Actually I am quite uncertain about the form hlarunya; as far as I know we don't have examples of u-duals with possessive suffixes, but my lambetyáve tells me the plural/dual marker should be the last; so should dthat be hlarunyat, with a duplication of the dual marker?

+Tamas Ferencz That's a very fair question. I agree that the dual inflection would typically go after possessive suffixes; using u as a connecting vowel for nonsingular is a possibility (as with elenillor). Perhaps PE21 has the answer somewhere…

Maybe when there is a pronominal suffix present, the speakers just switch to the t-dual inflection. After all, "hands" are eminent candidates for an u-dual and yet we have máryat attested.

+Robert Reynolds there are forms like nentau and nenuntau attested in PE21 which may point to the t-marker and u-marker coexisting in one word? Or we assume the marker reallocating to the end of the word after the pronominal suffix, like *hendenyau. It's looking weird, but the above forms show they may not be unimaginable.

eldamo.org - Eldamo : Middle Quenya : nén

A really interesting question. I'd still go with forms like hlarunya, hendunya (with no dual marker reduplication). If we constructed something like máryat, what will be the connecting vowel? the most logical would be i, but for me **hlarinyat would sound rather awkward, as if it had a plural -i marker inside.

+Andre Polykanine but wouldn't that be the case with normal plurals as well? karinyar "my heads"