This second instalment of Z.'s Budé, replacing Guillemin's set, follows the same principles as the first, for which see E. Lefèvre, CR 61 (2011), 149–50, and Z.'s spirited apologia in BMCR 2009.08.15. Again the text is newly constituted but prudent, with fresh punctuation and a cleaner apparatus. There is no further introduction; the hundred-odd pages of commentary are crisp and tuned primarily to Realien. In all this is a welcome addition to the Budé collection and, given the lack of detailed modern work on Epistles 4–6, to Plinian scholarship. The criticisms that follow range mostly from minor to minute, apart from an opening lament at typos in the text (5.6.45tibi for ibi, 5.10.2rupe for rumpe, 5.19.6peregrinatinem, 6.10.3sino nomine).

Epistles 1–9 has fared well enough in transmission and Z. is justifiably conservative, defending the paradosis at, for example, 6.18.2piis and 6.22.4defenditur. At 5.6.15 he prefers prominulam (M) to pro modo longam (βγ), with good reason, but some comment on rhythm (inconclusive on its own) and stemmatics is needed, given that both support the latter reading: Z. nowhere informs his reader that α (represented here by M) and γ share an archetype against β, making this passage a test-case for contamination (cf. G.P. Goold, Phoenix 18 [1964], 324). At 5.6.21 he again follows M, now less persuasively, with its phantom porticum aliam: we are surely looking back at the colonnade of §§15–18 and its views (cf. 2.17.5; to the alternative solutions discussed one can add Sherwin-White's excision of aliam).Footnote ¹ At 5.8.11, by contrast, Z. opts for et hoc ipso diuersa quod maxima (γ), ‘subobscurum’ at best,Footnote ² especially with the translation ‘et que suffit à séparer leur importance’ (part of a non-sentence in the French); better, surely, quo maxima (M) ‘and different in the very respect in which they are greatest’. In 5.16.2nondum annos XIIII impleuerat he appeals to the ‘unanimity’ of the MSS and assumes that Pliny got the age wrong (an inscription shows that it should be XIII), when we need only excise a single stroke to fix it. G. Liberman (BMCR 2009.07.16, 2009.09.44), though overenthusiastic for intervention and misguided in many of his criticisms, was right that unanimity hardly rules out error in the archetype (see too M. Deufert, Hermes 136 [2008], 68–71); in any case, unanimity is a less compelling argument after 5.6, by which time VBF have all given out. The MSS are agreed too in 6.24.1quam multum interest quid a quoque fiat, given by Z. with the translation ‘Qu'il est important de connaître les actes dont chacun est l'auteur!’; but the letter and common sense clearly demand the opposite, ‘what difference it makes to an action who performs it!’. G. Carlsson, Zur Textkritik der Pliniusbriefe (1922), p. 70, found defence against emenders in Quint. Inst. 1.8.1 quid quoque flexu … dicendum (‘with what inflection each phrase should be delivered’), the Latinity of which stands or falls with Pliny's here (quo quidque Spalding).

In orthography too the MSS merit a little less credence (here, as in many respects, Z. keeps company with Schuster's Teubner): which is more likely, that P. wrote for example 5.17.6adolescentes but 6.20.7abeuntis (both acc. pl.), or that spelling has been inconsistently ‘improved’ in transmission? The same goes for the gen. sg. of nouns in -ium/-ius (e.g. 5.8.6studii), where rhythm demands -i often, -ii never (S.E. Stout, Scribe and Critic at Work in Pliny's Letters [1954], pp. 138–9). Conversely, the indexes in B, an interesting peculiarity, deserve more consistent reporting in the apparatus: compare for example 4.27 ‘Pompeio F : om. codd.’ with 5.2 ‘Calpurnio F (B in indice) : om. MBγ’, where the indexes have ‘Adpompei•falconem’ and ‘AdcalpurN’ flaccum•' (as reported by F.E. Robbins, CP 5 [1910], 477–8); and why not mention that in 4.3 B has hadrianio in the text, ‘adrianum’ in the index?

The translation, supplied for this second volume by N. Méthy, is flexible, lively and precise, and Pliny's adventurous turns of phrase are generally well captured; tutoiement is now the rule. Just a few quibbles: 4.14.6et sane is not ‘et pourtant’ but ‘and indeed’ (as 1.14.9, 2.19.6, etc.), adding another reason for not sending just selected highlights; 5.2.2 ‘une simple lettre, sans plus’ misses the pun in steriles; 5.3.7mihi modestior constantia est means ‘my resolve is too restrained’ (a pointed blend of determination and diffidence), not ‘j'ai un force de jugement trop limitée’; 5.6.20contra: not ‘en face de’ but ‘aligned with’ or simply ‘at’ (cf. 2.17.5, 21, OLD contra 12b); 5.6.22 ‘un platane tout proche’: surely ‘le platane le plus proche’; 5.6.40argutior ≠ ‘trop minutieux’; 5.6.41 ‘les choses’: better, ‘ces choses’; 5.17.3commendabat ≠ ‘soulignait’; 5.19.9quantum sufficiat eunti in tua: ‘une somme suffisante pour arriver chez toi’ is improbably bland; better a sententious parting compliment, ‘as much as necessary for a guest of yours’ (‘vox animi amico fidentis’, as Gierig saw); 6.16.19 ‘jeunes esclaves’: seruolis is affective, no marker of age (cf. 2.17.22, 3.16.8); 6.19.5quoque sint plura uenalia efficiunt: ‘et le résultat est que le prix de ce qui est à vendre monte’ renders the old conjecture pluris, adopted by Guillemin but not by Z.; senators are reinvesting in Italy ‘and causing a rise in the number of properties for sale’; 6.33.11 ‘mon Discours pour Ctésiphon’ misses (as do most) the qualification ut inter meas (OLD ut 22), ‘an On the crown, so far as any of my speeches could be’, and ignores iterum dicam (cf. §1 ut inter meas pulchram). Finally, a glance back at Z.'s fine translation in vol. 1 to note that 2.1.3reseruatus modifies Verginius, not Nerva, and that 2.13.8daret is not present tense.

The commentary is likewise accurate and well judged, with a conservative leaning. Sherwin-White's commentary is a constant resource and punchbag (‘quoique dise Sherwin-White’ becomes something of a refrain), complemented by selective bibliographical aggiornamento. A few specifics: 4.21.1 ‘Helvidius Priscus le Jeune’ should read ‘Helvidius le Jeune’ (the cognomen is not attested); 5.1.3–4 ‘assez sèche’ seems arbitrary for the tone of §5 ut uoles: uoles enim quod aequissimum; 5.6.31 Sherwin-White was right on porticus ante medium diem hiberna, inclinato die aestiua (warm a.m., cool p.m.: cf. 2.17.10); 5.6.33 (on cupressis ambitur et tegitur) ‘le premier verbe suggère plutôt des cyprès laissés à l’état naturel' seems fanciful; 5.8.9–10 ‘illa désigne l'art oratoire, haec l'histoire’: the wrong way round, I think, and in any case this famous crux demands bibliography (see now A.J. Woodman, From Poetry to History [2012], p. 234); 5.19.6 Zosimus' longa peregrinatio cannot have been ‘un simple aller-retour’: he left ante aliquot annos and returned nuper; 6.2.3 irony is not self-evident; 6.4.1 Calpurnia's miscarriage in 8.10–11 is scant reason to speculate ‘qu'il s'agisse, dans le cas présent aussi, d'une affection de manière gynécologique’ (!); 6.11.3 scions of the old aristocracy like Fuscus and Quadratus inhabit ‘un milieu auquel il n'a lui-même pas vraiment accès’: a striking statement, but is there any substance to it?; 6.16.20 the Suetonian rumour that Z. quotes has Pliny the Elder committing assisted suicide, not being assassinated; 6.22.4ille ‘péjoratif’: not just a change of subject?; 6.32.1continentissimus: Z. writes about Stoics, but fiscal prudence is a stock Plinian virtue (e.g. 2.6.5, Pan. 3.4).

Literary comment is sparser, though Z. shows a good sense of Pliny's style (e.g. 6.22.4 on false parallelism). On intertextuality he is variable: allusivity is stated as fact in 4.15.1–2 (Cic. Lael.) and 6.31.15 (Rut. Nam. De red.), neither conclusive to my eye, while in 4.11.7nunc ad Vestam, nunc ad ceteros deos manus tendens Z. wonders optimistically about a ‘souvenir lointain’ of Sempronius Gracchus fr. 61 Malcovati quo me miser conferam? quo uortam? (one might rather think forward to Tac. An. 12.65.3 modo ad deos, modo ad ipsum tendere manus). Yet the striking similarity of 4.11.2facis enim ex senatoribus professores, ex professoribus senatores to Juv. 7.197–8 si fortuna uolet, fies de rhetore consul;| si uolet haec eadem, fiet de consule rhetor (with analogous commutatio) meets with determined caution (‘rien n'est moins sûr’) and 5.3.2facio non numquam uersiculos seueros parum passes without comment (cf. Mart. 1.35.1, 10.20(19).1–2 [the Pliny epigram]; also Cat. 16.8, quoted in 4.14.5). It is odd that most but not all addressees earn comment: nothing, for instance, on 4.2 Attius Clemens or 4.8 Maturus Arrianus (the latter is noticed, now as Arrianus Maturus, in the notes on 6.2). Last and least in a very clean volume: ‘III, 13, 2’ > ‘II, 13, 2’ (p. 129), ‘II, 15’ > ‘II, 11, 15’ (p. 135), ‘Aug<o>ustakis’ (p. 142), alia minora.