Now ἐπ’ ἀριστερὰ is commonly interpreted ‘on the left.’ But if it means on the left in Il. xiii., then the passages are contradictory: because this would place Paris on both wings, whereas he obviously is described as on the same wing of the battle throughout. But if we construe ἀριστερὸς as meaning the west in all the three passages, then we have the same meaning at once made available for all the three places, so that the account becomes self-consistent again; and if the meaning be ‘on the west,’ then we may understand that Idomeneus most naturally betakes himself to the west, because that was the quarter of the Myrmidons, where the Greek line was deprived of support. If, however, it be said, that the Greek left is meant throughout, then the expression in v. 765 is both contrary to what would seem reasonable, and at variance with Homer’s own precedent in the Fifth Book. Thus there is considerable reason to suppose that, in Homer, ἀριστερὸς may sometimes mean ‘west.’ So that _if_ ἐπὶ in Od. v. 277 really means ‘upon,’ the phrase will signify, that Ulysses was to have Arctus on the west side of him, which would place Ogygia in the required position to the east of north. ~_The force of ἐπὶ in Homer._~ The point remaining for discussion is at once the most difficult and the most important. What _is_ the true force of the Homeric ἐπί? I find the senses of this preposition clearly and comprehensively treated in Jelf’s Greek Grammar, where the leading points of its various significations are laid down as follows[665]: [665] Jelf’s Gr. Gr. Nos. 633-5. 1. Its original force is _upon_, or _on_. 2. It is applied to place, time, or causation. Of these three, when treating of a geographical question, we need only consider the first with any minuteness. 3. Ἐπὶ, when used locally, means with the genitive (_a_) _on_ or _at_, and (_b_) motion _towards_ a place or thing. With the dative (_a_) _on_ or _at_, and (_b_) _by_ or _near_. With the accusative (_a_) _towards_, and (_b_) ‘extension in space over an object, as well with verbs of rest as of motion.’ Of this sense examples are quoted in πλεῖν ἐπὶ οἴνοπα πόντον for verbs of motion, and ἐπ’ ἐννέα κεῖτο πέλεθρα for verbs of rest. Both are from Homer, in Il. vii. 83, and Od. xi. 577.