The Clausula in Amniiatiiis Marcclliuus. 181 



^f 12 4 eminuit ut docebimus 4 



<y^ 6 2 hactenus super Alpibus 2 



obsidia praeter solituin 



<5i9- 1 5 fVy- lidentius absolutus est 3 5 



Y iS 5- ita ut opinati sunt 1 



(^6^ restituit ut praeceptum est 1 



(possint obseruantibus) 



(commentum excogitatum est) 



1(76^) posse quod ablatum est 1 1 



(missus est ad Hispanias) 



On the evidence of the statistics from the second collection the 

 <[uestion of the length of the first word in the clausula may be dis- 

 posed of summarily. In any type of any form the first word may 

 begin either with the accented syllable (as in nosse confiiigil) or 

 with an}' number of unaccented syllables (as in percidsus ahscessit, 

 etc.). Moreover, there is no indication of any eftbrt to make the 

 first word either balance the last or contrast with it in length. The 

 length of the first word therefore appears to be a matter of com- 

 plete indifference to Ammianus, and we may dismiss it from fiirther 

 consideration. 



For the ' typology ' of Ammianus' clausula, the first point to claim 

 our attention is that there is essential agreement between the in- 

 terior and the final clausulae.^ Both exhibit the same types, and, 

 roughly speaking, in the same general proportions. One or two 

 types, to be sure, are somewhat more common before strong pauses 

 than before weak ones, and vice versa ; but these are minor stylistic 

 features which we ma}' leave out of account for the present.^ 



The normal clausula in Ammianus is composed of two words ; 

 four-fifths of his cadences are constructed in this way. In forms I 

 and II the only possible two-word types are /:?, 7 and (\, all of 

 which occur. The 7 types {periere coiiplnrcs, propcrabat intrepidiis) 

 and the 6 types {passibiis citis, opifex callidns) are very frequent, 



1 Of course the statistics from Book XXI include the final clausulae 

 there, but they form only a small proportion of the whole, and do not 

 affect the general conclusions drawn from a comparison of the two 

 collections. 



- See p. 202. 



Trans. Conn. AcAn., Vol. XVI. 12 Oct , 1910. 



