Astronomicon

as a representative of α is acknowledged by G and Ven. themselves, for they much oftener side with L against one another than they side with one another against L; but it is sometimes further confirmed by the adherence, entire or partial, of M, representing α's brother β. For example: II 958 consummat LM, consumat G Ven., III 635 status M, satus L, situs G, satis Ven., IV 23 an] ad LM, aut G, id Ven., 453 tu L, tri M, tum G, tunc Ven., 489 septima fertur M, sexta fertur L, sexta feretur G, sexta refertur Ven., 524 dote M, docte L, nocte G Ven., 525 cum profert unda M, profert cum unda L, profert cum uincla G, profert quando unda Ven., 580 babiloniacas sum mersa profugit M, babilonia casum mersa profugit L, babilonias casus profugit G, babilonia casum mutauit Ven., 677 orbes LM, urbes G, orbem Ven., v 57 decuma lateris] deus malateris L, deus mala terris M, summa lateris G, om. Ven., 137 ingenita est] ingeniest M, ingeniem L, ingenium G, ingentem Ven., 487 rorantis] rotantis LM, portantis G, potantis Ven. Beside these places, where the reading of α, preserved by L, was true or close to the truth, there are others where it was false or inferior, but where it is equally possible to see, or reasonable to suspect, that L has preserved it, and that any true or superior reading presented by G or Ven. is derived either from conjecture or from some source other than α. Such are II 259 pede G, per de LM, pes Ven., 449 per Ven., fer L, fers M, fert G, IV 223 peragrant Ven., perarant LM, pererrant G, v 41 rector G, pector M, pectus L, uectus Ven., 392 senibus GM, sensibus L, sensus Ven., 403 merces est parua GM, parua merces est L, parua mercede Ven., 407 mutat GM, mittat L, mittit Ven., 525 imperat GM, imperit L, imperio Ven.

Before I proceed to G I must interpose a narrative.

In 1903 G was the sovereign MS. manuscript Bentley had called it so; and although he did not treat it as such, but made more use of V in restoring the text, his words outweighed his actions, as words generally do, and as they did again when he preached one thing and practised another in his dealings with the Blandinianus uetustissimus most ancient Blandinian manuscript of Horace. Jacob in 1846 pronounced it interpolated, degraded it to the third place, and exalted V to the first; but he gave no reasoned and ordered proof of his assertion, and succeeding critics Ellis in 1891, though placing G first, allowed much weight to V; and in 1893 both he and Breiter recognised, though not sufficiently, the value of M. one after another, from Breiter in 1854 to Bechert in 1900, condemned his judgment and upheld the primacy of G. Bechert’s devotion exceeded all precedent, and M stood low in his esteem:

‘I have used the codices in such a way that, believing one must start from the book G in establishing the text, I received the testimony of this codex into the text, presenting the evidence of the other codices only where they were more plausible. But where the reading of G seemed depraved to me, I sought out the LC books; but since even these were not sufficient, I approached the VM books, which are not to be despised, especially in the final part of the poem.’