[5-7]: The text as it stands in the mss is difficult. There seems to be a clear statement that Eiríkr raided Gotland, but the mss differ as to whether strandir Gotlands ‘the coasts of Gotland’ belong together. The remaining nominal elements might well form a kenning for ‘man’, ‘warrior’ or ‘ruler’ as grammatical subject, but the base-word is elusive and other elements ambiguous, especially the duplicated vala (ll. 5, 6). The problems are hardly to be solved without emendation (cf. Jón Þorkelsson 1884, 60). (a) Adopted in this edn is the solution reached by Finnur Jónsson (Hkr 1893-1901; Skj B). This involves emendation of ‘vala’ in l. 5 to Váli, while vala in l. 6 is interpreted as gen. pl. of valr, a heiti for ‘horse’. Though not attested elsewhere in skaldic poetry, the god’s name Váli occurs in eddic poetry (LP: Váli 1) and is probably ancient (McKinnell 2009a, 190-1). (b) Kock (NN §553) points out that only the elements garð(r) ‘enclosure’ and Virfils ‘of Virfill <sea-king>’ are needed for the shield-kenning, which he appears to combine with Finnur Jónsson’s Váli to form a warrior-kenning. Kennings for ‘shield’ with the base-word garðr are exemplified in Eskál Vell 14/7, 8 and 27/2, and the combination with Virfils is possible, though the name of a legendary hero would be more usual: cf. Meissner 170-2. Kock takes l. 6 as a single phrase, Gotlands Vala strandir, which he translates as gotlänningarnas stränder ‘shores of the Gotlanders’. He does not explain gen. pl. Vala, but presumably takes it as a generalised application of the ethnic name Valir, often used of the Franks. Such a phrase would, however, be unparalleled. (c) Bjarni Aðalbjarnarson in ÍF 26, while acknowledging that the helmingr can barely be interpreted, likewise starts with garð- Virfils ‘shield’, suggesting that its vǫlr ‘staff/stave’ (represented by vala in l. 5) might be ‘sword’. (d) Hkr 1991 tentatively suggests that vala in l. 6 represents an indeclinable adj., equivalent to einvala ‘choice, excellent’ and describing the Gotland coast.