[1-4]: The main problem in the first helmingr is that the verb var (normalised vas) ‘was’, the reading of all mss, makes sense in ll. 1-2 but appears to leave acc. pl. sínar hendr ‘his hands’ unaccounted for. (a) The interpretation adopted here agrees with those early eds who retained var/vas and took sverja hendr sínar á skríni together as ‘swear (with) his hands (placed) on the shrine’ (e.g. Munch and Unger in Fsk 1847, 101-2, 194; Sveinbjörn Egilsson in LP (1860): sverja and SHI 6, 48). This rather strained assumption was taken up by C20th eds of Hkr (ÍF 28, Hkr 1991) and Fsk (ÍF 29) and by Gade (2000, 113). Bjarni Aðalbjarnarson, however, improves the solution by translating as batt hendur sínar með eiðum ‘bound (over) his hands with oaths’ and noting that this is not dissimilar from constructions like sverja e-m land (‘make over land to sby on oath’; cf. a similar rendering in Hkr 1991). Given that hendr ‘hands’ can refer to the whole person, as in Hǫgni varði hendr Gunnars, lit. ‘Hǫgni defended Gunnarr’s hands’ in Akv 19 (NK 243), there could be a double sense here, literal and figurative: Sveinn is placing his hands on the shrine while swearing himself in allegiance to Magnús. (b) Emendation of var to bar provides the necessary transitive verb, and some corruption would be understandable given that the presumed cl. bar sínar hendr is suspended over ll. 1 and 4. This yields: [[ALT]]Prose order: Mǫgr Ulfs bar sjalfr hendr sínar at skríni austr við Elfi ok hét fǫgru; Sveinn réð at sverja þar. Translation: Úlfr’s kinsman [= Sveinn] himself placed his hands on the shrine east at the Götaälv, and promised fine things; Sveinn did swear there. The emendation was suggested by Konráð Gíslason (Nj 1875-8, II, 298 anm.), and was adopted by Finnur Jónsson in his eds of Hkr (1893-1901) and Fsk (1902-3) and in Skj B. As well as being an emendation against the whole paradosis, however, this leaves sverja awkwardly intransitive, whereas the verb is normally followed by a noun object such as eið(a) ‘oath(s)’ or a cl. such as (þess), at ... ‘that ...’. (c) Konráð dismissed an alternative emendation, of hét in l. 2 to helt, hence helt hendr sínar at fǫgru skríni ‘held his hands over the beautiful shrine’, for the good reason that helt (inf. halda) would normally govern the dat., not the acc.