CRITICISM AND INTERPRETATION 211
of from three to nine words, and another line
of different rhythm. Of the eighty seven-stress
lines, twenty-two depend on departures from the
present text; but several consecutive seven-stress
lines^1 are discovered without any alteration of
the Hebrew text.
As a last example of Sievers' metrical analysis
I select Genesis i. on account of the peculiar
interest of the reconstruction of the text involved
in it : at the same time it is right to add that
Sievers expressly states that his analysis of this
particular chapter is one of the most uncertain
and tentative of his results. According to the
analysis the chapter contained forty-nine seven-
stress periods interrupted by one line (in v. 20)
of three stresses and by what is regarded as a
gloss of two lines in v. 16. Of the forty-nine
seven-stress periods no fewer than thirty-two
rest on textual alteration—a far larger proportion
than in either of the previous examples that have
been given here. But a large number of the
textual changes are of one type: in order to
obtain rhythmical regularity Sievers found that,
in every case where Myhlx, God, occurred, rhythm
required either one word less or one word
more: in the former case he omits Myhlx in,
the latter he prefixes hvhy, Yahweh; so that in
respect of the use of the divine names, Genesis i.
1 E.g. eight such lines occur in v. 42 (from jwy Mx) to v. 46 (to htw);
seven such lines in v. 47 (from ym-tb) to v. 51.