Bisyllables ending in -ex/ix: Does the accusative stress always fall on the same part?

Question

Bisyllables ending in x. Does the accusative stress always fall on the same part as Felix-Felīcem Helix-Helicem(strangely in Italian it results Hêlica) Fenix-Fenīcem (Fenîce) Syrex-Syrīcis (Sírice in Spanish)?

Answer 1

There are no native words in -ĭx, only in -īx, with a long vowel. It's not necessarily clear whether that's a regular development or just a coincidence—almost all of these words are feminine agent nouns (genetrīx, amātrīx, &c.) or other feminines (iūnīx; probably also fēlīx, which, though it is an adjective, must have a complicated etymological history) containing the same suffix continuing PIE *-ih₂- > -ī-, and AFAIK the remainder just have no clear etymology (pernīx). Either way, the long vowel means these words will have a heavy penultimate syllable in the "inflected" forms other than the dative and ablative plural, and because of the way Latin stress works, that syllable will be stressed:

	sg.	pl.
nom.	cornīx	cornīcēs
voc.	cornīx	cornīcēs
acc.	cornīcem	cornīcēs
gen.	cornīcis	cornīcum
dat.	cornīcī	cornīcibus
abl.	cornīce	cornīcibus

The same does not apply to words in -ĕx (there are no words in -ēx): the short vowel (which degrades to -ĭ- in the "inflected" stem) and the fact that no nominal ending (other than the nominative singular) starts with a consonant means all "inflected" forms have a light penultimate syllable, so the accent falls on the antepenult:

	sg.	pl.
nom.	cīmex	cīmicēs
voc.	cīmex	cīmicēs
acc.	cīmicem	cīmicēs
gen.	cīmicis	cīmicum
dat.	cīmicī	cīmicibus
abl.	cīmice	cīmicibus

For both words in -īx and words in -ĕx, this applies generally, not just for bisyllables.

Two of your words aren't native Latin words: helix is from Greek ἕλιξ, and fenix (more commonly phoenix) is from Greek φοῖνιξ (syrex, though it looks like a Greek loan, is actually properly sirex, and the genitive is sirĭcis, not sirīcis). Greek has a different history and the -ι- in both ἕλιξ and φοῖνιξ is short. Whether this translates to a short -i- in Latin varies: L&S claim yes for helix but no for phoenix, with commensurate implications for the locations of the accent.