________________
MAY, 1877.)
SANSKRIT AND OLD CANARESE INSCRIPTIONS.
137
vowel e or das written abovo a preceding con- to make up the metro ;-e.g. in 11. 6-7 of No. II. sonant. Compare, for instance, in the facsimile of my Dêvagiri-Yadava inscriptions, Jour. Bo. Br. plate of No. I. of this series, the form of the R. As. Soc., No. xxxiii., vol. XII., p. 11, we are vowel & in kald kaļa, 1.20, and in sarvván-etan, and obliged, for the sake of the metre, to pronounce, as partthivindran, l. 21, with the sign for the Virdma written, Art-Ved wydaa-Kutsar=Jamadaguni-Va. attached to the final n twice in étan and partthiven- sishthar, &c., and in l. 16 of the same, Adigo-dantidrán, 1. 21 ; and compare, in the facsimile plate of darta-sakal-orviyan, &c., and in l. 27 of No. IV. of No. II. t, the form of e in saluttam-ire, l. 12, the same set Sanu-margar baring-iral, &c. In order with the sign for the Virdma attached to the
to have constantly in view the importance of this final in Gayeyol, 1. 35. The modern Canarese vowel-sign u as used for the Viráma, I shall resign for the Viráma is derived directly from this present it in my transliterations by the italicized modification of the vowel 6 or 8.- And 2, more U, and, vice versd, by the ordinary 'u' in italicized commonly, by either of the two later signs used quotations. In metrical passages, it may then be for the vowel # as written with a preceding con- vocalized or not according to the exigencies of sonant. Usually the form of u made use of in this the metre; while, in prose passages, it will be way is that which is written entirely on the line, treated as an orthographical sign, or as a vowel, as in the modern characters; as instances, see according to the opinion formed by the reader of
e Virdma attached to the final in nclo-virtinol, the linguistic stage of the inscription. 1. 25, and to the final l in paduval and ba lagal, Oue sign remains to be noticed, the Avagraha, 1. 40, of No. IX.S, and to the final t of brimat in which indicates the clision of an initial a. It is 1 6 of No. XXXIII. below. It is but rarely that
of very rare occurrence in inscriptions, and the second and older form of u-which is some- the only instance that I can call to mind is the times entirely subscript, and sometimes partly Gadag inscription published by me at Vol. II., p. BO, commencing below the line and running up to 299. Wherever it occurs, it will be most conthe top of it is used for the Virdma. Instances venient to represent it by its own Dêvanågarf sign, of it will be found in 11. 9 and 12 of the fol- $, as in that transcription. lowing inscription, No. XXXII. And as used to
No. XXXII. represent the vowel, it will be found in the facsimile plate of No. I., in kula-tilakan and Ohd
This is an inscription in the Old Canarese ļuky-abharanaria, 1. 4, and Vishnu-nilayam and characters and language from Katt agerit Muni-gana-nil (Lt)yam, 1. 27, and in the facsimile in the Dadami Taluka of the Kalådgi plate of No. II., in Sántaļige-sdsiramumasa Mando- District. It is on the front or north face of one of li-sdsiranura Padineriut-Agrahdramumam dushta- a row of stones forming the entrance to the nigraha, 11. 22-4. As a peculiarity in the use of
steps at the north end of the small tank the vowel-sign w for the Virdma, I may mention
under the wall of the fort. The emblems at that I do not find it used with a final m. Having regard to the number of forms in which
the top of the stone, which is fixed upright in occurs as a termination in modern Canarese, while
the ground, are:- In the centre, a cow and calf; in the older dialect the same forms were shortor by
above them, the sun, with a linga above it; and one syllable and ended in a consonant, this
on the left of the centre, a curved sword, with use of the vowel u to represent the Viráma the moon above it. They are only roughly cut, must have played an important part in the de- in outline. On the right of the centre there are velopment of the language. In Old Canarese, the no emblems. The inscription covers 1'6}'' high scansion of metrical passages shows that, with by 1'4'' broad. The average sizo of the letters very few exceptions, it was only an orthographical
is !". The characters are well-formod types of sign. In prose passages, where no such test
the period to which they belong, and call for can be applied, I have hitherto treated it as an orthographical sign, or as a vowel, according to
no remark, except that the subscript form of the the age and general style of the particular in. !
vowel u is used to represent the Virána in lines scription. Bat, in some fow cases, the metrical
9 and 12. test shows that, even when used as an orthogra. The inscription is dated in the twenty-first phical sign, it was liable to be vocalized, if roquired year of the Western Chalu kya Vikrama• Vol. IV., p. 179.
Vol. V., p. 15. t Vol. IV. p. 208, and Vol. V., p. 356.
The orthography of the name is very doubtful; but, 1 The third and oldest sign, Scrook attached to the as far as I could ascertain, this is the correct form. Keri, onsonant and tarned downwards, occurs in some of the 'street', and kere or keri (old form, kece), 'tank', are oldest Old Canarese inscriptions, but it is strictly, a cha- common terminations in the names of Canarese villages, racter of the Oavo-alphabet class. I have never found it and, owing to carelessness in writing, are very liable to be used to represent the Dir&ma.
confused one with the other.
