________________
226
EPIGRAPHIA INDICA.
[Vol. VII.
show, Lattoor' as the name of a territorial subdivision ranking on equal terms with Bidar, Kalbarga, Sholapur, Vairåg, and Pandharpur. Along with Pratishthâna-Paithan and Tagara. Tår, Latur' is in that part of the Dekkan, watered by the Godavari and its tributaries, which has been indicated by Dr. Bhandarkar as a favourite region of early Aryan settlement;' and it is, in fact, only about twenty-eight miles east-by-Dorth from Ter, and three miles south of the "Manjera,' which is an important feeder of the Gôdâyari. And I strongly suspect that local inquiries would result in finding that Latur' is the ancient Lattalúr, Lattanûr. If so, there will, perhaps, be found at Latur' some notable temple or remains of such a temple, either of the goddess Durgâ in the form of Chåmunda, or of Vishnu in the form of the man-lion, or possibly temples of both those deities; since the Sitâbaldi inscription further describes Dhadibhadaka as "he who obtained favour by a boon of the goddess) Chåmundi," and Vasudeva as "he who obtained favour by a boon of the god) Nárasimha."
The matter must, of course, depend a good deal upon what is the actual spelling of the modern name which the maps and gazetteers present as 'Latur, Lathur, and Lattoor. We need not trouble ourselves abont the h which appears in one of these forms; it is as easily accounted for here, as in some other instances referred to by me elsewhere. For the rest, I feel no doubt that inquiries on the spot would shew that the real name is Látûr, with a long á followed by a single dental ti And, if that is so, the modern name is distinctly derivable from the ancient name, through steps the rules for some of which have been given to me by Professor Pischel. We start with the form Lattalûra, of A.D. 866, of which Lattanúr, appearing first at present in A.D. 1208, is plainly only an optional variant attributable to the interchangeability of I and n. The first step would be the dropping of one t in the second syllable, which would give us Latalure, and eventually the Latalaura which we actually have in the Sit&baldi inscription. The next step would be the omission of the short a of the second and final syllables, which would give us Latlûr. The next step would be the assimilation of the 7 to the preceding , which would give us Lattûr. And, finally, the nexus tt would be dissolved into the simple t, and the preceding short a would be lengthened by way of compensation ;8 and this would give us the ultimate form Latûr.
1 See the skeleton map on p. 951, and the key to it on p. 953, sub-divisions Nos 66 to 71.
See the Gaselteor of the Bombay Prerideney, Vol. XIII., Thana, Part II, p. 423, note 4, and Early History of the Dekkan (id. Vol. I. Part II.), p. 135 fr.
Compare another epithet of the Kadambas of Hangal, namely Jayanti-Madhukdoaraddoa-labdha-para. prasáda, he who obtained the excellent favour of the god Madhukesvars of Jayanti (Banawasi) ” (Ind. Ant. Vol. X. p. 252, text line 25); also, another epithet of the Guttas of Guttal, namely Ujjeni- Mahdkdladératd. Labdha-para-prasdda, "he who obtained the excellent favour of the god Mabákala of Ujjayani” (P. 8. 0.-C. Iwers. No. 109, line 10). But the records do not always present a title of this kind, in connection with the title commemorating the place of origin. And in some cases the epithet indicating family-god, refers to 8 god who was not the god of the alleged place of origin; for instance, though, like the Kadambas of Hangnl, the Kadambas of Goa were styled "supreme lord of Banavast, the best of towns," their other title was frf-Saptakt{fraradera. Labdha-para-prasdda," he who obtained the excellent favour of the holy god Sapta kotisvara” (Jour. Bo. Br. B. 18. Soc. Vol. IX. p. 804, text line 11-12, and compare Ind. Ant. Vol. XIV. p. 290, text line 27), and Saptakóțiś. vars appears to have been a god at 'Narven' in Goa (see Dyn. Kos Distro. p. 566, note 7).
• See Jour. R. As. Soc., 1901, p. 548 ff.
• See Prof. Pischel's Praksit Grammar, (260. As instances of the interchange of land , we may quote the place names Lañjigésara-Nandikoshwar (sce Ind. Ant. Vol. XIX. p. 817 a) and Balisa. Wapesa' (nee id. Vol. XVIII. D. 266, and Vol. XXXI. p. 897), and the proper name Linga pa-Ningapa, well known in the Kanarese country, and the ordinary words nahás for lahan in Gujarati (see id. Vol. XVIII. p. 266, note 5) and jalam-asttami for janm-dohtami in Northern Indis (see id. Vol. XX. p. 89, note 2).
See id. 148.
1 See id. $ 279, 296, and Beames Comparative Grammar of the Modern Aryan Languages of India, p. 282 (2)
See Beames' Comparative Grammar, Vol. I. p. 152, 4 41, and p. 281, 78 (1).
