PROBLEMS IN GERMANIC PHILOLOGY 301 



If we combine these three points it follows that in the first and third 

 person singular the Old Norse preterite presents a mixture of indic- 

 ative and subjunctive endings, in such a way that in the third person 

 the two endings have become identical by regular phonetic trans- 

 formation, while in the first person the ending of the subjunctive has 

 been adopted for both moods. 



A word remains to be said about the pre-Gothic vowels supposed 

 to have been retained in Anglo-Saxon. We are concerned here with 

 a theory advanced in an important article in the fifth volume of 

 Paul and Braune's Beitrdge 1 by a well-known scholar whom we are 

 fortunate enough to have with us at this meeting. His contention 

 is that original i in final syllables, which is regularly dropped in 

 Gothic, has been retained in Anglo-Saxon after a short stem-syllable, 

 while it was lost after a long stem-syllable. For example, A. S. wini, 

 "friend" = Goth. ga-wins, and A. S. stedi, " place "=Goth. sta^s ; 

 but A. S. wyrm=Goth. waurms, and A. S. giest = Goih. gasts. 



1 have no objection to the fact that in Anglo-Saxon the appearance 

 of final i depends on the quantity of the stem-syllable. But I am not 

 convinced that the i contained in forms like wini, stedi, can be traced 

 back to the Primitive Teutonic and Indo-European periods. I rather 

 believe that we have here a new ending of recent date, due to the 

 analogy of the /a-stems. This seems apparent from the fact that the 

 final i of the above and of similar forms goes hand in hand with the 

 loss of the old i-declension and with the confusion of i- and /a-stems. 

 Not even to the Gothic language is this confusion entirely unknown; 

 but there the process is seen only in its very beginning and in the 

 inflection of adjectives rather than of nouns. Old High German is 

 less conservative than Gothic, but has kept the i- and /a-declensions 

 separate, at least in the case of nouns, and has accordingly introduced 

 forms like wini, stedi, only sparingly. Frequently, however, do forms 

 of this description appear in Old Saxon, where the old i-declension 

 is becoming extinct. In Anglo-Saxon, finally, where the last trace 

 of the i-declension has disappeared, the apparent preservation of 

 the final i has become a definite rule. 



We observe in several instances that inflectional differences which 

 at first had nothing to do with quantity are at a later date made to 

 depend on the quantity of the stem-syllable. In Gothic grammar 

 we have, e. g. the rule that feminine /a-stems with a short stem- 

 syllable (like sunja, halja) have in the nominative singular the ending 

 -/a, while /a-stems with a long stem-syllable (like bandjd) form their 

 nominative in i. Yet actually we are in this case concerned with two 



2 Ed. Sievers, Zur Accent- und Lautlehre der germamschen Sprachen. m Zum 

 vocalischen Auslautsgesetz : Beitrage zur Geschichte der deutschen Sprache und 

 Literatur, vol. v (Halle, 1878), pp. 109-163. 



