Accentual Cursus in Byzantine Greek Prose. 431 
words the meter sometimes requires an ictus on the syllable without 
written accent, while the accented syllable stands in arsi; this license 
is generally limited to relatively unimportant words, such as_pre- 
positions; we find x«ré scanned as an accentual trochee, om as an 
-accentual iambus, a fact which plainly suggests that these words 
were pronounced with only a slight accent, or with an accent which 
varied according to the nearness of stronger accents on other words. 
The situation in the Political Verses may be stated in general terms 
thus: the accentual rhythm is plainly based first of all on the written 
word accents, and these accents alone are sufficient to mark clearly 
the pattern of the rhythm; there are enough written accents falling 
in the metrical theses to define exactly where all the verse accents 
must lie. Some written accents are allowed to stand in arsi, but 
not all accents indifferently ; the accents so treated stand on words 
which must have had slight, if any, independent stress in spoken 
discourse, except in certain special cases. The metrical excellence 
of the verse consists in the regulation of the licenses which the 
writer assumes in order to fit his language to the metrical scheme. 
Just so in the cursus, which has likewise an accentual basis, the 
best writers may be expected to show the highest art in the limi- 
tation set on the licenses which they permit, rather than in a me- 
chanical conformity to a rigid pattern. The treatment in the cursus 
of the weakly accented words such as the article may be illustrated 
from Zonaras, who makes about 75°/) of his clausule conform to 
the cursus forms by their written accents alone, and seems to allow 
about 10°/, plainly to transgress the law. To begin with the article: 
cvaoondy 6 Aupid. 
maoov Tov prior. 
anéoterke TH) AaBid. 
These cases are exactly similar as regards the cursus; neither 7 
nor z#y has an independent accent any more than o in the first 
example. Different cases are: 
zivOvvEvovoens Tis x3wtov, 
Onjdot tovto tH Borel. 
Here if the accents of rj and 7 are counted as before, we have 
in both cases a clausula of the 4 form; on the other hand, by 
reading them with stress we have a clausula of the 2 form. What 
is to be done? We are not bound to consistency in counting out 
the accented forms of the article; the usage of the Political Verses 
makes so much clear. But as to whether the cases in question 
should be read as conforming to form 2 or form 4 is for the present 
Trans. Conn. Acan., Vol. XIV. 29 Apri, 1910. 
