according to the new light thrown on them. In the Gupta period, the addition of a new variety of the alphabet is now possible, owing to the discoveries of the remains of ancient Indian civilisation in the deserts of Central Asia. Fresh discoveries have also made it possible to trace the gradual displacement of the Eastern variety of the Northern alphabet by the Western one, in the 5th and 6th centuries A. D., and to determine the exact epoch of the final displacement. Finally, new materials have facilitated the determination of the type specimens of each variety, in each particular century, with a nearer approach to accuracy.
From the 7th century onward, it has been found impossible to follow the arrangement in Dr. Bühler's work, as the development of the Eastern variety from 600-1100 A. D. has not been clearly shown there. In the following pages, the alphabets of the North-Eastern inscriptions of the 6th and 7th centuries A.D. have been separately analysed. In the 8th century, we find three different varieties of the alphabet in Northern India, or more strictly four, if we count the alphabet of Afghanistan, which is as yet but little known. The Western and Afghanistan varieties were developed from the old Western variety, while the Central and Eastern varieties were evolved out of the old Eastern. The Eastern variety lost ground and its Western boundary gradually receded eastwards. The development, of the Eastern alphabet only, has been followed in these pages. It has become possible to show, that proto-Bengali forms were evolved in the North-East, long before the invasion of Northern India, by the Nāgarī alphabet of the South-West, and that Nāgarī has had very little influence upon the development of the Bengali script. The chronology of the Pāla dynasty of Bengal, and specially their relations with the Gurjjara-Pratīhāras have been settled from
