LATIN INSCRIPTION AT CIRTA. 391 



accent is in equally violent opposition to the ictus. Let us take this one other illus- 

 tration. We know from abundant testimony that deinde, perinde, prSinde, eieinde were 

 accentuated in the manner indicated. Servius among others notices this fact in his com- 

 ment on Aeneid vi, 743, 



Quisque sues patimur mattes; 6xinde per dmplum. 



The accents of exinde per dmplum exactly correspond to those of milti comitantum ; 

 and yet how different the rhythmical effect of these two endings. Lucretius again terminates 

 a verse, vi, 1017, with unde vaceftt. Now we have the most conclusive evidence that 

 vacefit, and all cognate words, tepefit etc., were accentuated on the last syllable. Yet I 

 believe that to Lucretius the movement of these words was the same as Virgil's unde La- 

 tinum: to Servius or Priscian it was doubtless otherwise. 



As the caesura is of vital importance in the hexameter, and the metrical beat of the 

 dactylic rhythm is on the first syllable of the foot, and the Latin accent is such as we 

 have described it, it is perfectly true that in general those verses will be smoothest and 

 easiest in their movement, in the first three or four feet of which ictus and accent are 

 opposed ; the most impetuous and violent those in which there is the greatest amount of 

 agreement in the first four or all the six places. In the iambic and trochaic for cognate 

 reasons we found the contrary to be the fact ; the metrical beat of the iambus falling 

 on the second syllable, and the caesura of the senarius occurring, as in the hexameter, in 

 the middle of a foot ; the metrical beat of the trochee falling on the first syllable, and 

 the caesura of the trochaic tetrameter always coming at the end of a foot. That the 

 rhythmical movement however, and not the accent, is the occasion of this, may be shewn 

 from many considerations, and also by this fact which should never be forgotten, that 

 the Latin hexameter is entirely borrowed from Homer and Homer's Greek imitators, and 

 any notion of accent having the least influence on his rhythm is belied by every line in 

 the Iliad and Odyssey. Many verses of Lucretius in which accent and ictus have exactly 

 the same relation to one another which they have in many most easy-flowing verses, are 

 more violent and unusual in their rhythmical effect than any of the verses quoted above 

 in which ictus and accent coincide throughout. Out of hundreds of examples take these: 



Et membratim vitalem deperdere sensum. 

 Quidve tripectora tergemini vis Geryonai. 

 Quidve superbia spurcitia ac petulantia? quantas. 



Take again the following; 



Sed bona magnaque pars servabat foedera caste. 



Here the accents are arranged exactly as in a verse of this kind: 



S6d veterum b6na pars servabat etc. 

 Yet how different is the metrical effect of the two. The following line of Virgil is quite 

 unexceptionable : 



Thesauros ignotum argfinti pondus et auri. 



Substitute fSrri for argenti. The accents remain identically the same ; yet instead of a 

 Vol. X. Part II. 50 



