person singular and plural. The first person, however, also undergoes changes. Thus, in the present participle used as a future, इंदुसि “I shall be,” m., becomes इंदोसां, and इंदिअसि “I shall be,” f., becomes. So that we get forms—

m.

इंदोसांद् “I shall be to thee,” f. इंदीसांव “ I shall be to you,”

इंदोसांसि “ I shall be to him.” इंदीसांनि “I shall be to them.”

66

So, also, the plurals इंदासों m, and इंदिजंसों we shall be," become respectively हंदासूं and इंदिऊंसूं . The second person remains unchanged, merely affixing the personal suffixes. In the third person m. इंदो is shortened to इंदु, and f. हंदी to इंब to, ga or इंद ; pl. m. इंदा becomes इंद, except with the suffix of the first person plural, as “they shall be to us," but हंदासूं इंदव "they shall be to you;" pl. f. remains unchanged.

The past participle used as a perfect tense undergoes analogous changes. Thus

1Sing.”. होसि “I was,” becomes होस, as होसांदू “I was to thee.”

66 we were to them."

1Pl. m.हुआसीं “we were,",, हुआसूं,,, हुआसूंनि

The second and third persons remain almost unchanged. In active verbs, however, where only the 3 sing. is used, owing to the objective construction, a somewhat different system prevails. The subject, which in other languages is put in the instrumental, may in S. be indicated by a suffix, and the object being also shown by a suffix, it arises that the verb may have two suffixes at the same time. Thus "I forsook thee," would be in H. aga a de, lit. "by me thee forsaken," where the subject is in the instrumental, and the object in the accusative, case, the verb (i.e. participle) being left in the masc. singular, because there is no neuter in H. In M., which has a neuter,

the Bháva or impersonal construction is used, as म्यां तुला सोडिलें "by me to thee released," as though it were a me tibi relictum (est). Sindhi expresses this sentence by one word छडिओमांद, i.e. chhadio-mân-i="forsaken-by me-thee." Thus there arises a long string of forms for every possible combination of the agent and the object. A few may be given as examples; a full range will be found by those who desire to pursue the question further in Trumpp (p. 371):

छडिओोमांसि “ I have forsaken him.”

छडिआईसि “ he has forsaken him.”

विहारिआईसि "he caused him to sit.”

fauicifa "they said to her."

faifa" she said to them.”

The suffixes denoting the agent are sing. and pl., which Trumpp considers to be shortened from "by him," instr. of "this," and "by them," instr. pl. of ☎ “that,”

respectively.

A curious proof of the antiquity of these complicated forms with suffixes is afforded by the fact, that in connection with them the 3 sing. aorist of appears still in its old Pr. form of (f), § 59. This form exists only in combination with the pronominal suffixes, whereas the ordinary form is used both with and without suffixes. Thus they say आहेमि and अथमि “ there is to me,” as in the line—

गुझी अथमि गाल्हडी आउ ओहे तां ओरीआं ॥

"There is to me a secret matter, come near, then I will tell it." -Trumpp, p. 350.

It is used just as in Latin "est mihi," in the sense of "I have,” as अथमि “I have,” अथी (for अथइ) “ thou hast,” अथसि “ he has,” अधूं (for अथऊं) “we have,” अथव “ye have,” अथनि they have."

66

It is incorrect to say with Trumpp (loc. cit.)

that atha has in S. "been transferred to the plural." The verb remains in the 3 sing. throughout, and takes suffixes of both numbers and all three persons.

In the compound tenses the suffixes are attached to the auxiliaries, leaving the principal verb unchanged. Both single and double suffixes are used in this way, just as with the simple and participial tenses. Thus किओ होर “thou hadst made,” जंहिं खिदमतगार (K) खे अगे सेखारिओ होईं (for हुआईं) तहिं "He looked towards that servant whom he had previously instructed," literally, "Which servant previously by him instructed, to that (one) by him looked" (Trumpp, p. 379).

It is tempting to look for the origin of this habit of using suffixes to the Semitic languages, which, from the early conquests of the Arabs in Persia and Sindh, may have had an influence upon the speech of those countries. On the other hand, however, the presence of a precisely similar habit in Italian and Spanish, seems to show that there exists a tendency to such constructions even in the Aryan family; for I suppose that even if we see in the Spanish forms a trace of Arab influence, no such motive power can be argued for any part of Italy, unless it be Sicily.

In Italian there are separate forms for the suffixed personal pronouns, and when used with a verb in the imperative or infinitive, these suffixed forms are incorporated into the verb; thus they say rispondetemi "answer me," parlategli "speak to him," datele "give her," imaginarlo "to imagine it," offriteci "offer us." Double suffixes are also used, as assicuratemene assure me of it," dateglielo "give it to him," mandateglielo "send them to him."

66

So also in Spanish, vino á verme "he came to see me," vengo á soccorrerte "I come to help thee," quiero castigaros "I will punish you," dejeme "let me go," pasandome "as I was passing," escribale "write to him," diles "tell them." Here, also, double

suffixes may be used, as decirtelo "to tell it thee," mostradnosla "show her to us."

It is noticeable, however, that this habit in Italian and Spanish is modern, and does not exist in Latin, any more than it does in Sanskrit. Is it, then, a result of the confusion of forms that sprung into existence simultaneously with the decay of the old synthetic system, or is it an adoption of a Semitic principle? Diez finds the origin of the suffixed pronouns in shortened forms of the dative and accusative of Latin, which were already in use in the classical period. It remains, however, to be explained how this peculiarity arose in the Romance group, in one member of the Iranian, and two members of the Indian group, only, and nowhere else in all the wide range of the Indo-European family.

§ 77. Having now gone through all the forms of the modern Indian verb, the subject may be closed by some remarks on the way in which the terminations are added to those verbal stems which end in a vowel. So many of these terminations begin with vowels, that a hiatus necessarily ensues, and the modern languages, though they do not, as a rule, object to a hiatus, do in this particular make occasional efforts to avoid it.

Hindi stems end only in long vowels—â, î, û, e, o. Some grammarians call those stems which end in a long vowel open roots, and those which end in a consonant close roots. This terminology has nothing to recommend it, and there is no advantage in retaining it. The tenses whose terminations begin with vowels are the aorist, future, imperative, and past participle.

Before terminations beginning with û or o, no attempt is made to soften the hiatus, but before â and e there is sometimes inserted a or. As types may be taken the stems “go,”

1 Gramm. d. Romanischen Sprachen, vol. ii. p.85, et seqq.

पी “ drink,” छू “ touch,” दे “give,” and बो “sow.” The aorist of these five is as follows:

The common stems de and le usually suffer contraction by the elision of their final vowel, and one more commonly hears do "give!" lo "take!" dûngâ “I will give,” lûngâ “I will take,” than the full forms.

व्

The future and imperative follow the same rule as the aorist. In the past participle of stems ending in â, is inserted before the à of the termination, as आ " come," p. p. p. आ-य-आ (आया), पा “find,” पाया, खा “eat,” खाया. But in the poets, especially in Tulsi Das, instead of we find commonly inserted. Thus, य् एहि बिधि राम सबहि समुझावा “ In this way Ram explained to all” (Ay-k. 457). आवा “came,” बनावा "made,” पावा “found,” गावा “ sang,” for आया, बनाया, पाया, गाया respectively. Kabir uses both forms indifferently. Thus in Ram. 48, i. पढाया “caused to read," पाया “found,” but in the next, Ram. 49, पावा and आवा. In the fem. sing., however, and in the pl. m. and f., the junction-letter व is not used, thus आई " she came,” बनाई “she made,” fem. sing., आये, पाये masc. pl. As all