All these peculiarities of outline are the result of denudation, of the specific action of which agent we are very ignorant. The remarkable difference between the steep cliffs on the south face of the range, and the rounded outline of the hills on the northern slopes, may be explained on the supposition that when the Khasia was partially submerged, the Assam valley was a broad bay or gulf; and that while the Churra cliffs were exposed to the full sweep of the ocean, the Nunklow shore was washed by a more tranquil sea. The broad flat marshy heads of all the streams in the central and northern parts of the chain, and the rounded hills that separate them, indicate the levelling action of a tidal sea, acting on a low flat shore;* [Since our return to England, we have been much struck with the similarity in contour of the Essex and Suffolk coasts, and with the fact that the tidal coast sculpturing of this surface is preserved in the very centre of High Suffolk, twenty to thirty miles distant from the sea, in rounded outlines and broad flat marshy valleys.]whilst the steep flat- floored valleys of the southern watershed may be attributed to the scouring action of higher tides on a boisterous rocky coast. These views are confirmed by an examination of the east shores of the Bay of Bengal, and particularly by a comparison of the features of the country about Silhet, now nearly 280 miles distant from the sea, with those of the Chittagong coast, with which they are identical. The geological features of the Khasia are in many respects so similar to those of the Vindhya, Kymore, Behar, and Rajmahal mountains, that they have been considered by some observers as an eastern prolongation of that great chain, from which they are geographically separated by the delta of the Ganges and Burrampooter. The general contour of the mountains, and of their sandstone cliffs, is the same, and the association of this rock with coal and lime is a marked point of similarity; there is, however, this difference between them, that the coal-shales of Khasia and limestone of Behar are non-fossiliferous, while the lime of Khasia and the coal-shales of Behar contain fossils. The prevalent north-east strike of the gneiss is the same in both, differing from the Himalaya, where the stratified rocks generally strike north-west. The nummulites of the limestone are the only known means we have of forming an approximate estimate of the age of the Khasia coal, which is the most interesting feature in the geology of the range: these fossils have been examined by MM. Archiac and Jules Haines,* ["Description des Animaux Fossiles des Indes Orientales;" p. 178. These species are _Nummulites scabra,_ Lamarck, _N. obtusa,_ Sowerby, _N. Lucasana,_ Deshayes, and _N. Beaumonti,_ d'Arch. and Haines.] who have pronounced the species collected by Dr. Thomson and myself to be the same as those found in the nummulite rocks of north-west India, Scinde, and Arabia. CHAPTER XXX.