216 Moore I have acted the following cases in which attributive sum fol- lows its noun: Jjaet hy on ealond sum eagum wliten (Whale 12); On his heortan cwaeS unhydig sum (Psalms LII, 1); seo hi deafe deS, dytte'5 hyre earan, J>aet heo nele gehyran heahgaldor sum (Psalms LVII, 4); & eac heora scipu sume Jmrh oferweder wurdon tobrocene; 30 & J?a burgware hie gefliemdon, & hira monig hund ofslogon, & hira scipu sumu genamon. 31 Since these parallels show that hilderinc sum is not "gegen den sprachgebrauch, " the reading of the MS should be restored. 32 SAMUEL MOORE. University of Michigan. 30 Chronicle, 794, Laud MS., ed. Plummer, p. 57. 31 Chronicle, 895, Parker MS., ed. Plummer, p. 88. 32 From the metrical point of view hilderinc sum cannot be condemned, for it is paralleled by morpor-bed stred (Beowulf 2436b) and geomorgidd wrecen (Andreas 1548a), which are rare examples of the type E2 (Sievers, Beitrage, X, 267, Bright, Anglo-Saxon Reader, p. 239). Sievers, it is true, does not recognise 3124a as belonging to this type (Beitrdge, X, 314), but regards hilde- rinca as the correct reading of the line. In his discussion of type E in the second half of the line, however, he cites 2436b and says: "Vers 2437 steht zwar fur sich allein, ist aber an sich nicht auffallig. Er stellt die regelrechte umkehr des untertypus | x - (DII) dar " (Beitrage, X, 267) . Sievers supports his emenda- tion of 3124 also on stylistic grounds ("auch stilistisch ware der vers anstos- sig, " Beitrage, IX, 144). But stylistic considerations are not a sufficient ground for discrediting a MS. reading which makes good sense and which is not in
conflict with linguistic and metrical usage.
