500 PROCEEDINGS OF THE AMERICAN ACADEMY. 



In Ussing's next paragraph there is but one sentence that calls for 

 attention : ' It is certainly unclassical to employ the subjunctive in an 

 indefinite relative clause, as 158, 5: quorum utrum ei acciderit, merenti 

 digna constitit poena.' — While the subjunctive in this use probably does 

 not occur in the classical period, yet it is found not infrequently in the 

 Elder Pliny (Frobeen, Quaestiones Plinianae, p. 33), so that, if it were 

 found in Vitruvius, the phenomenon would be no proof of late authorship. 

 But in fact, I do not believe that acciderit is a subjunctive. The truth 

 probably is that constitit comes not from consto (as Nohl takes it in his 

 Inaex), but from consisto, the perfect of which is not infrequently used 

 in a present sense. For this use, see the grammars of Kiihner (II, p. 95) 

 and Lane (§ 1607), and for numerous examples, Munro's note to Lucre- 

 tius 1, 420, where he cites Cicero's letters, the two Senecas, Vergil, Ovid, 

 and Horace. This present meaning of constitit makes acciderit allowable 

 as a future perfect. Of course, however, the really remarkable thing in 

 the sentence is the employment of utrum where there is a choice of more 

 than two things (see the context). For this use I know of no parallel, 

 early or late. 



Ussing's last observation is as follows. ' Finally we shall briefly 

 mention the position of the words. We have already noticed the incli- 

 nation to put the negation foremost in the sentence. Similarly the 

 auxiliaries, esse, posse, and velle, etc., are preferably placed before the 

 infinitive to which they belong, as 10, 10: ut possint . . . disciplinas 

 penitus habere notas ; 91, 5: qui metopas aequales volunt facere. In 

 sum, the governing verb is very often put before its object, whether a 

 word or a whole sentence.' — And he begins his summary, which imme- 

 diately follows, with this sentence: 'These features and many others 

 point to the decadence of the Latin language and to its transition to the 

 Romance tongues.' — As for this argument, I am not aware that sufficient 

 collections have ever been made regarding the position of the auxiliary 

 verbs to warrant the use of it in fixing the date of a literary work. 

 This was the reason why Sittl published nothing on the order of words 

 in his treatise on the African writers, where he says : ' Die Beobachtung 

 der Wortstellung ergibt ebenfalls viel interessantes, aber da hier iiber die 

 uichtafrikanische Literatur fast keine Beobachtungen vorliegen, wage ich 

 es vorlaufig noch nicht, unseren Provinzialen etwas zu vindizieren ' {Die 

 lokalen Verschiedenheiten, p. 135). If now we examine the case of volo 

 in Vitruvius, we find him placing it 22 times before the infinitive and 6 

 times after the infinitive. But the Rhetor ad Herennium has it 42 times 

 before and 18 times after (see Marx's Index) ; in the Bellum Africum 



