the long e: (2) occasionally to denote the rough breathing—the use for which the Western alphabets regularly reserved it.
But our inscription presents new modifications of these uses:—
(1) In 
κηβόλῳ, by itself denotes epsilon with the rough breathing.
(2) In Δεινοδικο, for Δεινοδίκου, it perhaps serves, as M. Homolle suggests, to aspirate the κ. While koppa was in use[1], it, not kappa, was preferred before o and u. Where kappa was so placed, the need of a complementary sound may have been felt. As, however, we have ϙόρη, it is not easy to see why we have not Δεινοδιϙο. does not strengthen
ο to ου, for we have simply το Ναξιο for τοῦ Ναξίου.
(3) In Νασίο ἔσοχος, σ stands for ξ, which in the older inscriptions is normally expressed by χσ. Thus alone stands for an aspirated κ, just as above for an aspirated ε.
(4) Most remarkable of all is ΑΛΟΝ. No one, I think, who examines the facsimile given by M. Homolle will have any doubt that the word is rightly read thus. The letters are, indeed, clear. The preceding ἔξοχος is clear also. After ἔξοχος (which must be fem.), in hexameter verse, ἀστῶν is the only alternative which presents itself, and the
- ↑ The mere presence of the koppa is a point on which it is unsafe to insist here. In Kirchhoff's opinion (op. cit. p. 39) the known evidence does not compel us to suppose that the koppa had falledn into disuse so early as about Ol. 60 (540 B.C.).
