Bibliotheca Arnamagnæana - 01.10.1979, Qupperneq 174

148
above). Everything therefore suggests that JV copied the whole of his
text from the one exemplar O, whose text ofBevers saga was complete
in the late seventeenth century.
Before ending this section I must venture into the first of a series of
calculations which, although very tedious in themselves, will help us to
form a clearer picture of the original composition of O.
Figures for the ratio of O pages to S46 pages can be calculated for
approximately the first two-fifths of the Bevers text. They will provide
supporting evidence for similar calculations below involving Ivents
saga and will enable us to work out with some degree of accuracy
where in O Bevers saga began.
If we compare the citations for 0150-51 in F.d.9 (see p. 142 above)
with their counterparts in S 46 we can see that the division between
0150 and 0151 must correspond to a point in S46 between items 2
and 3. In the first calculation we will assume that this point is 86r.24,
the line in which item 2 appears in S46. We can work out a ratio by
noting the interval between this point and another word cited in F.d.9.
It will be convenient to take Palmari (item 7). This gives a reasonable
coverage of text and a terminal point for the calculation which has a
line reference to O. Palmari is quoted as being on line 37 of 0153 and
appears on 101v.30 in S46. There were probably 49 lines to the page
throughout O19, and there are 34 lines on each fully written page of
S46. The interval between 86r.24 (our starting point) and 101v.30
(item 7) in S46 is therefore 31,2 pages20, and there are 2,8 pages
between the division 0150/151 and 0153.37. 1 O page therefore =
[31,2:2,8=] 11,1 pages in S46.
In the second calculation we will assume that the division O
150/151 corresponds to 87r.l5, the line in which item 3 appears in
19 49 is the figure arrived at by Jonna Louis-Jensen on the basis of the material
relevant to Trdjumanna saga (EdAM A8, xiii); this she compares to Godel’s result of
50 lines per page (GodOS, 369). Both their findings relate to the first half of O and are
therefore not automatically true of the second half, which we are concemed with here.
In F.d.9 we do, however, find reference to line 47 on 0116, and line 37 on 0153 (item
7 in Bevers saga). It is therefore not unlikely that an average as high as 49 lines per
page was maintained throughout O.
20 Here, and in all subsequent calculations, a comma is used to indicate a decimal
point; the symbol 4- is used as a minus sign and : as a division sign.