Íslenzk tunga - 01.01.1963, Blaðsíða 58

56
BALDUR JÓNSSON
cusses the etymology of two Swedish place names, Álvsborg and Lödöse. He
maintains that the first part of Álvsborg must he the gen. of alvos (not of
*alvöse), and he rejects the theory that -öse in Lödöse is to be identified with
Norw. dial. öse n. Most likely, -öse is the result of a development from öse, the
dat. sing. of the noun ös n.
Section VI deals with the question whether ölj- comes from *alj- or *£//-.
It is pointed out that in the southwest of Varmland (Nordmarks hd and Jösse
hd), close to the borders of Norway and Sweden and practically in the center
of the elf-area, the form alvos occurs instead of alvos. In these districts tlie
opening of [æ] to [a] is quite common before ‘thick’ / and r and before supra-
dentals. The dating of this opening is unknown, but some place names occur-
ring in Old West-Norse sources indicate that the opening may he of a very
early date and have been somewhat more widespread than it is now. Sucli
names are, e. g., Aljarheimr (now Elverum) in Norway and Alfheimar, the
ancient name of the coast district between Göta iilv (Gautelfr) and Glomma
(Raumelfr). The traditional opinion, first proposed by Sophus Bugge, is that
alfar is an old genitive form of elfr, that this word was an z'-stem, and tliat alf-
is the regular un-umlauted form of this stem. But the present author finds it
more probable that these spellings arc evidence of a very old opening of [æ],
resulting in [a] or at least in a very open sound, which in West-Norse was
identified with /a/. Öljus, then, can finally be reconstructed as *alfös. The
word originates from Southeast-Norway or West-Sweden and has been trans-
ferred to Iceland by a settler or settlers some time around 900.
Section VII presents some reflections on the place name Álfsóss in the Land-
námabók, as well as on the story about the settler, Alfr inn egðski, and on the
I.andnámabók itself as a historical document.
In section VIII conclusions are summarized.