Accentual Cursus in Byzantine Greek Prose. 425 
rhythm is present with writing in which such a rhythm can not 
exist.! Such a test must be made with absolutely nothing taken 
for granted; every clausula must be counted and classified, not 
merely as regular or irregular, but according to its form: 0, 1, 2, 
3, 4, 5, 6. Further, no deviation must be admitted from the 
traditional written accents of the grammarians. This will certainly 
introduce a slight error, but this error can be corrected with the 
help of what appears certain after a preliminary attempt has made 
general principles clear. Every scholar can satisfy himself, but no 
one else, as to just what monosyllables can stand in arsi. Whose 
understanding of the matter can be confidently adopted when all 
disagree? It is far more satisfying and convincing to stick to the 
grammatical accents in marking out the general lines which the 
cursus follows. Everything is thus made definite and open, and the 
influence of subjective feeling completely eliminated. If the written 
accents alone can be found to distinguish cursus writers from others, 
then a vast deal is gained, and we may start from a plain scientific 
fact to work out details of usage. In the following count, therefore, 
every written accent is counted of equal value, whether it stand 
on a conjunction, preposition, noun or any other word, and whether 
it be grave, acute or circumflex. The different kinds of accent must 
be separated ultimately, and the question of their relative importance 
answered. Since this is a question of detail, it may be disregarded 
for the present. Further, although the cursus law is observed be- 
fore every sense pause, we can not be sure that our commas mark 
the pauses as understood by the Greeks themselves, and must 
begin by counting only the pauses before the stronger marks of 
punctuation—the period, colon and interrogation point. The writers 
chosen for comparison on this basis are the historian Zosimus 
(5 century) and Demosthenes, the passages being chosen at 
random, and the clausulaz counted consecutively. The figures are 
striking and can best be shown on a chart (p. 426). The forms of 
clausule found with greatest frequency in Zosimus are those desig- 
nated as 2 and 4, while forms 1, 3, 5 and 6 are strikingly rare; 

* In Greek we may feel on perfectly sure ground in comparing writers 
who wrote in an accentual cursus with those who have a quantitative 
cursus, if the theory of Havet and Meyer is true as to the origin of the 
Latin accentual cursus. The case is not as it would be in Latin where 
the comparison would be between earlier and later stages of a con- 
tinuous development, for the word accents of the Greek quantitative 
clausulee should contain no suggestion of the accents of the accentua! 
cursus. 
