sacrificial fire. When Mahâkâli saw that they were dying of thirst she gave them suck from her breasts, and named them Mâgadha and Sûta. The Mâgadha Brâhmans settled in the East and the Bhât Brâhmans are their descendants; the Sûta settled in the West, and from them sprang the Bhâts.” By another legend, when Kâli destroyed the Râkshasas she formed a figure out of her sweat and breathed life into it, so that it might record her victory. 2. Again, according to Mr. Nesfield, the Bhâts are an “offshoot from those secularised Brâhmans who frequented the courts of princes and the camps of warriors, recited their praises in public, and kept records of their genealogies. Such without much variation is the function of the Bhât at the present day. The ancient epic known as the Mahâbhârata speaks of a band of bards and eulogists marching in front of Yudhishthira as he made his progress from the field of Kurukshetra towards Hastinapur. But these very men are spoken of in the same poem as Brâhmans. Nothing could be more natural than that, as time went on, these courtier-priests should have become hereditary bards, who receded from the parent stem and founded a new caste bound together by mutual interests and sympathies. There are several facts in support of this theory, that one of the sub-castes is called Baram or Biram Bhât; that some Gaur Brâhmans still act as bards and genealogists; that the Bhât still wears the sacred thread, and is addressed by the lower caste by the Brâhman title of Mahârâja; and lastly, that by an obvious survival of Brâhmanical titles, the Bhât’s employer is called jajmân, “he who gives the sacrifice,” while the Bhât himself is jagwa jâjak or jâchak, “the priest by whom the sacrifice is performed.” To this Mr. Risley objects that “if the Bhâts of the present day are descended solely from a class of degraded Brâhmans, if, in other words, they are a homogeneous offshoot from the priestly class, how do they come to have a number of sections which are certainly not Brâhmanical, and which appear rather to resemble the territorial exogamous groups common among the Râjputs? Brâhmans, however degraded, hold fast to their characteristic series of eponymous sections, and I know of no case in which it can be shown that they have adopted section names of a different type. On the other hand, there is nothing specially improbable in the conjecture that Râjputs may have taken up the profession of bard to the chiefs of their tribe, and thus may, in course of time, have become incorporated in the Bhât caste. It will be seen that this solution of the difficulty in no way conflicts with Mr. Nesfield’s view, but merely modifies it by introducing a second factor into the formation of the caste. Mr. Nesfield regards the Bhâts as a homogeneous functional group thrown off by the Brâhmans. I look upon them as a heterogeneous group made up of Brâhmans and Râjputs welded together into one caste by virtue of their exercising similar functions.