never loses and often gains by leaving those questions alone. While I am suggesting, may I add that we may all do one small service to our science, for which no new knowledge is wanted? We may cease to call it philology.'

T. C. SNOW.

Some points of Roman Law in Prof. Tyrrell's Edition of Cicero's Correspondence. Vols. I., II. Dublin Univ. Press. £1 48. THIS edition has been so heartily and generally welcomed that I occupy no space with praising it. It is a pleasant and scholarly work. But I wish even now the editor would put the various readings at the foot of the page, and not assume the correctness of the chronological order adopted by his predecessors, In using it I have noticed a few passages where he seems to me to have mistaken the law, and some other slips which may be more briefly dealt with. take the more important passages first, out of their proper order.

I

Letter cxxi. Fam. vii. 13. Tantum metuo ne artificium tuum tibi parum prosit : nam, ut audio, istic non ex iure manum consertum sed magis ferro rem repetunt,' et tu soles ad uim faciundam adhiberi: neque est quod illam exceptionem in interdicto pertimescas,

quo tu prior ui hominibus armatis non ueneris'; scio enim te non esse procacem in lacessendo.

Trebatius, the lawyer, is attending Caesar in his camp in Gaul, and Cicero chaffs him on his position. 'I am only afraid, that your professional craft is of little good to you, for, I am told, where you are (to quote Ennius) "men don't join issue in due course of law, but effect a recovery sword in hand," and actually you are now called in to use force! Well, there is no ground for much fear of your being troubled with the plea of having been the first to come with a force of armed men, for I know you are not forward in attack.'

Trebatius, if dispossessed of some land by armed violence, would apply to the praetor for an interdict which would state the issue in some such words as these, addressed to his opponent: Unde tu, C. Trebatium ui hominibus armatis deiecisti, eo C. Trebatium restituas. If, however, Trebatius had himself previously turned out his opponent by the like means, the opponent would urge the praetor to insert in the formula after 'deiecisti' quo ille prior ui hominibus armatis non uenerit; and the praetor would naturally consent. The matter would then be

referred to a judge to decide with these instructions. If the judge found that Trebatius had been ejected by armed force and had not himself been the aggressor by the same means, the injunction would be made final, and the defendant probably be condemned in damages: if his prior aggression were proved or his dispossession not proved, the injunction would drop and in some cases the plaintiff would pay a forfeit (cf. Gai iv. 141, 161-165).

There is I think no doubt that the interdict referred to is de ui armata (which in the Digest is consolidated with that de vi Dig. xliii. 16). Prof. Tyrrell erroneously takes it to be the interdict uti possidetis, and naturally finds difficulties. He has been misled by the language of Cic. Caecin. 22, § 63 and Gai iv. 155, whence he infers that no exceptiones were allowed when armed force had been used. It is not necessary to assume in either passage that such a plea as we have here was in question. But that such a plea was allowable is, I think, clear (1) from this passage itself; (2) from the analogy of the interdict de ui' (cf. Cic. Caecin. § 92); (3) from the reason of the thing, supported by the language of the Digest.

The use of armed violence in matters of ejectment was rightly held to be so contrary to the dignity of legal procedure as to require peremptory prohibition. Accordingly a person who had himself acquired possession from his opponent by force (ui, not ui armata) or by stealth or by sufferance was yet entitled to immediate restoration, if his opponent ejected him by armed force. Obviously the same principle applies against him, if he has himself used armed force. His own armed violence disentitles him from claiming the peremptory protection of the law, on the same principle on which armed violence disentitles his opponent from pleading the wrongful possession of the former. Indeed the two acts may Iwell have been successive events in one day's struggle. If Trebatius (in the supposed case) had brought a body of armed men to dispossess his opponent, he could not be aggrieved by his opponent's resorting to the same means to dispossess him in turn. Eum qui cum armis uenit, possumus armis repellere, sed hoc confestim, non ex interuallo; dummodo sciamus non solum resistere permissum ne deiciatur, sed, et si deiectus quis fuerit, eundem deicere non ex interuallo sed ex continenti. (Ulpian in Dig. xliii. § 16, 1. 3, § 9; and cf. 1. 1, § 27, 28.)

usual meaning to inductis; (for inducere, 'to cancel,' see Dirksen's Manuale, s.v.). I believe the nomina cancelled were the nomina of the three augurs and the two consulars who were to give false evidence, and it does not matter for this purpose whether the nomina were names or entries or items, and whether the pactio was a separate written document or not. It may have been such, but as the pactio is said to have been made nominibus et perscriptionibus multorum tabulis, it seems better to suppose no such separate written document containing the whole plot to have existed. It was not made verbis, i.e. by a formal stipulation such as was usually applied to bind a bargain. How then was it made? That in some degree depends on the meaning which we attach to perscriptionibus, and, as the word occurs in several passages, it is worth while to examine it.

Perscribere is used of writing a matter out in full (e.g. Cic. Cat. iii. § 13, &c.), and perscriptio is thus 'a writing of some transaction.' But the precise nature of the writing so denoted is not easy to determine. It evidently had a business meaning, but whether this meaning was wide and general or specific is not clear. (a) In Rosc. Com. §§ 1, 2, 5 it appears clearly to be used of entries in the ledger, descriptive of the person credited or debited with sums of money. In Flac. § 44 the meaning is the same, only that the entry gave a description of the particular purpose of the payment as well as of the name of the creditor or debtor. Similarly in Verr. v. § 48 and Ter. Phorm. 923 (if we read with most MSS. perscripsi instead of the Bembine discripsi). (b) In the tablets found at Pompeii (Bruns' Fontes ii. cap. 9) perscriptio is the title ('docket') affixed to some documents which in form are certificates of receipt of money. And yet (c) a third mercantile meaning seems to be given by Cic. Att. xii. 51 Tiro narrauit perscriptionem tibi placere and xvi. 2, § 1 quod perscribi oportet which clearly imply some mode of settling a debt-probably something in the nature of a bill of exchange drawn upon a banker at some distance of time, e.g. a six months' bill. And in Liv. xxiv. 18, § 14 a quaestore perscribebatur we have clearly bills drawn on the quaestor. (For this use of the preposition a cf. Cic. Flac. § 44). Perhaps these three meanings may be reconciled by supposing perscribere, perscriptio to mean originally simply 'expression of the debt in writing' and to be specially applied to entries in a ledger or to receipts or to orders on a banker. It is possible to com

bine the two last, as in warrants on the English Treasury, which contain a receipt for the money, to be signed by the payee and exchanged by his banker against payment. In Att. ix. 12 § 3, where Cicero is describing how the world generally is going on in the regular routine (notwithstanding the political crisis implied by Pompey's being actually besieged by a Roman army), he says viri boni usuras perscribunt 'honest men are entering up their interest,' where the viri may be the creditors (according to Mommsen, Hermes xii. p. 112) or the debtors. So Cic. Orat. i. 58 de tabulis et perscriptionibus may well be merely 'about ledgers and entries,' but may also be about documents and receipts (or cheques).' In Phil. v. § 11 Antony is described as squandering public money falsis perscriptionibus donationibusque 'by fictitious book entries (or receipts or cheques) and giîts'; fictitious, because professing to be in accordance with Caesar's orders. Suet. Caes. 42 deducto siquid usurae nomine numeratum aut perscriptum fuisset; 'less anything which had been paid in cash or by cheque (or book entry) under the head of interest.'

But

It is thus hard to say the precise meaning of perscriptionibus in our passage. whether perscriptionibus is the fuller description of the nature of the debt and nominibus the mere entry of money paid to, or received by, a person named, or again nominibus includes the whole of the book entry and perscriptionibus denotes a number of warrants for payment in some shape or other, in either case the bargain with the consuls was contained in a number of books and documents.

I do not see how the dependence of the payments on the consuls' good faith could be secured by these documents. And for this purpose a pactio ne peteretur (to guard against the payments being enforced if the consuls played false) may have been necessary and may have been put in writing and may have been read out in the senate.

But it is not likely that the agreement was made in blind confidence that some three augurs and some two consulars would be found ready for the plot. I think particular persons were first secured; book entries or money orders in their favour were made and provisionally (at any rate) executed. But as Memmius or some one blotted out the names, Cicero could describe only the general character of the plot and the number of augurs and consulars concerned.

If it be insisted that nominibus should mean the same in both places (I do not think

it at all necessary), we may suppose that the whole or part of the book entries were blotted out, but that yet from the remainder or from the context the nature of the entry may have been sufficiently discoverable to support the evidence afforded by the perscriptiones or by Memmius' confession.

[Mommsen (Hermes xii. p. 111) considers perscriptio to be a 'payment accompanied by a formal receipt' and endeavours, not successfully, as I think, to reduce all the passages under this meaning].

Letter ix. Att. i. 4. Cui cum aequi fuissemus, tamen &c. Tunstall, Boot, and Tyrrell all take this as meaning though I might have taken a lenient view (had I so willed).' I cannot think Cicero would have so used a simple pluperfect subjunctive instead of a periphrasis with possum or the future participle (cf. my grammar, § 1521). Why not translate, 'Although I had been favourable to him' The verdict was that of the judices. Cicero gained general applause, because the verdict was approved and he had presided as praetor in an able and impartial manner, notwithstanding his feeling for the defendant.

Letter xx. Att. i. 14, § 3. Cicero says that the support given him by Crassus in the senate was the more remarkable quod meis omnibus litteris in Pompeiana laude perstrictus esset. Prof. Tyrrell inserts orationibus after meis and calls (in the second edition) omnibus litteris an ablativus mensurae. The ablative of measure is however only used with comparatives or expressions of precedence or distance. Still omnibus meis litteris may mean (cf. my grammar, § 1170) throughout all my letters' or 'throughout all my literary compositions,' and in either case offers no difficulty requiring addition to the text. A person writing to his intimate friends does not always so express himself as to leave nothing obscure. to persons 1900 years after in ultima Thule.

In the next section Aperte tecte is explained rightly in the beginning of the note and wrongly in the end. Cicero would never put two such adverbs to qualify, instead of to contrast with, one another.

letter i. § 3). Quintus had at the request of a creditor ordered Flavius' agents not to impair the estate of L. Octavius Naso, to which Flavius was heir, till the debt was paid. Marcus points out to his brother that he had rashly assumed that the claim was good and had thereby done injustice to Naso. He requests Quintus therefore to withdraw his order. 'I beg you to make a concession to Flavius' agents in the matter of impairing the estate.' Deminuere is used somewhat technically. Cf. Cic. Flac. § 84; Dig. xviii. 1, 1. 26; xxviii. 8, 1. 17.

=

Letter cxxvi. Fam. vii. 23, § 2. Tu autem ignarus instituti mei, quanti ego genus omnino signorum omnium non aestimo, tanti ista quattuor aut quinque sumpsisti. Prof. Tyrrell (see also the Corrigenda) apparently misunderstands the passage, for he proposes to omit non, to read aestimem for aestimo, to treat tanti and quanti as not correlative, and to regard quanti as quantuli. I do not see the difficulty. Fadius Gallus had bought for Cicero some statues of Bacchae and one of Mars, and had given a high price. Cicero wanted some cheap statues for his palaestra of a totally different kind, perhaps something respectable, as philosophers, &c. In your ignorance of my ways (or purpose) you have bought those four or five statues at a price at which I don't value all the statues in the world.'

Letter cliv. Att. iv. 18. Sulla non dubitans quin foris esset, (Gabinium) postularat. Prof. Tyrrell doubts what he says is the common interpretation of foris esset was in debt,' but still inclines to a metaphorical use 'was out at elbows,' was a defaulter': see p. 183 and pref. xii., xxi. In the other passage where it is used, also of Gabinius, (is. § 2) Madvig reads sordidissime and treats this supposed use of foris esse with scorn. Why may not Sulla have thought Gabinius to be away from home'? I see no occasion for a metaphorical meaning in this letter. If the words be retained in the speech, it is perhaps possible to take it as 'on the world,' i.e. not a steady citizen living at home on his own means.

Letter clx. Q. Fr. iii. 9, § 8. Prof. Tyrrell translates 'tabulas obsignare' 'sign the will.' The expression is incorrect. Roman wills were not executed by 'signing.' They were written (by any one), declared in the presence of witnesses to be the testator's will, and then the tablets were sealed up (obsignantur) by the witnesses. The sealing was not for execution, but to protect the contents from being altered. The same with other documents see Cic. Flac. 9, § 21 and the regu

lations for opening wills in Dig. xxix. 3. Even subscribere, subscriptio, when used in such matters, did not mean what we moderns understand, viz. the mere writing of the executant's name, but a declaration, e.g. Decretum. Fieri placet. Jubentius Celsus promagister subscripsi (Wilmann's Inser. no. 312). The 'subscriptio' of the testator to wills was first required by a constitution of Theodosius II. anno 439: see the elaborate essay of Bruns' Die Unterschriften, &c. in his Kleine Schriften, bd. ii.

Letter clxii. Fam. i. 10 (misprinted 20).

Illo si ueneris, tamquam Ulixes, cognosces tuorum neminem. Prof. Tyrrell, to save Cicero from the charge of not knowing his Homer well, accepts Klotz's conjecture of cognoscere nemini. But is this Ciceronian Latin Ovid, no doubt, could say and said lugebere nobis, but Cicero would not use a dative of the agent with a finite verb, except where 'for' a person is as suitable a meaning as 'by' him, e.g. N. D. ii. 48 bestiis cibus quaeritur; Q. Fr. i. 1, 25 aes alienum contrahi civitatibus. H. J. ROBY.

Novum Testamentum, etc. curante F. H. A. SCRIVENER. Editio Major. Deighton, Bell et Soc : MDCCCLXXXVII. 48. 6d.

THIS is a third, and greatly enlarged and improved edition of a well-known work, the first edition of which appeared in 1859, and the second in 1876. As before, the text of Stephens of 1550 is taken as the standard, while the chief variations are given in footnotes. The main features of this new edition are that (1) to the readings adopted by Beza, Elzevir, Lachmann, Tischendorf, and Tregelles have been added those of Westcott and Hort and of the Revisers; (2) carefully selected references to parallel passages have been inserted in the outer margin; (3) for the convenience of those who use this volume in collating MSS., the Ammonian sections and Eusebian canons are given in the inner margin, a table of these being added to the Preface. The paging is the same as in the second edition; room for this additional matter being obtained by increasing the size of the page. But the volume may still be carried in a fairly capacious pocket. In a few paragraphs added to the former Preface the learned editor speaks with gratitude, admiration, and severity of the work of Westcott and Hort. There is no need to praise their fine intelligence, immense learning, industry, and sagacity. These rare gifts, while they excite our envy, fail (in Dr. Scrivener's judgment) rdv 1⁄2тTW λóyov kрEITTW

ποιεῖν.

Still half enslaved by the principles of Lachmann, they have gone wider of what Dr. Scrivener believes to be the truth than even Lachmann nimself, and have produced a magnificent blunder rather than an everlasting possession (splendidum peccatum, non Kтñμa eis àel, in lucem emiserunt).

The Resultant Greek Testament, exhibiting the Text in which the Majority of Modern Editors are agreed, by R. F. WEYMOUTH, D. LIT. Elliot Stock: pp. xix. 644. 12s. 6d.

THIS work agrees with the preceding one in giving a Greek Text of the New Testament with a careful digest of the various readings adopted by editors; but it proceeds on a very different plan. Instead of adopting the exceedingly faulty text of Stephens as a standard, the editor constructs one for himself; not at first hand from MSS., Versions and Fathers, but at second hand from the most important editions of the last fifty years. His principle, therefore, is the same as that which for some years has been adopted in the Cambridge Greek Testament: but it is worked out in a much more comprehensive way. The editors

of that series have confined their attention mainly to Tischendorf and Tregelles, with occasional reference to Lachmann. Dr. Weymouth lays under contribution, not only these three great critics, but also Alford, the Bâle edition of 1880, Westcott and Hort, and the Revisers, together with Lightfoot and Ellicott for S. Paul's Epistles, and Weiss for S. Matthew. This eclectic text, therefore, 'is intended to exhibit in a compact and intelligible form the latest results of textual criticism.' Hence the somewhat peculiar title, The Resultant Greek Testament.'

Of the materials used only two require separate mention. The Bâle edition was edited in 1880 for the Bible Society of Bâle by Stockmeyer and Riggenbach. It is largely influenced by Tischendorf, and where it differs from him frequently anticipates the readings of Westcott and Hort. 'Das Matthäusevangelium' of Bernhard Weiss was published at Halle in 1876. Its text also is greatly under the influence of Tischendorf.

The work has been executed with the greatest care the proofs having been corrected, not from the 'copy sent to the press, but from the editions themselves from which the 'copy' was taken. The accuracy of the readings quoted has, therefore, been doubly secured.

A student with either Scrivener or Weymouth in his hand can at once see what reading has been adopted by Lachinaun, Tischendorf, Tregelles, Westcott and Hort, and the Revisers in any given passage; but with Weymouth he has the suffrages of other editors as well. Nor is this the only advantage. Weymouth's text is as clear and as free from interruptions as that of Westcott and Hort, and like theirs is spaced in accordance with the sense. Scrivener the frequent changes of type, and still more frequent insertions of letters and numerals, are trying to the eyes. Besides which, it is much pleasanter to work with a text to which all the best modern critics would in most cases assent, rather than with one from which nearly every one, excepting Dean Burgon, would be perpetually dissenting.

In

The two works have appeared almost simultaneously, and Dr. Weymouth was evidently unaware that Dr. Scrivener had a third edition in preparation. It is interesting to contrast his opinion of Westcott and Hort with that of Dr. Scrivener, quoted above. He is convinced that critical judgment will more and more converge towards most of the conclusions arrived at by WH,' and he speaks of their Greek Testament as a work beyond all praise, both for the erudition displayed and for the simple