[6] Vallandi ‘France’: This is the hitherto unrecognised ms. reading (refreshed). Above the first vowel, which as refreshed might be either <a> or <i>, is a stroke sloping downwards and rightwards that appears to have been part of an original <a>. A pen stroke runs across the <lld> of valldi, descending rightwards somewhat to merge with <d>. It is distinct from the ascender of <d>, as is shown by comparison with I 35/7 veldi and I 66/8 skildi, and might be interpreted as a mark of contraction, thus ‘landi’. A similar pen stroke, reaching leftwards from the ascender of <d> so as to cross the <l> or occasionally placed above both <l> and <d>, is seen in such refreshed readings as II 13/2 ‘landi’, II 15/4 ‘Bretlands’, II 19/4 ‘landreki’ and II 31/4 ‘landher’. This stroke can usually but not always be distinguished from the stroke seen elsewhere in association with preceding <l> or <ll> which is merely a run-in to the <d>, and where no contraction is involved (e.g. I 38/10). Sometimes, however, the abbreviated form of land lacks any mark of contraction, as in the refreshed readings II 1/6 and II 2/2, and this may have been the norm in the original hand (cf. I 12/6, I 23/4, I 28/4 and I 37/4). Regardless of how the pen stroke is interpreted, then, expansion to ‑landi seems secure and there are no indications that the refreshing is other than accurate on this occasion. Earlier eds, not recognising the reading, are led into tentative emendations. Scheving is reported in Bret 1848-9, without further explanation, as conjecturing valdastar ‘mightiest’ or vildastar ‘choicest’, in reference to the eikr (f. acc. pl.). Bret 1848-9 emends to valskar ‘French’, likewise qualifying eikr. Finnur Jónsson reads villdi a (Skj A) and emends to vildjá.. (sic: left untranslated), with ellipses to show that the line was metrically deficient (Skj B, followed by Skald and Merl 2012).