The heart of this important and scholarly book is an edition of one of the glossed manuscript copies of Ovid's Ars amatoria, Book 1. The ninth-century manuscript, Oxford, Bodleian Library, Auct. F. 4. 32, is among the earliest of a group of manuscripts containing glossed texts of this key Ovidian work, and although the Oxford manuscript copy (fols 37r–47r) is imperfect, it is distinctive because it is heavily glossed not only in Latin but also in Old Welsh. This makes it an object of considerable historical, cultural and linguistic interest, and R. makes the vital point that no-one has previously studied the two sets of glosses together. R. himself, working at the intersection of these disciplines and trained in both, is one of the very few scholars who could successfully take on a comparative study of the glosses in their manuscript context and what this can tell us about the composition of the manuscript, the accumulation of the layers of glossing, the linguistic features of the glosses and the significance of Ovid's work in the literary tradition of medieval Wales.

R. begins by locating the Oxford manuscript in the context of Ovidian glossing as a scholastic tradition and its use in the classroom, emphasising the primarily pedagogic function of the glosses to teach learners ‘what they needed to know to read Latin verse’ (p. 2). He also draws attention to the importance of Celtic-language glosses on Latin texts as the earliest linguistic evidence for Old Irish, Old Welsh, Old Cornish and Old Breton. One of R.’s most suggestive conclusions is that the Old Welsh glosses in the Oxford text of Ars amatoria 1 (labelled the O text) represent the final layer of glossing resting on top of the layers of earlier Latin glosses, at least one of which shows idiomatic traces of Celtic Latin. Three sections of the Oxford manuscript showing Celtic features, including the copy of Ars amatoria 1, ended up in Glastonbury Abbey by the tenth century, and it seems likely that the O text of Ovid was copied and glossed somewhere in south Wales in the ninth century. As R. notes, this makes it ‘one of the oldest surviving manuscript fragments [of the text], and the only evidence that the text was known in early medieval Britain’ (p. 13).

Chapter 2 presents a survey of scholarship relating to the O text, including the nineteenth-century tradition of Celtic linguistic scholarship and its keen interest in Old Welsh glosses of the kind found in the Ovid manuscript. The nature and appearance of the Welsh and the Latin glosses are considered in relation to the overall presentation and layout of the manuscripts, with a complete account of the four main scribal hands, two of which copied the glosses and marginal notes as well as the main text. A large section of the chapter concerns the linguistic and grammatical features of the glosses in both languages, particularly nominal case marking, as part of the pedagogic function of the glosses. Altogether, this is the most detailed study of the O text yet published, and it provides us with an authoritative description and assessment of its value.

The third chapter discusses other glossed manuscripts of Ovid's Ars amatoria 1, comparing the extent and nature of their glossing with that of the manuscript here edited and listing glosses which are common to more than one of the manuscripts. This survey, bolstered by useful tables and graphs, is important, not just for the codicological and linguistic detail but for what these comparisons can tell us about the cumulative layers of glossing. Much of the discussion in this chapter makes a valuable contribution to our understanding of the practice of manuscript glossing in general, such as the way in which copyists chose to include existing glosses while also adding others, depending on the intended readership. R. also makes a distinction between glosses that are ‘predictable’ and those that are ‘sufficiently distinctive’ (p. 112) to have been created uniquely at particular points in the transmission. The patterns he identifies in this group of glossed manuscripts reveal possible connections between manuscripts that share some of the same glosses.

The edition of the O text and its glosses in two languages is clearly presented with an idiosyncratic but serviceable editorial methodology and apparatus. The text has line numbers in the margin and footnotes at the bottom of each page of text; the footnotes contain the glosses (giving the Latin or Welsh gloss with an English translation) but are also used to make occasional editorial comments. The notes to the edition, grouped together in a separate chapter, are largely grammatical, which is appropriate in an edition concerned mainly with linguistic issues. There is a minor difficulty in the fact that the reader does not know in advance which lines of the text have a note attached to them. The layout assumes that the reader will have a copy of the text and a copy of the notes side by side and will work through the text line by line, a difficult feat to accomplish when using a bound book. The two indexes of the glosses, by line order and then by alphabetical order, are extremely useful, enabling word searches as well as an overall view of the glosses from the beginning to the end of the text.

The final chapter of the book is an excellent summary of the reception of Ovid's work in medieval Wales. Given the very limited evidence for Latin manuscripts held in monastic libraries in Wales and the Welsh March, R. concludes that the late-medieval Welsh poets who celebrated ‘Ofydd’ both as a love poet and as a signifier of ‘love poetry’ in general knew about him ‘from their school education and not from some higher-level perusal of surviving manuscripts of Ovid's poetry’ (p. 220).

The book is densely written, heavy on detail and rather too light on synthesis of argument. Although there is an index, it is not always easy for readers to find their way around the book or to recover significant information that was stumbled upon by chance and then lost to view, such as the intriguing level of knowledge of the Roman world (including its theatre) evident in the Welsh glosses (p. 81) or the patchy nature of the glosses and what this might mean (p. 51). There is also a surprising lack of detail about the Oxford manuscript itself and its provenance: the few references to Glastonbury and south Wales seem to assume that readers are already well informed, but more could be said, even speculatively, about how the manuscript (or at least parts of it) came to be copied in Wales and in which monastic foundation(s) it might have been copied. The final section on the reception of Ovid in Wales, while noting the lack of direct evidence for texts of Ovid in Wales after the ninth century, offers no discussion of the singularity of the O text, whose provenance and location in Wales are left unexplored.

Nonetheless, this book is an impressive feat of scholarship. While the Oxford manuscript has been the subject of previous scholarly interest, R.’s edition breaks new ground in providing a full account of the glosses in relation to the actual text, and there is every likelihood that it will become the standard edition of this particular manuscript copy of Ars amatoria 1. R. has achieved what he set out to do, which is to create a text that speaks to both Classicists and Celticists, but his book has a wider relevance still for the study of medieval glossing practices and Classical reception in medieval Britain.