Bibliotheca Arnamagnæana - 01.06.1957, Page 455

SPECIMEN
435
Eyvindus, alias Eystenus. These definitely indicate the use of
two sources, one of which is no longer known. 2) AJ has one or
two additions which are not known in other Landn.-redactions,
see III 18330 Isafiørd; 21820 prædium Hørdubol; 24031-33 sinus
Watna-fiørd ripam orientalem; quum Occidentalem ejusdem
sinus ripam Isthmus VVatz-nes supra memoratus constituat. (To
this in Landn. corresponds only “Skagastrpnd”). The first addi-
tion is admittedly also found in x 11 and Skålh., but in these it is
probably derived from AJ. The second, as far as its subject-
matter is concerned, comes from Laxdæla saga, but the form of
the name is wrong. The third can, at least in part, have ori-
ginated with AJ himself, but amongst older sources only Vatns-
doela saga records the name VatnafjprSr. 3) A number of names
show such divergences from all known Landn.-redactions that
they cannot be dismissed out of hånd as due to AJ’s inaccuracy
or misreading, see e.g. III 19235 Gisserum (Gnupr), 19331 Thor-
laugå (Svanlaug), 19831 Thora (AsgerSr), 20036 Tryhirnings
(Einhyrnings), 2053 Stvvffs-hollt (Suluholt), 2061 Thorgrimi
(Kolgrimr), 21025 Gudmundus (Gunnvaldr), 21433 Ormerus
(Brandr), 21510 Kolgrimus (Lorarinn), 22214 Thormodus (Qrn-
olfr), 2469 Staffaa (Sti'fla), 25113 Thorriida (GuSrfår), 25229
Saur-vviich (SkprSuvi'k), 25331 Riinaar-os (Tjgrnes), 26311 Borg-
argierdis- (HreggsgerSis-).—Finally, AJ quotes directly from
his S-text III 24721-27 (the beginning of the quotation is not in
HSkL, which also have a different wording in the remainder).
The quotation does not agree exactly with S (see notes ad loc.) ;
most of the discrepancies can be due to AJ, but it is worthy of
note that in AJ’s text Helgi inn magri’s son is named LårSr and
the usual Hrålfr only appears as a variant (probably from Sk).
We should most naturally conclude that AJ read LorSr in his
source here, and certain of the other variants thereby gain in
significance, e.g. 24721-22 hann trude a Christ a lande (å lande
omitted S) ; 24724 hann skylide land (land skylldi S; AJ’s read-
ing = HSkL); 24724 Lå viisade friettinn (eN fretten visadi S) ;
24726 friettinn (Lor S).
It follows from what has been said above that it is extremely
probable that for his first draft of Specimen AJ used a now lost
manuscript of S and not the Resenian vellum which Bjorn å
28*