maistre des cinges qui seoit en son trosne audessus des autres appella le flacteur et luy demanda: "Qui suis je," dist il, "et qui sont ceulx qui me seruent?" "Tu es," dist celluy, "vng empereur, et ceulx cy sont tes princes, tes ducs et tes barons." Icelluy fist le maistre singe grant honneur et luy feist moult de biens. Quant celluy qui ne sauoit flater ne mentir vit ainsi honorer son compaignon pour mentir, il dist en luy mesmes: "Ce mon compaignon pour flacter et mentir a este ainsi honore. O! comme le seray je haultement pour dire verite." Le maistre singe l'appella et luy demanda qu'il luy sembloit de luy et de ses gens. "Tu es," dist il, "vng cinge, et tous ceulx d'entour toy sont cinges." Lors tout incontinant le cinge et ses subgetz luy rovirent sus et fut tout desclue, esgratigne et malmene.

Par lequel exemple nous est donc a entendre que la verite n'est pas tousiors bonne a dire; car les prelatz et les princes ne veulent ouyr dire que li coses qui leur plaisent. Bien sont singes ceulx qui font ou seussient faire les cingeries en leurs maisons, et qui croient plus tost vng flacteur que vng homme veritable.

JOHNS HOPKINS UNIVERSITY.

GEORGE C. Keidel.

REVIEWS AND BOOK NOTICES.

Prohibitives in Silver Latin. By WILLARD K. CLEMENT. Reprint from A. J. P., vol. XXI 2. 1900.

I did not suppose that I could be tempted into writing anything more on the subject of the Latin Prohibitive, but Professor Clement's method of criticism is so unusual, and his comments would be so utterly misleading to the casual reader, that I must, in justice to myself, say one word more. It is unfortunate that Clement did not devote greater care to the preparation of his article, as he has undoubtedly collected much valuable material. In its present form, however, the article is, in most respects, quite without value, so far as its criticism of my own views is concerned, on account of its numerous inaccuracies and its utter lack of discrimination between relevant and irrelevant matter.

In my articles on the Latin Prohibitive I made the claim that, prior to the time of Livy, the perfect tense in prohibitions differed from the present in being a more energetic form of expression. Clement's method of combating this claim is to cite some instances from Silver Latin which he does not think in harmony with it. One might as well try to refute the grammar-rule that quamquam takes the indicative in classical prose by citing from Silver Latin the numerous instances of the subjunctive.

Whatever might have been shown to be the usage of Silver Latin, my claim as to earlier times would have remained quite unaffected. However, after examining such of Clement's statistics as really have a bearing upon my theory, I am now quite ready to assert, as I could not have done when I wrote my Latin Prohibitive, that the claim I made for earlier times holds also, in all its essential features, for Silver Latin, and to treat Clement's article as though my claim had originally been framed in such a way as to cover the latter period also.

The distinction I made between the two tenses was made solely with reference to independent prohibitions introduced by ne (neve) and cave. In attempting to prove false this distinction, Clement cites a curious mixture of subordinate clauses, undoubted contingent-future ('potential') subjunctives, mere conjectural readings, and subjunctives with nec, minime, nullus, nemo, nihil, numquam, non, and vide. When Clement professes to discuss a certain claim of mine, I surely have a right to insist that he shall take that claim as it stands, without any additions or modifications. In that part of the Latin Prohibitive which prompted Clement's

paper, not a word was said by me regarding the subjunctive uses with nec, nihil, numquam, etc., except an emphatic statement to the effect that they lay entirely outside of my theory and had characteristics very different from those of the instances I was going to discuss. What possible objection can Clement have to doing in reality what he professes to be doing? What possible objection can he have to separating (at least temporarily, for the purpose of testing my claim), in his discussion as he does in his headings, the instances of ne (neve) and cave from those of nec, nihil, numquam, etc.? Such a separation could not by any possibility affect his discussion or his conclusions in any way detrimental to the truth. If the instances with the latter words present the same general characteristics as those with the former, then his conclusions would not be affected at all by the fact that he had temporarily separated the two sorts of instances. On the other hand, if the two classes of instances are found to show important differences in usage (differences that can not be accounted for by mere chance), then surely it would be quite inexcusable in any one not to recognize the justice of treating the two sorts of clauses separately. In either case, then, such a method of procedure as I suggest would have been perfectly fair to Clement's side of the case, and it would have had the additional advantage of being fair to mine.

In the following discussion I will confine myself, as I did in my original article, exclusively to the instances of ne (neve) and cave. Clement cites 25 such instances of the perfect tense (pp. 156 ff.). Of these 25 instances, he admits at the outset that 17 are in accord with the distinction I made. At first he classifies all of the remaining 8 instances as being not in accord with my theory; but a little later he decides (pp. 164-5) that 5 of these 8 instances are not necessarily against it, after all. In other words, he finds, according to his own admission, only 3, out of a total of 25 instances, which he considers as distinctly opposed to my theory that the perfect tense indicates energetic utterance, prompted by alarm due to fear that the prohibited act will be performed. Let us examine these 3 alleged exceptions:

Phaedr. App. 26, 3 ne timueris. Just as I was on the point of admitting that this is a clear violation of my theory, I discovered that ne timueris is merely one of several conjectures, and has not the slightest authority of any kind whatever. All the other conjectures have the present tense, which would be in complete accord with my theory.

Tac. Ann. VI 8 ne ultimum Seiani diem, sed sedecim annos cogi

averitis. Here I fear that Clement neglected to read the context. Failure to heed this prohibition will inevitably result in the speaker's condemnation and death. The speaker is on trial, charged with being a friend and accomplice of Sejanus. He says, in effect, to his judges: 'Do not think of

me as intimate with Sejanus on his last day, and for that reason condemn me as implicated in his crimes; but think of me rather as his friend of former years, when all men were proud to claim Sejanus as their friend.'

Mart. II 68, 3. If this prohibition is not complied with, the speaker's former rex and dominus will call him insolent, an act which might or might not prompt energetic prohibition, according to the speaker's feelings regarding it.

These are the 3 instances upon which Clement depends for the refutation of my claim that the perfect tense with ne (neve) and cave indicates energetic utterance! Surely, further comment is unnecessary on this part of his paper.

The most unfortunate part of Clement's discussion is his treatment of the present subjunctive. He has here classified his instances in a hit-or-miss way, sometimes apparently without even so much as testing a given expression to see whether it can be construed as a prohibition or not. He has included in his list of prohibitions instances which no amount of violence could distort into prohibitions. This statement may be verified by a mere glance at pages 161 and 163, for example, where instances of nec possis are repeatedly cited as prohibitions. Who ever heard of such a prohibition as 'And do not be able' (as though 'being able' were something that could be ordered or prohibited)? Similarly nec adsequare, cited (p. 163) from Tac. Ann. 6, 8, is not a prohibition and is not regarded as such by any editor or commentator. Again, cases of the subjunctive introduced by neque enim are classed by Clement as prohibitions. They have, of course, the same modal force as that illustrated in nec enim numeraverim (Cic. Brut. 47, 173), neque enim fugerim (de orat. III 38, 153), etc. Neque enim is confined to explanatory and illustrative statements, and is not used with a prohibitive subjunctive for the same reason that it is not used with the imperative mood.

Further evidence of inexcusable carelessness will be noticed in Clement's free intermixture of subordinate clauses (oro ne facias, etc.) with prohibitions proper. If Clement read my own discussion as carefully as he ought to have done before attempting to criticise me, he must have noticed that I said on pp. 135 (3) and 149 (17) of The Latin Prohibitive that clauses of the type oro ne facias were, as a matter of course, excluded from my discussion (with the exception of some four or five instances in which an accompanying imperative, the order of words, or some other consideration made it probable that the ne-clause was independent), and that I had not even attempted to collect the very numerous instances of this use. And if he did notice this, I am surprised that it did not seem to him quite unjust to me to cite against me all the numerous instances of such a usage in Silver Latin, and thus to give the impression that they belong to the phenomena

that I myself discussed. In discussing the distinction between tenses in prohibitions, the type of expression represented by oro ne facias should not enter into consideration, for the reason that one can never say that the ne-clause is not a subordinate clause. Indeed, all such clauses in Cicero (and he is full of them) have almost uniformly been regarded as subordinate. If a single one of them could be positively proved to be independent, the theory that ne with the present subjunctive in prohibitions is foreign to Ciceronian prose (except when addressed to an indefinite second person) would be dead without further discussion. It is a grammatical commonplace that, in the process of subordination, distinctions observed in independent clauses are very frequently obliterated. Such an obliteration of tense-distinctions has occurred in the type oro ne facias. Many such instances of the present in Cicero are full of emotion and involve acts that are regarded with great alarm. But what has all this to do with my distinction between tenses? The perfect tense is, as far as I am aware, quite unknown in clauses of this type, with the exception of a few cases in early Latin, where they may have been felt as quasi-independent clauses.

It may be that the levelling influences of subordination are discernible even in the cave-constructions. At any rate, cave was used in early Latin with both tenses. But before the time of Cicero, the use of the perfect tense with cave had practically disappeared. The only instances I know of after Terence are Hor. Sat. 2, 3, 38 and Curt. Ruf. 5, 2, 21.

The subordinate clauses which must be excluded from Clement's collection of instances are the following: Phaedr. App. 26, 3-4; Curt. Ruf. 6, 3, 12 ('I say this, lest, etc.'); 9, 2, 28; Il. Lat. 330; 724; Apul. Met. 22 (19, 3); 8, 8 (169, 24-26; this is of course a clause of proviso introduced, as frequently, by modo (= dummodo)); Dracont. 5, 276 ('lest'); Sil. Ital. 17, 367 (oro occasionally takes subjunctive without ut at all periods); Incert. (Baehrens 3, p. 273). Most of the remaining instances of ne and cave with the present, that are cited against me by Clement, support my theory so conspicuously that I can not account for his classification of them except by supposing that, through some oversight, they got into the wrong column. Certainly no one could seriously regard the acts prohibited by them as being of an alarming character that would in any way be likely to call forth vigorous utterance. That each reader may convince himself of the truth of this statement, I cite below all the instances of ne (neve) and cave with the present, adding after each reference the result that will follow a failure to comply with the prohibition. It will be seen that the result in the following instances will never be disastrous:

Pers. 3, 96: the speaker's life will or may be saved by his friend's insisting that he shall care for his illness.