This monograph is the reworked version of P.’s doctoral dissertation and offers the first comprehensive treatment of the figure of the witch Canidia, Horace's bête noire, who is mostly known for the spine-chilling depictions in Epod. 5 and 17 and for her comic presentation in Sat. 1.8. P. focuses extensively on these poems as well as on the shorter references to Canidia in Epod. 3, Sat. 2.1 and 2.8. The poems and the passages in which Canidia appears are given in the original and in P.’s own translation into English. The central point developed is that Canidia should not be understood as an unchanging character, since Horace readapts her figure according to different literary settings. Thus, P. argues that to understand fully the function played by Canidia in each of these poems, it is necessary to contextualise her role by exploring the relevant themes and sources on which Horace might have drawn.

The volume consists of five chapters divided into further subsections, which include useful introductions and conclusions. Two main topics are addressed in Chapter 1: the possible historical figure behind Canidia and her profile as a witch. P. dismisses the identification of Canidia with Gratidia, as attested in the commentary written by Pomponius Porphyrio in the early third century ad. According to Porphyrio (Hor. Epod. 3.7–8), who probably read this in the lost second-century commentary on Horace by Helenius Acro, Canidia was a woman from Naples who worked as a producer and seller of cosmetics (unguentaria), whom Horace slanderously called ‘witch’ (venefica). Instead of discarding a relatively small piece of information, which does not seem to affect our literary appreciation of Canidia, it might have been interesting to explore the possible reasons that led Porphyrio – or maybe Acro – to include this account in his commentary some 200 years after Horace's death. P. then discusses the ambiguous Latin terminology indicating the female practitioner of magic, a question to which he has devoted an extensive and informative study in CQ 64 (2014), 745–57. Through examining several sources, P. concludes that the fluctuating Latin terminology does not enable us to categorise what we would now define as witch. Therefore, instead of resorting to this modern label, one should focus on Canidia's role and characterisation in the specific contexts in which she appears. P. begins to do so in Chapter 2, where he delves into the ironic description of Canidia in Sat. 1.8. There, the vivified statue of Priapus guarding Maecenas’ gardens – formerly a graveyard – describes Canidia's and Sagana's attempt to collect bones and magical materials at night, until the witches are scared away by the statue itself. P. discusses the description of Canidia and Horace's possible models at length, yet little attention is paid to the figure of Sagana. P. argues convincingly that the witches’ presence pollutes the gardens and that Priapus’ statue has an apotropaic function. Furthermore, P. suggests that the role of Priapus might have been inspired by Callim. Ia. 7 and 9, where we also find personified statues.

Chapter 3 deals with the gloomy description of Canidia in Epod. 5, where she buries a boy alive in order to starve him to death and extract his organs for a love philtre. P. compares Canidia to the Lilith, a third-millennium demon from Sumer, which P. sees as the possible model for Greco-Roman child-killing figures like the Lamia and the strix. Although suggestive, the hypothesis appears rather speculative given the different historical and socio-cultural settings to which these figures belong. P. then proposes that Epod. 5 should be seen as a response to Virg. Ecl. 4, expressing Horace's pessimistic thoughts on the outcome of the Civil War. The idea is interesting to say the least and, although some of the parallels traced might not always seem persuasive, it could have significant implications for our understanding of Augustan literature.

Chapter 4 is a discussion of Epod. 17, the final poem of the Epodes, written in the form of a dialogue: Horace begs Canidia for mercy but his words remain unheeded, since she keeps tormenting him, who dared to stick his nose into her magic (ut ipse nosti curiosus, cf. Epod. 17.77) and to divulge and mock her occult activities through his poetry. Here one might have expected a mention of the theme of curiosity about magic, which has a key function in a source on which P. often draws, Apuleius’ Metamorphoses, and a reference to M. Leigh, From Polypragmon to Curiosus (2013). P. highlights the irony in the words of Horace's persona aiming to subtly insult Canidia, who is characterised as an Empusa, a female demon preying on handsome youths for their blood, who can be repelled with insults (cf. Philostr. VA 4.25). P. covers similar figures in his analysis, including Apuleius’ Meroe (already discussed in Chapter 1, especially pp. 14–19), yet a comparison between Apuleius’ Pamphile and Canidia is not offered. Canidia is considered as the anti-muse of the Epodes, and her evil influence is ultimately averted – similarly to that of an Empusa – through Horace's veiled irony.

Chapter 5 offers an analysis of Canidia's brief appearance in Epod. 3.7–8, Sat. 2.1.47–8 and Sat. 2.8.94–5. P. argues that these passages and their context show how Canidia's role is malleable and that she is not always characterised as a fearsome evildoer, although the close association with Medea (cf. Epod. 3.9–14), with the venenum of a certain Albucius known for poisoning his kin (cf. Sat. 2.1.47–8) and with the deadly African snakes (cf. Sat. 2.8.94–5) might have also inspired a sense of horror in Horace's contemporaries.

The book closes with endnotes, a bibliography and an index. The endnotes provide further interesting discussions of passages, intertexts and useful examples, which assiduous readers may happen to miss; thus, this information might have been more easily accessible had footnotes been used instead. In the bibliography there are a few minor misprints (e.g. p. 200 for ‘Betz, Hanz Dieter’, read: ‘Betz, Hans Dieter’; p. 201 for ‘Burris’, read: ‘Burriss’; p. 203 for ‘Fedeli, Paulo’, read: ‘Fedeli, Paolo’; p. 204 the reference to Graf (1980) should precede that to Graf (1997); on the same page, at ll. 30–1, for ‘Allessandro Barchiesi’, read: ‘Alessandro Barchiesi’). The index allows readers to browse easily through the topics and the ancient writings discussed, yet an index locorum might have been useful to locate specific passages from the works of other authors that are often mentioned, for example Catullus, Virgil's Eclogues, Tibullus, Propertius and Apuleius’ Metamorphoses. This, however, certainly does not detract from the quality of P.’s attractively produced book.

Although P.’s discussion might have been enriched by engaging with the recent studies by M. Reif, De arte magorum (2016), pp. 123–78, and M. Labate, ‘The Night of Reason: the Esquiline and Witches in Horace’, in P. Hardie (ed.), Augustan Poetry and the Irrational (2016), pp. 74–96, this is a very welcome contribution to the scholarly debate on Horace's poetry and Canidia. None of the points I have raised is significant enough to weaken the considerable value of P.’s work, which offers an original insight not only into the literary figure of Canidia, but also into the poems in which she features and their interpretation. P. is, therefore, to be congratulated on this stimulating volume, which is also commendable for the careful signposting and the clear style that make this study accessible to a wide variety of readers, from experienced scholars of Horace to those with little or no prior background knowledge.