f. Oxford, 6 St. Margaret’s Road.