Astronomicon

with L. In the few places where M agrees with G in the older form and L presents the later, it is a still more certain inference that G is preserving, and L has altered, the spelling of α; for L sometimes modernises, giving affectus and similes where M has adfectus and similis, but G never gives the old forms where M has the new. Therefore at v 38 in ignis G, unguis M, in ignes L, the spelling of G comes from α; and so too at III 269, 663, v 98 inp- GM, imp- L.

The adverb idcirco therefore appears in 12 places. In all of them it is so spelt by M, and in 11 by L; and when G in 8 of the 11 offers iccirco we set its testimony aside. When therefore at III 525 L gives iccirco and G together with M idcirco, against its usual inclination, we infer that G and not L is here true to α.

The nominative or vocative of Iouis Jupiter occurs 10 times. G and L agree once in Iuppiter and once in Iupiter. In 4 verses L has pp and G p; and M (except once where it is quite corrupt) agrees with L. In 4 other verses L has p and G pp, and in all of them M agrees with G, though in one it has ippiter. It is to be concluded that in all 8 verses α had the true spelling and that G's pp is thence derived.

Even outside the department of orthography there is at least one place where the agreement of M with G in a true reading conducts us to the text of α. I 520:

nec motus puncto curuat cursusque fatigat.
nor does the motion curve with a point and weary the courses.

puncto curuat M, ponto curuat G, ponto currat L, ponti curat Ven. G's curuat is no conjecture, for it needs puncto to make it intelligible: it is the reading of α, diversely corrupted in L and Ven.; for I already assume, what I shall presently prove, that Ven. also is independently derived from α.

Mr Thielscher himself allows, or rather did allow in 1907 (Philol. LXVI pp. 114 sq.), that some of G's capitula chapter headings must have come not from L but from the archetype. Most of L's capitula are of later date than G, and some of G's differ from L's and agree with M's. The two most noteworthy examples are II 297 "concerning the measure of them, sign-wise and part-wise" GM, "concerning the measure of triangles and squares into parts" L (not only at the verse but in the list prefixed to book II), and IV 866 "that the reason of the fates can be perceived" GM, "how we may be able to perceive the reason of the fates" L. In 1926 (LXXXII pp. 174 sq.) he preferred to hope that these capitula may have been on the first leaf of L, now lost, and that G went there to find them.Bechert de Man. em. rat. p. 10 n. 9 had more reasonably suggested that this leaf contained a list of the capitula of book I like that prefixed to book II.