¹ In this passage, only one unimportant MS. (β) supports trecenti. On the other hand, in 42. 7 the editor (with Luchs) reads supra septingentos, where P has supra septingenti. There only, it seems, is the full adverbial use of supra with numerals well attested in Livy. In 30.6. 9, where Luchs reads supra duo milia septingenti, P has only a numeral sign, and the nominative is supported only by βHF. It remains doubtful whether supra was used (like ad) with numerals without affecting their case. Cf., e.g., Neue-Wagener, Formenlehre der lat. Sprache, II, p. 926.