Page images

II 65.

11 37. cui nihil absit B, quo nihil absit ACEPV. The original must have had quoi altered by B and misunderstood by the others.

II 61. ipsa B, ipsa vis ACEV, ipsa vi V!. It would seem that the vis of the previous line had got wrongly inserted here, B alone representing a.

planius quam B, planius quem AV', planiusque V'E, pleniusque C. Quam was no doubt abbreviated in a and misread by all but B.

caelo B, melo A'CEPV, celo A”. Is B an emendation or the correct reading of a misread by the others ?

11 69. deflagravisse [CEP]V’, deagravisse A, deam migravisse V', demigravisse B (mi by corr.). Here I should suppose that A comes closest to the original, the letters ft being lost or obscured in a : V would then be a bold emendation, and the true reading conjecturally restored in P.

11 112. cujus propter laevum genu CV, c. p. laeum genum A, at propter laevum genus omni ex parte locatas parvas B', with cujus for at B', with cujus and geny E. It is plain that E follows B and that B is taken from the Aratea, see n. in loc.

11 114. Chelis B (probably corrected from Aratea), cetis AC.
11 117. sublimis sed B, sublimi sed ACPV.

11 126. alvos ibes [P] , alvos ibis CE with a, alvo sibis A, alvo sibi B, alvos hibis V. Here B is evidently an emendation.

11 131. varia et tam V°, variae tam AV, varie tam B, varia tam CEP. Probably AV represent a altered by B and the others.

11 134. molitur B with a, mollitur ACEPV*

11 145. omnisque sensus-—antecellit [APC], omnesque 8.antecellit EV', omnesque 8.antecellunt B, perhaps right.

ex quo videlicet quid (ACPV], ex quo videmus quid BE, no doubt an emendation to explain construction.

11 159. fabricarier ensem et AE (er in ras. A) Vo, fabricari ferens emet C, fabricariferensem et B (with re superscr. after fer), fabricari ferro ensem et V'. Here it would seem that A'B'C all had substantially the same reading which must have been that of a. Bo and V are attempts to improve on this.

II 147.

It appears from the above examination of the readings that we might arrange the mss in a scheme as follows.

* Deiter l. c. says the true reading of B is mollitur.

[ocr errors][ocr errors][merged small][ocr errors][merged small][merged small][merged small][ocr errors][merged small]

This agrees fairly with Forchhammer's view (Nordisk Tidskrift for 1880, p. 23 foll.), except that he does not recognize any connexion between E and the 1st line of descent (B), whereas many instances will be found above in which E agrees with A and C against B. I think also he is wrong in speaking of the group (y) as a familia deterior. Almost all the inferior codices may I think be affiliated to P. Again, where B alone has the true reading, Forchhammer will not allow that this is due to its being itself copied from the archetype or even to its belonging to a better tradition : in all such cases he holds that the true reading is an emendation. I do not think any one who has taken the trouble to look through the readings given above will accept this view of B. On the other hand I cannot agree with Müller (Neue Jahrbücher 1864 pp. 127—147, 261-281, 605—631) and Vahlen (in the introduction to his edition of the De Legibus) in ranking B higher than A*. It is undoubtedly less trustworthy, and though it is more often the sole representative of the archetype, this may be only because, A being closely allied with CPV, where A is right, these are usually right with it.

It may be worth while to add that the texts used by Priscian and Probus seem to have been in some respects better than our archetype, cf. 11 117, 118 and 91 for Prob. 106, for Priscian.

* See notes in my vol. 1. p. lxx,



[Reprinted from the Journal of Philology, Vol. xii. pp. 248–255.]

As I have had occasion to spend a good deal of time upon this Codex, which was kindly lent to me by the authorities of Merton College with a view to my edition of the De Natura Deorum, I think it may be useful that I should put on record what I have learnt as to its history and character.

We are told in the fly-leaf that it was given to the Merton Library by William Reade, a Fellow of the College, who was Bishop of Chichester from 1368 to 1385. He purchased it from Thomas Trilleck, who was Bishop of Rochester between the years 1364 and 1372. It is curious that the British Museum contains a us volume of Latin Sermons (Royal mss 10 A XI) similarly purchased from Trilleck by Reade, and presented by him to the College de Sancta Trinitate,' founded by Richard Earl of Arundel at Chichester. Both volumes have Bp Reade's library mark, and the fly-leaf in both shows the same handwriting, probably Reade's own.

The Merton Codex, which I have denoted as Oxf. in my 2nd and 3rd volumes, is a neatly written parchment volume consisting of 134 leaves or 268 pages. There are two columns in the page, each column containing 37 lines, and each line containing on the average 8 words. The words and sentences are divided. Abbreviations are frequent. It contains the three books of the De Officiis in 68 pages, two pages of Epitaphia Ciceronis edita olim a duodecim sapientibus, the three books of the De Natura Deorum in 64 pages, the 1st book of the De Divinatione (here called the 4th De Natura Deorum) as far as $ 106 duros ulta labores in 21 pages, and the first four Philippics as far as iv $ 15 quem habebat amisit. The 3rd Philippic is made to end at 111 27 victurum neminem, the latter half (from § 28 hodierno die to censuerint) appearing as the 4th, and our 4th as the 5th. This completes the original codex ending at p. 200. The last 68 pages, which are occupied with Palladius De Re Rustica in 13 books, are written in a different hand belonging to the 13th century. Mr E. M. Thompson of the British Museum has kindly examined the volume for me and informs me that the original Codex was written in England towards the end of the 12th century. As there are only three mss which are definitely stated to be of an earlier date, viz. the Vienna Codex V of the 10th century, the imperfect Harleian K, and the Leyden Vossianus A of the 11th, to which we may perhaps add the Palatine (P), called perantiquus in Baiter's ed., it is evidently deserving of a full collation*. The only other ass which can rival it in age are two others in the Leyden collection, Orelli's B and C (Baiter's H) both of the 12th century, and two French mss which seem never to have been collated, one in the National Library at Paris no. 15085, said to be written at the end of the 12th century, and one in the Library at Tours no. 688, said to belong to the same century. I mention these last in the hopes that I may be able to learn further particulars about them from some of my readers, and also that I may perhaps hear of other mss of equal antiquity which have escaped my notice. Halm in his preface to the Orellian edition of the philosophical treatises of Cicero mentions a Codex of the 11th century contained in the Munich Library (Ms 528), but I am informed by Mr Reid that he can learn nothing further about this from the present Librarian.

Notwithstanding his neat writing, the scribe is undoubtedly very careless (1) in the division of words and sentences, (2) in mistaking uncommon for familiar words, (3) in repeating words or clauses, (4) in omissions arising from the recurrence of similar words or syllables. As examples of (1) I may cite quid jus for quidvis 1 39, in situ for nisi tu 1 57, video for in deo 1 67, feres for se res 111 66, hoc diceretur pius esse for hoc diceret turpius esse 1 70, invidia for in India 11 42, quid doceam for quid Oceani 111 24, tam utiles for tutelae 111 74. Examples of (2) are carnales for Carneades 111 29, triformis for Trophonius 111 49, celso for caesios i 83, teximus eo ede for Thelxinoe Aoede 111 54, et amet for Aeetam et 11 55. Examples of (3) are 111 17, where, instead of plurima a te Balbe dicta sunt, we read plurima cum pulchritudine mundi B. d. s., the words cummundi being taken from a few lines below, where they followed pulchra, which the copyist probably confounded for the moment with plurima, and never corrected his mistake, if indeed he ever became

* For P and the Laurentian Codex 257 see the preceding Essay on Orelli's Mss.

aware of it. In the same way in 11 33, instead of nullum igitur animal aeternum est, we read n. i. a. appetit quaedam aeternum est, without any attempt at sense, the words appetit quaedam being inserted from below, where they followed another animal ; but the copyist writes on, apparently quite unconscious of his mistake. So in il 34, instead of quin id intereat, etenim ea ipsa, the copyist looking back a few lines sees another intereat followed by necesse est, and accordingly writes necesse est for etenim here : in 111 71 (inita subductaque ratione nefaria scelera meditantes), the copyist on coming to ratione allows his eye to stray to another ratione some lines below and goes on there qui in amore summo summaque inopia, returning then to nefaria.

The last kind of carelessness specified was the omission of clauses owing to the recurrence of similar words or syllables. The following may be quoted as examples. II 21 after non utitur om, nihil autemutitur.

after esse mundum om. similiteresse mundum. 32 after pluris esse om. necesse est-pluris esse. 36 after non sit deterior om. mundihomine deterior.

43 after praestantem intellegentiam om. in sideribusintellegentiam.

46 after nihil sit melius om. mundoid sit melius.
47 after absit extremum quantum, om. idem a summo-eruditum.
64 after vacare voluerunt om. ea partevoluerunt.

III 9 after facerem in om. causis-facerem in, which is however superscribed in the same hand.

18 after esset aliquid om. in rerumesse aliquid.
29 after omne animal om. tale est-omne animal.
35 after corpora intereant om. non-cum intereant.
79 after valere sic om, non-nemo sit.
90 after penis om. luendis-poetis.
1 70 after alterum utrum om, esse verum-concessit before esset.

So in i 95 we read nisi nunquamne vidisti, instead of nisi numquam vidi solem aut mundum beatum. Quid ! mundum praeter hunc umquamne vidisti ?

In the great majority of the above quotations, if not in all, and in many similar cases the Merton Codex stands alone. While they show the carelessness of the copyist, they also show that he does not go wrong of malice prepense, like the writer of the Cambridge Codex, with the idea of improving on his original. He does not try

« PreviousContinue »