vári ‘defender’: A point of difficulty. The reading ‘vare’ is partially secured by the versification, as [v] is required by the alliterative pattern and [r] by the internal rhyme. The vowel is almost certain to be long, since of all Arnórr’s hrynhent ll. only st. 2/2 fails to begin with a long syllable, and hnika there is doubtless corrupt, so that if no emendation is made the word is vári. (a) Vári may be related to the verb verja and hence have the sense ‘defender’ (so AEW: vári 1). It appears in ÚlfrU Húsdr 2/4III, where it is probably to be construed with ragna ‘of the gods’ or ragna rein- ‘gods’ land’ to form an appellation for Heimdallr, who figures elsewhere as guardian of the gods (e.g. Grí 13, Lok 48). In the present st. vári ‘defender’ would be an apostrophe addressed to Magnús and would be qualified by virðum kunnr ‘renowned to men’. In meaning it would be similar to other heiti for ‘prince’ such as skyli ‘defender’ and vísi ‘leader’. This has been adopted here as the most satisfactory interpretation of vári, but there are alternatives, of which the following are the most plausible. (b) Konráð Gíslason suggested that the original text had a word meaning ‘son’ or ‘descendant’ which would be defined by the gen. grams ins digra hence ‘son of the stout lord [of Óláfr]’ (l. 6) (Nj 1875-8, II, 352 n.), and Björn Magnússon Ólsen (1903, 108-9) took up this suggestion, proposing that vári ‘relative’, etymologically linked with várar fem. pl. ‘faith, compact’, was the word in question; cf. also Vár, goddess of truces. Finnur Jónsson, in Skj B, postulated verja as the word meaning ‘son’. (c) Kock also proposed emending to verja, but read valkǫstr verja ‘corpse-heap of men’ (NN §816). This expression, however, would be tautologous and it has no parallel in ON, except for the still more unlikely construction proposed by Kock for Arn Magndr 11/4 (see Note below).