Manilius: Astronomica

Vat. 2, regat p M. 770 regal M. 772 tempore Vat. 2 M, cedunt Vat. 2 p, caedunt M. 773 sub Vat. 2 M. 774 imperium F p Vat. 2 M. 775 positas eidem. 776 cum fratre Rhemus hanc Vat. 2 p M (l2). Lib. I 560 sex p, sed Vat. 1. 2. 561 finis Vat. 1 p M. 562 uariantia celum Vat. 1. 2. 563 flammeis Vat. 1. 2. 565 aretophilaca Vat. 1, aretophelica M, patens Vat. 1. 2 M. 612 hidram Vat. 2, idram M, sub astris Vat. 1 p. 613 ad uersus Vat. 1. 2 p M, rursus mare Vat. 1. 2 (in are urb. 2). 614 coetum p, cetum M, certum Vat. 2.

We do not know whether any other books followed the Madrid codex besides those we named above. The fact that the coats of arms of princes are painted on folio 2r shows that the Urbinates were both beautifully written and decorated with painted initials, and were once the ornament of some prince’s library. I have explained (Wochenschr. f. kl. Phil. 1904 p. 21) that they sometimes differ from each other, though rarely, and I could add several things to the examples given there to prove that the Urbinas (667) most faithfully expressed the Madrid exemplar, and that the Urbinas (668) and much more often the Vossianus 2 inserted their own corrections. Bechert (d. Man. em. r. p. 12 sqq.) shows that this book was written in 1470 by a very careless and quite inelegant hand. Moreover, I have not hesitated, in Book I, 1–82, where these verses are missing in the Madrid [codex], to insert the variant reading of both Urbinas into my notes. I have noted with the letter 'u' that both agree with each other. We know of no books that follow the Gembloux codex. Nor is it proven by sufficiently certain arguments that it originated from the same source as the Leipzig and Madrid [codices], so that it seems to constitute its own family. Although Scaliger had claimed for it the primary place among Manilian books, and Bentley had attributed such great authority to it that he nonetheless quite often preferred the readings of the Vossianus, Jacob (1846) neglected it as openly interpolated. I (1854) and Bechert (in d. Man. em. rat. and in the edition) contradicted this, and recently Housman (preface XXIII seq.) did so in a very rich discussion. Today it is agreed by everyone that the Gembloux [codex], although it often corrected ancient readings for the sake of meter or sense, nevertheless preserved genuine readings in as many places as the Madrid and Leipzig [codices], and where it differs from these books, one must carefully weigh, according to the nature of the passage, which [books] to follow.

I have arranged my variety of readings so that I would diligently note the corrections of the Leipzig [codex], mark the consensus of the four books (l g m c) with the letter 'o' (= all), and the disagreement of three or two with the letter 'ω' (= the rest); I indicated with the letters (E) or (L) where the collations of the Madrid book by Ellis and Loew differed from each other. But I took care, where the text...