^{page 169 note 1} Including the new edition of Georges (1913) ‘diūutius bei ältern Dichtern’ in which we also read ‘di turnus.’

^{page 171 note 1} natricis for nūutrīicis Bacch. 434.

^{page 172 note 1} The gender may be due to the sense intended, as in the mea of Catullus 56. 7 or the haec of Petronius 24 fin.

^{page 172 note 2} Professor Lindsay, whose attention I drew meter, to this gloss, thinks it may be a contamination of two glosses ‘natricem serpentem’ and ‘explodit excludit et expellit’ or, if a single gloss perhaps a scrap from a Christian poet's hexa-meter.

^{page 173 note 1} For the form ess compare MrHodgman's, A. W. collection in Classical Quarterly I. 1907, p. 107Google Scholar.

^{page 174 note 1} Transactions of the Cambridge Philological Society, vol. 3, p. 178 (1890)Google Scholar.

^{page 174 note 2} Pompei commentum artis Donati (Gramm. Lat. V. 225 Keil) ‘si autem i non habeat ante o sed e habeat, futurum tempus in -bo mittit, exeo exibo; exiam non dicimus; soloecismus est.’

^{page 174 note 3} Handbuch, p. 537, ‘das entsprechende Futurum hat schon Tibull. I. 4, 27 einmal unter dem Zwang des Metrums gebildet.’ The ‘stress of metre’ then stopped Tibullus from using transibit in a hexameter; but Propertius (2. 11. 5) was presumably more fortunate.

^{page 175 note 1} ei for later i is well attested in this for early Latin; cf. Lindsay, Latin Language p. 456Google Scholar and Hodgman, A. W.Classical Quarterly I. p. 122Google Scholar, who says there are thirty-three instances in Plautus.

^{page 175 note 2} The nominative of Nerienem given as the name of the wife of Mars, GelliusNoct, Att. 13. 23 (22)Google Scholar with the passages there cited, including Plaut, . Truc. 515Google Scholar, is not known for certain, gellius suggesting ‘Nerio’ or ‘Neriences.’ But as the balance of authority is clearly in favour of the quantity Nĕriĕnem; or Nēriĕnem, not for the Neriēnem which a number of scholars adopt, the word does not in any case help us here.

^{page 175 note 3} Its Latin origin is doubted by some on the ground that the inflexion -ō, -ēnis Postulates an Ablaut without example in Indo-European, Sommer, op. cit. p. 360Google Scholar with the references given there. The original declension seems to have been Aniō, Aniēnis, but Anien is as old as Cato, and it is clear from the examples in the Thesaurus s.u. that both forms were in use. A later inflexion was Aniō, Aniōnis after the normal pattern of legiō, legiōnis.

^{page 176 note 1} With this would agree the Sanskrit plīhan- (plihan-), the only one of its Indo-European congeners that stands very near to it.

^{page 176 note 2} The stem of this word, as the gen. plur. iuuenum (Ennius) and the cognates iuuen-tus, iuuen-cus indicate, was originally consonantal. The singular appears to be later than the Plural, the nominative first apperaing in Catullus, 64. 58 (sommer, op. cit. p. 370)Google Scholar. The word and its relations to senex, senis have been discussed by Brugmann, Archiv 15. pp. 1 sqqGoogle Scholar. He considders that the e was preserved through its associations with senex (these are we may observe especially strong in the plural where iuuenes, earlier iuniores, and senes, earlier seniores, is a standing military distinction). But inasmuch as ῐ and a nasal is not found in Latin except where the i is original as in uinco, uincio, neither iuuenis nor iuuencus stand in need of apology.

^{page 176 note 3} Cf. Walde Lateinisches etymologiches Wörterbuch s. u. ren. He says ‘Formell ist Kreuzung mit Lien zu erwägen.’

^{page 177 note 1} It is worth noting that according to Marius Plotius Sacerdos (Gramm. Lat. V. 474. 25 Keil) there were no Latin feminine substantives that ended in -n.

^{page 178 note 1} In his treatise De Analogia.