KHE 211 stereotyped in literature. This argument would not apply to Shiurájpur, the ancient seat of the Gaurs, to Sitapur, Gadaipur, Bhojpur, and many other towns over which ages have passed, and the termination remains unchanged. There is yet another interesting phase of the question. If the term in- ation is Sanskrit, the first portion of the name would surely also often recall some well known feature of the social life or religious belief of Sans- krit-speaking races. Certainly nine-tenths of the urban names, which donow end in unchanged pur, pagar, kot, are derived from gods of the Arian Pan- theon, as for instance Durgápur, Shiupur, Bishnpur, Krishnagarh, Brahma- pur, Bhawanipur, Sitapur, Ganeshpur: Hundreds of other names similarly compounded are found in Kheri and all Oudh districts, but these Sanskrit nameş never preface the eri and oli terminations referred to; therefore we must suppose that in the age wlien all these towns were founded, not only was language so liquid and unstable as to allow such remarkable changes, but at that time the Sanskrit-speaking races knew nothing of its great triad, or else it would have applied them in its urban nomenclature. However late a development of the Arian theogony, it is at any rate generally supposed that Bishn and Shiu were its chief gods centuries be- fore the colonists were in a position to build cities in the remotest corners of Hindustan. Nor, indeed, are the earlier dieties of the Rig Veda, Agni, Vaya, Surya, Mitra, ever associated with these terminations. Rudra and Mitra are the only now recognizable possible components of these old for- mations, and even if Rudauli and Mitauli indicate the culture of those gods by their founders, it is by no means certain that these were Arian godst at all. The fact is that the Sanskrit language has prevailed over and been adopted by many Turanian races, who now use fashionable terminations in the nomenclature. The great Chhattri tribes of Rajputana in every case have a city or province exhibiting these or similar terminations from evidently allied roots. The three chief tribes no doubt now come from Udaipur, Jaipur, and Jodhpur, but those provinces or towns formerly bore the name of Mewar, Amber, or Dhundhar, and Márwár or Mandore. In fine, we can come to no other conclusion than that towns whose names exhibit these terminations are Turanian foundations in nearly every case, otherwise we are forced to several impossible theories --one that none of the Turanians, the great builders of antiquity, and to this date the majority of the Indian people built any towns, or that if they did, their names have been entirely lost ; another, that the Sanskrit termination per has exhibited the wildest vagaries changing into oli, eri, wari, intro- ducing unlimited novelties and always eliminating the p; another, that for some reason in former times the Arians, never built towns in honour of eponymous deities as they have universally done as far as history extends. From these and many other difficulties we escape by holding that these terminations are identical in origin as they are in form with the at, ul, ur, wára, uli, which we find to this day in common usei among Turanian races, meaning a town, village, dwelling place. Similarly, Mr. Lyall in the Berar Gazetteer points out that the original name of
- Muir'a Sansorit Texte, Vol. IV., page 832.
See Muir's Sapecrit Texts, Vol. IV, pages 57, 71,
