292 GERMANIC LANGUAGES 



following a of the O. H. G. word, but it is derived from forms which 

 have also in Gothic the vowel a or o in the final syllable, e. g. Gothic 

 fuglos = O. H. G. fogala. 



Holtzmann's theory had been almost universally accepted when 

 towards the end of the sixties the discovery of the "European e" 1 

 put a sudden stop to it. Holtzmann as well as Grimm and, in 

 fact, philologists of their time generally were of the opinion that 

 the Germanic languages possessed originally, like the Sanskrit, only 

 three short vowels a, i, and u, and had remained, as regards their 

 vowel-system, in an earlier stage of development than, e. g. Greek 

 and Latin. But now it was shown that, e. g. the i in Goth, niman, 

 was not directly derived from the a of Sanskr. namdmi, but corre- 

 sponded more nearly to the e in Greek vefjua, and that the i in Goth. 

 itan had passed through the stage of the e, as seen in Greek ISa>, Lat. 

 edo, etc. Nothing more natural than to assume that the forms to be 

 presupposed for Goth, niman or itan had been retained in O. H. G. 

 neman and ezzan. 



The latter opinion has remained to the present day the current 

 view in Germanic philology, and hence the Gothic vowel-system is 

 generally believed to be in many respects more recent than that of 

 the other Germanic languages. For we should have to infer that, e. g. 

 Modern German words like nehmen, essen, would still exhibit earlier 

 vowels than the corresponding verbs in Gothic. A theory leading to 

 such consequences could not have attempted to rival Holtzmann's 

 view, were it not that the latter has failed to explain an important 

 exception, in which apparently i is found as a regular vowel of 

 "ablaut" before a following a. In attempting to prove, there- 

 fore, that Holtzmann's hypothesis is correct and preferable to the 

 current view, it will not suffice to show that this hypothesis is not 

 necessarily in conflict with the " European e," but we shall have to 

 account also for the exception which Holtzmann has left unexplained. 



As regards the European e, the inference on our part of course is 

 that forms like Goth, niman, itan, have passed through the stage of 

 * neman, *etan. The transition, however, from e to i has taken place 

 in the Primitive Teutonic period. Old High German itself, as well as 

 the other Old Germanic languages, presupposes forms like niman, 

 itan. This is shown by the fact that under favorable conditions, i. e. 

 without a following vowel or when followed by one of the two vowels 

 i or u, the i appears also in Old High German (e.g. nimis, nimit) . Only 

 in case the next syllable contained an a or o, or one of the diphthongs 

 ai or em, this i became e through a-Umlaut. 



Parallel to the relation of i to e is in the Germanic languages that 

 of u to o. In some respects, however, the matter is somewhat simpler 



1 As regards the discovery of the European e, compare especially F. Bechtol, Die 

 Hauptprobleme der indogermanischen Lautlehre seit Schleicher, Gottingen, 1892. 



