ancient versions. I have considered that Perugia is more likely to be known to readers than Perusia, and Modena than Mutina. Most controversially of all, I have occasionally used the names of countries that did not exist in the same form in classical times, such as Turkey, Albania and Afghanistan. I have employed ancient names only in cases where either there is no human habitation now where towns or cities once stood, or where the historical name is the one that is generally understood. It would be perverse to speak not of Troy but of Hisarlik, the Turkish village that now stands on the site. The numbering of the poems (if they originally had any titles, these have not survived) follows the order in  medieval manuscripts, our earliest source. Cases where the manuscripts give as one poem what are now thought by scholars to be two account for numbers like ii.26a and ii.26b. The opposite situation accounts for such numbers as iii.24–25.

This book could not have come about without Michael Schmidt, Director of Carcanet Press and General Editor of PN Review, which first published some of these poems. He was willing to gamble on someone whose only previous contribution to world literature had been wire-service news stories with a maximum life expectancy of twenty-four hours. As I made the transition from news agency journalist to verse translator, he also gave me the benefit of his tradecraft acquired over more than forty years as a practising poet, critic and editor, suggesting improvements and turns of phrase.

I need to acknowledge those who taught me Latin and ancient Greek, in particular the late Robert Levens of Merton College, Oxford – himself no mean translator of some of Catullus’ more outrageous epigrams.

In my teens, two books fired my interest in Latin poetry. One was Poets in a Landscape (1957) by the great classicist Gilbert Highet. Some of his literary judgments now look dated, but his enthusiasm for his subject (including the links

251