This is a substantial study, the result of many years of assembling data and writing reports on brooches from excavations, and its publication will undoubtedly delight enthusiasts not only in Britain but throughout north-western Europe. It joins the widely used volumes of R Hattatt (based mainly on his extensive private collection), M E Snape (for northern England) and J Bayley and S Butcher (for the large Richborough assemblage). These works, and very many others, are gratefully acknowledged by the author. So too is the almost mythical corpus by M R Hull of which, after many vicissitudes, only the Iron Age part has appeared (edited by the late Christopher Hawkes), though the Roman section as edited by the late Grace Simpson and by Nina Crummy is promised in the near future.

An assemblage of 15,000 items is daunting and is surely by far the largest ever assembled at least in Britain. It has allowed the author to divide the material into quite discrete and often regional groups to a degree not attempted before. Apart from the very simplest brooches, employed more or less like modern safety pins, brooches were showy items of jewellery even when not made from precious metals. This is scarcely acknowledged save in the last chapter. Although there is some consideration on how they were worn, which appertains not only to jewellery but also to provincial styles of dress, this vital aspect is not fully examined. The manufacture of brooches is briefly discussed but not illustrated.

It is rather telling that our Fellow Catherine Johns's chapter on brooches in her book on Romano-British jewellery (Johns Reference Johns1996, 147–85) is not cited, because it is the work above all others which could have placed Mackreth's great assemblage in its proper context. If it had been, the clear relationship of enamelled disc-brooches and especially the gilded brooches with glass (sometimes intaglio) settings to the gold brooch set with gems, fibula aurea cum gemmis, mentioned on the ‘Marble of Thorigny’, would have been emphasized. Some brooches were, indeed, set with hardstone intaglios and cameos. Moreover, the marked changes in brooch design in the third century, which Mackreth notes, were not simply the result of Tetrarchic regulations; they began earlier, and are matched in changes in design involving all aspects of jewellery (Henig Reference Henig1981). That is not to say that the most developed crossbow brooches in precious metal were not official insignia worn by military officers and civil servants, but I am sceptical as to whether every base-metal crossbow signified rank rather than simply being indicative of the wearer's taste. And incidentally, it would have been good to see a discussion of the possible relationship of post-Roman fibulae to their Roman models, whether crossbow, disc or pennanular.

Some of the most interesting brooches are those bearing figural designs, sometimes religious. The rare representation of Cernunnos with a stag on the fan-tail for a brooch from the temple site at Thornborough (pl 30, no. 9995) is iconographically of the greatest importance and it is only a pity that we do not have a better illustration. Horse-and-rider brooches (pl 124) can be safely identified with a local martial, rider-god, unique, it seems, in Britain, especially in the light of an example figured by Mackreth where he obligingly flourishes a spear (no. 7902). Many examples of such brooches come from temples, though whether they were deposited as votives or intended as souvenirs for pilgrims is not known. I agree with Mackreth that it is hard to ascribe the various animal types to any particular cult, but some of them could have specific meanings. For instance, there exist cameos showing hounds chasing hares which bear the legend ‘you are caught’ and these suggest a possible amatory significance for the hound-and-hare brooches figured on pl 125. Did young men buy such brooches as gifts for their girlfriends at rural fairs?

The question arises as to the extent Mackreth's work will recommend itself to the general student of Roman material culture. The text is over-detailed and sometimes turgid and does not make for easy reading especially as at times a jarring colloquial tone intrudes. But how is it possible to pass over a book which deals definitively (even if too often at inordinate length) with every type of brooch likely to be encountered on an excavation in Britain, even though in practice it is quite difficult to relate the individual brooches listed and illustrated to the sites where they were found? The many drawings in volume ii, while clear enough for general identification, are sadly lacking in finer detail and so do not allow proper appreciation of the minutiae of decoration in niello, tinning, gilding, enamel or intaglio. Such deficiencies in illustration could and should have been made good with photographs; their absence is certainly a pity with regard to the finer brooches where artistry may be of more moment than questions of typology and dating. The only photograph in the book, that of a splendid dragonesque brooch on the front cover of both volumes (compare it with the drawings of similar brooches on pl 129), makes that point very well.

Outside the small coterie of those who are entirely devoted to brooches, most of us, I fear, will continue to use Richard Hattatt's catalogues until something better comes along. That work, absolutely inevitably, will be dependent on Mackreth but at the same time will integrate this vast body of knowledge with other aspects of material culture, particularly advances in the study of Roman jewellery.