Editiones Arnamagnæanæ. Series B - 01.10.1968, Blaðsíða 46

XLIV
could be explained as scribal inaccuracies. When
the B text is taken into consideration, however,
it appears that the relationship is probably different,
for 457 and B share a nurnber of variants from A:
8 ok hans kona sett A, settur 457, B.
8 þeim A, þann 457, B.
11 bygðu A, biuggu 457, B.
25 þer ok mer A, mier og þier 457, B.
36 i 4, á 457, B.
42 hui A, þui 457, B.
78 skrapa A, skafa 457, B.
78 moður A, +þinnar 457, B.
82 syst A, leist 457, B.
83 etið A, +af 457, B.
85 þessur A, þesse iij 457, B.
87 vij underlined A, þriá 457, B.
89-90 J^-foður A, +457, B.
It is impossible for 457 to be a copy of B, and the
text of the A redaction that was used by the com-
piler of the B text cannot have been 457, for A and B
agree in a number of instances against 457:
42 spurði AB, sagde 457.
46 suarar AB, sagde 457.
66 aptr AB, +annann týma 457.
72 þá sá hann AB, +457.
84 honiim 4B, +457.
87 verða AB, vera 457.
112 allar iartegnir AB, allskinz Jiardteikn 457.
117 oflangt AB, +edur of miked 457.
Although there is little material to build on and
the possibility of chance agreements cannot be ex-
cluded, it seems likely that a stemma was as follows:
A
x
457 B
The possibility that x does not derive from A
but from a sister MS cannot be excluded.