The pure Italian group stands out clearly-M and its five derivatives,1 URHV. Between M and URH there lies at least one now lost copy (m); for so only since they are independent of one another-can we explain the fact that they all conjoin Manilius and Serenus. R seems from its readings, so far as these are known, further removed from M than either Uor H. Between M, again, and V2 there lies at least one now lost codex (x); for, as I have pointed out elsewhere,' V was copied from a MS. containing 26 lines to each page. These facts have importance only for lines 1-82 of Book I.

The pure Belgian family consists of GL Ven. (all saec. xi), C (saec. xii), and the five fifteenth-century MSS., V1, v, c, V3, Pith.,

1 For the relation of these MSS. to M see C. Q., i. 4, p. 297, iii. I, pp. 57-9; Breiter, Praefatio, pp. vi-vii. The relation of μ to URH must remain uncertain till that rather inaccessible MS. has been examined.

of which the first three are (as I have already pointed out) probably copies of C, while the last two are, it seems likely, copied from c. The beginnings of the process of italicizing the Belgian family are to be seen already in the corrections of c. So far as can be judged, all the codices of the mixed, or Hungarian, family are descended from c+m (m standing here for one or more MSS. of the stirps Italica). This is certainly true of P, f, g, r—these four MSS. form a clear group. Another more or less defined group is F+ b, r2. Their close interrelation may be inferred from the fact that in their tituli they all (1) give the dedication to Augustus (so v however); (2) add 'poetae' after the name 'Manilii'.

There remain B, o, r, r, which there are no data for classifying. B seems to be the latest of all our MSS.-several printed editions had already appeared when it was written. o contains some curious transpositions which I have not seen reported from any other codex. But it seems unlikely that from any of the MSS. of the Hungarian family there will emerge anything of value for the text of Manilius.

1 C. Q., iii. 1, p. 599.

2 By c+m I mean that the basis is c, the admixture of m varying considerably in the different MSS.

3 See above, pp. xliv-xlv. I think it not unlikely however that P is descended from c through Pith. P and V3 both omit II. 716-17.

4F, or some MS. closely akin to it, must have been employed by Regiomontanus and by the ed. Bon., as may be seen by comparing my App. Crit. with the citations from Fin Bechert.

III

MANILIUS AND HIS EDITORS

'Wölbt sich der Himmel nicht dadroben,
Und steigen freundlich blickend

Ewige Sterne nicht herauf?'-GOETHE.

2

I have called this section Manilius and his Editors, and if there is rather more in it about the editors than about Manilius that is because the former are so numerous and of the latter so little is known. The very name of Manilius is not certain, and his generation who shall tell? Our earliest authorities for the name are Poggio1 and Francesco Barbaro, of whom the former speaks of 'M. Manilium Astromicon' (sic), the latter of 'Manilium Astronomum'. Consonant with this is the subscriptio to Book I in the Codex Matritensis, M. Manili Astronomi con* (sic) Liber Primus Explicit.' But the same MS. has elsewhere 'M. Manlii Boeni', and again 'M. Milnili'. The Codex Gemblacensis, which has lost its titulus, gives no name in its subscriptions; while the Leipsic MS. assigns the poem to Aratus ('Arati Philosophi Astronomicon Liber Primus Incipit Prelibatio').

6

But a scholar who lived somewhat later than Poggio and Barbaro is perhaps a better witness than either to the name Manilius. Politian refers to the ancient codex of Manilius which he found in Padua 5 as a MS. of 'M. Manilio astronomo e poeta antiquo'. It is no doubt an over-statement when Sabbadini says that this reference establishes the name with certainty. Yet it seems more than probable that Politian found this name

1 C. Q., 1909, p. 58; cf. C. R., xiii, p. 125.

2 L'Enfant, Poggiana, ii. 314.

The titulus of M is in a seventeenth-century hand.

4 i. e. Astronomicon. Scaliger at i. titulus, and Comment. p. 5 has misunderstood this Greek genitive.

5 See above, pp. xlii-xliii; Sabbadini, Le Scoperte, &c., p. 169. I. del Lungo's Prose Vulgare, to which Sabbadini refers, I have not been able to see. (The Bodleian has no copy.)

M. Manilius in the titulus to his MS.-at any rate he says nothing which would discountenance such a supposition.

The 'M. Manlii Boeni' of M has provoked a good deal of speculation. The name 'Boeni', in other MSS. of the Italian family, occurs as 'Boetii', and again as 'Poeni'. From the form 'Poeni' it has been conjectured that Manilius was a Carthaginian, and that this accounts for his bad Latin. The fact is that his Latin is extremely good (Kraemer1 justly remarks that no one has ever attempted to show in detail the 'Africitas' of Manilius' style), and that 'Poeni' is a bad emendation (as 'Boeti' is perhaps a good one) of 'Boeni'. A Carthaginian could not have written, to take only one example, anything so signally Roman in temper as the Exordium to Book IV.2 Bechert, in a single sentence, has truly appreciated this essentially Roman temper of Manilius: Quotienscunque enim poeta res Romanas tangit, dictio eius animum hominis uere Romani spirat.' And Manilius' temper is not merely Roman, but it is markedly anti-Carthaginian. Lanson in this connexion has aptly directed attention to IV. 112 sqq. It is a 'singularis ac nouus patriae amor', says Lanson, which can find nothing worthy of note in its native country save 'belluarum omne genus ac monstrorum'.

3

'Boetii' deserves a more civil consideration. The great Gerbert (Sylvester II), astronomer and humanist, writes to the monk Rainard for copies of 'M. Manlius de Astrologia'. 'Age ergo et te solo conscio tuis sumptibus fac ut mihi scribantur M. Manlius de astrologia, Victorinus (Victorius codd.) de Rhetorica, Demosthenes Ophthalmicum.'" In a letter," again, to Archbishop Adalberon, he speaks of 'viii uolumina Boetii de astrologia' which he had discovered. It has been conjectured that this Manlius and this Boetius of whom Gerbert speaks were one and the same person, and this person it has been sought to identify with our Manilius. This identification scholars as prudent

1 Kraemer, Ort u. Zeit der Abfassung der Astronomica des Manilius, Frankfort, 1904, p. 1o; see below, p. xcviii. Scaliger's remarks, Proleg., pp. 2-3, on the same subject are excellent and unanswered.

2 See IV. 41.

3 De Manilio, &c., p. 10.

4 De Manilio poeta eiusque ingenio, p. 9, Paris, 1887.

6 Havet, Lettres de Gerbert, p. 117, 'te solo conscio,' because astrology was a science forbidden to the vulgar in this period. See Cumont, Astr. Cat. 5, p. 85.

6 Havet, ed. cit., p. 6.

as Ellis1 and as hasty as Housman 2 seem alike inclined to accept. Others understand Sylvester to be speaking in both passages of Anicius Manlius Torquatus Severinus Boetius, the author, that is, of the famous work On the Consolation of Philosophy. Now if we had only the reference in Sylvester to 'viii uolumina Boetii de astrologia' it is quite certain that no one would ever have dreamed of finding in it a reference to our Manilius. Of our Manilius we have to-day, after all, only five uolumina; and the criticism which would identify these with the 'viii uolumina' of Sylvester would be too frankly Procrustean to be worth a serious consideration were it not for Sylvester's other reference to 'M. Manlius de Astrologia'. Sylvester, it is urged," could never have spoken of the author of the Consolation as 'Manlius': he must have said Boetius. I have elsewhere shown, as I think, conclusively that this is a false assumption. For Sylvester's pupil, Richer, uses 'Manlium' without qualification for 'Boetium'. And he so uses it in a passage in which he is actually describing a lecture of Sylvester's; and a passage again in which 'Manlius' is coupled with the very writer with whom Sylvester, in our letter, conjoins him the rhetorician Victorinus. We can hardly doubt that, in describing Sylvester's lecture, Richer says 'Manlius' just because he had often heard the lecturer so speak of Boetius. Moreover, the fourteenth-century Sorbonne Catalogue, N. LIII. 9, has, as I have also pointed out elsewhere, Anicii Manlii', without the addition 'Boetii', for the author of the Consolation.

46

How, then, are we to explain the addition 'Boeti' after the name of Manilius in the MSS. of the Italian family? It might seem that there was something to be said for the ingenious suggestion of Kraemer, that Boeti, Boeci, Poeni represent an original 'poetae clarissimi'. This description is added after the name of Manilius in several of those MSS. from which 'Boeti' is absent. 'Poetae clarissimi' was perhaps written in contracted form 'poe. cl. '—and this might have passed into 'Poeci' or the like. But against Kraemer it is to be urged that those MSS. which have this addition 'poetae clarissimi' seem all to be 1 N. M. 230. 2 Introd., pp. lxix, lxxii. 4 C. Q., 1909, p. 56.

3 See particularly Ellis, N. M., p. 230. Histt., ed. Waitz, 1877, iii. 46.

• See Excursus III in W. v. Voigt's paper (Philologus, 1899, pp. 171 sqq.), Unter welchen Gestirnen wurden Caesar Agrippa und Tiberius geboren? I had not seen this paper when I wrote in C. Q.