Gripla - 01.01.1995, Page 144

142
GRIPLA
three times, as do Sth. 15 (80v32) and AM 655 XIV (2rll). In another
passage Sth. 3 skips a word, so that the passage „vrdv margar j ollvm
stodvm“ (233:16-17) only makes sense in reference to „ok vrdv margar
inrteign/>“ in Sth. 2 (47rb30; Hms 302:29) as well as AM 661 (17vl5);
this reading is supported by AM 655 XIV (2v39) and AM 655 XXII
(lr5).
Conversely, readings in Sth. 3 reveal the omission of words in Sth. 2.
When Juliana asks Bishop Cyrillus that she be permitted to remove
the remains of her husband, the bishop refuses her request in Sth. 3
„þvi at hann qvezt eigi vita hvors likamvr være enns helga Stefanno
eda bonda hennar“ (234:17-18). Whereas the corresponding passage in
Sth. 2 omits mention of the husband: „hvors likami var eda ens sæla
stephani" (47val4-15; Hms 303:14-15) - as happens in AM 661
(18v3-4) - the reading in AM 655 XXII supports the variant in Sth. 3:
„[lijkami var stephanus. eða bvanda heNflr“ (lv4), as does NRA 67e:
„stefanus eða bva[nda]“. Some lines later a similar phrase occurs:
„hvor kistan er enns helga Stefane edr bonda þins“ (Sth. 3, 234:31-32).
This time Sth. 2 contains the full text: „hvflr tr kista ens helga stepha-
ni. eda bonda þins“ (47va31-32; Hms 303:29-30), but AM 661 writes:
„huor hier tr kista bonda þins“ (18v20-21).
A comparison of the manuscripts shows that the readings in Sth. 3
vacillate; they agree now with one, now with another manuscript,
which Ole Widding had also remarked upon (1952, p. 151). Now and
then Sth. 3 agrees with an original reading in Sth. 2, at other times with
a corrupt variant. For example, Sth. 3 agrees with Sth. 2 (45rall; Hms
293:30) in writing „<H>eyrit menn brædr og fedr“ (220:26). Unger re-
jected this reading in favor of „HEyrit mik fednr ok brædnr“ in AM
661 (7v6), but the corresponding passage in the Acts of the Apostles
attests that Sth. 2 and 3 transmit the original translation: „Viri fratres
et patres, audite" (Acta 7:2). Similarly, Sth. 2 (45va40-41; Hms
296:14-15) agrees with Sth. 3, where we read: „eda giorda eigi min
avnd alla þessa hlvte“ (224:2-3); AM 661 writes hond instead of avnd
(10r22), to the contrary, and reference to the Acts shows that this time
the correct reading is transmitted in the last: „Nonne manus mea fecit
haec omnia?“ (7, 50). A third example of agreement between Sth. 2 and
3 is similarly misleading. Both manuscripts write: „Hvar skvlvm vær þaa
þin leita“ (Sth. 3, 228:28-29; Sth. 2, 46val; Hms 299:14-15), but the di-