both for active and neuter verbs, as bolâ "he said," kiya "he did." In the medieval poets, however, and to this day in the rustic dialects of Oudh and the eastern Hindi area, there exists a preterite with terminations retaining traces of the incorporation of the old substantive verb. Before these terminations the long â and of the p.p. masculine and feminine are shortened, and the vowel of the masculine is often replaced by e. Thus we have (mâr "strike”)—

Also in m., etc. In the sing. 2, 3, the syllable få is often added, as मारेसि, and variated into हि, as मारेहि m., मारिसि, मारिहि f. Thus चलेउ हरषि हिय धरि रघुनाथा “ he went rejoicing, holding in his heart Raghunâtha" (Tulsi Das, Ram. Sund-k. 4), देखेउं नयन राम कर दूता “I have seen with my eyes the messenger of Ram" (ib. 12). Tulsi does not observe the gender very closely, if at all, — पुनि परीहरेउ सुखाने उ पर्णी ॥ "Again she gave up even dry leaves" (Bal-k. 155), qofe पूछेसि ana ale cele "She asked the people, why is this reलोगन काह उछाह ॥ joicing ?" (Ay-k. 87). But the feminine is kept in fuenfafa सिष दीन्हिउ तोहि ॥ “ The flatteress has given instruction to (has prompted) thee" (Ay-k. 101). The type ending in si, though used for both 2 and 3 sing., more strictly belongs, I think, to 2 sing. from Skr. asi; but in this tense the traces of the substantive verb are so much abraded that it is difficult to speak with certainty about them. The following handful of instances, taken at hazard from one page of the Sundarakânda of Tulsi's work, will show the various senses in which this affix is used : खाएसि फल अरु बिटप उपारे “He eats the fruit, and tears up the bushes" (S-k. 40), dg mîfa dig affa कछक मिलायेसि धुरी ॥ " Some he slew, some he trampled under foot, some he caused to mix with the dust,” कह लंकेश कवन तें

कीसा । केहि के बल घालेसि वन खीसा । कीधौ श्रवण सुनेसि नहिं मोहि । मारेसि निशिचर केहि अयराधा । " Saith the lord of Lanka, who art thou, and what? By whose strength hast thou torn to pieces the forest, hast thou never heard of my fame, . . . for what fault hast thou killed the demons?" (ib.) Panjabi throws no light on the subject, as it does not use this form, but employs the participle simply as a tense, as maî, tû, uh mâriâ, "I, thou, he, smote." Indeed, to such an extent in H. and P. has this custom of using the bare participle as a preterite tense prevailed, that it cannot now be used in any other sense, and if we wish to say "smitten," we must not use H. mârâ or P. mâriâ alone, but must add the participle of the modern substantive verb, and say H. mârâ hûâ, P. mâriâ hoïâ. The only trace in P. of the old substantive verb is to be found in a dialectic form which I have often heard, though it does not seem to be used in writing, as atat "he did," which is कीतोस probably to be referred to S. aatsfa. The grammarians,

however, suppose that kitos is in some way a metathesis of us ne kita, so that kita+us kitos. The instrumental, however, of uh "he," is not us ne, but un; us ne is Hindi, and would hardly have been resorted to in the formation of a pure dialectic type like this. Moreover, in the 1 plural we have such expressions as tet, which is evidently khânde+'sâ, for asâ=asmâh.

Different from modern, but strikingly similar to medieval, Hindi in this respect is Sindhi, which does not employ the participle singly as a preterite, but, except in the 3 sing. and pl., has relics of the substantive verb incorporated with it, thus (hal "go")—

By comparing these terminations with those of the S. future,

which is based upon the present participle (§ 41), it will be seen that they are absolutely identical, thus:

halandu-si corresponds to haliu -si.

and the terminations may, in the case of the preterite, therefore, as well as in that of the future, be referred to the old Skr. verb as in various degrees of decay.

Marathi exhibits the same analogy between the present and the preterite; to its p.p. in m., ƒ., n., etc., it adds the same terminations as to the present p. in π m., ☎ƒ., à n., etc.

The forms exactly agree with those of the present, as will be seen by turning to § 42. There is no conditional as in the present. When it desires to use this form in an adjectival sense, M., having apparently forgotten its originally participial nature, adds another एला, thus we get मेलेले जनावर “a dead animal," "a made (i.e. experienced) man." The fact so well established for S. and M. may help us to understand, if we cannot fully explain, the preterites of O. and B., From a p. p. देखिल, 0.

which are formed in the same way.

constructs a preterite, thus

where the terminations correspond exactly with those of the conditional, which is similarly formed from the present participle.

Bengali does the same (pace the Pandits), as

with a final, as in the

Here the 3 sing. has also imperative and future, concerning which see § 53. The 1 sing. in nu is frequently heard in speaking, and is very common in the old poets, as अपरूप पेखनु रामा “I saw the fair one looking woe-begone" (Bidyapati, vii. 1), where some read fag.

In Gujarati the participle is used alone as a preterite in both forms, that in yo and that in elo, but more frequently a modern substantive verb is added for greater clearness. This language has no traces of the old incorporated Sanskrit as.

§ 50. In the past tenses of all but B. and O. the prayogas or constructions mentioned in Vol. II. p. 264, are employed. In most of the languages, indeed, their use is restricted to the past tenses. The direct or kartâ prayoga is used with neuter verbs, and requires the subject to be in the nominative case, while the participle, which does duty for a preterite, changes with the gender of the speaker. Thus

And so through all the persons except 1 and 2 plural, where no

distinction of gender is necessary, as the speaker is known. In the active verb, however, the karma or objective construction is used, where the subject is put in the instrumental, the verb agreeing in number and gender with the object. Thus, H. मैं ने तुम से बड़े कठोर वचन कहे हैं “ I have spoken very harsh words to you" (Sak. 33). Here the subject mai ne is in the instrumental, the verb kahe hai is masc. plural, to agree with the object vachan.a à âu tì sîìu fær feet (ib. 39) “Destiny has joined just such a joining" (has brought about such a marriage).

So also in M. the p.p. is declined for all three genders in both numbers so as to agree with the object, as aft "he read the book," where vâchili is fem. sing., to agree with pothi. In M. and S. many verbs are both active and neuter, in which case the preterite has a double construction, direct or kartâ when the verb is used as a neuter, objective or karma when it is used as an active. So also in G. The distinction appertains to syntax, and not to formlore, and need not be more than mentioned here.

There is also a third or impersonal construction technically known as bhava, in which the object is not expressed, and the verb, therefore, remains always in the neuter. In M., however, this construction is used even when the object is expressed, as eiði en anfâ "he beat him," literally "by him to him beaten."

§ 51. The participle of the future passive, which in Sanskrit ends in, plays an important part in the modern verb in some languages. It does not, like the two previously noticed participles, form modern participles, but rather various kinds of verbal nouns, such as in Latin grammar we are familiar with under the names of gerunds and supines, also the infinitive. The Latin gerund itself is, however, closely allied to the participle of the future passive, for amandi, amando, amandum, are