This handsome book complements David Wright's earlier work on the Vergilius Romanus and the Vaticanus.[] The aim of the latest study, as the preface says, "is to present the Roman Vergil in such a way that the reader can appreciate its art in its historical context" (5). In order to do so, Wright begins with historical background, then presents each of the illustrated pages (this chapter forms the bulk of the book), and finally considers how the manuscript was produced, its iconographic sources, and the style of painting.
The first chapter very briefly outlines the political situation in Rome in the late-fourth and fifth centuries (the place and time Wright says the manuscript was written). It then analyzes five ivory diptychs from the fifth century, each of which is reproduced in monochrome at 80 percent actual size, the same ratio as the manuscript illustrations. Here Wright's discussion focuses on whether the artists fulfilled classical requirements for representing human figures. By doing so he shows the change from the earlier diptychs, which represent classical norms, to the later ones, which evolve away from them. This discussion establishes the criteria that Wright uses in the next chapter to discuss the manuscript's paintings. More historical context would have strengthened the background presented here. For example, the Saturnalia which Macrobius would have composed at roughly the time the Romanus was written, shows how central Virgil was to fifth-century pagans looking back to the golden age of the Praetextati, as John Matthews put it. In addition, the excellent discussion from Chapter 5, "Style of Painting," which compares the paintings in the Romanus with the mosaics at Santa Maria Maggiore in Rome and in Ravenna and the Vatican Virgil, could profitably have been added to this chapter. The ivories establish one set of criteria Wright uses to analyze the manuscript paintings, but discussing the mosaics would have been useful here as well. It might also have given him an opportunity to address the views of Kenneth Dark, who partially based his argument that the Romanus was written in Britain on Roman mosaics found there.[]
The "Presentation" chapter is the heart of the book. In it Wright offers a close reading of each illustrated page in the manuscript. The goal here is "to give the reader as nearly as possible the experience of turning the pages of the Roman Vergil" (13). To do so, the publishers have taken care to reproduce many of the pages correctly on recto or verso as they appear in the manuscript and at a consistent size. Of the original 410 folios, 309 survive and this chapter reproduces twenty-two pages: the twenty that survive with paintings (counting the framed opening of the Aeneid as a painting) and two to illustrate a colophon and an introductory argumentum. I have not seen the manuscript, but the reproductions seem well done. In one copy there were registration problems in the reproduction of folio 136r (36), which gave the script a slightly fuzzy appearance; in another copy, however, the color plates were in register.
From the surviving leaves, it appears that the scribe treated the Eclogues differently than the Georgics and Aeneid. There are no argumenta for the individual Eclogues and none of the surviving illustrations is a full page. By contrast, each of the Georgics and each book of the Aeneid seems to have opened with an argumentum and a facing pair of illustrations and closed with a colophon. Three of these paired images survive and are reproduced in the book as facing pages; the single sheets that survive from five other pairs are reproduced as full pages with explanatory text opposite. The Romanus is the only extant late-antique Virgil manuscript to integrate the argumenta and at times they face the colophons of the previous book and at times they were the recto and verso of the same leaf. The example of these features given on pages 36 and 37 is somewhat misleading since it shows the colophon of Aeneid 5 on folio 136r facing the argumentum for Aeneid 7 on folio 162v.
The design of the book also reflects this difference in treatment by reproducing the illustrations from the Eclogues not as full pages but as details (with two exceptions) and by including explanatory text on the same page as the reproduction. In this way it gives the sense that the Eclogue illustrations represent a much smaller pause in reading from poem to poem than the elaborate breaks between each Georgic and each book of the Aeneid. This comes at a cost, however. On the facing pages 16 and 17 of the book, folio 3v of the manuscript faces the upper part of folio 6r, an unidentified amount of whose lower half has been replaced with the printed discussions of the two images. Likewise, on pages 18 and 19, folios 9r and 11r face each other, with parts of the upper and lower margins removed for the discussion. This is done to help the reader's imagination make reading Wright's book like reading the manuscript. Of course reading any study of this sort requires an exercise of the imagination to move from printed book to manuscript and Wright is a helpful guide in this. My own preference would have been to require the reader to imagine the rhythm of the book from full-page illustrations, not details.
Wright is an observant guide in describing the paintings, taking pains to compare image and text. The paintings in the manuscript seem to have various styles, which Wright sees as the result of a single artist unaccustomed to working in manuscripts improving as he went along. Perhaps this is so, but might there also be more than one hand at work? In any case, most technical aspects of the painting come in for withering criticism. Wright carries this sharp eye to his discussion of the script as well, and his description of the scribe's practice is a model of clarity. It is odd to see, however, that even in such a luxurious manuscript the text has literally gone to the dogs in Eclogue 4. In place of Virgil's declaration, "If we sing of woods, let the woods be worthy of a consul" (Si canimus siluae, siluae sint consule dignae), canimus has become canibus) (46).
The chapter on iconographic sources focuses on possible models for the paintings in the Romanus. More specifically, Wright is interested in discovering whether there was a set of stock illustrations from which the artist of the Romanus copied. After comparing the illustrated scenes in the Romanus and the Vaticanus, Wright finds only one case in which they may have had a common model, the episode of Ascanius shooting the stag in the Romanus and the wounded stag returning to Silvia in the Vaticanus. By working out the ratio of text to picture on a page, Wright says that the Vaticanus and Romanus could have shared a model with paintings separated by ten lines of poetry. He rejects this possibility, however, as "quite unprecedented and most unlikely" (50), saying that there must have been two sets of Aeneid illustrations. Yet there may be independent evidence for just the sort of manuscript Wright rejects.
R.A.B. Mynors remarked in the preface of his Virgil edition that the ninth-century manuscripts Bern 172 and Paris 7929 had such a close relationship with the Romanus that they must have been copied from it and he used readings from the two medieval manuscripts to supply Romanus readings where it was lacking.[] Louis Holtz noticed that the Bern manuscript has a capital letter every eight lines of text, regardless of whether the sense of the poem seems to demand it. He argued that these capitals come from the practice in many classical and late-antique manuscripts of beginning new pages with an enlarged letter (also a feature of the Romanus, Wright says). From this, Holtz argued that the exemplar of the Bern manuscript could not have been the Romanus but must have been a book with eight lines of text per page, the rest of the space occupied by a painting.[] These eight lines of text correspond with tantalizing closeness to Wright's estimate of ten lines per page and would seem to strengthen the case for the existence of just such a heavily illustrated copy of the Aeneid.
In the last chapter, Wright summarizes his view of the manuscript and shows how it and other late antique manuscripts provided models for scribes of the Middle Ages. For Wright, the Romanus points the way forward in its planned sequence of display pages executed by an artist fundamentally incompetent in drawing the human figure. One might ask whether the artist chose to paint the way he did, rather being condemned to it by his lack of skill, but that would be another book. This attractive book by one of the foremost students of late-antique Virgil manuscripts will be welcomed by those acquainted with them as well as by those seeking an introduction.
[] See Codicological Notes on the Vergilius Romanus, Studi e Testi 345 (Vatican City, 1992), which I have used in writing this review. On the Vaticanus see The Vatican Vergil: a Masterpiece of Late Antique Art (Berkeley, 1993); and Vollstaendige Faksimile-Ausgabe im Originalformat von Codex Vaticanus Lat. 3225 (Graz, 1980; commentary volume, 1984).
Wright spells the name of the poet Vergil in English, which I've retained in quoting him. When I've written the name, I've followed the more general practice of Vergilius in Latin, Virgil in English.
[] Kenneth R. Dark, Civitas to Kingdon: British Political Continuity, 300-800 (London, 1994), 187-191.
 P. Vergili Maronis Opera, ed. R.A.B. Mynors (Oxford 1969), x.
 Louis Holtz, "La redécouverte de Virgile aux VIIIe et IXe siècles d'après les manuscrits conservés' in Lectures médiévales de Virgile: actes du colloque organisé par l'École francaise de Rome (Rome, 1985), 9-30 at p. 25.