Once more, Demeter is a goddess of Earth, not of Dawn.  How, then, does the explanation of a hypothetical Dawn-myth apply to the Earth?  Well, perhaps the story, the unseemly story, was first told of Erinnys (who also is ‘the inevitable Dawn’) or of Deo, ’and this name of Deo, or Dyava, was mixed up with a hypokoristic form of Demeter, Deo, and thus led to the transference of her story to Demeter.  I know this will sound very unlikely to Greek scholars, yet I see no other way out of our difficulties’ (ii. 545).  Phonetic explanations follow.

‘To my mind,’ says our author, ’there is no chapter in mythology in which we can so clearly read the transition of an auroral myth of the Veda into an epic chapter of Greece as in the chapter of Saranyu (or Surama) and the Asvins, ending in the chapter of Helena and her brothers, the [Greek]’ (ii. 642).  Here, as regards the Asvins and the Dioskouroi, Mannhardt may be regarded as Mr. Max Muller’s ally; but compare his note, A. F. u.  W. K. p. xx.

My Theory of the Horse Demeter

Mannhardt, I think, ought to have tried at an explanation of myths so closely analogous as those two, one Indian, one Greek, in which a goddess, in the shape of a mare, becomes mother of twins by a god in the form of a stallion.  As Mr. Max Muller well says, ’If we look about for analogies we find nothing, as far as I know, corresponding to the well-marked features of this barbarous myth among any of the uncivilised tribes of the earth.  If we did, how we should rejoice!  Why, then, should we not rejoice when we find the allusion in Rig Veda?’ (x 17, 1).

I do rejoice!  The ‘song of triumph,’ as Professor Tiele says, will be found in M. R. R. ii. 266 (note), where I give the Vedic and other references.  I even asked why Mr. Max Muller did not produce this proof of the identity of Saranyu and Demeter Erinnys in his Selected Essays (pp. 401, 492).

I cannot explain why this tale was told both of Erinnys and of Saranyu.  Granting the certainty of the etymological equation, Saranyu=Erinnys (which Mannhardt doubted), the chances against fortuitous coincidence may be reckoned by algebra, and Mr. Edgeworth’s trillions of trillions feebly express it.  Two goddesses, Indian and Greek, have, ex hypothesi, the same name, and both, as mares, are mothers of twins.  Though the twins (in India the Asvins, in Greek an ideal war-horse and a girl) differ in character, still the coincidence is evidential.  Explain it I cannot, and, clearly as the confession may prove my lack of scientific exactness, I make it candidly.