36 Colorado College Studies. 



assigned by Vahlen to B. C. 14. Suetonius, de cjramm. 11, 

 says of Cato. 'vixit ad extremam senectutem,' so that 

 'emendare parat' might be literally true if the lines were 

 genuine. Marx claims that the words need mean only 

 'emendare studet, emendationi operam dat, emendaturus 

 est,' comparing Juv. 8, 130, 'per oppida curvis unguibus 

 ire parat nummos raptura Calaeno.'' Moreover, he main- 

 tains, the author of these lines pronounces upon the whole 

 recension of Cato, implying that it was already finished, so 

 that they were not necessarily composed in the time of 

 Horace. 



Keller objects even to the sentiment of 'teste Catone' 

 that (1) Horace required no one's authority to confirm his 

 opinion of Lucilius, and (2), in view of Epp. I. 19, 39-40, 

 it is not likely that he would have ajjpealed to the author- 

 ity of any grammarian."^ This he regards as another evi- 

 dence of interpolation. 



Vs. 3. Some editors punctuate with a period after 

 'versus,' and another after 'doctissimus,' verse 8, With 

 this punctuation 'hoc' would most naturally be taken as 

 accusative after a finite verb understood. It seems better 

 to point with commas and supply such a participle as 

 'facturus,' taking 'hoc' as the ablative corresponding to 

 'quo.' 



Vs. 4 is certainly corrupt. 



(a) It is strange that 'melior' should be given as a 

 reason for 'lenius.' It must have been this difiiculty that 

 gave rise to the variant 'lenior.' Cato's moral character is 

 not at all concerned. All that is required of him is ability 

 to correct metrical errors and halting sense in Lucilius' 

 verses, defects which had probably been multiplied even 

 in his day by mistakes of the copyists. Nor does 'sub- 

 tilior' suit 'lenius,' for Lucilius' verses are 'male facti.' 



(6) There is a false quantity in 'vir.'" 



'"'non ego, nobilium scriptorum auditor ot ultor, 

 grammaticas ambire tribus et pulpita dignor.' 

 '"The Italian dialects show that the 'i' in 'vir' was once long {veir) : cp. 

 Buocheler, Lex. Hal. p. 30. 



