¹ About 1325 Metr. Hom. runnande (NED).

² Sievers, Angels. Gramm. ³, par. 386, anm. 2, 3.

³ Part. Pres: 1330 Arth. and Merl. arnand, 13 … K. Alis. (Laud MS.) arnyng (NED).

⁴ The form renna is a later development; the e is taken over from the causative; rhymes in Scaldic poems testify a rinna as well as a brinna.

⁵ Jordan, Handbuch der mittelenglischen Grammatik (Heidelberg, 1925), par. 36.

⁶ It may be observed again and again that certain sound changes recur at different periods in the same language group.

⁷ Jordan, op. cit., par. 34, 2.

⁸ Noreen claims that the e in brenna, renna as well as the i in brinna and rinna can be explained as an old change n:nn within the inflectional system. See his Geschichte der nordischen Sprachen ³ (1913), par. 235.—This is not quite convincing, since a Germanic verb could probably, according to its structure, be derived only from the plural form ṇvánti and not from the singular, which has an element -eu- (in ṇóti). The influence of the causatives brenna and renna is to be seen here already.

⁹ Walde-Pokorny, i, 796.

¹⁰ The same root with the vowel i in: Middle Low German triseln, Westphal. triseln “rollen, taumeln,” Dutch trillen “zittern.” Walde-Pokorny, i, 796.

¹¹ Noreen 272, Anm. 2(rn>nn); however, spenna and sperna probably unrelated.

¹² As speornan is rather late, it may be a late formation, perhaps by analogy to eornan.

¹³ One might think of a reduction dre- from a dissyllabic base dere- (compare Homer att. ; see Specht, K. Z. 59, 117). That dere- is originally disyllabic is made probable by drū- in OIcel. trúÐr, “Gaukler,” OE. trúÐ, “Possenreisser” = OIrish drūth, “Narr.”

¹⁴ Various attempts at explanation are given in Walde-Pokorny, i, 589.

¹⁵ rinnan might also be (in part), but not very probably, a development of OInd. ríṇvati, (by the side of ).

¹⁶ Concerning the forms cf. Walde-Pokorny, ii, 168; i, 138.