In Latin and Greek dactylic hexameter poetry, a bucolic pause is a pause at the end of a 4th foot which is a dactyl, for example:
Virgil, Georgics iii 54
tum longo nullus lateri modus; omnia magna
“meanwhile [a good cow]'s long flank has no limit; all [points] are large”
I have noticed that in most bucolic pauses that I have seen, the 4th foot ends in “short vowel, consonant, pause”. There seems to be a tendency for a syllable boundary near a pause to be attracted from its usual place to the pause; this here changes the 4th foot from a dactyl (-i modu-) into a cretic (-i modus), causing a tension or suspense, which is resolved when the word after the pause (omnia) comes and starts with a vowel (or h+vowel), thus showing that the 4th foot is a dactyl after all.
I think you’re confusing syllabification with prosodic continuity. What you’ve observed is a paraphonological phenomenon. You have it even between sentences and at changes of speaker, and it’s not confined to dactylic hexameter. Plenty of examples in Greek comedy, for instance.
I realize that difference. Here by the usual rules the syllable divisions are:
tum lon|go nul|lus late|ri modu|s; omnia| magna
but a major pause, at least in English, in my experience tends to cause a speech hesitation:-
tum lon|go nul|lus late|ri modus;| …
and, during that speech hesitation, “-ri modus” sounds like a cretic, which is wrong and causes a tension, until the word “omnia” starts, and prosodic continuity reasserts itself, and the tension resolves in the hearer’s mind as the remembered syllable boundary reverts to the usual place “-|ri modu|s; omnia|…” .
To make your point more strongly you could have adduced corpora praecipue matrum legat. optima torvae just 3 lines earlier, or te sine nil altum mens incohat. en age segnis a little before that at 42. But no, I’m afraid you’re wrong. Nothing “sounds like a cretic” in dactylic hexameter, not even momentarily. A pause after modus or legat or incohat doesn’t compromise the dactylic rhythm so long as a vowel follows, as we know it will. (The horizons of expectation are conditioned by the fact that this is dactylic hexameter.) I hate to say it, but what you need is some metrical theory. Syllable division is a trickier business than “the usual rules” would have you believe. Think about the caesura at (aptly enough) 110 nec mora nec requies; at fulvae nimbus harenae.