386 ' Mr MUNRO, on A METRICAL 



quite in accordance ; and we feel that this ought to be so. We might read for instance 



the second of the verses above quoted, thus : 



Ecce Caesar nunc triumphat qui devicit GalUas. 

 But we feel that 



Ecce Caesar nunc triumphat devicit qui Gallias 



would be inadmissible, though that would give precisely the same rhythm for the fourth foot 

 which we had in the verse before quoted from Plautus : 



Eadem haec Latine Mercator Macci Titi. 



Such progress in popular poetry had the desire of agreement between accent and ictus 

 already made. In the song of Galba's soldiers a century later, 

 Disce, miles, militare ; Galbast non Gaetulicus, 



quantity is accurately observed, but accent agrees with ictus in every place, and we feel that 

 such a rhythm as 



Disce, miles, militare ; Gaetulus non nunc adest 



would not have been tolerated. 



That the popular taste for this agreement between accent and ictus was already very 

 decided, may I think be inferred no less certainly from its ostentatious avoidance by a 

 learned writer of iambics. I allude to the tragedies of Seneca. He is most strict in his 

 observance of the regular caesura ; and this, as he always has iambi in the even places, 

 necessitates an agreement between accent and ictus in this place. If then he concluded 

 the verse with the same kind of fall as Nicomedes Caesarem, the writer must have felt that 

 he would be conforming himself to the vulgar taste, and therefore in the fifth foot of his 

 verse which, if not always, is almost always, a spondee or anapaest, he contrives that ictus 

 and accent shall be nearly always in violent opposition. The Hercules Furens thus 



opens : 



Sorer Tonantis, hoc enim solum mihi 

 Nomen relictum est, semper alienum Jovem 

 Ac templa summi vidua deserui aetheris, 

 Locumque, caelo pulsa, paelicibus dedi, etc. 



In the eleventh line we first come to an apparent exception, but only an apparent one which 

 really proves the rule. 



Passim vagantes exerunt Atlantides, 



where Atlantides is a Greek word and accentuated on the penultimate ; and we know from 

 Quintilian (and the unanimous statements of the later Grammarians confirm what he says) 

 that in the time of Seneca the Romans, when they adopted Greek words, always gave them 

 the Greek accent, though Quintilian adds that he remembers when a youth that the most 

 learned old folks pronounced such words, Atreus for instance, with the Latin accent. Indeed 

 in looking through a good deal of Seneca, I have been surprised to find how many of the 

 apparent exceptions consist of such Greek words. At v. 495 of the same play. 



Umbrae Creontis et penates Labdaci, 



