The Paper Museum (‘Museo Cartaceo’) is a collection of thousands of watercolours, drawings and prints, assembled in the seventeenth century by the Roman collector Cassiano dal Pozzo. It represents perhaps the most significant attempt before the age of photography to document ancient art and culture. The collection was sold by Cassiano's heirs to Pope Clement XI in the early eighteenth century, acquired by King George III in 1762 and transferred in 1834 to the Royal Library at Windsor Castle. The catalogue of this huge visual encyclopaedia has been in the course of publication since 1993 in 36 volumes, divided into three series: Antiquities and Architecture, Natural History and Prints.

The 160 drawings catalogued in this sixth volume of Series A are taken from five ancient manuscripts, among the oldest surviving Latin codices: the famous Vatican Virgil from c. a.d. 400 (Biblioteca Apostolica Vaticana, Vat. lat. 3225), the Roman Virgil from the late fifth or sixth century (BAV, Vat. lat. 3867), the Vatican Terence from c. a.d. 825 (BAV, Vat. lat. 3868), the Palatine Agrimensores (BAV, Pal. lat. 1564) and a now lost Carolingian copy of the Calendar of the year 354. They were copied for Cassiano between 1632 and 1634.

Introductory essays by C. about Series A and Classical manuscript illustrations, and the Paper Museum in general, are followed by a short description of the five manuscripts (with some nice reproductions from the original manuscripts) and their history down to Cassiano's day. All the drawings are reproduced as full-page colour plates, with detailed descriptions, transcriptions and translations of the texts and a commentary on the iconographical content.

Most interesting is H.'s chapter, ‘Late Antique Manuscripts in Early Modern Study: Critics, Antiquaries and the History of Art’, which focuses on the two Virgil manuscripts and the Terence, which enjoyed a common fortune. H. illustrates how interest in these manuscripts changed over time, mirroring ‘a certain shift in scholarly preoccupations that may be understood as reflecting the general transformations of intellectual paradigms’ (p. 52). Humanists looked at these manuscripts from a philological and palaeographical viewpoint, exploiting them for variae lectiones and comparing their capital letters to the ones found in antique inscriptions. It is no coincidence that these manuscripts figured prominently in Mabillon's pioneering De re diplomatica (1681). For a long time, philologists debated the precise age of these manuscripts, but merely on palaeographical grounds, since they had no interest in the splendid miniatures which decorate them. The discovery of the miniatures was the work of antiquaries like Fulvio Orsini and indeed Cassiano dal Pozzo, whose collection of drawings was used by many seventeenth-century scholars in their antiquarian publications. They tried to reconstruct ancient civilisation, customs and institutions, for which the study of material culture was indispensable. Pre-eminence was given to the analysis of costume, which was considered a reflection of social identity (as it was in the seventeenth century). The Paper Museum was therefore conceived as a corpus antiquitatum, covering all aspects of ancient life. It often illustrates both Cassiano's obsession with taxonomy and his copyists' tendency to ‘improve’ the images and impose a classical canon on them. Unlike textual critics, antiquarian scholars were not particularly interested in dating the manuscripts, given their view of antiquity as static. Only in the late eighteenth century did the first glimmerings of an art-historical point of view manifest themselves, emancipating the history of art from the history of documentation, especially in Séroux d'Agincourt's Histoire de l'art.

It is quite reasonable that the authors have not aimed this volume primarily at a readership of Latinists. It is more regrettable, though, that they appear not to have called on any Latinist in its preparation, as no occasion for gaffes seems to have been missed.

When C. wrongly calls the use of pingi in pingi curavit ungrammatical (p. 20) and H. prints awkward renderings of the few Latin quotations translated in his chapter – for example, the models ‘not yet lacking the majesty of the early Empire’ for nihilque in iis exhibetur quod primam Romani imperii majestatem non redoleat (p. 60; and in n. 98 on p. 87 con teguntur should be one word) – this is just a prelude to the disaster awaiting the reader in the pages about the Roman calendar of 354 (pp. 94–126). The non-existent gabet (for habet) transcribed on p. 112 is no mere keyboard slip, as the illustration on p. 97 has an h that might indeed be taken for a g – as happens in seventeenth-century handwriting – and on p. 124 quamius should be quam vis. More problematic methodologically is C.'s decision, while offering an allegedly diplomatic transcription of the verses that accompany the illustrations of the months in Cassiano's calendar model, to quote the English translation by M.R. Salzman (On Roman Time: the Codex-Calendar of 354 and the Rhythms of Urban Life in Late Antiquity [1990]), which is based on a thoroughly emended version of these corruptly transmitted verses. On p. 122, for instance, C. prints Devotusque satis incola Memphi deis and then adopts Salzman's rendering ‘your’, which reflects Shackleton Bailey's emendation tuis for satis (as printed by Salzman). Elsewhere Salzman's translation is more explicitly adapted, but not improved: on p. 118, for Captivam filo gaudens religasse lacertam we read ‘he rejoices at the lizard, held captive on a string (trans. adapted from Salzman, p. 103)’, whose ‘he rejoices at having tied up the lizard’ did make sense. On p. 332, (Judex de finibus et controversiis refert), de quibus consulendus est imperator, ut iis decidendis eius iussa sequatur is translated ‘about which the emperor is consulted, so that he can decide about them and his judgment may follow’. In point of fact, the judge should consult the emperor, so that in his ruling he may follow the latter's orders.

Finally, one can only be appalled at the comedy of errors on display in the transcription of the ‘Outline plan for an edition of the Vatican Virgil with its miniatures’ from Vat. lat. 10486 on pp. 373–80. Besides a myriad of obvious typographical errors (quea for quae; littoral for littora; innupteaque for innuptaeque on p. 373 alone), this edition strangely combines a superfluous use of ‘[sic]’ next to forms that are perfectly acceptable in humanist Latin (fenilia, premia, caetera, exequitur, peana, extantem) with ‘transcriptions’ of non-existent words that do call for a [sic]. The reproduction on p. 374 of the folium transcribed on p. 373 shows that this edition's pontis stat in agmine castics in the manuscript reads positis stat in agmine castris, as it should – all these lines are easy to check in Virgil – and that peragros is per agros, while reginis is regnis. No manuscript facsimile of the remaining six pages is provided, yet unchallenged readings such as siut, mavibus, imode pectore, sommus, iumdudum, nnumen, sabctae, gravatyum, bultus and nuque speak for themselves. And why not resolve abbreviations like qd, a.io and p.terea?

Maybe the editors of a book about Latin manuscripts that ships at over a hundred pounds should not have skimped on the modest expenditure of hiring someone with the necessary skills to avoid such an editorial nightmare. In the end, however, the neglect of basic philological and palaeographical concerns exhibited here might in itself be considered a most telling, yet unintended, instance of the shift in scholarly preoccupations that these beautifully illustrated ancient manuscripts have undergone through the centuries.