This is a reworking in English, with updated bibliography, of Giorgi and Buzi, Bakchias. Dall'archeologia alla storia (2014). It reviews the results of the excavations from 1993 to 2012 by teams from Bologna and Lecce, then Bologna and Sapienza (Rome), interwoven with historical interpretation. It distils data from the many interim reports, articles and monographs already published by the teams - though not the specialist reports on coins, glass, amulets and so on - so a reader interested in the details of, for instance, the large granary will need to consult the 2019 study by Tassinari. The volume is abundantly illustrated with plans and photographs (many in colour), and despite some issues of presentation, provides a very useful introduction for anglophone readers to the fascinating results of these excavations.

The mound (kom) of Umm al-Atl lies in the Fayyum, the ancient Arsinoite nome, 12 km east of the better known Karanis. It was much damaged by sebakhin into the 20th century, and now suffers from the rising water table. Its Egyptian toponym Ghenut - whence the name of its main deity Soknobkonneus, 'Sobek lord of Ghenut' - goes back to the NK (discussion of this is rather verbose), but the earliest sparse traces of settlement are of the seventh to sixth centuries BC (11, 24). After a long break in occupation (but see 62), a small settlement and temple were built in the early third century BC, soon followed by a larger village and temple, as part of the Ptolemaic development of the Fayyum. The settlement, named Bacchias (advertising the Ptolemaic link to Dionysus), was sited on the north bank of the new high-level feeder canal, to which its main streets ran parallel. The village continued to grow, and its public buildings were added to and enlarged, through the Ptolemaic to Roman periods, reaching a peak in the second century AD. Contraction is visible in the third century, and inhabitation ceased in the fourth century; the last dated text is of AD 313. Two churches of the fifth to sixth centuries were built on the 'south' kom across the canal, re-using stones from temple C, and there are later traces of activity, including fences for corralling animals and burials, on the main kom too.

The excavations focussed on the public buildings, especially the main temples. The complex of Soknobkonneus, the local version of the crocodile god of the nome, began with a small mudbrick temple (B) facing south. This was soon overshadowed by a massive mudbrick temple (A) facing east in alignment with the canal, built as part of a grander settlement doubtless related to the more ambitious irrigation scheme under Ptolemy II, to which the Zenon archive from nearby Philadelphia relates. In the first century this was replaced by a smaller south-facing temple of stone (C), which incorporated temple A into its complex. Two new south-facing mudbrick temples were built further east in the late second-century BC (D and E). Of these, temple E was rebuilt in stone in the early Roman period; it is known from a group of papyri dealing with its administration to have been dedicated to Soknobraisis, another version of Sobek. East of these temples a residential area developed in the second century BC, along with a granary (BSE 225), presumably the public granary of the documents, which by the second century AD had grown into the largest yet known from the Fayyum. South of the residential area, along the canal, was an area with kilns. Further west, south of the main temple complex, lay a public bath-house (BSO 313), which was extensively rebuilt in the earlier second century AD. It is exceptional for its lavish decoration - pebble mosaics, painted wallplaster, marble elements, glass windows (well illustrated in a reconstruction) - and because it is a sequence of communal hot and cold rooms of Roman type rather than the far more common Hellenistic pair of circular chambers for men and women with individual bath-seats. In the northern area, one two-storey residential building (BNO 360) was explored: built in the third/second-century BC and rebuilt in the second/third century AD, it produced various artefacts related to worship of Isis, and so has been named the 'house of the priestess of Isis'. It is also proposed that the adjacent BNO 356 might be a shrine of Isis. The two churches, lastly, on the south kom were both of similar build, with three aisles and wallpaintings, and they are therefore taken to be of a single monastic complex.

Historical interpretation could sometimes be more cautious. The customs 'gate' (pule) at Bacchias was an institution, not a structure (55). The desert road simply entered through a gap in the house walls (protection against sandstorms). The function of the adjacent building (BNO 354) is quite unclear, but the tower block to the south (BNO 355) might well have been a base for the village guards (cf. the Tebtunis tower complex). The occupants of the house with Isiac artefacts (BN 360) were clearly devotees, but why a priestess? The idea that it was the village 'birthing house' is fantasy (58); women used a hidden 'birthing space' in their own homes. The granary was indeed large, but the calculations of capacity are flawed (76-7); Bacchias did not supply a tenth of the annona to Rome! As regards presentation, the translation into English and proof-reading leave something to be desired. Apparently 'dieticians' accompanied consignments of tax grain (8), and two altars at Tebtunis were dedicated by 'strategists' (34 n. 45). The sense, however, is usually clear. Most frustrating is the poor labelling of the many plans, which makes it extremely difficult to locate buildings. Most plans just use the site numbers, which are hard to correlate. The general site plan (104) has no labels at all. A single plan (31) labels the key structures by name as well as site number (though without the letters A-E for the temples), but it also names hypothetical structures (e.g. 'royal road', 'temple of Amun'). These irritations, however, can be remedied in future and are outweighed by the sheer usefulness of having a summary in English of the settlement archaeology of another Fayyum village to set beside others such as Karanis, Tebtunis and Narmouthis, as part of a uniquely rich regional picture.

Dominic Rathbone