Editiones Arnamagnæanæ. Series B - 01.10.1983, Blaðsíða 86

LXXXII
3:25 velkominn A3, W 108, 1159] vel vid kominn A4, 423;
sentence lacking SAM 6.
8:34 prova natturu steininn A3, W 108, reina steininn ok hans
náttúru 1159] profa natturu steinsinz SÁM 6, 423, 272.
12:57 lands A3, W 108, 1159] meiginlands A4, SÁM 6, 423.
If the stemma on p. lxxxi is correct, the variants to
the right of the bracket can be explained most
naturally as being the original *Ab-readings. If so, we
are led to assume the existence of a lost intermediate
link (*Ax) between *Ad and the three MSS which
share the supposedly secondary readings. Such an
intermediate link might also help to explain the
distribution of the variants at 3:16, where A3 and W
108 seem to share an error: vitnu, acc. sing. of a noun
not recorded in the dictionaries: *vitna. The readings
of the other *Ac-MSS are vitund SÁM 6, 423; vissu
1159 (= B1, doubtless by chance agreement). A4’s
reading is uncertain, since it can be read either as
vitnu (= A3, W 108) or vitun, which is not found in the
dictionaries either, but which might be a scribal error
for vitrun (cf. vitran A1-2). If A4’s reading is vitun, the
distribution of the variants might be explained as
follows: *Ab, *Ac and *Ad all had this reading. It was
interpreted as vitund by the scribes of SÁM 6 and
423 (or its exemplar, see below) but as vitnu by the
scribe of *Ax. The scribe of 1159 (or its exemplar)
substituted vissu for the non-existent vitnu and in this
way brought about secondary agreement with B1.
423 and 272 share a small number of innovations
which may be taken to suggest that they derive from
a lost common exemplar (*Ay):
4:109 Fracklande A3, SÁM 6, W 108, 1159] frans 423, 272.
5:76 lid A3, SÁM 6, W 108, 1159] land 423, 272.
16:21 hertugar, jarlar, barönar A3, SÁM 6, W 108, 1159] 4-
423, 272.