The Clausula in Ammianus Marcellinus. 201 



Chapter ii. The Clausula ix Composition'. 



Having discussed in detail the nature of the clausulae which 

 Ammianus employs, we must now give brief consideration to the 

 manner of their use in composition. 



Each clausula terminates a group of grammatically connected 

 words, which, in conformity with ancient usage, we may call a 

 kolon. The kola vary considerably from a minimum of the two 

 words necessary to contain the clausula -to a seldom-reached maxi- 

 mum of about ten words. Knowing the number of clausulae in 

 Book XXI, we may readily estimate from this the average length of 

 a kolon ; it is three-fourths of a line of the Teubner text. Thus the 

 typical kolon of Ammianus corresponds very closely to the ideal kolon 

 of Cicero, the length of which he sets at the length of an hexameter 

 verse. ^ It is natural to be sure, that they should correspond, for the 

 length of a kolon is not entirely dependent upon individual caprice ; it 

 is conditioned by the physiological necessity of pausing for breath. 



The average length of a sentence in Book XXI is 3.8 lines ; there- 

 fore in that book there are 5 kola in the typical sentence. There, 

 however, the sentences are short; an estimate in the first six and 

 the last three books shows an average of 5 lines to the sentence. 

 Consequently we may fix the number of kola at 6 or 7. 



This must not be construed as a violation of the dictum of Cicero, 

 repeated by Quintilian, that the period should consist of 4 kola.^ 

 For the word period (periodus) and its Latin equivalents (ambitus, 

 comprehensio, continuatio, etc.) to Cicero and Quintilian of course 

 do not mean what the Germans call a Periode and what we call 

 a sentence.-^ They mean a combination of closely connected kola 

 which may be a whole sentence or only part of one. 



condition and not a theory. That it began considerably prior to his 

 time and was general is, it seems to me, clearly enough shown in the 

 attitude of the grammarians toward quantity. The little treatise De 

 Finalib2is by Servius (K IV. 449 ff.) is nothing but a guide to quantity, 

 whicli takes the ignorance of tlie public for granted. It is well worth 

 reading. See also what he says about sjdlables in his commentary on 

 Donatus (p. 423), especially the phrase plane quom'avi difficilis est depre- 

 heusio circa syllabas naturaliter longas^ idcirco primum debemus considerare 

 quern ad modum naturaliter proferantjir. There is a great deal more evid- 

 ence of tills sort to be had, and I cannot understand whj" It has been 

 ignored in the past. 



1 Or. 222. ■>■ Cic. /. .-. ; Quint. IX. 4. 125. 



^ Tor a discussion of the meaning of the word pei-iod see Blass, 

 Rkythmen der asian, ti. rom. Kunstprosa, p. 'd ff. 



