PREFACE ix
India. Lastly, the names, being common to the Iranian and Indian languages, might be assigned to the Indo-Iranian period when the two branches were still one people living in Iran. This theory would still allow two centuries for separation, migration to India, and the commencement of Vedic literature in the north-west of India.
The lower limit of the Vedic period is the epoch of the rise and spread of Buddhism, or, roughly, 500 B.C. The Brāhmaṇa literature to be exploited was assumed to be undoubtedly anterior to that date. The boundary line would, however, to some extent have to be overstepped by drawing on the Sūtras for evidence where the Vedas and Brāhmaṇas fail. But though the Sūtras are roughly contemporaneous with the first three centuries of Buddhism, they are practically an epitome of the practices of the Brāhmaṇa period, and are thus often of great value in illustration or corroboration of the facts of that period. They are also important as representing the Brahminical evidence for those three centuries, especially as it is somewhat uncertain how far even the earliest Buddhist literary sources go back in an authentic form to the three centuries following the death of Buddha. Names and practices not referred to before the Sūtras were, however, to be mentioned only incidentally if at all: the few cases to the contrary that actually occur are not real exceptions, because they are derived from Vedic verses quoted in Sūtras, or from Brāhmaṇa parts of Sūtras such as Baudhāyana.
Method Pursued.—Such was the scope of the work on which I finally decided before it was begun, and the plan has been adhered to in its execution as regards the contents. The manner in which those contents were to be presented was the next question to be settled and acted upon. Though both Dr. Keith and myself are familiar with the literature of the Vedic period from which the facts collected in these two volumes are drawn, and the mutual check exercised by two workers sifting the same material acts as a safeguard, it is
