In their recent article on this subject, Rose and Sullivan suggest that we read: super scorpionem locustam, super sagittarium oculatam, super capricornum caprum et cornutam, etc. This is attractive and can, I suggest, be supported by a further argument.

I propose that locustam be regarded as a pun on the celebrated lady poisoner of the period, Locusta. According to Tacitus, she was already infamous by the year 54, and continued to be a valuable tool of Nero throughout his reign. Petronius is fond of puns, and the passage under discussion contains more than one. And the suggested play on locusta may be confirmed by Trimalchio himself, a little later on, in his lecturette on astrology: in scorpione venenarii et percussores.

Editors are divided on the interpretation of nuda. According to Butler and Barber ad loc. it denotes the absence of defensive armour. D. R. Shackleton Bailey, Propertiana, 217 understands it primarily with reference to offensive weapons. ‘Battles fought with sharpened sticks were nuda by comparison with swords and pila.’ Camps compromises: ‘nuda presumably = inermia, which may be used to indicate absence either of offensive or of defensive arms, so that probably both ideas are present here.’

It was a generally accepted tenet of ancient literary criticism that an excess of sibilants was cacophonous. To discover if and to what extent this antipathy is discernible in the actual practice of the main Latin poets, random samples of 50 lines from each were analysed. The results of this analysis are set out in Table I.

The purpose of this paper is twofold: to comment on certain remarks of E. Koestermann, and to examine briefly some passages adduced by K. Wellesley as evidence for the alleged independence of the Leidensis (hereafter L).

In a paper in CQ N.S. XV (1965), 299–322, I attempted to demonstrate by various arguments that the readings of L are not such as to support the claim that this manuscript has authority independent of M. Those arguments may be summarized as follows: that the majority of L's readings fall into a pattern of systematic normalization, that they give virtually no help in solving really deep corruptions, that L is particularly unreliable in transmitting proper names, that many readings of L show clear signs of being derived from M, that few (if any) of the good readings in L could not have been extracted from the corrupt text of M, and that other fifteenth-century manuscripts of Tacitus contain good corrections, not inferior to corrections found in L. No defender of L has yet, as far as I know, answered the arguments I put forward, and E. Koestermann has not even understood them, for he writes as follows (vol. iii of his commentary on the Annals [Heidelberg, 1967], p. 21): ‘[Goodyear] legt den Finger darauf, daß die abweichenden Lesarten in L in ihrer Diktion stärker ciceronischen Charakter tragen, demnach eher als Konjekturen zu verstehen seien. Aber dies Argument ist nicht durchschlagend, da die späteren Annalenbücher … eine gemässigtere Tendenz aufweisen und damit näher an Cicero heranrücken.’

The meanings collected by Mr. Lee (CQ 1968, 382 f.) seem very hard to extract from the Latin, neither do they seem to reflect the author's meaning. Surely the sense of the chapter is:

The Cherusci ruined themselves with a long peace … when it comes to a fight, moderation and justice are … For example, the Cherusci were once virtuous and just, but now are called idle and foolish, and the success of the Chatti who conquered them has become prudence.

Details of poetical expression have received only incidental mention in my earlier articles on Valerius Flaccus. The purpose now is to fill this gap by outlining what has struck me most forcibly about Valerius' use of language and metre. This is offered not as a final assessment, were such a thing ever possible, but rather as a supplement or epilogue to what has already been published, with the emphasis on aspects unnoticed or not elaborated by others.

The uneasiness caused by deberent loqui is reflected in the variant readings and the emendations put forward. Is there in fact a personal subject? To understand iuvenes is possible but clumsy, but if the tragic poets themselves are intended, then these great writers are strangely selfish. Petronius surely is talking about the proper language for good oratory (cf. 2. 7), which depends on a healthy literary language in general. This rules outWilamowitz's too particular and Fuchs's uglier supplements: easier than deceret would be deberet, paralleled almost exactly in Petronius’ contemporary, Seneca (de Ira 3. 3. 1). Deberent and deberemus were doubtless due to a desire to provide a personal subject for the misunderstood impersonal verb.

The desperate straits to which commentators are driven in attempting to explain inter impotentis et ualidos falso quiescas: ubi manu agitur, modestia ac probitas nomina (Puteolanus: nomine MSS.) superioris sunt are illustrated by a recent contributor to this journal (1968, 382 f.). In the decent obscurity of a review of Büchner's fourth volume of Studien zur römischen Literatur I hazarded a suggestion that has escaped notice. The crux may be removed by reading non superioris and supposing a confusion between N ═ non (4th century) and Ñ ═ nomine (9th century). In war legality is dumb, and amid the squabbles of competing tribes restraint and honesty are not the marks of the winner. The Chatti were proved wiser than the children of light, for they had fought and flourished: Chattis uictoribus fortuna in sapientiam cessit.

Shortly before the death of Augustus, Tiberius dedicated the celebrated ara numinis Augusti, thus formally enshrining the numen of Augustus within the Imperial Cult. The step was a radical one, fundamental to the whole development of the emperor's ‘divinity’. Whereas the official cult of the emperor's genius had continued a traditional Republican practice, if with significant differences, to ascribe numen to the princeps was to establish Augustus as a through whom divinity could function as an intermediary. For to pay cult to the numen Augusti was to ascribe to the human emperor the quintessential property of a god.

T. D. Barnes has recently impugned the authenticity of these verses and calls for a defence of their genuineness. Although I agree with Fergus Millar that ‘the problem of the Historia Augusta is one into which sane men refrain from entering’,2 yet I think we can at least counter Barnes's objections.

Barnes musters four arguments which he naturally calls ‘quite conclusive’. He first points out that the verses are omitted in the epitome of Dio by Xiphilinus, who is our sole source for Dio here, and claims that it is unlikely that Xiphilinus could have omitted such a ‘striking poem’. This is putting an extraordinarily high value on Xiphilinus, who is quite capable of omitting things of greater moment than five Latin verses which he would presumably have had to translate into Greek; as an epitomator of Dio, he is inferior to Zonaras, and his account of Hadrian's reign is particularly poor.

In 1883 Alexander Enmann demonstrated the existence of ‘eine verlorene Geschichte der romischen Kaiser’. Not all of his arguments or conclusions were valid, but one fundamental postulate is undeniable: Aurelius Victor in 359/60 and Eutropius a decade later independently used a common source, a lost Kaiser geschichte of relatively brief compass. This lost work (it ought now to be clear) went down to the death of Constantine in 337, and traces of it can also be discovered in other writings of the late fourth century: in Festus’ Breviarium, in Jerome's revision of Eusebius’ Chronicle, in the Epitome de Caesaribus—and in the HA. If the HA used the Kaisergeschichte, its composition postdates 337— as Otto Seeck stated plainly in 1890.