452 H. B. Dewing, 
This procedure leaves Procopius with a much lessened percentage 
of regularity —54 + as against 63+ in the former test—but the cursus 
characteristics are still apparent in the proportions of the 2, 3 and 4 
forms. Leaving the question of elision, apheresis and crasis for the 
sake of argument undecided, there is sufficient ground for maintain- 
ing that the cursus law is operative in the prose of Procopius. It 
remains to define the extent of its application; the principles followed 
are those deduced above from the usage of Zonaras and Zosimus. 
In the following an attempt is made to enumerate all the cases 
in which the regular forms are to be recognized. It is safest to 
make a start from the formal classification by written accents alone. 
I. Perfect clausule of the 2 and 4 forms. 
In these cases the rhythm is plainly marked by written word 
accents, the written accents, namely, of the last two words, except 
in case an enclitic or proclitic intervenes; hiatus, if present, is weak. 
Form 2: 1734 (2 Us2eih2 
3 , ~ C 
EPEVETO THE. 12,8. 
\ é THe tts 
To maodnay avdvos. 225, 26: 
yodpew ove Eyw. 447, 10. 
Form 4. 310 (4 I; 4 I) 
UEETHS TEM OLNUEVOY. LSI 6: 
dooupogors 6 Me&tuivos. 506, 23-24. 
In 59 cases the thesis is marked by what would seem to be a 
weak spoken accent. 
Form 2. 59 (2 Ill) 
moopcAdc chat Ev taig EvuBoacis. 267, 6. 
élvat tovs mokeutous. 387, 23; 388, 1. (Form 4?) 
olol Te avtiteively Eiol 176, 10. 
In certain cases the force of the first accent is weakened by a 
heavy accent immediately preceding it. 
Form 2. 65 (2 IV) 
ovde yoortoi Edogey Elva. 455, 23. 
ndovns yéyovev aiodrats. 190, 15. 
ZUZAOTE QOS ovens MVOUCOT CL. 131, 15-16. 
Form 4. 18 (4 II) 
Tequava@ lnuvtées mooseyworou. 505, 13. 

+ The number of cases included under each head is given; the numbers 
in parentheses refer to the tables of forms (see Appendix). 

ie tte a Pee BE ce 
di te aa a at a lal hi a ES a - _— 
> 


