64, and may suppose both to have been similarly stated by Posidonius, though he expressed his own assent to the latter; that C.'s motive for maintaining the other view in § 85 was probably the wish to give his own experience on the subject; lastly that the anti-Stoic remarks are no more than were required in order to give the proper colouring to a speech put in the mouth of an Academic; that they occur sometimes in purely Stoic passages; that in general the Stoic writers form the store-house from which C. borrows his arguments against Epicurus, whilst he attacks the Stoics themselves with weapons forged by the Academy, as in the De Finibus; that in the present treatise this is foreshadowed by the language used of the Epicurean doctrines in § 3, of the Stoic in § 4; that Euhemerism is not the same as Stoicism, and that the observations about the mysteries are an interpolation of Cicero's (see my nn. on § 119). See further, as to the difference between the undoubted criticism of Carneades and that contained in this section, my note on § 92 under habebit igitur.

§ 5. TEXT AND ORTHOGRAPHY.

The text which I have given agrees in the main with that of the latest editor, C. F. W. Müller, Teubner, 1878, but I have endeavoured throughout to weigh the evidence, internal and external, for each reading to the best of my ability; and I have in some instances retained the reading of the MSS, where it had been altered by Müller in common with all the recent editors. Thus I have thought it unnecessary to insert a second eadem before requiro in § 21, and I have three times ejected a non which they had inserted, before potest in § 21, before nihil in § 93, before pudeat in § 111. Elsewhere I have ventured on transposition of sentences as in §§ 5, 30 and 97; and on emendations of words, as in §§ 26, 49, 71. In the critical notes my object has been to put the reader in possession of the requisite data for forming an independent judgment on the text. As a foundation I have given the more important of the readings contained in the 2nd ed. of Orelli, brought out under Baiter's supervision in 1861; but, though the MSS (ABCEP)1 there cited supply the principal material for determining the text of the 1st book of the N.D., they do not seem to me to

1 For a description of the мss see the note prefixed to the text.

possess such a transcendent superiority, either in point of accuracy or of age, as to make it unnecessary to weigh carefully the evidence furnished by other MSS. I have therefore thought it my duty to examine, as far as was in my power to do so, all evidence which could throw a light on the condition of the text up to the end of the 15th century. Thus, besides the critical editions of Orelli, Heindorf and Creuzer, I have had in constant use the Ascensian ed. of 1511, and two MSS (U and Y) most kindly lent to me by S. Allen Esq. of Dublin, whose father's name will be familiar to students of Cicero under the Latinized form 'Alanus.' I am further indebted to J. H. Swainson, Esq., late Fellow of Trinity College, Cambridge, for the use of his very careful collation of eight MSS. seven belonging to the British Museum, and one to the Cambridge University Library, as well as of the two earliest printed texts. This collation is given in an abridged form at the end of the volume, Another MS (0), recently purchased by the Trustees of the British Museum, has been collated for me by a member of the staff of the MSS department there: but in this, as in the other cases, I have myself compared the collation with the MS, wherever special importance attached to a particular reading. In like manner the readings of four Oxford MSS given at the end of the Oxford 4to edition of 1783 have been tested for me, e o u by H. P. Richards, Esq., and by J. S. Reid, Esq.. Beside the мs readings, I have also mentioned the differences between my text and those of Schömann, Müller and Baiter, both in his earlier and later editions.

In order to show that the record preserved to us in Orelli's MSS is at any rate not so complete as to dispense either with emendations or with a careful comparison of other MSS, I have given below, 1st, a list of passages, in which the text is supported exclusively by what are considered the inferior MSS in opposition to all Orelli's MSS: in many of these cases the true reading had been independently restored by conjecture, and it is of course open to question how far the MSS themselves are to be considered as witnessing to a traditional reading or merely giving the scribe's emendation; 2nd, a list of passages in which the received text is supported by one only of Orelli's MSS; and 3rd a list of passages in which the

2 I may mention as an illustration of the danger of trusting to negative evidence in the case of мs readings, that scarcely one in ten of the inferences which I had drawn ex silentio on the part of the Oxford collator of 1783 was verified on examination of the мss themselves.

received text rests solely on conjecture unsupported by any existing MS. Under each head I have mentioned only those readings which are accepted (except where otherwise stated) by Baiter, Schömann, and Müller in common with myself.

1. True reading preserved by inferior MSS in opposition to all Orelli's MSS.

§ 18.

oculis (om. animi), Asc. V (Schömann dissents). § 26. continentem for incontinentem, Asc. UCHMRV. § 28. commenticium for conventicium, Asc. INOV.

earumque for eorumque, Asc. CMV.

§ 29.

§ 39.

fatalem vim for f. umbram, El. (Baiter dissents).

$61.

ceteroqui for ceteroque, Oxf. u Moser's E and M.
consessu for consensu, Asc. Ry.

$72. nihil ex for nihil ne ex, UHO.

§ 77, quasi sui for quam sui, I of Moser.

§ 81. defendes for defendens, UYL.

§ 82. Aegyptio for Aegypto, Asc.

§ 83. laudamus Athenis for l. esse Ath., I of Moser.

§ 86. aliquid esse for aliquid iste, El. Oxf. e.

§ 87. lustrationem for illustrationem, GC El. Reg. Herv. ·

§ 97. $99.

at figura for ad figuram, UCHLO. (Baiter dissents).

ad speciem nec ad usum for speciem nec usum, G. (Baiter

and Müller read by conjecture specie nec usu.)

§ 103. oportet et for oportet, CN of Moser.

§ 113.

§ 116.

se ipsa for se ipse, Asc. CR.

nam etiam for nam enim, G.

allicere for elicere, GHRV.

voluptate for voluntate, ULMNORV.

§ 123. homunculi for homunculis, MR Herv.

2. True reading preserved by inferior MSS in opposition to all but one of Orelli's MSS.1

ut before magno argumento, B3 Asc. UILO (Baiter dissents). turpius for fortius, C Asc. El. Herv. Pal. 3.

descendisset for descendis (or descendens) sed, B Asc. CNMR. si di for sic di, EUCV.

§ 28. reprehenditur for reprehenderetur, E (and by correction in

B) CLMN.

§ 1.

§ 18.

§ 25.

§ 37.

sententia est qui for sententias qui, A3UCM Asc.

1 I have not thought it necessary here to distinguish, as I have done in the critical notes, between positive evidence and ex silentio inference, as regards the readings of Orelli's Mss.

M. C.

e

§ 122. nulla re for in n. r. CUY Asc. HO.

3. True reading restored by conjecture in opposition to all MSS. Greek for Latin in πρόνοιαν § 18, στεφάνην § 28, ασώματον § 30, εἱμαρμένην § 55, μαντική § 55, &c.

§ 33.

cingat for cingit.

revocet for revocat.

a magistro non dissentiens. Other conjectures are given by Sch., Ba. and Mu., but all alike change the MS reading. modo deus moveri for modo mundus moveri. (Sch. gives a different conjecture.)

§ 39. universitatemque for universam atque.

§ 45. vitae actionem mentisque agitationem for vitam et actionem mentis atque agitationem.

§ 49. ad nos for ad deos. Ba. gives a different conjecture. § 64. om. aut before Neptuni.