thus addressed to masses of men, are they not difficult to follow, and do they not moreover possess enormous prolixity?" The answer of Clinias shows that hardness to be understood by the common man (τῷ δυσμαθεῖ there, τοῖς ἀνθρώποις above) and a long discussion are feared. Ast's version, therefore, otherwise good, οἱ ξυνακολουθεῖν λόγοις persequi oratione, is inadmissible. That expression relates not to the difficulty which the legislator finds. of expressing himself, but to the difficulty which the citizens find of following what he says.

Page 21, line 3. 892. D. “The common reading is εἰ καθάπερ. We have ventured to make the change [to xavánɛg ei] from the exigency of the place and on the authority of Stephanus." But Stephanus merely offered a conjecture, which the MSS. do not confirm. The exigency of the place requires ei, whether, after oxo7ɛitɛ, as much as it does ei, if, before de. As the sentence is constructed, xavánɛg seems to embarrass it, but was inserted on account of the comparison, which is not fully expressed. The construction without xavάneo is clear if a, if, is supplied. "Think whether, if we three had to cross a river-and I made a proposal, I should seem to have reason in what I said." Does not xαváno imply something like this: "Think whether I shall seem to have reason in what I say, just as, if we had to cross a river, etc. -I should seem to have reason in what I said." The thought for which xavάneg is inserted follows in the next words xai 8ǹ xai viv, etc.

The editor here says

Page 26, 9. 894. C. Tíra пooxoívayev. that "it is evident that av should be supplied." As there are numerous examples of the optative in a simple interrogative sentence without av, it is better to follow the MSS. See Hermann de particula av, 3. § 5.

Page 30, 16. 896. B. We are unable to see any attraction in this passage, and cannot perceive the justice of the extended remarks upon it on page 203. Пolhoory violates no grammatical Πολλοστὴν construction, nor could any other form of expression stand as well in this place. If we just supply the nominative olhoorn in thought, all will be clear. The word is to be taken twice; once in the relative clause, and once with 7oooúzor, and in one or the other instance good usage would suppress it. Page 40, 10. 899. B. Mr. Lewis prefers Böckh's emendation, εἶθ' for εἴθ' to Ast's έσθ'. But as ὁμολογών is found in five of Bekker's MSS., and three Florentine ones of Stallbaum, it is, we think, on the whole, to be preferred to ouoλoyei, and then Ast's cor

1845.]

Philological Criticisms.

545

rection o' follows by inevitable necessity. For the rest, ov beginning a sudden interrogation, is quite as strong as the era of indignant questioning.

Page 42, 13. 899. D. Zvrrévaa here and p. 44, 3, is no doubt relationship. Πρὸς τὸ σύμφυτον τιμᾶν cannot be to a natural honoring, but to honor something akin to you, or towards that which is akin to you, viz., to honor it.

Page 43, 6. 900. A. "Orav, inserted after ragúrta rà võv, on the authority of Eusebius and two Florence MSS. by Stallbaum, will bring order into this sentence. It is wordy and careless in its structure, like the talk of old men.

To make two sentences

of it, is to make Plato repeat the same identical thought, not by way of explanation, but as if he were saying something new. 18ov denotes observation in general, which is divided into that suggested by the reports of others ( aiovóμeros) and that derived from one's own eyes. (î avròs avrónīys, etc.

μέμφεσθαι.

Page 44, 2. 900. A. Años el μéμqɛoda. "A peculiar Graecism," says Mr. Lewis, " equivalent to dŋλóv ¿ori oé μéμqova." But δῆλος εἰ μέμφεσθαι is not Greek. The construction is δῆλος εἶ οὐκ ἂν ἐθέλων μέμφεσθαι. "It is clear that you cannot consent to

blame."

Page 44, 8. 900. B. Iados cannot, we think, be in apposition with rò vuv nagòv dózμa, but, if a part of the text, must be taken with ἐπὶ μεῖζον. "In order that your present opinion may not grow into a more considerable or pronounced state of feeling leading to impiety."

Page 44, 9. We see no so very great strength of meaning in ἀποδιοπομπήσασθαι. It is found both before and after καθήρασJa, as being nearly synonymous. It is used (Laws, 9. 877. E), in just the same way as άqootów in Laws, 9. 873. B. 874. A. Its verbal is used by Phrynichus (p. 306. Lobeck) as the opposite of ¿nréov. The metaphor, however, as the word is here used, gives the passage an intensity of meaning.

Page 46, 14. 900. Ε. Καὶ τῶν μὲν προσήκειν, etc. The editor justly finds fault with Ast's construction, but his own seems not to be unobjectionable. The sentence will have none of that complication which he gives it, if we take noоońжɛ and μɛróv, (i. e. μɛròv elva, comp. Soph. Electr. 459), as parallel, and supply zooαῦτα, the antecedent of ὅποσα with τῶν. Των then refers to both αἰσχρά and καλά. “ And we will affirm that of the things mentioned, so many as are bad pertain to us, if they do to anybody, but

that the gods have no share in any such things (i. e. çλavoor) great or small.”

Page 46, 5. It might be added in support of ovveterálouer, that Eusebius has that reading. (Praepar. Evang. 12. 52.) But ovvežetάCortov, which has the authority of the MSS. in its favor, can be borne with, if taken as a participle.

Page 48. 3. 901. A. The editor here supplies eos as the subject. But against this there are serious objections. It is not in the near preceding context, and the author names tov je dɛòv first a few lines below. But worse than this, it is unmeaning. The sense would be, "such a deity would be to us all, i. e. in our opinions, indolent, careless and lazy, etc." Such a deity as what? Either such a one as is indolent, careless and lazy, which is nonsense, or such a one as is vicious, which is not necessarily true. Apparently the text is imperfect. Ast supplies morrós, μισητός, and Ficinus may have found a similar word in his authorities, since his version is odio nobis habetur. But this does not explain the answer. Possibly the word of Hesiod, veuεontós, with another reference to him, may be wanting after quiv. The sense is, a person, if indolent, careless and lazy, one whom the poet declared to be just like dock-tailed drones, would be to all of us [an object of indignation, as he says]." To which Clinias replies in a common formula, "and very correctly too." To this the answer is, "then we must not say of God, at least, that he has a character of that very sort that he himself hates. The words avròs puoɛi contain an allusion to Hesiod's words, τῷ δὲ θεοὶ νεμεσῶσι καὶ ἀνέges; and лãow quiv, if that, and not nas quiv, be the true reading, alludes to soi xai άvéges in the same passage.

Page 50. 3. 901. B. How dŋ;-answer λéyouεv. Better as Eusebius has it, ποίω δη λέγομεν ;—answer ἢ διαφέρον, etc.

Page 52. 11. 902. A. The reading from Eusebius preferred by Ast and Stallbaum, τοῦ γιγνώσκειν instead of τὸ γιγνώσκειν, seems to suit the ensuing context best. The sense is, "what remains but the opposite of knowing," (i. e. not knowing that they ought to attend to all such things). With the vulgar text, which the editor retains, the sense is, "what remains but that they know the opposite ;" and it must be confessed that a good sense can be derived from this reading.

Page 52, 13. 902. A. The construction of this sentence might be made clear by remarking that two forms of expression, both common after words of saying, are used together: Aéyorra ós àyνοοῦντας, and λέγοντα γιγνώσκοντας οὐ ποιεῖν.

1945.]

Philological Remarks.

547

Page 53. 7. 902. B. We prefer ovnɛo the vulgar reading, which Mr. L. retains, to wonɛe, Ast's and Stallbaum's reading, which is found in one MS., in the margin of another, and in the version of Ficinus. Rone denotes merely comparison; one contains an argument from the greater τὸν οὐρανόν, to the less τὰ θνητὰ ζῶα. Page 54. 6. 902. D. iarog dý, etc. Mr. Lewis gives a far better construction to this sentence than Ast does. We are surprised that he can hesitate for a moment as to the question whether προστεταγμένον can be ever used as an impersonal absolute.

Page 55. 11. 902. E. We see no use for the mark of a broken sentence at the close of this passage, and we doubt if the structure be anacoluthous, common as that structure is in these books of the Laws. Supply, if anything, not elva but ovra with qavλότερον. In τὸν δὲ θεόν, δὲ has its usual place after a parenthesis, and may mark contrast likewise. The construction proceeds thus. "Let us not think that God, inferior to mortal workmen,— that God, I say, takes no care of things small, but [does take care] of the great." In this sentence a révy is beautifully contrasted with oogárazov, the single art of human artists with the boundless wisdom of God.

Page 57, 2. 903. A. In separating oods from un héyev and assigning it to another speaker, the editor is obliged to give to un Léyew the sense of "saying nothing to the purpose," which he affirms to be common in Greek. We wish that he had given one example, for although μηδέν, οὐδέν, λέγειν are often so used, we are not aware that the other formula occurs with this sense. 'Oμoλoyeiv μǹ héyev can only mean, we believe, to own that he does not talk, or to promise not to talk. The same words with goos have the sense desired, viz. to own that he is not right in what he says.

Page 57, 10. 903. B. ovvreraɣuéva cannot mean, " so arranged as to coöperate with the universal guardian," but arranged together in a system.

Page 57, 14. 903. C. ànegɣaoμévot. This word seems to us incapable of receiving the sense of appointing or constituting an officer, as Mr. Lewis understands it. But the meaning is not clear. If zélos could denote perfection here, we might translate thus: "having brought about perfection even to the ultimate division, i. e. having carried perfection into things the most minute." Cousin's version is something like this.

Page 58. 6. 903. C. The reading of Eusebius, approved by Stephanus, Ast and Stallb. πρὸς τὸ κοινῇ ξυντείνων βέλτιστον, referring

them, or making them all aim together at a common good, seems to us much better than the vulgar text, πρὸς τὸ κοινῇ ξυντείνον βελ Totov, which the editor (on p. 291) prefers, and which has indeed the manuscripts in its favor. For besides the harshness of ourτείνων κοινῇ instead of εἰς, ἐπί, οι πρὸς τὸ κοινόν, the highest good, which is aimed at and is only a result, is awkwardly said here to aim or tend. For the active use of ovvreiro comp. Gorg. 507. D, Republic 591. C, ὅ γε νοῦν ἔχων πάντα τὰ αὑτοῦ εἰς τοῦτο ξυντείνας βιώσεται.

Page 59.7. 903. E. We think that the editor has well comprehended the scope of this difficult passage, in note 53—far better indeed than Cousin; yet there are several points in his explanation of it which do not satisfy us. We have strong doubts about vdwo upuyor,cooling water, not frozen, as he has it,-which is only an emendation of Cornarius. uwvzor, animated, is favored by u úzove noάžas just below, and by 895. C (iar idoμer etc.). In his version of rs... xooμnows he neglects the article; nor is it possible to supply μετασχηματίσματα here. To make sense as well as grammar we seem obliged, with Ast, to supply và before tns, which had also occurred to the writer. The sense is, "the altered arrangements"-i. e. the changes of arrangements-" would be endlessly numerous."

Page 60. 6. 904. B. 'Ayavòv after opɛheir deserves to be expunged, as wanting in four Florence MSS. and in Eusebius; and as hurting the sense, which is "and that whatever good there is pertaining to the soul naturally tends to benefit, and evil to injure."

Page 61. 8. 904. B. We like the view given in note 57 that σμικρότερα τῶν ἠθῶν is contrasted with μείζων δὲ δὴ ψυχή, and that Thɛiw d is a subordinate member of the first clause. This had

also occurred to the writer. But the words μείζων, ματαλάβη, have scarcely a shadow of manuscript authority and must give way to μείζω, μεταβάλῃ. The meaning however will not then be essentially altered. The principal divisions of the sentence still begin with 1. smaller traits of character when they change less, and when they change more and for the worse; 2. when the soul undergoes greater changes, i. e. when its leading or larger traits of character are altered. In the latter part of the passage, the mention of some better place immediately after that of a "place surpassingly excellent and altogether holy," together with the use of the aorist uɛrazoμodeioa are difficulties which trouble us, and which the editor does not remove. As for the rest, Mr. Lewis