76 SOCIAL SCIENTIST colonization of virgin lands. Later myths al>out this process counted Parashnrama, Bashistha and Naraka among such pioneers. Other legendary figures like Bhagadatta, Vajradatta and Banasura also belonged to thi^ early phase of Indo-Aryanization. If we gp by Bhattaswamin's commentary on Arthgslwfra Magadha was already importing certain items of trade from this Valley in Kautilya's days 8 Thickly forested under heavy rainfall conditions as it was, the Brahmaputra Valley had only sparse jhum'centred settlements in the neolithic times. These presumably thrived on the banks of hill streams, lakes and river-confluences, where the natural process of erosion and flooding cleared land for thenr by uprooting forests. Their agriculture was characterized by land rotation, slash-and-burn and simple tools like hoes, axes, digging sticks and choppers, all made of stone. It was the Indo-Aryan newcomers, equipped with shaft-hole iron axes and cattle-powered, iron-tipped traction ploughs, who cleared forests on an extensive scale for permanent wet-rice cultivation. This in its turn led to an increase in population and laid the basis for the State or Janapada of Kamarupa to emerge.9 The mixed population absorbed Sanskrit culture, and the latter also, in its turn, absorbed many local cultural traits. Kamarupa moved from protohistory to history in the 4th century A D. If the wet-rice cultivation and the Indo-Aryanization of the autochthons had started so early, how is it that some dry-rice producing tribes and their tribal ways of life continued to persist in several pockets of the alluvial yalley even in the medieval and early modern times ? If the population was steadily increasing all the time, how is it that this valley, about 20,000 sq miles in area, had a population of only one to two million souls throughout the 19th century ? These basic questions clarify the nature of the continuity of underdevelopment in the region's history. It was the persistent tribal factor th^t, time and again, had presumably halted the process and even temporarily reversed it. The point may b^ further clarified with some hard archeological evidence, relevant to the threshold of our period. This evidence was found at the two adjacent villages of Sarutafu and Marakdola, both located at a distance of only 16 miles from Ambari in the city of Guwahati.10 AtSarutaru, an undated neolithic site, seven shouldered stone celts, resembling present-day iron digging hoes of the Khasis, were found together with crude cord-marked pottery of the southeast Asian neolithic types. At Marakdola, its adjacent post-neoMfhac site, driy one such stone celt was found together with fine wheel-turned. Ambari-type pottery and other objects, definitely datable within the 7th-13th centuries by 8. For the approach to the problem, see my "A historiographical perspective for north-east India"^ Man.w India, Vol. ^2, Sept., pp. 221-33. - 9. Ibid. 10. S.N. Rao, "Excavations at Sarutaru : a neolithic site in India" and "Continuity and survival of neolithic traditions in northeast India", both cited, ibid.