『Abstract
　A knowledge of Himalayan erosion history is critical to understanding crustal deformation processes, and the proposed link between the orogen's erosion and changes in both global climate and ocean geochemistry. The most commonly quoted age of India-Asia collision is 〜50 Ma, yet the record of Paleogene Himalayan erosion is scant - either absent or of low age resolution. We apply biostratigraphic, petrographic, geochemical , isotopic and seismic techniques to Paleogene rocks of the Bengal Basin, Bangladesh, of previously disputed age and provenance. Our data show that the first major input of sands into the basin, in the >1 km thick deltaic Barail Formation, occurred at 38 Ma. Our biostratigraphic and isotopic mineral ages date the Barail Formation as spanning late Eocene to early Miocene and the provenance data are consistent with its derivation from the Himalaya, but inconsistent with Indian cratonic or Burman margin sources. Detrital mineral lag times show that exhumation of the orogen was rapid by 38 Ma. The identification of sediments shed from the rapidly exhuming southern flanks of the eastern-central Himalaya at 38 Ma, provides a well dated accessible sediment record 17 Myr older than the previously described 21 Ma sediments, in the foreland basin in Nepal. Discovery of Himalayan detritus in the Bengal Basin from 38 Ma: 1) resolves the puzzling discrepancy between the lack of erosional evidence for Paleogene crustal thickening that is recorded in the hinterland; 2) invalidates those previously proposed evidences of diachronous collision which were based on the tenet that Himalayan-derived sediments were deposited earlier in the west than the east; 3) enables models of Himalayan exhumation (e.g. by mid crustal channel flow) to be revised to reflect vigorous erosion and rapid exhumation by 38 Ma, and 4) provides evidence that rapid erosion in the Himalaya was coincident with the marked rise in marine ⁸⁷Sr/⁸⁶Sr values since 〜40 Ma. Whether 38 Ma represents the actual initial onset of vigorous erosion from the southern flanks of the east-central Himalaya, or whether older material was deposited elsewhere, remains an open question.

Keywords: Bengal Basin; Himalayan erosion; Barail formation; Bangladesh; detrital thermochronology; Surma Basin』

1. Introduction
2. Geological background
　2.1. Himalayan geology
　2.2. Existing Paleogene records of Himalayan erosion
　　2.2.1. The orogen's western region
　　2.2.2. The orogen's central and eastern regions
　2.3. Tertiary strata of the northern Bengal Basin, Bangladesh
3. Approach, methodologies and results
　3.1. Biostratigraphy
　3.2. Seismic data
　3.3. Petrographic and heavy mineral data
　3.4. Zircon U-Pb data
　3.5. Zircon fission track data
　3.6. White mica Ar-Ar data
　3.7. Sm-Nd bulk rock data
　3.8. ¹⁸⁷Os/¹⁸⁸Os bulk rock data
　3.9. Cr-spinel geochemistry
　3.10. Clay mineralogy and illite crystallinity data
4. Interpretations
　4.1. age of the Paleogene rocks of the Surma Basin
　4.2. Provenance of the Paleogene rocks of the Surma Basin
　　4.2.1. Characteristics of the potential source regions (Table 2 and supplementary item 1)
　　4.2.2. Provenance of the Paleogene Surma Basin rocks: interpretation and integration of the data)
　　　4.2.2.1. A cratonic provenance for the Barail Formation?
　　　4.2.2.2. A Himalayan provenance for the Barail Formation
　　　4.2.2.3. Subordinate arc/ophiolitic input to the Bengal Basin
　4.3. Regional applicability: correlation and comparison with data from the southern Bengal Basin, Bengal Fan and Himalayan foreland basin
　　4.3.1. The southern Bengal Basin
　　4.3.2. The Bengal Fan
　　4.3.3. The foreland basin
5. Discussion
6. Conclusions
Acknowledgements
Appendix A. Supplementary data
References