Versification:  a7’ b7 c7 d7’ e7 f7’ in odd stanzas and b7 a7’ d7’ c7 f7’ e7 in even stanzas (Frank 864:6). Nine coblas doblas. However, the poem uses derived rhymes that link morphologically the ‘a’ and ‘c’, the ‘b’ and ‘d’, and the ‘e’ and ‘f’ rhymes in each stanza and between stanzas, thus (see Dominique Billy, L’Architecture lyrique médiévale: analyse métrique et modélisation des structures interstrophiques dans la poésie lyrique des troubadours et des trouvères, Montpellier, 1989, p. 185):

Chambers (Frank M. Chambers, An Introduction to Old Provençal Versification, Philadelphia 1985, p. 54) suggests the rhyme scheme may in fact be represented ‘ababcc’, but a more accurate representation (see the table above) is ‘a’bab’cb’’ for odd and ‘ba’b’ab’c’ for even stanzas. The use of derived rhyme means the melody would need modifying for even stanzas, though this is not as big a problem as Chambers thinks (Introduction, p. 54); indeed if the tune was strongly rhythmical, as the term tresc/tresca suggests, the modifications needed would be slight, affecting only the end of each line. The patterning of rhymes indicates material has gone missing in transmission: stanza VII in the Ms. is followed by a stanza which apparently has the rhyme scheme of odd stanzas, though it is clearly corrupt as in the first line it has a rhyme in -ans and a rhyme-word that is not morphologically linked to the rhyme-word of the third line of the stanza. Whereas previous editors assume a complete stanza is missing and emend twice (our lines 46 and 51; lines 43 and 45 in Dejeanne and Paterson) to produce a rhyme-pair that makes the stanza conform to the rhyme scheme of odd stanzas, we assume that the corrupt stanza in the Ms. has been confected from two stanzas. The song concludes with a two-line tornada.