eskeinie. PE22:103 per ending (where okoine is prob. for *okoinie).
Start per ibid.:112. TQ exeinie.
eskéye not excluded, also eskeitie if xíta, not xéya.

+Александр Запрягаев hm, why not *eskítie, *eskínie, wouldn't having a stress on the syllable push it towards de-diphtongisation (if that's a word:)?

√SKEY xíta- to pass > √KEY cíta- to suppose?

+James Coish you mean that as an analogy? Yes, that may be, unless one wants to distinguish that to pass is intransitive and suppose is transitive. But I guess that may be going too far.
But the question for me wasn't really the aorist (I can accept either of them as valid) but the shape of the perfect.

+Tamas Ferencz I mean that perhaps xíta- might follow the same pattern as cíta-. Maybe I used the ">" incorrectly. :)

+James Coish perhaps, but it does not help with the perfect ☺️

+Александр Запрягаев I agree with you, except that in later Quenya sk usually metathesized to ks even medially:

eldamo.org - Eldamo : Quenya : [sp], [sk] metathesized to [ps], [sk]

Medial sk is more common that ks. So I would use exeinie or exeitie

Right. So to piggyback on +James Coish 's idea, is the prefect of kíta then ekeitie?

Erg. It seems that ei is not a normal diphthong in Quenya, and usually ei became í, as in níta < neitā. Thus the perfects should be exínie or exítie, or possibly reformed to ixínie or ixítie.

Hmm. I checked PE19/106, and there the rule is that ei became ē in stressed monosyllables and stressed non-final syllables, but otherwise becomes ī. Apparently this change was after the Quenya stressed shifted. So I’m not sure why the attested form is cíta- and not céta-. Apparently Tolkien didn’t quite make up his mind on the development of ei in Quenya.