go Accidence [§ »o6 restricted to the feminine and neuter, whereas in the West. Germanic languages it became restricted to the masculine, as OE. guma, OS. gumo, OHG. gomo, man, from ·ο, beside OE. tunge, OS. tunga, OHG. zunga, tongue; OE. ēage, OS. ōga, OHG. ouga, eye, from -on. In the inflected forms the stem-endings had originally the following phases of ablaut: ace. sing, and nom. pi. -on-, loc. sing, -en-, gen. dat. sing, and ace. gen. pi. consonantal •n-, dat. and loc. pi. vocalic ·η·. These distinctions were not faithfully preserved in the historic period of any of the Indg. languages. Owing to levelling out in various directions the different stem-endings were extended to cases to which they did not originally belong. Thus in Gothic the •an- in the nom. pi. hanans from prim. Germ. *xananiz was extended to the gen. The old form is still found in ab-n-ē, of men ; aúhs-n-ē, of oxen ; man-n-ē, of men. In tuggō the ō of the nom. sing, was extended to the other cases. And similarly in OE. the -an- of the ace. sing, of guma, man; tunge, tongue, was extended to the gen, and dat., so that all three cases became the same: guman, tungan. The masc. and fern, η-stems were originally declined alike, as in Latin, Greek and Sanskrit, but already in the prehistoric period of the Germanic languages they became differentiated in some of the cases by partly generalizing one or other of the forms. Thus, as we have seen above, the nom. sing, originally ended in -o or •on in both genders, Gothic restricted -ð to the feminine and -on to the masculine, but in the West Germanic languages the reverse took place. In the feminine Gothic O.Icel. OS. and OHG. levelled out the original long vowel of the nom. sing, into the oblique cases, whereas OE. had the same forms as the masculine except in the nom. sing, tunge from prim. Germanic -on. From a morphological point of view the η-stems should