AMERICAN

JOURNAL OF PHILOLOGY

VOL. XXIV, 1.

WHOLE NO. 93.

I.-LIMITATION OF TIME BY MEANS OF CASES IN EPIC SANSKRIT.

In JAOS. xxiii, p. 150, I have given an instance of a predicate nominative taking the place of a space-accusative: yojanānāṁ sahasrāņi caturafitir ucchritaḥ, (the mountain is) "eighty-four (nom.) thousands of leagues high" (nom.). So in time, although the accusative is the norm, the nominative serves, but in parataxis, to indicate duration. Thus iii. 296. 26: saṁvatsaraḥ kiṁcid ūno na nişkrāntā 'ham åçramāt, "a year (nom.) almost, I have not gone out from the hermitage". Usually, when an event is described as subsequent, such a clause is filled out with its verb, and atha follows: tato na 'timahān kālaḥ samatita iva 'bhaval, atha, etc., "a short time passed, then ", etc., i. 39. 3; māsajātas pita bhavati atha jagmuḥ, “thy father becomes a month old, then came they", xiv. 70. 13. But there is another construction, which has been credited with the implication of preterite time. It is common enough in the epic, though our Sanskrit Syntax illustrates the case but scantily. This is the nominative with a dependent genitive. The examples following, showing that the time does not depend on the case, may introduce the general question as to the kind of time indicated by

THE GENITIVE.

The examples are: şaşṭho hi divasas tedya praptasye 'ha, "it is to-day the sixth day (since) you got here"; saptāṣṭa divasās tv adya viprasye 'ha 'gatasya vai, "a week (since) the priest came hither", xii. 359. 5; 360. 13; sagraḥ samvatsaro jātaḥ tava payataḥ, "a whole year has been born (while) you (were)

looking on", xiii. 19. 48; tasya vahyataḥ kalo munimukhyān abhyavartata, "a long time elapsed (while) he (was) making the saints his bearers", xiii. 99. 12; catvāriñçad ahány adya dve ca me niḥsṛtasya vai, (it is) "now forty-two days (since) I set out", ix. 34. 6; evaṁ vimṛçatas tasya dirghaḥ kālo vyatikrāntaḥ, a long time passed (while) he was ", etc., xii. 267. 44; kadācid bhramamāṇasya hastinaḥ... kalo 'gacchan niçāniçam, “time passed night by night (while) the elephant (was) wandering", xii. 117. 6; kālo mahāns1 tv atīto me çūrasūnum apaçyataḥ, "a great while has passed since I saw" (without seeing), xiv. 50. 20. Compare mṛtakasya tṛtīyāhe and mṛtasya daçarātreṇa, in xiii. 136. 15 ff.

2

The so-called "genitive of time after" (the only meaning given by Professor Speyer, Syntax, § 128) is confined exclusively to indefinite phrases and (as is usually said) may have been developed from the construction trir abdasya, "thrice a year," etc. But it does not really, so far as I have observed, connote time after, but time in which, interchanging with the "within" of the ablative (in this sense, see below) and instrumental, dirghasya kalasya = dirgheņa kālena. The epic phrases are few and stereotyped, cirasya, kasya cit kalasya, and (accompanied with the word for "then") tataḥ katipayahasya, i. 8. 17; xiv. 71. IO, "then within a few days".

:

Of course, in the ptc. construction (above), a preterite idea lies neither in the ptc. (as it is also present), nor in the genitive, which is either absolute or loosely dependent, e. g. "forty-two days (of) me gone", or "me being gone". At any rate, "since" is only a convenience suited to our idiom and does not imply that the genitive gives time after. A strong argument against "after" is that we never find a case where definite periods of after-time are so construed, but only a vague time-notion; e. g. 'after three days' is never expressed by trayāṇāṁ divasānām. On the contrary, a definite time in the genitive never means "after" but "within".

This is the right form. In xiv. 56. 18, mahākālaḥ must be changed to mahān, as in C.

"This may be expressed (but not often) by cira: mătulaç ciradṛṣṭo me tvayā, "it is long since you saw my uncle", xiv. 53. 45. The idea here is really "long ago", as in svargato ‘pi pită ciraṁ tava, "died long ago", i. 157. 28. But with a ptc. this cira (cirarātra) usually means for a long time, cirasupta, long asleep, cirarātroṣita, i. 168. 3.

In this instance, as in the case of the ablative, to get the real meaning we must not start with a theory of the genitive, then make it apply in examples so indefinite as to explain nothing (for whether we say "in" or "after" a long time is indifferent), and finally force this meaning upon the definite cases or regard them as exceptions; but start with the definite time-words, establish their required genitival significance, then apply this to the indefinite time-words, and if it suits these also, construct a theory. Thus we may take the definite time indicated by a fortnight. There are two in a month and the second is called the "other wing" of the month. Now in the ritualistic rule of Āp. ii. 7. 16. 4, there is something mentioned which must be done every month in the "other wing" of the month, and this is expressed by māsi māsi kāryam aparapakṣasya, and it is quite impossible to translate it otherwise than by "month by month in the other wing this is to be done" (not "after"). So in Manu xi. 260 and iii. 281, trir ahnaḥ and abdasya is "three times in a day, in a year", and though we may dispose of this as a partitive genitive, we cannot do so when "year" stands alone in the genitive. Thus in Manu v. 21, samvatsarasyāi 'kam api must mean "at least one in a year"; and in Manu vii. 137, varşasya means "in a year" or yearly (not "after a year", as given by Professor Speyer, Ved. u. Sk. Syntax, § 65). So when we find in the epic, i. 152. 8, upapannaç cirasya 'dya bhakṣaḥ, we should render it "food furnished now in a long time", rather than "after"; and sudirghasya 'pi kālasya, Manu viii. 216, etc., "in a very long time", however well, in these general time-words, "after" will suit our idiom. The genitival relation is really adjectival, samvatsarasya = samvatsariņa; samvatsarābhiçasta, "a year-accused" is "accused within a year", Manu viii. 373. Parallel stands antar with the genitive of space, sarvabhūtānām antaç carati, “permeates all", iii. 76. 34 (tasyā 'nu, “after”, is incorrect).

When this genitive is formally identical with the locative, it is impossible to tell which is intended. Thus in Manu ix. 104, jwvatoḥ means "in the life-time of the two" (parents), and may be either genitive or locative, though probably the former.

Again, time within, or during, which is the only time-meaning implied when other words are used, as, for example, when one says "he shall not study while his hands are wet", prodakayoḥ panyoḥ, Āp. i. 3. 10. 25.

Finally, in housvastos, the meaning is not "after" but "in, within", the dawn-time; and Toλλ érv should also be taken in the same way, "in many years," that is, in the course of many years, rather than "after". Compare the Avest. gen. "in the night, day," etc. So nahts, in Gothic, etc. In emphasizing in and within as the force of the temporal genitive, I wish merely to protest against the meaning "after". The fundamental meaning seems to me to be rather the loose adjectival relation usually implied by the genitive, best expressed in English by "of". Old English offers good parallels, as well as our colloquial "of old", "of a morning". Compare (cited in Cent. Dict.) "earn ten shillings of a night" (Mayhew); "Sir, I moste go, and of long tyme ye shul not se me ageyn", (Merlin, E. E. T. S.); "not wink of all the day" (Shakespeare).

THE LOCAtive.

As in tasminn ahani, “on that day", aparedyuḥ, “on the next day" (i. 63. 20; iii. 65. 35); samaye, "in good time", iii. 192. 38, the locative gives the time "at or (with) in". The notion "after" lies not in the case but in the idiom of translation. In xii. 122. 16, "after a thousand years it fell" is the natural English of "on the thousand years completed it fell" (as he sneezed, ksuvato 'patat; after the words "he carried it many years", acc.). The "after" idea is formally expressed when required, tato 'stame tu divase, "on the eighth day from this", i. 129. 20; tataḥ samvatsarasyā 'nte (kāle), “at the end of a year from this", i. 139. 1; 167. 4; pūrņe samvatsare tataḥ, xiii. 111. 70; or, as above, simple completion is expressed, ib. 136. 16, dvādaçãhe vyatite, "on the twelfth day completed". So sthite Parthe (ajagmuḥ), xviii. 3. 1, "on (or at) his standing". So xv. 1. 6; 3. 12 and 34, etc. In kale bahutithe, iii. 65. 2; bahutithe ‘hani, i. 108. 2; kāle bahutithe vyatikrānte kadācana, "once on some time being passed," iii. 296. 1; dvimasoparame käle vyātīte, "at the expiration of two months' time," xii. 282. 26; xii. 360. 1, etc., time at and not (as generally rendered) time after is expressed. In iii. 61. 12, PW. renders this "viele Tage hindurch", although the verb is "saw", and the action immediate; the meaning being "in (within) considerable time he saw some birds." In i. 173. 14 and 31, the same event is narrated, sthite tasmin... jagāma viprarşis tada dvādaçame 'hani, "on his standing there (night and day), then on the twelfth day came

the priest"; and again, dvādaçarātre . . . samāhite, (he came) "on the twelfth day completed". So ib. 45, tasmin pravişte pravavarşa, “it rained on his returning". The locative absolute connotes the same time as does the simple locative.

The asterism at (under) which is usually locative; but with Puşya, Tiṣya, regularly, and with other asterisms occasionally, Hasta, Abbijit, the instrumental is used.

Puşyeṇa samprayāto

'smi Cravane punar āgataḥ, ix. 34. 6, is a typical case.

The durative sense, though found, as in ÇB. vi. 1. 3. 20, “recite during a year or two", samvatsare, dvayor (vā), and ib. xi. 4. 2. 19-20 (where "during a year" is expressed by both acc. and loc.), is often injected into a locative because the accompanying verb seems to require it. Thus in i. 3. 35, "on going to the teacher's house at the end of the day, divasakṣaye, after guarding the cows in (during) the day," ahani gā rakṣitvā (ahni, i. 185. 29), like ekāhnā, "in the course of one day." So in i. 148. 17, "they stood guarding the house in (during) the night", rātrāu. In the repeated locative, māsi māsi çatam samāḥ, "month by month, a hundred summers", i. 79. 6, the pure locative sense is not lost, though the concurrent construction, anudinam, "day by day," i. 185. 15, equal to divase divase, is durative. The translation "during" is often given for the meaning "at some time in the course of". Thus, strictly speaking, though too pedantic for practical use, when in Manu iii. 28, yajñe tu vitate sutādānam is translated (the gift of a daughter) "during the course of the sacrifice", the translation is incorrect; for the gift is not given "during" the sacrifice but at some one time in the course of the sacrifice, and "at the sacrifice begun" is the literal meaning. In the subsequent stanza, iii. 108, väiçvadeve tu nirvṛtte, the translation "after the offering has been finished" puts the meaning of the verb into the case. It is not quite wrong, but it is not quite right; since, as far as the time goes, it means "on the completion". The compass of the locative is formally expressed by antar, antardaçähe, "within ten days", M. v. 79, etc.

It is customary, since Gaedicke, to say that the locative means "after", when it marks the culminating time and (Delbrück, Syntax, p. 225) to compare the Sanskrit use with évavre, “after a year", woλ xpóv, "after a long time ", etc. But in reality the "in" sense not only suffices but is the real meaning in all these