χωρὶς τό τ᾽ εἰπεῖν πολλὰ καὶ τὰ καίρια.
The τε … και puts το ειπειν πολλα and (το ειπειν) τα καιρια in tandem. These two things are χωρίς, “apart” from each other. (MarkAntony misunderstood the syntax.)
χωρις is an adverb (unlike πολλα and καιρια, which are adjectives). If it were used prepositionally it would take genitive, but the syntax here is different: lit. “Apart (are) both speaking πολλα and (speaking) τα καιρια,” i.e. “Talking a lot and to the point are two separate things.”
PS. I now notice there’s a textual point here too, which nicely illustrates issues discussed (off-topic, to please Paul) just a couple of days ago at http://discourse.textkit.com/t/anabasis-1-9-16-singular-becomes-plural/14254/1 The OP’s text, which we’ve all been following, gave τὰ καίρια, and that is the reading of all the surviving manuscripts of the play. But the Suda—a 10th-cent. lexicon drawing on earlier sources—quotes the line with τὸ καίρια, and that gives better syntax, since it puts two articular infinitives in parallel, just as in the passages quoted by jeidsath: το (ειπειν) καιρια balancing το ειπειν πολλα. With τα καιρια, we have to understand not only ειπειν, which is unexceptionable, but το ειπειν, which is not. And the article of τα καιρια serves no good purpose here; without it, καιρια balances πολλα. So το καιρια stands to be the right reading, and this is a case where the manuscripts of the play (the “direct tradition”) are all wrong.
The corruption is easily explained, the utrum in alterum question easily answered. το καιρια, with no infinitive on the immediate scene, would inevitably have become τα καιρια (the lectio facilior, therefore).
Errors that have permeated the medieval tradition can be exposed as such either by older manuscripts, esp. papyri, or by the “indirect” tradition, such as quotations in other authors, as here with the Suda. (Only one papyrus with fragments of this play has yet come to light, and it is missing here, so we can’t know whether it had τα or το.) For most passages, however, there are neither papyri nor indirect sources available to us, so it’s more than likely that there are hidden corruptions in the transmitted text of Sophocles and of other authors. That’s where conjectural emendation (aka “tampering with the text”) comes into play. If all sources had τα καιρια here, probably no-one would have questioned it and proposed το καιρια instead. But they probably should have.