OE., is clear from these examples taken from Ælfric's Grammar. He translates the Latin sentence: Uideo te docturum esse, ic geseo, þæt du wylt tæcan1. So, again: docturus sum cras pueros, ic wylle tæcan to merigen þam cildum2. This example will illustrate the use of sculan as a tense-auxiliary: LS. 2. 28. 406 pa ic sceolde in on pa dura gangen, þa ongunnan hi butan ælcere lættinge ingangan.

The meaning of motan, and its difference from mugan, is very well seen in examples of this kind, which are frequently met with, especially in BIH. and Wulf.: BIH. 95. 25 Forpon we sceolon nu geþencean, þa hwile þe we magon & motan, ure saula þearfe, pe læs we foryldon þas alyfdon tid, & ponne willon ponne we ne magon; Wulf. 27. 5 ac do nu manna gehwylc, swa him mycel þearf is, geswice yfeles and bete his misdæda þa hwile, pe he mage and mote; ÆH. 1. 268. 31 gif he ær geswican nolde, þaþe he mihte and moste. Once we find the conjunction omitted, probably a mere error of the scribe, and the auxiliaries stand together: BIH. 115. 20 Uton we ponne þæs gepencean, þa hwile pe we magon moton, þæt we us georne to gode þydon.

By way of summary, then, the prevailing mode in the OE. temporal clause is the indicative, save in clauses with connectives rendered by Modern English before.

When we find the optative in clauses other than these, it is due to an imperative in the main clause, or to its being in an indirect question or in an objectclause, or more rarely to the general indefinite character of the sentence. The indicative with connectives of the ar-class occurs only in clauses of fact in past time, and is due to this circumstance. The

1

150. 18.

2

152. 10.

auxiliaries mugan, sculan, motan, and willan in temporal clauses, almost without exception, have their full independent meaning, though in the case of willan and sculan there is already a noticeable tendency toward the later use as tense-auxiliaries.

CHAPTER III.

POSITION OF THE CLAUSE AND WORD
ORDER.

The position of the temporal clause in OE. is very free. Indeed it may occupy any one of the three possible positions: that is to say, it may precede the main clause; it may follow it; or the temporal clause may stand between different members of the main clause.

This interposition of the temporal clause is met with less frequently than either of the other positions, and a clause so placed is likely to be of a parenthetical nature. This is true of all the clauses except those with oo (det), which always follow their main clause.

I have nothing to add, of a general nature, in regard to the order of words to the results set forth by Dr. C. A. Smith,1 but I have noticed that in all the clauses introduced by swa followed by a superlative, hraðost, oftost, lengest, the subject, and nothing else, is always placed between swa and the superlative.

My study has led me to believe that this principle, laid down in Bosworth-Toller, is much more freely violated than is usually supposed: 'When the word 1 Order of Words in Anglo-Saxon Prose.

da stands at the beginning of a clause, and may be translated by then, the verb generally precedes its subject; if it is to be translated by when, the subject generally precedes the verb.'

SEQUENCE OF TENSES.

As in all other languages, the general principle is the familiar one of Latin Grammar-principal tenses depend upon principal tenses, and historical upon historical.

But this fact is to be explained by the logical relation of the action of the two clauses, and not by any influence of the verb of the main clause on that of the dependent clause. Indeed, Professor Hale 1 showed that this is true for Latin, for which the doctrine of the sequence of tenses has been taught religiously for centuries. We are not surprised to note, then, occasional violations of sequence in OE. I quote only one example to illustrate: O. 220. 10 Þonne hie from gesælgum tidum gilpað; þonne wœron þa him selfum þa ungesælgestan.

NEGATIVE.

The negative of the temporal clause is ne or na, or both.

1 The Sequence of Tenses in Latin (Am. Jour. Phil. 7. 8).

CONCLUSION.

In closing this study, it may be desirable to recapitulate some of the most obvious results arrived at:

1. The surprising number and variety of the connectives used to introduce the temporal clause.

2. The fact that da, the most common of the connectives, is used only with the preterite tense of the indicative mode.

3. The fact that certain connectives, or groups of connectives, are found, for the most part, only in a definite group of writings.

4. That Ælfric seldom omits de from the conjunctional formulæ.

5. That the syntax of the temporal clause is essentially the same throughout the OE. period, save that different writers, in the same period as well as in different periods, use different connectives.

6. That ar and its variants have the indicative in about one-fourth of the cases in which they occur.

7. That or, when used with the indicative, usually occurs in narratives of fact in past time, and that the mode has the effect of emphasizing the actual occurrence of the action of the clause.

8. That the indicative is the prevailing mode in all clauses, save in the ar-clauses.

9. That the occurrence of the optative in clauses other than those of the ar-type is usually due to peculiarity of the main clause (the presence of an imperative, &c.), and that hypothetical or indefinite character in some measure leads to the use of the optative in such cases.1

1 Cook, A First Book in Old English, p. 104.

10. That the use of the optative does not decrease perceptibly in the later period of OE., so far as the temporal clause is concerned.

11. That the so-called modal auxiliaries mugan, sculan, motan, willan have their full verbal content, and are not used as a mere paraphrase for the optative.

12. That these auxiliary verbs take the optative under precisely the same circumstances that ordinary verbs do.

13. That the use of the auxiliaries does not increase noticeably in the later period of OE.

14. That sculan and willan, though preserving their full independent meaning, show a tendency to pass over into the use of tense auxiliaries.

15. That the position of the temporal clause is free, save that the oỡ (dat)-clause always follows the main clause.

m