^{page 90 note 1} The form in Aryan (Indo-Iranian) was *penth-: the more widespread *pent- is in this paper cited alone as the IE. root, and should be understood to represent *penth- also, as occasion requires.

^{page 90 note 2} For an explanation of this meaning, see ‘The Name of the Euxine Pontus’, C.Q. xxxiv. 2/3. The sense ‘sea’ is there derived from an earlier ‘way’.

^{page 90 note 3} This uncertain Irish cognate will be discussed later.

^{page 90 note 4} This meaning is plain from the idea of a way frequently used, much trodden on: e.g. Hom, . Il. vi. 202Google Scholar, πάτον νθρώπων λείνων, of Bellerophontes avoiding human intercourse.

^{page 90 note 5} This is the common abbreviation for ‘speakers of IE.’, which it is convenient to use.

^{pate 90 note 6} Similar reasoning from another example may be seen in K.Z. lxiv. 62–3. There Specht connects ἂγυια with ἂγομαι ‘go, be led’, and so arrives at the conclusion that in the IE. period the road was the track of cart-ruts, made by the merchants or migrating bands with carts and herds of cattle.

^{page 91 note 1} This Latin word clearly has a further semantic history of its own. There is the suggestion of Täubler, E., Terremare und Rom (mentioned by Specht with approval, K.Z. lxii. 245–6)Google Scholar, that pons ‘*way’ acquired its meaning ‘bridge’ from the way leading from the mainland to a terramara settlement: this opinion certainly agrees well with the picture of watery surroundings to which we have been led.

^{page 91 note 2} Specht, , K.Z. lxiv. 62–3Google Scholar, also cites Lat. viam secare. But I do not know of an early Latin use of this idiom, and believe that Virgil, (Aen. vi. 899)Google Scholar has the earliest example of it: that author's propensity for copying Greek idiom makes it likely that in this particular phrase we have not a native Italic metaphor. There is a possibility that Lat. secta might be compared. The etymology of this word is disputed, one may connect it with either seco or sequor (Ernout-Meillet prefers sequor). But the meaning secta ‘faction, principles’ is the predominant one, and such an early metaphorical development makes the derivation from seco unlikely, if we have regard to the date of the viam secare idiom. It is unsafe to settle the derivation of secta by appeal to the ancient conjunction of sectam and sequor, for that fact may show only that, to the ancient Roman mind, the words sequor and secta were related.

^{page 91 note 3} Since I do not see such a good reason for this special connexion of *pent- with watery surroundings in three branches of IE., if we start from the meaning ‘go’.

^{page 91 note 4} He is not merely a man who goes, a traveller in a general sense, but one who treads as he goes, who goes on foot.

^{page 91 note 5} See Becker, O., Das Bild des Weges und verwandte Vorstellungen im frühgriechischen Denken, p. 35 (Berlin, 1937)Google Scholar. The words are used especially of hill paths.

^{page 92 note 1} Schrader, , K.Z. xxx. 466Google Scholar, would derive Ir. étaim ‘I go, find’ from *pent-. But this seems unlikely, and it is preferable to derive it from *ei-, of which it is then a frequentative form: see Lewis, and Pedersen, , Concise Comp. Celtic Grammar, p. 361Google Scholar.

^{page 92 note 2} So, too, OED. s.v. find, though the descent of that verb from *pent- is not there definitely accepted.

^{page 92 note 3} Except Lith. ràsti quoted by Walde-Hofmann from Trautmann, , B.B. xxix. 308–9Google Scholar. ràsti has the sense ‘find’, and Trautmann compares cognate words with senses developed from ‘go’. But the etymology is uncertain: Brugmann, , I.F. XXX. 38Google Scholar, prefers to link ràsti with Gk. εύρίσκω and a different root.

^{page 93 note 1} This is the derivation given in Lewis, and Pedersen, , Celtic Grammar, pp. 398–9Google Scholar.

^{page 93 note 2} I appeal to others who have hunted in long grass in the outfield for lost cricket balls, who are able to appreciate both parts of this argument.

^{page 93 note 3} K.Z. xli. 198–9.

^{page 94 note 1} K.Z. XXX. 466.

^{page 94 note 2} Hesych. δροπα· δρεπτά. Σοϕοκλς Παλαμήσῃ.

^{page 94 note 3} K.Z. lxvi. 26–7.

^{page 95 note 1} See a similar metaphor introducing σίκη in Hesiod, , Works, 225–6:Google Scholar

οῖ δ δίκας ξείνοισι κα νδήμοισι διδοσιν

ἰθείας κα μή τι παρεκβαίνουσι δικαίου…

^{page 95 note 2} It is a good thing, for your own side, to do to the enemy in war. πάτη is used of a stratagem in war, Thuc. ii. 39.

^{page 95 note 3} Indeed, his zeal throughout is somewhat exhausting and occasionally unnecessary: e.g. his exposition (pp. 44–9) of the simple fact that in Homeric similes a way (όδός κτλ.) is just a place where things happen, and the idea of it does not include the intention of the people, who made and used the way, to get somewhere by using it. And this he precedes by promising to deal with the subject kurz.

^{page 95 note 4} The earliest uses of πορία ἂπορος are metaphorical: this may help to account for the comparatively late appearance of the words in Greek (not before Pindar?).