[2] meiðar †seiðs grafins† ‘the trees … [GOLD ? > MEN]’: The noun meiðar ‘trees’ appears to be the base-word of a man-kenning, but the phrase seiðs grafins, of which the most obvious translation would be ‘of the engraved/buried coalfish’, does not provide the necessary determinant. Several interpretations of the two untranslated words, which probably form a kenning for gold, have been proposed, but none of them are entirely satisfactory. The emendation eiðs gramnis ‘of the isthmus of the snake [GOLD]’ was adopted by FoGT 1884, Skj B and FoGT 2004, but gramnis is unmetrical, as a disyllabic word with a short first syllable is expected here. The emendation of grafins to gramnis was first proposed by Jón Ólafsson of Svefney (1786, 61), and has been followed by all subsequent eds except Kock (Skald and NN §2354), who retains grafins. This unattested ms. form may be a variant of the snake-heiti grafningr (cf. Þul Orma 2/3). Ms. W’s seiðs ‘of the saithe/coalfish’ could also form the base-word of a snake-kenning, but would be otiose in this sense if grafins also denotes a snake. Hence eds have emended seiðs to eiðs ‘of the isthmus’ to provide a base-word that will produce a gold-kenning following the model ‘land of the snake’. Jón Ólafsson (1786: seidr) suggested that seiðr here might mean ‘sorcery, magic’ (specifically, pulpitum, cui insidebant venefici ‘a platform on which wizards sat’), giving the gold-kenning seiðr gramnis ‘platform of the snake’, but such a usage of seiðr is unprecedented (cf. FoGT 1884, 248). Sveinbjörn Egilssson (SnE 1848-87, III, 155) saw a sword-kenning in seiðr gramnis, glossed as ignis clypei ‘fire of the shield’. Seiðr in the sense of ‘sorcery’ is not used in gold-kennings either as base-word or determinant; cf. Meissner 223-43.