Remember 1.174 Ac primum silici scintillam excudit Achates? Last month I said I was inclined to take silici as ablative, not dative, and after some discussion I said I’d consult Horsfall on 7.668 capiti next time I was in the library. I’ve now done so, and can report he takes that as dative (which I welcome), explicitly rejecting ablative, for which however he cites “Cat.68.124, Tib.1.1.72, NW 1, 366” (I don’t know what NW is, unless a misprint for NH, the Nisbet-Hubbard Horace comms.). So I turn to Fordyce on the Catullus, and he says the abl. in –i is “not uncommon in Lucretius, and later poets sometimes find [it] convenient,” instancing capiti, cineri, lateri, and nemori (which all have the same metrical shape as silici) and citing Vergil and Propertius for the archaizing capiti. By “convenient” I first assumed he meant metrically convenient, but evidently not, since capite would scan equally well in both the Catullus and the Tibullus (I haven’t looked up others).
I still favor ablative here, now more strongly than before. I have little confidence in my Stilgefühl when it comes to Vergil, but I feel dative would be just too odd. Then the question becomes Why silici rather than silice? Sound? Archaic tone? I fancy it may be no more than instinctive discomfort on Vergil’s part with having a short vowel before sc-, a collocation I think Latin poets tend to avoid because of its metrical equivocality.
Sorry this is such a small and esoteric matter.
I don’t think I need to defend what I said (whatever it was) about Homer reading faster than Vergil. Homer, simple and direct, reads fast because the phrasing is repetitive and largely predictable; in at least half the lines you know what the last word will be before you get to it, and rarely do you have to think. Not so with Vergil, where the forms of expression are infinitely more complex, and sound and metre infinitely more important. Every word is telling, every line and every scene crammed with resonance and intertextual subtlety. Do I exaggerate? Perhaps.
Vergil can be read quite fast. (After all, Vergil read parts of the Aeneid to Augustus, we are credibly told—but then Augustus knew Latin, and we don’t). I’ve just read through most of Aen.7 at a fair clip. But I think I’d have done better not to.
And as to translating them: to translate either one is to wreck it, but the wreckage is worse with Vergil. Far far better to read him without translating. We’ll never learn to read Latin if we insist on translating all the time.
Incidentally, going back to ater, Aen.7.456f. has atro / lumine!