Tlie Clausula in AniDiiamis Marcclli)iiis. 233 



treatment of ^/— namely, that syllabication occurs also after r in the 

 words seruo and obseruo (I. 254. 8 ; 263. 11 ; II. 117. 3; 269. 23; 273. 

 23), and possibly in paruus (II. 269. 5 ; 33. 1 7). Herein, of course he again 

 contravenes poetical usage, in which the only approximation to a 

 parallel is the ante-classical lania, laruatiis, and /(T/nW/s (^Friap. 32. 13j. 

 Neither initial nor intervocalic u appear to be sul)ject to syllabi- 

 cation in his pronunciation. 1 



QUANTITY. 



For the sake of completeness mention should here be made of 

 the fact already established that Ammianus was to all intents in- 

 different to vowel-quantity but not to syllabic quantity produced by 

 position, which within certain limits is observed in his clausulae.^ 

 It has also been demonstrated previously that in his observance of 

 positional quantity syllables ending in -ns and -nt are not considered 

 long.'^ 



FORMS. 



In regard to the form of the genitive singular of nouns witli -io- 

 stems the clausula reveals nothing new in the usage of Ammianus 

 and simply confirms the testimony of the manuscript to the effect 

 that in proper names both the old form in -/ and the newer form 

 in -// are used indifferently, while in common nouns only the newer 

 form is employed.* 



In proper names the old form in -/ is demanded Ijy the clausula 

 (sometimes against the manuscript) in 



Constant! I. 67. 6; 25. 26; 38. 15; 284. 13 



Antoni I. 319. 8 



Eusebi I. 153. 22 



Theodosi II. 185. 26 



Vadomari I. 233. 27 



Danubi I. 139. 1 



The new form is demanded, either with or without manuscript 

 sanction, in 



* There are so many instances in which he does not resort to it tliat 

 we can hardly employ it to explain the two clausulae Masso Vetemeusi 

 (I. 42. 30) and saevum reperton-m (IT. 193. 8). 



» p. 187 ff. 



» p. 192 and note 3. 



* Neue, Formenl. I. 154. 



