Omar Khayyám, a Persian poet who died in 1131, wrote a number of quatrains in Farsi which are regarded by some as representing the very summit of Sufism (that is, of the mystical dimension of Islamic thought) and by others as being essentially agnostic and hedonistic in nature. Those who are of the latter view are often strongly influenced by the ‘translation’ into English of some of these quatrains by Edward Fitzgerald, a British poet and writer whose first edition of the Rubáiyát of Omar Khayyám appeared in 1859, at the height of the Victorian era. Although there have been several other translations of Khayyám’s quatrains, none has been as popular or, perhaps, as highly regarded as an artistic work as that of Fitzgerald. It has rarely, however, been regarded as a work that is faithful to the intent of the original. In deciding to translate into Māori Fitzgerald’s rendering into English of some of Khayyám’s Farsi quatrains (5th version), Pei Jones was faced with a peculiarly complex set of problems (linguistic, literary, cultural and religious). Pei Jones’ translation, a translation of a translation, is generally regarded as being faithful to Fitzgerald’s version of the Rubáiyát. It would appear, therefore, that he decided to treat Fitzgeralds’s text, in spite of the reference in its title to the original text, as his source text. This gives rise to a number of questions, including questions about what it means for a translator to be faithful or unfaithful to a source text. With particular reference to Pei Jones’ translation of Fitzgerald’s Rubáiyát of Omar Khayyám, this thesis explores the concept of ‘fidelity’, a concept that, it is argued here (see Chapter 3), is often treated in the literature on translation in a way that belies its extremely complex nature. The thesis proposes a new approach to the concept of fidelity, one that is based on nine fidelity types: grammatical, lexical, informational, metrical, imagistic, rhetorical, historical, didactic and functional fidelity. In terms of this nonagonal analytical model, twenty-five of Pei Jones’ quatrains are analysed in relation to the equivalent quatrains in Fitzgerald’s version (Chapter 4). The analysis indicates that Pei Jones’ translation has neither metrical fidelity (a consequence of the very different nature of the source and target languages) nor functional fidelity (a consequence of the very different expectations and sensibilities that a Māori audience has in relation to the verbal arts). Metre and function are both, however, fundamental to the enduring appeal of Fitzgerald’s quatrains. The overall conclusion is that since it is often impossible to achieve all nine types of fidelity, translators need to carefully consider what their primary aim is in undertaking the translation of artistic works and be prepared to sacrifice certain types of fidelity (e.g. historical and informational fidelity) in order to create a work that fulfils the aesthetic expectations of the target audience. Pei Jones was undeniably successful in achieving those types of fidelity that were possible. However, the work may have had more widespread appeal if he had sacrificed some of them in order to create a work that was more closely aligned with the aesthetic expectations of Māori readers.