^{page 272 note 1} O. Ir. raith might as well be derived from Celtic ratio, cognate with Latin reor, ratus.

^{page 272 note 2} I formerly explained (Am. Jr. Phil. 20, 167) comes as from com it s with an analogical e, as in index, but index, vindexand, above all, iudex may all contain -dak-, cognate with and and, were not Latin dico attested by dialect forms with -dic- (cf. especially meddix ‘magistratus’), I should be tempted to think -dico a composition form of daco; but both daco and dico are proper derivatives of DE(Y)-K- (cf. Am. Jr. Phil. 26, 395).

^{page 272 note 3} With i-sli ‘gang’ cf. Skr. ayutam ‘myriad,’ a tautological complex in which ay- means ‘gang’ and -utatm (cognate with u-tis ‘gang’) something like ‘company’, perhaps.

^{page 273 note 1} I notice in passing that O.Bulg. otu ‘von’, with the original sense of ‘weg-’ (so Miklosich, Woert. p. 228), cf. Eng. ‘away’, may be adequately explained as an o-grade derivative of ET- ‘errare, ire’.

^{page 273 note 2} It is curious how, in this example, lends itself to the interpretation of ‘dona aduenae errantis’; and in ∑ 104 Achilles, holding aloof from the fray, calls himself ‘a profit less burden of the earth’, ay, but also ‘qui arua aliena onerat’. Similarly, Hesiod's (Op. 411) may, in view of all the context, be interpreted as ‘qui opus alienum facit’.

^{page 274 note 1} With ‘errans’ we might compare ‘immense, vast, extraordinaryg’, but ‘circumerrans’ hardly justifies the developed sense of ‘immanis’; and ‘praeter-fas’ seems a good explanation of ( ‘iustus’). However, if the rough breathing of may be safely impugned, an adequate definition of is ‘herkommlich; from ’ (: Eng. from, see Skeat's Concise Etym. Diet., s.v. frame), which admits of the derivation of from ET ‘errare’: for the sense, cf. Lat. antiquos ‘bonus’, with the examples of the Thesaurus at ii. 179, 24 sq., rather than at 180, io.sq.

^{page 274 note 2} Since this was written Professor Buck has pointed out apparent evidence for tos in Greek (See Glotta, I. p. 128. But it is possible that is to be derived, not from but lather from with vowel lengthening in composition. This phenomenon, supposed to be limited to Sanskrit (cf. Brugmann, Grundriss, ii.², p. 80), is perhaps also found in Greek, e.g. in Hesychian . In explanation of -χερμος I note Italic herna ‘saxa’ (v. Walde, woert., s.v.), proand, though I have not been able to turn the entry up in Schmidt's smaller edition, the following lemma from Hesychius as cited by Liddell and Scott: ‘to throw out of a field, and so clear it for cultivation’.–Nor can we be sure that did not start from a stem newowot nor is the influence of on inconceivable.–Professor Buck's remaining evidence the Socrian adverb ‘horno’, may own its vowel color entirely to the synonym σâτεε ‘horno’

^{page 274 note 3} So far as regards the suffixes in WETOS and OTNO-S, note the syncretic veternus ‘dotage’ (:〈‘mentis〉 erratio’), from vetes-nnos. The syn cretism is more fully illustrated in (-es stem simplex in , nom. plur.): Gothic -airkns (-no- stem), both suffixes in .

^{page 274 note 4} For the sense of ‘loafs’ cf. the following excerpt (cited in The Century Dictionary) from Lowell's Biglow Papers: I once heard one German student say to another ‘Ich lauf (lofe) hier bis du wieder kehrst’, and he began to saunter up and down–in short, to ‘loaf’.

^{page 274 note 5} looks rather like a tautological compound, with a part from and a part from ‘proand, cedo’; cf. ‘butcher’, as explained in Class.Rev. 20. 253 fta. 3.

^{page 275 note 1} Is this passage, where the MS. reading is , the source of the gloss given above ? Or is in the gloss a mere lexicographical rhyme, characteristic of the Hesychian definitions? And the idea of εσμς might be approximately hit off by a derivative of ‘binds’ (cf. Eng. knot=band ‘’).

^{page 275 note 2} The only doubt arises from Osc. lilmtiiim ‘limi-turn’ (von Planta, Gramm. ii. No. 127. 29) where, however, the penultimate f, if not for e (‘limetum’, cf. likitud=liceto on the same inscription, 1. 37), may stand for e(cf. von Planta, Gr. i. § 31), with a special assimilation due to the antepenultimate it. Or is liimtum only borrowed Latin, after all ?

^{page 275 note 3} I feel great doubt about deriving semita ‘path’ from semita (:meo). Perhaps sedita (:seditio) was, under the influence of tramet-, changed to sim-ita. Or is semita from SE-M- ‘caedere’ (cf. Am Jr. Phil. 26, 183), and to be etymologically defined by ‘schneide’ (ibid. 198) ?

^{page 276 note 1} write e after the Thesaurus and, with that cowauthority, call in question the Hesychian gloss , but I do not know whether the syllable -es has long e, or long s (ss), after the current explanation of Plautine miles. For a flexion -es -etis we might appeal to pes pedis; and we shall presently see e in the confix -etum.

^{page 276 note 2} In Whitney's Eng.-Germ. Dictionary grove is defined as ‘die Baum-Allee’, and in Wessely's Dizionario Ingl.-Ital., by ‘viale d'alberi,’ but I find nothing to support these definitions in larger works.

^{page 276 note 3} The e of these words has affected bu-citum ‘cowauthority heath’, if we are to accept unchallenged the bucita of Varro, L.L. 5. 164, emended out of bucitatum; but it is not inconceivable, as Lindsay thinks {Lat.Lang. p. 335), that bu<c>etum quasi ‘cow-wood’, has, like senticetum, an adventitious <c>.

^{page 276 note 4} (on syncopein the rhythmical type see below.

^{page 277 note 1} We can restore the primitive sense to -etum in aspreta = ‘loca asperas-uias habentia.’

^{page 277 note 2} It is with se{y)- ‘to bind’ that we must connect Umbr. sevo-, Osc. sevo- ‘toto-’ (quasi ‘verbindung’); and perhaps, also, ‘thicket’=quasi .

^{page 277 note 3} Perhaps uanus and uafer are to be derived from the root wādh- ‘procedere, errare’.