Bulletin of the Geological Society of Malaysia, Volume 20, August 1986, pp. 289 – 310
Geological Survey of India, Shillong, India
Abstract: Recent discovery of stone age sites in the Quaternary fluvial terraces has brought Tripura on the prehistoric cultural map of India and Southeast Asia. Artifacts have been collected from 25 localities spread over 45 km radius around Agartala in course of Quaternary mapping in West Tripura. These findings demonstrate the existence of multiple dispersed colonies of stone age people on the stabilised river terraces.
The geological context of these sites is well established. The tools occur in open stations (primary sites) mostly on the dissected remnants of the upper most terrace (T1) lying between 5 m and 15 m above extent floor regime (20 m-83 m) above m.s.l. These are found profusely scattered in open fields, and as interlayers within the moderate yellowish brown light brown and grayish orange (10 YR 5/4, 5 YR 5/6 and 10 YR 7/4) loamy latosols up to a depth of 1.7 m. The soil mantle overlies a fining upward, polycyclic, fluvial sequence, comprising ferruginised, dark yellowish orange (10 YR 6/6), arkosic, medium sands. Generally, mottled soft latosols occur below the humus enriched A-zone. Locally, hard concretionary laterite is also observed. The above terrace sequence is unlithified and belongs to the Upper Pleistocene (34680 ± 2960 Y.B.P. by C14 dating). It lies over the folded Neogene strata with a demonstrable angular unconformity. The younger, unoxidised/unaltered, Holocene terrace sediments (T3) dated at 3450 ± 110 to 1100 ± 90 Y.B.P., and (T4 =nascent floodplain) at 165 ± 80 Y.B.P. by C14 method, occur as successive insets. T2 has very limited distribution.
Typologically, the artifacts include flaked axes, polished celts, chopping (bifacial) tools, adzes, scrapers, points, hammerstones, grooved stones, prepared cores, blades, pebble tools and a bulk of flakes. It is noteworthy that 98% of the tools are made of silicified wood embedded within or derived from the Neogenes by recycling. Interestingly, the flakes are removed from the silicified wood core, generally either oblique to or across the wood structure, possibly for controlling the size of the flakes. Controlled flaking is extremely difficult in petrified woods, except along an axis transverse to the growth rings. Consequently, the tool types are restricted mainly to tabular varieties. Rarely, levalloisian technique of flaking is seen. The tool assemblage points to a wood land ecology in the area.
The tool ensemble can be grouped into: (A) Preneolithic assemblage without polished axes but having typological affinity with neolithic tools and (B) Neolithic assemblage dominated by polished axes. The Preneolithic assemblage belongs to the Late Pleistocene(?) age (much younger than 35,000 Y.B.P. and older than 3,500 Y.B.P. as indicated by the dated stratigraphic records of the older and younger terrace deposits). In its totality, the artifact assemblage bears a close semblance with the Anyathian and Neolithic tools reported from the Irrawaddy valley of Burma. The present discovery thus establishes cultural linkages across the barrier of the Indo-Burma ranges during the prehistoric times.