Kluge's and Weigand's Etymological Dictionaries 375 whence the oules of Chaucer in the Sompner's Tale 22, With fteischhok or oules To ben yclawed. The NED brings this quota- tion forwards as proof in the word history of awl, but it has as little place there as the ^Elfric gloss 316 fascinula awul or the quotation from the Legend of St. Katherine 2206 Tuhen hire titles up of hire breoste. wid eawles of irne or the quotation from Owl and Nightingale 80, Thi bile is . . . scharp and hoked, Right so an owel that is crooked. These quotations belong under oul which is wrongly designated as an obsolete spelling of awl; it is the legitimate development of OE. dwel, dwul, which is just as wrongly designated as " variant" of OE. eel, al "not accounted for." It is an altogether different word. It is radically con- nected with Latin vellere whose root, according to Walde, is uel-s and its meaning is that of 'evulsor,' ^-being OE. prefix whose function is about the same as that of Latin ex. Its OHG. congener I find in ar-uuel-z-an 'evellere, eruere,' with which I think uualza 'pedica' is closely connected. If so, the radical idea of the word is ('foot) catcher.' What connection there is between OHG. aruuelzan 'evellere' and aruuelzan 'evolvere,' that is a question I expect to deal with at some later time. In the mean while I insist that OE. dwel, dwul, dwol has no stand- ing in an etymological discussion of modern English awl, which is the legitimate phonetical development of OE. al, al 'subula'; the development is on a par with that of al, eal 'ommV to modern all and smcel 'gracilis' to modern small. When E. Zupitza 1.1. tried to make out the a of OE. dwel must have been short, he overlooked the controverting fact that in a MS. of ^Elfric's Grammar preserved in the Worcester Cathedral Library, designated W r by Julius Zupitza, and assigned by him to the 12th century, the fascinula awel of the original reads fascinula owel and, significantly, is followed by uncinus hoc; the gloss is printed in Wright-Wucker 548. 20 If Zupitza Junior's assumption were true, the a of the original would have appeared as a also here. That it appeared as o is plain indica- tion of the length of the vowel. For the scribe of this MS. almost invariably has changed the long a's of his original to o, a clear proof that at the time he wrote the transition of long OE. a to Middle English o had already started. This phonetical change of OE. dwel to ME. owel is further testified to by a vocab-
ulary of the 15th century, preserved in the Trinity College
