Sindhi thus preserves the case-endings more strictly than the other languages. The latter mostly take the Prakrit locative, or ablative, and entirely reject the terminations.

Hindi has अनत " elsewhere,” Skr. अन्यत्र, निकट " near, ” also नेरे (dialectically नेरु and नेरौ) ; परे “on the other side,” भीतर “ within,” Skr. अभ्यन्तरे, बाहिर, बाहर " outside,” Skr. वहिस्, and others.

M., like S., has, but in the sense of "before," also g “before,” “beyond," "above,” “near,” which are peculiar to itself. In the other languages there is nothing deserving special mention; the subject has already been treated in Vol. II. p. 296.

Adverbs of manner.—While the adverbs of place, being also, in their nature, postpositions, and as such used to form cases, do not call for special mention, adverbs of manner are not so used,

and it is to them that the term adverb, in its more special sense, correctly applies. Such words as âge "before," pichhe "behind," and the others, may, indeed, be properly regarded as adverbs when they are used alone, but when in conjunction with nouns, they become true postpositions, giving to the relations of the noun a more extended application. Adverbs of manner, on the other hand, are, for the most part, adjectives used adverbially, and this practice is common in all Aryan languages. In Sindhi, which preserves distinctions obliterated in the other languages, adjectives may, as pointed out above, be used adverbially by being undeclined, or, in their true use as adjectives, by agreeing with the subject in gender and case. Thus, to quote the instances given by Trumpp:1

ओचितोई कुए हिकिडे उन भांडे जे भरि में बृद्ध किओ

"By chance one mouse made a hole near that granary."

Here ochito- is an adjective in the nom. sing. masc. with emphatic, and although by the accident of the construction it is in the same case as the subject kue " a mouse," yet it is evidently used adverbially.

तडिहीं जब हारे जाल घणो रोई माउ वरिओसि

"Then having shed tears much, having wept much, his mother returned."

Here ghano is an adjective in the nom. sing. masc., and clearly does not agree with mâü "mother," which is feminine; it must be regarded as used adverbially.

qai ayfa anfaet eier aufa zię

"The winds strike hot, the days burn fiercely."

In this sentence I do not think we should regard the adjectives as used adverbially; "winds," is a noun in the nom. pl. fem. and affi "hot" agrees with it, so also

1 Sindhi Grammar, p. 386.

"days," is nom. sing. masc. and “fierce, excessive" (Skr. ), agrees with it; so that we might more literally translate, "the hot winds strike, the fierce days burn."

In Marathi and Gujarati also, where adjectives have the full range of three genders, they are often made to agree with the noun in constructions, where in English they would be used adverbially. When intended to be adverbially used, they stand in the nom. sing. neuter, ending in M. and G. In Bengali and Oriya, where no gender exists, it is impossible to draw the same line of distinction, and this remark applies also to that numerous class of adjectives in Hindi and Panjabi, as also in all the other languages, which are indeclinable, or which, ending in mute a, do not vary their terminations. Those adjectives in H. and P. which end in a masc. and i fem., when used adverbially, stand in the former gender and do not vary with the noun.

In B. In B.,

§ 83. Conjunctions.-In Hindi the common word for "and" is और, Old-H. अरु, अवर, from Skr. अपर “ other.” and O. it loses the final consonant, B., O. E. अउ. however, a, pronounced ebong, is very common; and "also." P., often shortened into a, probably from Skr. "at the end," "afterwards," as well as (TC). S.,, or, in which the fondness of S. for the i-sound comes out. G. uses generally literally "then.” It has besides and for " also," which belong to the group from अपर· G., shortened, I am disposed to connect with P., and ascribe to Skr. . In the former case the has been lost, in the latter the 7. It may, however, be allied to M. fu, uct “and," from Skr.

"other." In all the languages, however, the shorter conjunction is in use, side by side with the words given above. The ordinary Skr.

has left no descendants. The Gipsies use te or ta, which

agrees with P., also u, which is Persian ý.

"Also."-H. ft, Skr. afg fg, Pr. fa f. The various steps from fa downwards are all retained in S. fa, fa, fu (for fafe), and t. The other forms q, yfu, fqfu, are from Skr. "again," and show a gradual change from the u to the i, in accordance with S. proclivities. P. means literally "near to," locative of "near." G. q, meaning also "but," is

from . B. uses, and O. fu. पुनर्.

"But" is very frequently expressed even in Hindi by the Arabic words लेकिन J, बलकि 4, and Persian मगर ...; and in the other languages also. Pure Sanskrit are ta and किन्तु, as well as बरण (Skr. वरं). In Hindi पर (Skr. परं) is also used, and in P., which also uses, and a strange form

, in the initial syllable of which we may recognize an incorporation of the near demonstrative pronoun, so that it means "rather than this." P. has also Яqi, at, probably corruptions of.. Peculiar to P. and S. are the forms P. qui, हथां, S. हथा, हथो, and emphatic हथाईं and हथोंई " but rather.” These are ablatives from, and the idiom may be paralleled by our English phrase "on the other hand." S. has also "but," and ч, in which latter it agrees with G. It also uses #f, having added a final i to.. G. and M., in addition to पण, use also परंतु, 0. किन्त and पर, B. किन्तु and वरं.

"Or."-H., T, which are Skr., with for. The Arabic is very common, and commoner still is fa, probably shortened from Skr. fait. This fa is colloquially common in nearly all the languages. G. writes it, where e is short. M. lengthens the vowel and retains the nasalization कीं, also using अथवा, as does G. P. के, अथवा, and या. S. के and को. B. uses several varieties of किंवा, as किम्बा, किबा, कि,

and बा, also अथबा• O. the same.

"If."-Skr. afa, generally pronounced and written af, is universal. H. shortens it into

and

(=, with loss of

and, by rejection of final i, and

goes out, its place

being supplied by and; thus. Persian is also very common. P. and . This last in

troduces कर for करि “having done.” S. जे.

S..

"Although."-Skr. af is used in B. O., also in H., but

even."

P. has, literally “one

B. also uses जदि जो " if

S. तोडे, तोणे, sometimes with जे prefixed, जे तोडे; this, too, is literally "if even," for alone is used to mean

"( either," ," "even," and is apparently really the correlative (a3af) with a diminutive affix.

=

other languages. G.

It is not found in the

correlative; thus to H. जो or जो भी answers तो or तो भी

"then," or "even then;" to P.

पण, answers तो पण, and so in all. H., the correlative is Skr.

answers at ft; to G.

In B. and O., as in written

f "yet.”

"Because."—H. f literally "for why." B. and O. use Skr. कारण “ cause,” and जे हेतु or जे हेतुक. G. माटे, probably Skr. f, which is also sometimes used in O. in the sense of "only," "merely," "for the simple reason that." G. has also केम के “ for why,” and shorter कांके. S. जेलां, जेल्हं, where the last syllable is for “for” (see Vol. II. p. 260). S. also uses a string of forms with छा “ why ?” as छाजो, छा लाइ, and छो ता, छो जो. a. The correlatives "therefore," etc., take the ☎ form तेला “etc.”

These instances may suffice to show the general principles on which the languages proceed in forming their conjunctions. There is, as in other respects, a general similarity of process, accompanied by variations of development.

§ 84. Interjections.-The various spontaneous or involuntary sounds, some of them hardly articulate, by which sudden emotions are expressed, are scarcely susceptible of rigid scientific analysis in any language. Everywhere we find ha! ho! or ah! oh! and the like. Surprise, fear, disgust, delight, and