Gripla - 01.01.1975, Qupperneq 73

MANIFESTATIONS OF RAGNARS SAGA LOÐBRÓKAR 69
that if, as Lönnroth suggests as a possibility, the variant texts are
based independently of each other on oral tales, then the verbal cor-
respondences between them are bound to be especially marked in
verse-passages or passages dependent on verses, or in passages of a
markedly repetetive, rhythmical or alliterative type, since such pas-
sages, as is well-known, survive longer and more easily in oral trad-
ition than ordinary prose.16 My reasons for using passages which do
contain a verse when I compare a passage from Hauksbók with one
from 147, on the other hand, will be made clear at the appropriate
moment.1'’ I have followed Mageröy’s example in counting as ‘words
in common’ all words which occupy the same or a closely correspond-
ing contextual position in the two texts compared, and which, while
basically the same, may sometimes differ from each other in case,
number, mood or tense. Also included are variant derivative forms of
the same word, and words forming an element in a compound.17 By
legible words’ in the 147 text I mean all those words which Olsen
was able to read in their entirety, and those words in which he found
enough letters discernible for it to be obvious from his text which
words are in question. The expression ‘theoretical legible total in
1824 b refers to the number of words in the relevant section of the
1824 b text which gives the same percentage of the total number of
words in that section as the actual total number of legible words gives
in relation to the estimated total number of words in the corresponding
section of 147.1S The estimated total number of words in each section
of has been arrived at by multiplying the average number of
words per legíble line in the whole of the 147 text of Ragnars saga by
the total number of lines in each section.19 Three passages have been
1,1 Mageröy, 237, 240, 252, shows that he has taken this into account, also, in
his remarks on folk-tale variants of the chain-tale type and on parallel passages in
the A and C texts of Ljósvetninga saga.
10 See pp. 72-73 below.
17 See Mageröy, 238-39.
18 The percentage of words in common in relation to the ‘theoretical legible
total in 1824 b is comparable to the percentage of words in common in relation
to the actual total number of legible words in the relevant section of 147.
19 It was found that the average number of words per legible line in the whole
°f ^he 147 text of Ragnars saga was 12.8. The first of the three passages from 147
chosen for comparison covers 26 lines of 2r and 16 lines of 2v (see Olsen, 177-78),