Page:Beowulf (Wyatt).djvu/97

ecg-hete, ēoweð,  ac him eal worold^[1]
wendeð on willan.  Hē þæt wyrse ne con,^[2]

XXV.

1740oð þæt him on innan  ofer-hygda dǣl
weaxeð ond wrīdað,  þonne se weard swefeð
sāwele hyrde;  bið se slǣp tō fæst,
bisgum gebunden,  bona swīðe nēah,
sē þe of flān-bogan  fyreniim scēoteð.
1745Þonne bið on hreþre  under hehn drepen
biteran strǣle;  him bebeorgan ne con
wōm^[3] wundor-bebodum  wergan gāstes;
þinceð him tō lȳtel,  þæt hē lange^[4] hēold;
gȳtsað grom-hȳdig,  nallas on gylp seleð
1750fǣtte^[5] bēagas,  ond hē þā forð-gesceaft
forgyteð ond forgȳmeð,  þæs þe him ǣr God sealde,
wuldres *Waldend,  weorð-mynda dǣl.Fol. 168b.
Hit on ende-stæf  eft gelimpeð,
þæt se līc-homa  lǣne gedrēoseð,

1. Grein ‘ne gesaca (adversary) ōhwǣr ecg-hete ēoweð (shows).’ On the whole I prefer to abide by the MS. reading, although examples are wanting of ēowan used intransitively, as its compound oðēowan frequently is.
2. 1739. The MS. has a stop after con, the usual space with the number XXV, and then a large capital O. But it seems impossible to begin a fresh sentence with oð þæt “until,” as Earle does. Grein makes the break in the middle of l. 1739, Heyne after l. 1744.
3. 1747. Heyne ‘wom’; cf. ll. 1758 and 3073. But wōm (Sievers § 295, N. 1) scans better and makes better sense. Bebeorgan takes acc. rei in 1758; but that passage alone is insufficient to settle its usual construction, and no other instance of its occurrence is known.
4. 1748. Zupitza: “to imperfectly erased between he and lange.” It is inserted in the text of all the editions.
5. 1750. MS. ‘fædde.’