The Kdbeiroi or Megd/oi Theoi 315
kersos 'He that cleaves with the Axe1,' and Axieros the 'Axe-Eros2'
or ' Axe-spirit3.' On this showing Axiokersa was an appellative
of the Thracian earth-goddess, Axiokersos an appellative of the
Thracian sky-god, while Axieros was the name given to an emana-
tion or rebirth of the latter. The same substratum of primitive
Thracian belief probably underlies several of the early Greek theo-
gonies. Thus Hesiod was aware of three primeval powers, Chaos,
Gaia, and Eros4. Pherekydes of Syros, not only recognised Zeus,
1 Hesych. Kepcrai " Ko\^ai, repLeiv, neipai, yafirjaai (so cod. A. Meineke cj. yaWiaai) and
Kepcrr/s- yd/xos (so cod. M. Schmidt cj. ydWos). The Hesycliian glosses were cited in this
connexion by Welcker Gr. GStterl. i. 329 n. 3 and after him by L. Bloch in Roscher Lex.
Myth. ii. 2526. Other views in R. Pettazzoni loc. cit. p. 693.
- That the second element in the compound 'A£t-epos is none other than Eros
(Homeric 'tpos) was perceived by F. G. Welcker Die Aeschylische Trilogie Pro7iietheus unci
dieKabirenweihe zu Lemnos Darmstadt 1824 p. 240 (after N. Freret), E. Gerhard Hyper-
boreische-rdmische Stiuiien Berlin 1852 ii. 209, C. Strube Studien iiber den Bilderkreis von
Elensis Leipzig 1870 p. 74. G. F. Unger in the Jahrb.f. Philol. u. Pddag. 1887 cxxxv. 57
assumes a word meaning ' earth ' connected with Zpa, Zpai'e. O. Crusius Beitriige zur
grieckischen Mythologie und Religionsgeschichte Leipzig 1886 p. 26 n. 2 divides 'diji-
iepos?'
3 The derivation of Ipwj, dpos is much disputed (see e.g. Prellwitz Etym. Wbrterb. d.
Gr. Spr.2 p. 152, Boisacq Diet. etym. de la Langite Gr. p. 270 f.) and need not concern
us. The essential thing is to realise that Eros was the form taken by the soul of the sky-
god on the occasion of his rebirth ; for the art-type of Eros is the art-type of souls in
general. Miss J. E. Harrison hits the nail on the head, when she says : 'Eros is but a
specialized form of the Ker ; the Erotes are Keres of life, and like the Keres take the form
of winged Eidola' (Harrison Proleg. Gk. Pel.2 p. 631). Eros is therefore the male Harpy
(Hesych. dpirw ipwra. AioXeis with M. Schmidt ad loc., et. mag. p. 148, 33 ff."Ap'irvs • 6
"Epcos' ws 7rapd Uapdeviw ev Kpivayopa, " dp.(pOTepois e7ri/3aj "Ap7ri/s eXyftaaro " :—irapa. to
ap-rra^eiv ras eppevas) or Erinys (Hesych. opva- 'Epivvs). On the relations of Eros to
Psyche see L. Stephani in the Compte-rendu St. PJt. 1877 pp. 53—219, A. Furtwangler
in Roscher Lex. Myth. i. 1370—1372, O. Waser in Pauly—Wissowa Real-Enc. vi 531 —
542 and in Roscher Lex. Myth. iii. 3237—3256, G. Nicole in Daremberg—Saglio Diet.
Ant. iv. 747—750, Gruppe Gr. Myth. Pel. p. 871 ff., id. Myth. Lit. 1908 p. 320 ff. The
soul of a double-axe might be represented also as a butterfly (infra § 3 (c) i (c)).
4 Hes. theog. 116 ff. as cited by Plat. symp. 178 B, Aristot.- met. 1. 4. 984 b 27 ff., de
Xenoph. 1. 975a 11 ff., Sext. adv. math. 9. 8, Asklepios in Aristot. met. p. 29, 30 ff.
Hayduck, Stob. eel. 1. 9. 5 p. 112, 20 ff. Wachsmuth, cp. 1. 9. 12 p. 115, 5 ff. Cp. Paus.
9. 27. 2, Theophil. ad Autol. 2. 12, schol. Ap. Rhod. 3. 26, schol. Theokr. 13. 1, and,
though Hesiod is not named, Serv. in Verg. Aen. 1. 664 = Myth. Vat. 2. 35. See further
A. Rzach ad loc. Similarly Akousilaos frag. 1 (Frag. hist. Gr. 1. 100 Miiller) ap.
Damaskios quaest. de primisprincipiis 124 (i. 319, 8 ff. Ruelle), cp. Philodem. wepl evcre-
f3eias 137, 13 p. 61 Gomperz, made Chaos come first, then Erebos and Nyx, from whose
union arose Aither, Eros, Metis: Plat. symp. 178c and Stob. eel. 1. 9. 12 p. 115, 7 f.
Wachsmuth state generally that Akousilaos agreed with Hesiod; schol. Theokr. 13. 1
asserts that, according to Akousilaos, Eros was the son of Nyx and Aither. Sappho
frag. 132 Bergk4 ap. schol. Ap. Rhod. 3. 26, Eudok. viol. 354 held that his parents were
Ge and Ouranos. Ibykos frag. 31 Bergk4 ap. schol. Ap. Rhod. 3. 26, Eudok. viol. 354
is said to have traced the descent of Eros from Chaos ; but T. Bergk ad loc. points out
that cod. Laur. of schol. Ap. Rhod. 3. 26 reads "IfivKos <.....> 6 5^ 'H<n'o5oj iic Xdovs
\eyei To^"Epwra, which leaves us in the dark as to Ibykos' view. In Parmenides frag. 13
kersos 'He that cleaves with the Axe1,' and Axieros the 'Axe-Eros2'
or ' Axe-spirit3.' On this showing Axiokersa was an appellative
of the Thracian earth-goddess, Axiokersos an appellative of the
Thracian sky-god, while Axieros was the name given to an emana-
tion or rebirth of the latter. The same substratum of primitive
Thracian belief probably underlies several of the early Greek theo-
gonies. Thus Hesiod was aware of three primeval powers, Chaos,
Gaia, and Eros4. Pherekydes of Syros, not only recognised Zeus,
1 Hesych. Kepcrai " Ko\^ai, repLeiv, neipai, yafirjaai (so cod. A. Meineke cj. yaWiaai) and
Kepcrr/s- yd/xos (so cod. M. Schmidt cj. ydWos). The Hesycliian glosses were cited in this
connexion by Welcker Gr. GStterl. i. 329 n. 3 and after him by L. Bloch in Roscher Lex.
Myth. ii. 2526. Other views in R. Pettazzoni loc. cit. p. 693.
- That the second element in the compound 'A£t-epos is none other than Eros
(Homeric 'tpos) was perceived by F. G. Welcker Die Aeschylische Trilogie Pro7iietheus unci
dieKabirenweihe zu Lemnos Darmstadt 1824 p. 240 (after N. Freret), E. Gerhard Hyper-
boreische-rdmische Stiuiien Berlin 1852 ii. 209, C. Strube Studien iiber den Bilderkreis von
Elensis Leipzig 1870 p. 74. G. F. Unger in the Jahrb.f. Philol. u. Pddag. 1887 cxxxv. 57
assumes a word meaning ' earth ' connected with Zpa, Zpai'e. O. Crusius Beitriige zur
grieckischen Mythologie und Religionsgeschichte Leipzig 1886 p. 26 n. 2 divides 'diji-
iepos?'
3 The derivation of Ipwj, dpos is much disputed (see e.g. Prellwitz Etym. Wbrterb. d.
Gr. Spr.2 p. 152, Boisacq Diet. etym. de la Langite Gr. p. 270 f.) and need not concern
us. The essential thing is to realise that Eros was the form taken by the soul of the sky-
god on the occasion of his rebirth ; for the art-type of Eros is the art-type of souls in
general. Miss J. E. Harrison hits the nail on the head, when she says : 'Eros is but a
specialized form of the Ker ; the Erotes are Keres of life, and like the Keres take the form
of winged Eidola' (Harrison Proleg. Gk. Pel.2 p. 631). Eros is therefore the male Harpy
(Hesych. dpirw ipwra. AioXeis with M. Schmidt ad loc., et. mag. p. 148, 33 ff."Ap'irvs • 6
"Epcos' ws 7rapd Uapdeviw ev Kpivayopa, " dp.(pOTepois e7ri/3aj "Ap7ri/s eXyftaaro " :—irapa. to
ap-rra^eiv ras eppevas) or Erinys (Hesych. opva- 'Epivvs). On the relations of Eros to
Psyche see L. Stephani in the Compte-rendu St. PJt. 1877 pp. 53—219, A. Furtwangler
in Roscher Lex. Myth. i. 1370—1372, O. Waser in Pauly—Wissowa Real-Enc. vi 531 —
542 and in Roscher Lex. Myth. iii. 3237—3256, G. Nicole in Daremberg—Saglio Diet.
Ant. iv. 747—750, Gruppe Gr. Myth. Pel. p. 871 ff., id. Myth. Lit. 1908 p. 320 ff. The
soul of a double-axe might be represented also as a butterfly (infra § 3 (c) i (c)).
4 Hes. theog. 116 ff. as cited by Plat. symp. 178 B, Aristot.- met. 1. 4. 984 b 27 ff., de
Xenoph. 1. 975a 11 ff., Sext. adv. math. 9. 8, Asklepios in Aristot. met. p. 29, 30 ff.
Hayduck, Stob. eel. 1. 9. 5 p. 112, 20 ff. Wachsmuth, cp. 1. 9. 12 p. 115, 5 ff. Cp. Paus.
9. 27. 2, Theophil. ad Autol. 2. 12, schol. Ap. Rhod. 3. 26, schol. Theokr. 13. 1, and,
though Hesiod is not named, Serv. in Verg. Aen. 1. 664 = Myth. Vat. 2. 35. See further
A. Rzach ad loc. Similarly Akousilaos frag. 1 (Frag. hist. Gr. 1. 100 Miiller) ap.
Damaskios quaest. de primisprincipiis 124 (i. 319, 8 ff. Ruelle), cp. Philodem. wepl evcre-
f3eias 137, 13 p. 61 Gomperz, made Chaos come first, then Erebos and Nyx, from whose
union arose Aither, Eros, Metis: Plat. symp. 178c and Stob. eel. 1. 9. 12 p. 115, 7 f.
Wachsmuth state generally that Akousilaos agreed with Hesiod; schol. Theokr. 13. 1
asserts that, according to Akousilaos, Eros was the son of Nyx and Aither. Sappho
frag. 132 Bergk4 ap. schol. Ap. Rhod. 3. 26, Eudok. viol. 354 held that his parents were
Ge and Ouranos. Ibykos frag. 31 Bergk4 ap. schol. Ap. Rhod. 3. 26, Eudok. viol. 354
is said to have traced the descent of Eros from Chaos ; but T. Bergk ad loc. points out
that cod. Laur. of schol. Ap. Rhod. 3. 26 reads "IfivKos <.....> 6 5^ 'H<n'o5oj iic Xdovs
\eyei To^"Epwra, which leaves us in the dark as to Ibykos' view. In Parmenides frag. 13