¹ Whitney, Roots s.v. śam ‘labor’, cites šamnīṣe, -īṣva, and -īthās for the Brāhmaṇas; Böhtlingk-Roth, s.v. 3 śam ‘Jmd ein Leid zufügen’, cite only šamnīṣe, -īthās, both attested in the same passage, Kāṭhaka 10.7.

² To my knowledge, the only scholar to doubt that támnō replaced a -nā-present is Kuiper, Nasalpr. 68 with note 4 (1937), who considers támnō a development of a PIE -en-present. The root of this verb is clearly tema-, as indicated by témakhos. A -nā-present of such a root is perfectly in order (cf. Meillet, Mélanges Vendryes 275–85 [1925]); and since támnō is like kámnō in formation and the latter corresponds to an Indic -nā- present, one is justified in concluding that támnō did replace a -nā-present.

³ Hirt, Idg. Gramm. 1.331 (1926), attributes the variation between skídnēmi and kídnamai to a PIE variation sk-/k-. But Debrunner, IF 45.183–85 (1927), has demonstrated that the variation in Homer is metrical.

⁴ In this and other passages where forms of krímnēmi occur, there are variants with the root form krēm-, which is from the participle krēmnós (see Specht, KZ 59.98 [1932]). Kretschmer, KZ 31.375 (1892), asserts that in all these cases one should read krim-.

⁵ In the following discussion, I shall omit zeínamen and thematic ézeinen (so Schwyzer 693; cod. ézinen-epesbénnuen, see M. Schmidt's ed., 1858). No other Greek form indicates a dissyllabic root; and the Slavic nasal present of gas- ‘be extinguished’ (e.g. OCS ugasnetÛ, Russ. gásnut') replaced an earlier je-present; see Tedesco, Lg. 24.368–9 (1948). Omitting zeínamen and éz[e]inen does not affect the outcome of this discussion, since for the latter the same conclusions apply that are drawn for édamnon (§2.2).

⁶ Étamon and étemon preserve the ablaut variation of an original athematic root aorist, the third singular active of which was ∗(e)tema[t]; see Wackernagel, Sprachliche Untersuchungen zu Homer 14 (1916), and Specht, KZ 59.56.

⁷ Wackernagel's observation contradicts Meillet's thesis that the languages which would distinguish ∗-nē-, ∗-nō-, and ∗-nā-presents show no traces of any but the last type (Mélanges Vendryes 276). The aorists ébalon and ézele represent different root grades of an athematic root aorist; see Specht, KZ 59.98.

⁸ As the source of the -m- of dámnēmi, Schwyzer cites damáō etc. (693 note 1), but the Nachtrag to 693 qualifies this note as doubtful.

⁹ I do not go into the problems concerning the origin of krémamai (beside krímnēmi and kremánnūmi), since they bear no direct relation to the substance of this paper.

¹⁰ The treatment -ll- < ∗-ln- in bállō, zéllō (instead of -l- with lengthening of the preceding vowel) does not indicate a later formation of this nasal present. This can be explained as due to the retention of ∗-ln- (influence of the aorist) past the time when the development -l- < -ln- took place.

¹¹ Most of these scholars seem to have felt difficulties in this explanation, however, for their formulations are quite vague. Thus Meillet writes, loc.cit., ‘Quelques présents en ∗-ne/o- résultent du passage de présents en ∗-nā- au type thématique, par ex. gr. kámnō…’ Chantraine is quite circumspect: ‘Certains présents thématiques comme kámnō ou támnō peuvent s‘être substitués à de vieux présents athématiques’ (§249 Rem. II). ‘Un certain nombre de présents en -nō risquent d‘être des arrangements de formes anciennes de racines dissyllabiques: kámnō …’ (§253). Although he sets forth the standard view most clearly, Schwyzer too seems to have had reservations about it, and qualifies the examples he cites in support of the view (see note 13).

¹² This represents the standard view, according to which athematic third plural actives were formed in PIE with the endings ∗-onti/-ont in addition to ∗-enti/-ent. See, for example, Brugmann, KVG §780 (1902); Hirt, Idg. Gramm. 4.§82 (1928); Meillet, Introd.⁸ 228 (1937). From the discussion presented in this paper it follows that to explain kámnō and támnō one need not operate with PIE athematic ∗-onti/-ont.

¹³ Schwyzer cites these forms with some reservation. While he cites támnousi as definitely representing an old ∗támnonti (362), he later states, ‘… -onti vielleicht … in kámnousi …’ (663); he qualifies édamnon as ‘vielleicht alt, mit ∗-onti …’ (693); and he considers épitnon ‘unsicher … [vielleicht wie édamnon u. ä.]’ (695).

¹⁴ Schwyzer himself hints at the answer to the question posed above, by expressing reservation about the antiquity of édamnon and épitnon; see note 13.

¹⁵ I omit the parts of this passage whose mention would bring up problems which, important as they are for the evaluation of épitnon, would take us too far from the central concern of our discussion.

¹⁶ Paley (The epics of Hesiod ², 1883) takes épitnon as meaning ‘covered’ and translates ‘strewed with corn’. There is no reason to attribute to pítnēmi the meaning ‘cover’; Euripides, Electra 713 thumélai d'epítnanto khruslatoi, which Paley quotes in support of this meaning (’the altar steps were carpeted’) is translated by E. T. Vermeule (in Euripides, Tragedies, Vol. 5, ed. by Grene and Lattimore; 1959) ‘the altars spread their wings of hammered gold’.

¹⁷ Another reason for preferring to consider épitnon as depicting something done to the grain is that in the Homeric description of the reaping scene on the Shield of Achilles, after which Scut. 286–291 is modelled, three actions are also depicted, all of them having to do with the grain, none with a threshing floor: Il. 18.550–51 … éntha d‘érithoi/mōn ‘the reapers cut’; 552 drágmata d‘álla … pîpton éraze ‘some handfuls (of grain) fell to the ground’; 553 álla d'amallodetêres en elledanoîsi déonto ‘others the sheavers bound in sheaves’. I hope to show in a future paper how Scut. 291 épitnon alōn is an adaptation of Hom. pîpton éraze. I wish to express my thanks to C. Dawson of Yale University with whom I had the privilege of discussing these passages. My debt to him will be more apparent in my discussion of épitnon alōn as an adaptation of pîpton éraze.