206 Austin Morris Harmon, 



are also cadences in plent}- which I have not noted, which do not 

 precede pauses but occur in the midst of a grammatical construc- 

 tion.^ Tliese cadences certainl}- cannot be intentional, for the}' con- 

 tradict the whole theory of the clausula, which requires that the 

 clausula should terminate a word-group and not break it. They 

 ought, therefore, theoretically speaking, to have been eliminated ; 

 they occur probably because Ammianus' choice and arrangement 

 of words was so determined b}' rhythm at the end of the clause 

 that ' the numbers came ' in spite of him where they were not in 

 place. To the same cause we must also assign prolonged clausulae 

 like pueriiiac ianpus extremum and aetdtem ingrcssiis adi'tltam, and 

 interlocked clausulae like insidiis fallcrctur occi'tltis and celsiore 

 fidi'icia repugnarent. The interlocked cadences are rather rare ; the 

 prolonged ' dactylic ' rhythms are numerous. 



There is great danger of error in ascribing to the clausula any 

 profound influence upon the manner of Ammianus. Inquiry would 

 better be directed into the influence of Ammianus' manner upon 

 the clausula. But for such an inquiry there is as yet too little 

 material to give any satisfactor\- results. 



INFERENCES BASED UPON THE CLAUSULA. 



Chapter III. Matters Of Proxun'Ciatiox. 



While the pronunciation of Ammianus in the main conforms to 

 that which we regard as classical, the clausula reveals many parti- 

 culars that are important. They are for the most part phenomena 

 either of accentuation or of syllabication, and we shall discuss them 

 under these two captions. The illustrative material cited is drawn 

 from the entire text of the Histories, although, in the case of 

 phenomena which occur frequently, no attempt has been made to 

 register every instance. 



ACCENT. 

 In the construction of Ammianus" cadences phrase or sentence 

 accent, as distinguished from word accent, plays very little part. 

 I find no instances in which this factor produces a shift of word- 



^ The reader will find examples in almost any page of Book XXI. 



