The Clausula in Aimiiiaiius Marcellinus. 221 



Ammianus of course employs many loan-words which are in no 

 wise connected with the Greek. To a specialist a study of the 

 accentuation of these words might reveal much. I shall cite only 

 a few which have seemed to me in one way or another especially 

 interesting. 



Hannibal I. 285. 3; Hannibalem 156. 14 



Saxones II. 151. 24; 152. 12; 222. 13; Santones I. 73. 18 



Vasatae I. 73. 19 ; in Ausonius Vasates (III. 5 ; XI. 2. 4 ; XV. 

 26. 8) 



Antennacum I. 147. 32 (Andernachj 



Haemimontum II. 90. 10 



Vagabanta II. 161. 9 



Dibaltum II. 253. 29 ; 264. 16 



Alatheus II. 45. 5 



Arintheus II. 71. 22 



Dagalaifus II. 15. 5 



Frigeridus II. 248. 25 ; 249. 4 



Richomeres II. 252. 10 



Libinonem v. p. 132: uncertain 



Saphrax, Saphracis II. 264. 28 



Tamsapor I. 90. 23 ; Tamsapore I. 155. 15 ; Sapori I. 123. 23 ; but 

 Saporem 212. 9 



Some of these examples show an accent which does not accord 

 with the penultimate law and which is probably influenced by a 

 foreign system. In dismissing them from consideration we may 

 quote Sergius (K. IV. 483. 29 ff.) : in barbaricis nominibus nulla sui 

 ratione sunt, sed quali volumus sane non aspero proferamus accentu : 

 which probably means no more than that Sergius could see that 

 the penultimate law did not hold, and in the lack of any tangible 

 system could only refer his readers to good use. 



Before leaving the subject of accentuation we should note that 

 Ammianus does not give us any light on the important question of • 

 secondary accent, for as we have seen already, he avoids the use 

 of the one-word clausula, which alone could be of any assistance 

 to us in this connection.^ 



There are two considerations, however, that point to the con- 

 clusion that the secondary accent must have l^een light. In the 

 first place Ammianus is reluctant to admit two dissyllables (par\-a 

 manus) as the equivalent of a tetrasyllable ( postulerunt) in the close 

 of form III.2 



' See p. 185 ff. « V. p. 184. 



