138

THE INDIAN ANTIQUARY.

Journal of the German Oriental Society, thus expresses the difficulty: “The old Prakrit end ing in -o has, in Sindhi, been split up into two great classes, one of which has corrupted the Prakrit -o into -u, the other has preserved it un changed. No rule seems to have influenced this separation, at least I have not yet discovered any, but daily usage seems to have decided in favour of the one or the other ending. It is however noteworthy that many words which in Sindhi end in o, in Hindi end in d, the same re mark holds good of Marăthi, Bengali, and Panjā bi, while on the other hand the short final u in

Sindhi has in those languges been thrown away, or become quiescent.” The rule which Dr. Trumpp professes himself unable to discover appears to me to be this. A Sanskrit noun in -a which bears the accent on

to writing.

[MAy 3, 1872.

Consequently they have retained

the accent which they bore in the older language. In

late

Tadbhavas however

the case

is

different; late Tadbhavas are those words which

had entirely dropped out of use, and were only resuscitated and brought into vogue again at a period when Sanskrit had ceased to be known to

the people. Being revived from books, they were spoken by the eye, if such an expression may be permitted; that is to say, they were

pronounced as they seemed destined to be, the accent generally lying on a syllable already long by nature or position. These words are recog nizable by the much smaller amount of corrup

tion they have undergone, and by the corrup tions which do exist being of a different nature from those demanded by the rules of Prakrit Grammar.

the last syllable, or, in other words, is oxytone,

Moreover, these late Tadbhavas are gene

generally ends in the mediaeval languages in au, and in the moderns in o or d ; while a noun in a

rally words which are synonymous with already existing earlier words. They are the grand, high flown words of the language, not so frequently

which has its final syllable unaccented, or is barytone, ends in the mediaeval languages in u, and in the moderns in u, or a, or entirely rejects the final vowel.

With regard to the practice in each language —Hindi, Bengali, Panjābi, Uriya, and Marathi take 6 in oxytones, Gujarāti and Sindhi take o. It cannot however be said that every oxytone substantive in Sanskrit gives rise to a noun in a or o in the modern languages. On the contrary, the exceptions to the rule are as numerous as the illustrations of it.

This leads to a further

definition of much practical importance. The class of words called early Tad b h a v as is, as a rule, faithful to the accent.

This class consists

of those words which were in existence in San

skrit, and continued to be used in Prakrit, and

have uninterruptedly retained their position in the mouths of the people down to the present time. These words may be recognized by their appearance. They have undergone the regular and usual phonetic corruptions and abrasions

of all Prakrit words, and are often now only recognizable as of Sanskrit origin by the application to them of the rules of Vararu chi or other Prakrit and Pali grammarians. Inasmuch however as their use has been con

used or so expressive of simple ideas as the early Tadbhavas. The proportion of these two classes to each

other varies in the different languages. In those which have been less cultivated, and which have

been most under Muhammadan influence, they are not so frequent as in the more cultivated and more Brahmanical languages. There are many other collateral and sub

sidiary considerations which further complicate this difficult question, a question which is ren dered all the more difficult by the absence of continuous literature.

When the mediaeval

poets began to write, the languages were already so far fixed as to have passed the stage of formation of either early or late Tadbhavas, and to have got into the stage when the vast crowd

of Tatsamas began to make its appearance. The line of investigation thus briefly sketched in outline is of the utmost importance in the elucidation of the origin and formation of the modern noun, and I hope on a future occasion to give examples and illustrations. It will be seen that it is in the determination of the treatment of the oxytone -a base that the

real crux of the question lies, because the bary

tinuous, and as they were derived from the

tones naturally lose their final vowel, and thus

Sanskrit at a time when it was still spoken, they have always, so to speak, been pronounced by

fall under the the same head as the late Tadbhava

ear, and were so long before they were committed

oxytones, except in Sindhi, where they retain the obscure final-u for masculines, and-afor feminines.

• Journal of Germ. Or, Soc, vol. XVI, p. 131.