method of adding the terminations.

language. A common verbal stem in each is given to exhibit the $ 37. Here follows a table showing the simple tenses in each

§38. The simple tenses in the Gipsy verb, as given by Paspati and Miklosich, differ very widely from the Indian type, and it is difficult to grasp their forms, so much have contraction and a slurring habit of pronunciation weakened the original terminations. The present among the Rumelian gipsies has the following endings: S. 1. a, 2. sa, s, 3. 7; Pl. 1. sa, s, 2. na, n, 3. na, n. Thus from keráva “to do”—

Of the two forms, those ending with a are the fuller and more correct forms, and those ending in the consonant which precedes the a are used in ordinary conversation. The S. 2 sounds also keresa, kerés. Here we distinguish two junctionvowels á and e, as ker-á-sa, ker-é-sa, a peculiarity which recalls the practice in Prakrit by which the e originally proper to the tenth conjugation is often used in verbs of the Bhû and other types, and as often omitted in causals; so we have yЯ and पुछेमि, रोचामि and रोचेमि, हारद् and हारेद्र, करद्द and करे. But with regard to the terminations, there is much difficulty; we recognize, indeed, the termination âmi of S. 1. in Paspati's áva, or âv, and asi in his âsa, or âs. So also anti, Pr. enti, reappears in ena or en. The ela, el of S. 3. may stand to ati in the same relation as the ila of O. and M. p.p.p. does to Skr. ita ; but if so, it is a strange confirmation, and from an unexpected quarter, of what is as yet little more than an unsupported hypothesis. In the P. 2. the ena, en may have been borrowed from P. 3, for we have seen similar cases in the other languages, but the P. 1, with its ending in s, is entirely inexplicable.

The Syrian gipsies have retained a fuller form of the S. 1, as jâmi "I go," âvami "I come," stûmi "I am," and the following almost pure Prakrit words, bihemi "I fear," chinemi "I cut,"

dâmi, demi “I give,” jánami, jánemi "I know," enemi "I bring " (from ), kinimi "I buy" (t), and others (Miklos. ii. 4).

The imperative is the only other simple tense, it has the forms ker "do thou," kerén, do ye," me kerel "let him do," me keren "let them do." The meaning and origin of this prefix me is not explained by Paspati, and I am not aware of anything in the Indian languages with which it can be connected. It is probably a construction borrowed from modern Greek, or Turkish, or some of the languages spoken in or near Rumelia. The imperative is, in its general form, precisely analogous to the languages of our group, but there is nothing specially noteworthy about it.

CHAPTER III.

THE PARTICIPIAL TENSES.

CONTENTS.-§ 39. DEFINITION OF THE PARTICIPIAL TENSES.-§ 40. THE PARTICIPLE ACTIVE. - § 41. TENSES FORMED THEREBY-THE SINDHI FUTure. -§ 42. MARATHI INDICATIVE AND CONDITIONAL PRESENT.-§ 43. BENGALI AND ORIYA CONDITIONAL.-§ 44. HINDI, PANJABI, AND GUJARATI PRESENT.§ 45. THE PAST PARTICIPLE PASSIVE.-§ 46. EARLY TADBHAVA PARTICIPLES IN SINDHI AND PANJABI.—§ 47. THE SAME IN GUJARATI AND MARATHI.— § 48. THE SAME IN OLD AND NEW HINDI.-§ 49. TENSES FORMED FROM THE PAST PARTICIPLE.—§ 50. THE PRAYOGAS.-§ 51. THE FUTURE PARTICIPLE PASSIVE.-§ 52. TENSES FORMED FROM IT IN SINDHI, GUJARATI, AND MARATHI.—§ 53. THE FUTURE IN ORIYA, BENGALI, AND EASTERN HINDI.§ 54. THE HINDI AND PANJABI FUTURE.-§ 55. MARATHI FUTURE COMPARED WITH THAT IN CERTAIN HINDI DIALECTS.-§ 56. SYNOPSIS OF THE PARTICIPIAL TENSES IN ALL SEVEN LANGUAGES.-§ 57. PARTICIPIAL TENSES IN THE GIPSY VERB.

§ 39. So widely has the modern verb diverged from its parent, that the simple tenses, in which there still remain traces of the ancient synthetic structure, are, as we have just seen, extremely few. Far more numerous in all the languages are those tenses which are formed by the aid of a participle derived directly from the Prakrit. These tenses may be divided into two classes, (1) consisting either of a participle alone, as in H. chaltâ "he moves," which is really "moving (he is)," or of a participle, to which are attached much-worn fragments of the old Sanskrit substantive verb, as in M. hasatos "thou laughest," which is really "laughing thou art," Pr. hasanto 'si (whether the remnant of the substantive verb still appear, or whether it have entirely dropped out, in either case the principle underlying the formation is the same, and words

like H. chalta, and M. hasatos, belong, therefore, to the same category): (2) consisting of a participle, to which is subjoined a substantive verb, the two words standing separate, but forming one phrase, as in H. dekhta hai "he sees," i.e. "he is seeing," M. lihat âhe "he is writing."

Between these two classes there is this fundamental difference, that in the former the traces of the substantive verb which do exist are still in the Prakrit stage of development, whereas in the latter the substantive verb, which is combined with the participle, is not in the Prakrit shape, but is a later form, evolved by the languages out of the Prakrit.

The first of these two classes I propose to call "participial tenses," and they will be treated of in this chapter; the second I shall call, following the example of the grammarians, "compound tenses," and shall reserve their discussion for another chapter.

The participle used in the formation of tenses may be traced back to the Prakrit equivalents of the following Sanskrit participles.

1. The present Active (Parasmai.), as in पचन् m., पचन्ती /, पचत् ”. 2. The past Passive (with inserted ), as in E m., gal ƒ., då n. (Pr. करिओ etc.).

3. The future participle Passive or verbal adjective, as in a m., • तव्या, °तव्यं ".

To these must be added certain much abraded forms of special past participles, which are used in a peculiar way in three of the languages, as will be shown hereafter, and it must be borne in mind that, especially in the case of the past participle passive (noted as p.p.p.), it is the Prakrit forms that are to be looked to, rather than those which occur in classical Sanskrit. The classical language does not prefer to insert the intermediate in the p.p.p., but the popular languages do prefer it to a very great extent, so much so, that it has almost