The beautiful fairy woman who transforms into a serpent after being betrayed by her human husband was already a familiar figure of written and oral lore by the time Jean d’Arras finished his Mélusine or La noble histoire de Lusignan (HM) around 1393. However, whereas earlier analogues relate a basic folklore motif in which the fairy is rarely given a name, Jean's detailed prose account transforms the story into an elaborate, semi-historical romance, which was to become so well known that it would forever link stories of hybrid serpent women with the name ‘Mélusine’. Around 1401, not long after Jean finished his romance, another French author – known as Coudrette or sometimes La Couldrette – created a verse account of the Mélusine story, which he entitled Roman de Parthenay (RP). Although it has been suggested that the two French versions were derived from a now-lost common source – often presumed to have been written in Latin – recent scholarship has shown that Coudrette's poem is almost certainly a reworking of Jean's version.

In fact, scholarly speculations about an earlier Latin source are based on a mistake: in the inventories of the library of Margaret of Austria a copy of the Castilian translation printed in 1489 is erroneously listed as written in Latin, a reference which some scholars took as evidence supporting the theory of a lost Latin source. Another oft-cited clue in favour of a possible Latin exemplar – Coudrette's comment that he made use of ‘deux beaux livres […] en latin’ (two fine books […] in Latin) which he ‘translater en franczois’ (translated into French) – is a similar red herring, as this claim is more likely a rhetorical device intended to give greater credibility and status to Coudrette's work than a real identification of his sources. A third clue – the mention of two books entitled Histoire de Lesignen ‘en latin’ in the inventory of the estate of the Duke of Berry – is unconfirmed: since neither book survives and the name Mélusine is not mentioned in the entries, we cannot be entirely certain whether these refer to copies of the HM. As this study can only effectively assess Mélusine versions whose existence is proven, I will leave aside such speculations about earlier sources.