Students of Roman pottery tablewares in the Mediterranean (and further afield) will long have been aware of the existence of multiple sources in the Roman provinces of Africa Proconsularis and Byzacena, roughly equivalent to modern Tunisia. The definition and description of these wares was first brought into an effective and readily usable form by Hayes (Reference Hayes1972). In the (nearly 50) years since then, our understanding of these wares has advanced impressively. Carandini et al. (Reference Carandini1981) brought another perspective to the topic and made strenuous efforts to differentiate both the principal and more subtle variations of fabric. Following this, field survey and research in Tunisia (stimulated in large part by the multinational excavations at Carthage in the 1970s) led to the identification of some of the production centres. Michael Mackensen, author of the volume under review, was the first to produce a monographic study of one of these, functioning from 320/330 until 640/660+ at El-Mahrine in NW Tunisia (Mackensen Reference Mackensen1993). Subsequent revisions to the Hayes typology and chronology were proposed by Pröttel (Reference Pröttel1996), based on material from Slovenia and the Upper Adriatic, and – more importantly – by Bonifay (Reference Bonifay2004), based on his studies in Provence, Tunisia and other parts of the Mediterranean.

When Hayes published his typology of ‘African Red Slip Ware’ in 1972, he could cover only cursorily vessels (particularly closed shapes) with applied relief decoration. The reasons for this were their evident variety (many shapes being represented by one example only) and the fact that most of the examples known to him were in museum collections and without secure find-contexts. Since then, the body of material from known and dated contexts has increased slightly, but at the same time an enormous quantity of intact vessels (devoid of archaeological context but surely derived on the whole from the robbing of graves in Tunisia) has flooded the art market. Some of these came into the possession in the 1970s of the Prähistorische Staatssammlung in Munich, where Mackensen was employed in 1981–82. Now, following further fieldwork in Tunisia and projects of chemical analysis which have enabled many of the visually distinct fabrics (and unprovenanced pieces) to be attributed to specific production sites, he has used the extensive private collection of Professor K. Wilhelm as the anchor for a new study of much of the African red slip ware that bears either stamped or applied relief decoration and which spans the period indicated in the title, from the second to the sixth century AD.

The work opens with a thorough account of studies to date on the African red slip wares (pp. 21–78), fully referenced and equipped with illustrations often drawn from earlier publications. The bibliography comprises approximately 230 entries, of which 30 list articles and monographs on the subject by Mackensen himself: the work under review represents a comprehensive synthesis of earlier studies together with a unified restatement of much that he has previously written in scattered articles. The text is supported by 1,251 references and footnotes (mercifully placed at the foot of each page!); one is left in no doubt that the author is master of his subject and acquainted with everything of any importance that has been written about it. Equally, his lucid summary of the ten most important production centres in north, central and SW Tunisia and NW Libya, describing the visual characteristics of their fabrics and the nature of their output, largely saves the reader from the necessity of seeking these details in other publications.

The second (and main) part of the text (pp. 79–248) is headed, deceptively, ‘Selected vessels and vessel-fragments of North African sigillata in the K. Wilhelm Collection’; however, as already indicated, these 240 pieces constitute the anchor for a study of a very much wider body of material. Some, but not all, of this has been previously published, but whether or not this is so, everything is illustrated to the highest possible standard, by means of both drawings and photographs to uniform scale. (It would be nice, here, to acknowledge by name those who have been responsible for the excellent drawings, photography and composition of the illustrations, but they are too numerous: the reader should pay due attention to the acknowledgements on pp. 9–11.) In this way alone, this work will constitute an invaluable source of reference when attempting to identify new pieces and to link them to what is already known.

On pp. 79–108, decorated sigillata (or red slip) vessels of the late first to early third centuries from NE Tunisia (so-called fabric A^1/2) are discussed. The precise source of these has not been found, though the unity of the production is assured by uniformity of fabric. One is immediately struck by the unexpectedly high quality of craftsmanship shown by the potters. One may have been accustomed, through comparing the undecorated wares of North Africa with their Italian and Gaulish counterparts, to thinking of the African products as somewhat heavier and less elegant; yet one can hardly fail to be impressed by the expertise which was required to transfer large and extremely delicate moulded reliefs such as the vine trails of cat. 16 and figs. 37 and 38 from the mould in which they were made to the curved surfaces of the vases to which they were applied. A new typology of 16 open and closed forms covering the ‘later’ phase in this group (last third of second century to mid third) is proposed and summarised in figs. 55–56. Many examples of the motifs used in this workshop are illustrated, but the author considers it premature to attempt a catalogue of them, since many more are likely to have existed.

Almost the entire remainder of the text (pp. 108–245), reflecting the content of the Wilhelm Collection (cat. 84–234), is taken up with the output of the central Tunisian production centre of Sidi Marzouk Tounsi (hereafter ‘SMT’). This site was first identified and reported by Peacock et al. (Reference Peacock, Bejaoui and Ben Lazreg1990, 59–84), but an intended study of the products with relief decoration by F. Bejaoui is still awaited. To a large extent, this lacuna is now filled by the work under review. The fabric of the products from SMT, which includes that named ‘El Aouja ware’ by Salomonson in 1960 and classified by Carandini et al. (Reference Carandini1981) as their C and E wares (with subdivisions), is generally identifiable without difficulty with the naked eye; in addition, many of the pieces discussed have now been chemically analysed and their attribution to this source confirmed. On the periphery of this production is the separate site of Henschir el Guellal, the products of which also travelled widely, but these too can mostly be distinguished both visually (Carandini's C/E ware) and by analysis.

First under discussion is the third-century El Aouja ware with applied relief decoration. A first typology of 28 forms was published by Salomonson in 1968. This is now enriched (but not wholly replaced) by the 24 forms illustrated in figs. 61–63. Some of these are duplicated (5, 6, 9, 11, 12 are succeeded by 18–22 and 24), but this repetition serves to distinguish the very thin walls of the main phase (C^1/2 ware) from the typically thicker walls of the late phase (C² late and C/E wares). A typology of handles is shown in fig. 60, with suggested chronological evolution, also a classification of four decorative schemes for the applied decoration (figs. 64, 65) and of 115 applied motifs/variants (figs. 64–71).

The next section (pp. 150–79) is concerned with the use of stamped motifs on plates and dishes from the fourth to the sixth centuries. Here Mackensen acknowledges that the essentials of chronology and the characterisation of different productions/sources were well established by Hayes and still call for little adjustment (except for the extension of Hayes’ Style D into the mid sixth century, rather than c. 500). Once again, the material in the Wilhelm Collection provides an enrichment of the known repertoire of stamped motifs, shown in figs. 81–83.

Applied decoration on the later wheel-thrown products of SMT (pp. 179–217), once thought to be marked by discontinuity with the earlier phase (notably by the cessation of the closed forms), has now been shown by overlap in the use of certain motifs to have evolved without a break. These motifs (drawn from both pagan mythology and Christian tradition) have by now been extensively catalogued and studied (with another 40 vessels in the Wilhelm Collection). They extend in date from perhaps 320 to 450, but as yet no attempt has been made to construct an internal chronology or – for lack of pieces from dated contexts – an absolute one.

A final section on the products of SMT (pp. 217–45) considers the large rectangular mould-made platters with extensive relief decoration both in the centre and on the rim, together with large circular platters which seem at first sight to have been wheel-thrown, but which also must have been made by the prior construction of a clay prototype and then a plaster mould for the upper surface. Associated with these are a small number of very varied closed vessels (including a hot-water bottle, cat. 231). Once again, the new fragments catalogued here enrich the available data and (through chemical analysis) confirm the origin of these pieces, but add little to the exiguous chronological evidence that can be adduced for this group (c. 350 to the mid sixth century, rather than terminating c. 430 as suggested by Hayes and others).

The first volume concludes with brief discussion of half a dozen late Roman vessels of unusual character from unidentified sources in northern Tunisia, and with four-page summaries of the entire text in German, English and French. The second volume contains the itemised catalogue of all the pieces (with extensive listing of parallels), bibliography, tables of chemical analyses, then 112 plates of drawings and 108 plates of photographs, some in black-and-white but many in colour.

In a review of Mackensen's volume on El Mahrine (Kenrick Reference Kenrick1994, 290-1), I wrote that ‘This book represents, it is to be hoped, the first of a new generation of studies of individual fine ware factories in Tunisia … One may hope that before long their output will be as clearly differentiated and understood as is now the case for the manufacture of terra sigillata in Gaul in the first three centuries AD’. It has taken a long time for a comparable study to appear; while Mackensen himself has been able to make only limited visits to the central Tunisian site of Sidi Marzouk Tounsi, what he has done here puts it on a similar footing to El Mahrine and provides for all a most useful source of reference of the very highest quality. The author has lamented to me in the past the fact that publications in German are not widely read by the archaeological community at large, and he has published substantial articles in English. One could not ask him to compose such a major work as this in any but his own language, but readers should make a strenuous effort (encouraged by the summaries) to make good use of it.