The Clausula in Anntiiaints Marcelliints. 231 



kept in mind not onl\- that the Histories were intended to appeal 

 to a more or less cultivated ])ublic, but that the}" were actually 

 read in sections before an audience and were received, if we are 

 to believe Libanius, with great applause. Consequentl}' we should 

 be chary of assuming the existence of any conspicuous idiosyncracies 

 or Hellenisms in the pronunciation of Ammianus ; the}- would not 

 have escaped attention and censure at the first reading. ^ Further- 

 more, there is some little outside evidence to show that this phen- 

 omenon was not entirely foreign to the Latin ear. 



Lachmann long ago pointed out that in the earl}' poets the word 

 aqua is in several instances trisyllabic (aqua).^ This is certainly the 

 case in Lucretius VL 552 and 1072, and again in VL 868 if we 

 follow Lachmann in substituting aquae for laticis on the authority 

 of an unknown grammarian.'^ The same scansion is found in Ennius 

 in one place {Ann. 1 68 Vahlen), and perhaps in another [Ann. 379j.* 

 Lachmann, seconded by Bergk, maintained its occurrence in the 

 drama also ; this Ritschl stoutly denied,^ and the Plautine scholars 

 have acquiesced in his view. However it may be with this matter, 

 which in the absence of any new evidence it would be out of place 

 to discuss, it is certain that u after q in Plautus was not entirely 

 without effect upon the metre, for Lindsay notes that words like 

 loqui resist ' iambic ' shortening of the ultima. He explains this by 

 saying" — "so to the ear of Plautus qu almost made a preceding 

 vowel long by position, unless ive say that loqui etc. sounded to 

 Plautus something like a trisyllable.'^ Surel}^, ceteris paribus, the 

 latter of his alternatives is preferable. 



After the time of Lucretius this license apparently is avoided by 



* Of course this argument does not apply with the same force to an 

 inconspicuous feature like the syllabication of intervocalic ?', which we 

 have hesitatingly ascribed to Ammianus' Greek origin, and wliich affects 

 so few clausulae that it may well have passed unnoticed. 



^ See his note on Lucretius VI. 552. 



* In the latter passage Lachmann's reading, vigorousl}' defended by 

 Bergk and as vigorously opposed by Ritsciil, has been adopted by 

 Bockemiiller and by Munro. Giussani retains laticis. 



* Cf. Bergk, Opusc. I. 309, 345. In Atui. 379 he would read erugit 

 for exerngit. In Ann 155 Vahlen rejects Tarcuini corpus (Servius) in 

 favor of exin Tarqiunium (Donatus). 



» Lachmann ad Liter. VI. 552; Bergk, Cpusc. I. 72; 345: Ritschl, Opine 

 II. 600, 604 ; Schroeder, Stndemund's Studien II. 20. 



* Lat. Lang. p. 87. The italics are mine. The fact that such words are 

 occasionally shortened does not invalidate the general truth of his remark. 



