308 S. Porter on the Vowel Elements in Speech, 
dropped or changed to e—in a succeeding rnleshact of inflection 
or derivation; as Wort-Worter, Hand- kurtz—kiirtzen ; 
and as Bett, Ende, from old badi, andi; ina of it in Tingiah 
are bed, end, men, sell (Goth. salyan), and other cases not a few ; 
—and in the Old N orse, for instance, a takes umlaut from u in 
the néxt syllable. The umlaut is believed to have come up 
through an intervening stage of diphthongation,—badi, for ex- 
ample, becoming first bazdi and then bedi or bed: but whether 
so or not, reference to it here is pertinen nt. 
I maintain that such change is to be explained almost wholly 
by palato-lingual action. Of course it is so in the case of e or @ 
from az. As for au, labial position would not determine its re- 
duction to o or d rather than toa labial d. In the # from wi or wu, 
there is seenly an admixture of a consonantal y, such as fae 
from the rapid utterance of the extremes in close succession 
than from an attempt to effect their simultaneous utterance. 
The developed product does not necessarily take the quality of 
either of the two extremes. 
Change from a diphthong to one of its components, by dropping 
the other, is also not uncommon. Thus from A.-S. ea, we have 
shall, sharp, hard, calf, &c., as well as deaf, head, red, ee ; from 
a eo, seven, heaven, devil, &e.; and we have benefit, parish, 
venison, comparison, &e., from the old or the later French bien- 
, Veneison, comparaison. 
Another process of change is from a simple vowel to a diph- 
ng, —common] si not vag aa a by prefixing or annex 
wife, porn t, ae. ion, Se od ‘ou meres 
—. cas S. hus, esis hund, &. In ee 
The 
” so common in Sander 3 is cae ‘of a <owel 
