[3]: This l. has significant variants. (a) The present edn follows the Mork version but replaces hríð ‘storm, onslaught’ (l. 3) (so Mork) with hirð (f. dat. sg.) ‘retinue’ (so A, B), which restores the internal rhyme (-irð- : ‑erð-) and retains the imagery from ll. 1-2 of Haraldr preparing a feast of carrion for the beasts of battle. It looks as though <ri> in Mork is a misreading of the original abbreviation for <ir>. The H and Hr versions (hǫrð veitk hilmi gerðu ‘I know that the leader performed harsh’ (so H); hǫrð verk hilmir gerði ‘the leader performed harsh deeds’ (so Hr)) are secondary and must represent attempts to restore the missing internal rhyme (see Louis-Jensen 1977, 153). (b) A and B provide stýrðu lit. ‘controlled, ruled, possessed’ (l. 3) for Mork’s gerðu (past inf.) ‘prepared, made’. The syntax of the resulting cl. is not clear. Faulkes (SnE 1998, I, 222, II, 404-5) suggests that stýra (stýrðu past inf.) here means ‘prepare’ although the verb is otherwise not attested in that meaning (see Fritzner: stýra; LP: stýra). Furthermore, stýra is regularly construed with a dat. object (see NS §102.e), and jól ‘yule-feast’ (l. 4) is n. acc. pl. (c) Skj B (and Skald) emends hilmi (m. dat. or acc. sg.) ‘the leader’ to hilmis (m. gen. sg.) qualifying hirð ‘retinue’ (l. 3), taken as f. acc. sg.: hykk hirð hilmis gerðu hugins jól translated as jeg ved, at kongens hird foranstaltede en fest for ravnen ‘I know that the king’s retinue prepared a feast for the raven’ (ll. 3-4). This reading entails an emendation that goes against all ms. witnesses and shifts the focus from Haraldr to his retinue. (d) Another possible reading would be hykk hirð gerðu hilmi jól Hugins ‘I believe the retinue prepared the yule-feast of Huginn (i.e. ‘carrion’) for the ruler’, but again, this interpretation destroys the otherwise consistently applied imagery of Haraldr preparing a feast for the beasts of battle.