^{page 162 note 1} CQ 27 (1977), 438–40.CrossRefGoogle Scholar

^{page 162 note 2} Orig. 20.12. My colleague, Dr. M. B. Walbank, warns me that Isidore can be very unreliable on technological matters. But surely anyone could distinguish a twowheeler from a four-wheeler–it involves no more in modern terms than telling a car and a bicycle apart.

^{page 162 note 3} Cod. Theod. 8.5.8.

^{page 162 note 4} AC 4.2.

^{page 163 note 1} Pan. 33.3: ‘nemo e spectatore spectaculum factus miseras voluptates unco et ignibus expiavit’’.

^{page 163 note 2} Dictionaries variously register it thus or as semiaxius or semiaxiarius.

^{page 163 note 3} This entire notice of Gellius is consecrated to the real and alleged distinctions of meaning between dimidius and dimidiatus; it is clear from Gellius and Varro that Romans used both words with some indifference.

^{page 163 note 4} ‘Juvenalian Emendations and Interpretations’’, Eranos 63 (1965), 33–5.Google Scholar

^{page 163 note 5} Referring to S. G. Owen's note on the passage in CR 11 (1897), 400–1.Google Scholar

^{page 163 note 6} Owen attributes the conjecture to Dobree, although the latter actually says he had it from adulescens quidam (very like Aulus Gellius, that!). Owen went on to conjecture and print dent lucis in his O.C.T. With taeda as the subject (cf. Madvig and Mayor on this), dat lucis would also be possible. Notice, finally, that retention of deducit arena does not dilute the Virgilian flavour, since the phrase occurs in Georg. 1.114.

^{page 164 note *} One afterthought: Lucretius 3.1017 enumerates the standard punishments thus: ‘verbera carnifices robur pix lammina taedae’’; this might enhance the claims of taeda in the Juvenal passage.