when the auxiliary 'was' (speaking, &c.) is not sufficient; or by adverbs like 'constantly,'' often,'' still,' 'gradually;' generally by our loose aorist.

Cf. 7, 33—6; 25, 3, 5, 15, &c. (25) 4, 21, &c.; (26), 13.

tive.

§ 30. a. Wherever a fact is stated directly, or referred Subjuncto objectively, the indicative must be used; where it is alluded to merely as an idea of the mind, or stated indirectly as in oratio obliqua, the subjunctive. All mere conceptions then, belong to the subjunctive, and in all its uses a mental conception is implied. Aims and objects are conceptions: so also causes not realized as facts: and wishes, and conditions-(though a condition may form such an obvious fact that it is expressed as such, e. g. si lucet, lucet). Cf. 1, 20; 3, 6; 27, 8, 15; 37, 40; 46, 5.

Of the six so-called pure uses of the conjunctive Pure conjunctive. (Potential, Conditional, Concessive, Optative, Dubitative, Hortative, cf. the Primer, p. 141), five are strictly elliptical, dependent on verbs (as fac ut, suadeo ut, &c.) suppressed: and dependent clauses are conceptions. The 'conditional pure use' (as in 19, 14; 26, 38) is of course also a conception (vellem ire) dependent on a condition often unexpressed. Cf. 19, 14; 40, 1, 4, 33; 44, 10; 53, 19.

tions.

B. The future too is strictly pure conception, and Concepetymologically is closely connected with the subjunctive; both the future and fut. perf. indic. are often almost identical in form with the pres. and perf. subj. and might almost as well be classed with the subj. tenses, which they often replace, or are replaced by; or if not, we may say that the future is and may be stated as a fact, as the past, though belonging to the region of fact, may be treated as a conception or hypothesis; though, even then, the ideality or uncertainty of the future often leads to the use of the subjunctive, the certainty of the past even in hypothesis to that of the indicative; e.g. si velis for si

Conjunctions used with either mood.

voles; manebo donec redeat (redierit); ausim; haud facile dixerim; fecit si potuit, cf. 27, 14.

7. Frequency (with temporal conjunctions, or si, or relatives), as an indefinite conception, may reasonably take the subjunctive, and does so generally in Livy and Tacitus. See Madvig, § 359.

Cf. 9, 5; 16, 33; 17, 13; 21, 10; 22, 25.

8. Conjunctions have no inborn predilection for indicative or subjunctive. We shall find that most may be used with either, and there are good reasons for the exceptions. We must not take then for our guide arbitrary rules, that they rejoice in this or that, but examine the idea of the sentence and see whether it is a conception or fact stated. The facts about their usage may be summarized as follows: reasons for the usage are added below:

1 Ne is used to negative final, imperative, optative, and some conditional clauses; non all other clauses.

i. Cause, as a statement of fact, is generally expressed by the Moods used indicative; causal conjunctions then are mostly found with the with causal conjuncindicative--compounds of quidem (which means in fact') nearly tions. always; but cause in your own mind or the mind of another is conception, not fact, and takes the subjunctive with quum (a kind of temporal conception implying sequence and so cause) or with any of the causal conjunctions given above. Cf. 7, 7, 15, 41; 22, 3, &c.

ii. All temporal conjunctions are found both with indicative with temand subjunctive. Simple juxtaposition of facts co-ordinately ar- poral, ranged requires the indicative; therefore use quum, postquam with the indicative where you wish to state two facts in some temporal connexion; but as soon as you get to pure conception (as often of a fact never realized e. g. priusquam rex veniret abiit) or inchoate and incomplete actions, these particles take a subjunctive, though postquam (from its meaning) very rarely does.

Cf. 3, 6; 7, 33, 34; 9, 24; 16, 14; 23, 7, 35, &c.

=

N.B. Notice that dum is found mostly with the pres. indic. even in oblique narration (as vividly descriptive), and ante quam, priusquam (if with the indicative), postquam, with the perfect rather than pluperfect; e.g. postquam venit after he had come; the point of time to mark (after which the other event happened) is venit not venerat. Where we wish to mark strongly a previously completed fact, and not merely to use it as a point of time, the pluperfect is found e. g. postquam occiderat; after he had first killed. ' Cf. 24, 40; 47, 2, 7; 51, 16.

6

Quum however (as marking chronological sequence with causal connexion more or less implied, or as stating facts allusively and not directly) takes the subjunctive in connexion with historic tenses.

iii. An aim must be a conception, and these particles are only with final, found with subjunctives. Cf. 3, 6; 9, 4, &c.

iv. Concessive, as conceding either a fact or a hypothesis, will with contake accordingly indicatives or subjunctives. Elliptical forms like cessive, licet (ut)-(fac) ut, really introduce dependent sentences which come under another rule, and are only apparent exceptions. Quamvis and quantumvis are, strictly, not conjunctions but adverbs qualifying an elliptical dependent clause, e.g. quamvis sit, &c. be it as much as you like, &c. Cf. 29, 19; 47, 9.

In Tacitus quamvis is usually found with indicative, quamquam with subjunctive, as also sometimes in Livy.

V. You may take either a certain fact or a conception as a with conditional, condition. These particles therefore are found with both moods.

Cf. 37, 21, 24, 39; 52, 6, 24, 28.

with consecutive,

with comparative.

Tenses.

vi. Consecutive clauses are found invariably in the subjunctive, as merely qualifying preceding statements and not stating (otherwise than allusively) a fresh fact. (The Greeks often seem to treat their wore as a mere copula itaque.) Cf. 7, 20; 20, 4.

vii. Comparison also takes indicative or subjunctive accordingly as you compare conceptions or facts.

Cf. 7, 50; 28, 3; 39, 27.

Subjunctives, preceded by ut, often stand alone parenthetically, or as subjects apparently of fit, abest, accidit, restat, &c. (Madv. §373), where a substantival infinitive could often be used. They doubtless are, or were originally, either final or consecutive classes and take for their negative ne or non accordingly: so that the negative may be used as a test; e. g. ne plura dicam, restat ut ne taceam, tantum abest ut non taceam, &c. Cf. 15, 9; 27, 14; 37, 40.

Subjunctive E. The tenses in the subjunctive follow the rules laid down in § 27, § 29, but differ slightly from the indicative.

They are sometimes used in a future sense, or where you would expect futures, perhaps from the connexion' in character and etymology of this mood with the future; e.g. metuo ne veniat; gratulerne tibi an timeam? dubito an faciam. Cf. 17, 11; 19, 11; 37, 60; 38, 4, 8, 14.

Even the past tenses (imperf. and pluperf.) seem to have a future sense in wishes and conditions, but they always imply at the same time something past and impossible. An act of the past, existing only in hypothesis, is hopelessly unreal', as the past is unalterable; e.g. si venisset, utinam adisset, veniret si posset. In final sentences (haec monui ut veniret) the past is used after the past because if the action is past, the preceding aim of it must be also, even though future with respect to the action. Cf. Madv. § 378.

:

1 The pres. and perf. subj. (as the Greek subj. throughout), in form are like the indic. future, and are used of future, probable, or possible conceptions: the imperf. and pluperf. of past, improbable, or impossible (like the Greek optat.), being more akin to the past tenses of the indicative.

conditional

. In oratio obliqua the pres. subj. is constantly used Dependent for the future: scribet si quid habebit frequently becomes futures. dicit se scripturum si quid habeat; but where it is necessary to express at once the future or conditional as well as the conceptive or dependent meaning of a clause, the subjunctive cannot do double duty without leading to ambiguity, and the future or conditional meaning has to be marked more fully and distinctly; e. g. (scribat) scriberet si quid (habeat) haberet, but accedit ut scripturus (sit) esset si quid (habeat) haberet: not ut scriberet, which would mean 'that he wrote.' The protasis it will be noticed remains unchanged. Cf. 40, 23.

Tenses.

7. In the sequence of subordinate clauses the present Sequence of and perfect of a main clause are followed by the present and perfect subj., the imperfect or the aoristic past-perfect by the imperfect and pluperfect subj. In these clauses the imperfect of the subjunctive is also aoristic in sense: while the perfect is less so than in its indicative (being used mostly of completed perfect actions); though in consecutive clauses, aoristic also.

e. g. i. ut veniat dat (dedit) talentum.

ii.

ut veniret (or quum venisset), dabat (dederat) dedit (aorist) talentum.

iii. tam stultus erat ut veniret (consecutive), or

ut venerit, of one completed or definite action.

Cf. 3, 10; 6, 7; 20, 4, 8, 36, 41; 41, 22.

Subjunc

0. The futurum exactum subjunctive, both in the Future active and passive, seems to have the same form as the tive. perfect subjunctive, not that of the futurum exactum indicative; e.g. tam segnis est ut jam redierim1ante

1 This seems in accordance with Madvig's rule, but surely it would have to be ut futurum sit ut redierim to avoid ambiguity wherever the context would allow of redierim being taken as an absolute perfect.