As the carboniferous limestone thins out in approaching the Indus, the Oolitic formation increases in importance and forms much disturbed hills, all the way from Moosa Khel to Kalabag. It is continued west of the Indus in the Chichalee Range and the northern end of the Speen Ghur; a little above Moola Khel it disappears under the alluvial, and does not reappear till Sheikh Bodeen, where, as we have seen, it attains a considerable thickness.

65. The salt and gypsum is continued on the west side of the Indus, in the hilly country of the Kuttuks, but it is there much covered by tertiary clays and sandstones. It crops out near Bahadoor Khel and along the course of the Teeree Towe. At the first named place the Saliferian forms an anticlinal arch; the salt, above fifty feet thick, is the lowest bed seen, and is very regularly stratified; above it is a thin bed of red marl, another of grey sandstone, also thin; then gypsum, about twenty-five to thirty feet thick; then a thin band of a limestone with minute debris of fossils, and which resembles lithologically the Oolitic bed of Kalabag and Maree on the Indus; then the dark, brown, sandstone which often forms the base of the nummulitic formation; some coarse and crumbling shales without fossils; and finally, a bed of limestone rich in nummulites, volutes, veneridae, &c., and about ten to twelve feet thick. This is at last covered by the marly lumpy clayey beds of Miocene. A fault running approximately W. E. through the Soordak Pass, has caused an upthrow of the beds on its southern side, and there the nummulitic limestone, much tilted up, forms a pretty high hill.

Along the Teeree Towe the Saliferian is immediately covered by Tertiary. As far as Lachee the rocks seen are Miocene sandstone, clay and conglomerates; thence to Peshawur the country is entirely covered by nummulitic limestone and shale, and the Miocene sandstone is only seen here and there in small detached beds and patches, which are evidently the remains of layers which have been mostly removed by denudation.

66. North of the Salt Range we have also a great extent of Tertiaries. Nummulitic limestone, shale and sandstone first covers in the secondary layers in the western portion of the range, but rests directly on the salt marl and gypsum in the eastern half of it. It attains a great thickness, where well developed, (4500 feet,) and forms

the summits of nearly all the highest hills of the Salt Range. It is continued to within two miles of Maree on the Indus where it thins out, but reappears near Kalabag, and is very well developed in the Chichalee Range and in the Speen Ghur. Near the Indus, all the beds of the Salt Range, excepting the Saliferian marl itself and the secondary strata where much locally disturbed, dip towards the N. E. On the western bank of the Indus, that is in the Chichalee Hills and the Speen Ghur, the dip is W. N. W. or N. W. This last dip is generally that of all the strata of the Kuttuk hills.

The nummulitic formation appears in the Salt Range as a thick belt which, beginning at the Mount Tilla near Jheelum, is continued to near Maree on the Indus, where it disappears for a little space, but reappears on the other side of the river, and is to be seen forming the bulk of the Speen Ghur to near Esokhel. The formation keeps a remarkably similar aspect the whole way. It is, from below upwards, composed of 1. Sandstone often coloured by iron, but generally dirty white or pale grey. 2. Very arenacious, thin bedded or lumpy limestone, with gasteropods, few and small nummulites and innumerable debris of oysters or grypheæ. 3. Shales of various colours, with beds of lignite and of alum carbonaceous shales. The alum shales are only developed where the lignite is situated close to the Saliferian formation, and appear to be patches of lignite metamorphosed. 4. Argillaceous limestone, full of large nummulites, chama, cardita, crassatella, ostroa, many gasteropods, very large echinodermata, &c., &c. 5. Shales often replaced by a clay-slate containing nummulites. The shales contain sometimes lignite and Rol (alum-shale), but the seams are made less well defined than in the lower shales. 6. Argillaceous limestone, extremely white in some places and containing the same fossils as layer 4; in the eastern portion of the Range it contains flints; it is often foetid. 7. Chert, hard limestone, weathering rough and pitted; pale yellow or flesh-colour, brittle and

Occasionally a bed of white soft fragile limestone is seen to form the base of the nummulitic formation. It is characterized by a planorbis which is tolerably abundant; but it contains neither nummulites nor any other fossil. It is found in lenticular beds of little extent, and rarely more than two or three feet thick. It suggests to the mind beds formed in pools or creeks among sandy islands and promontories at the mouth of a river. Whenever it occurs, I have found in the nummulitic limestone above it a great number of teeth and bones of fishes (sharks).

splintery. Shells fewer, nummulites small, but very abundant, especially the N. variolaria, whilst the flat and irregular N. pushi and N. laevigata, so abundant in beds 4 and 6, are not to be found here, or are at least rare. A nummulite about the size of the N. pushi, but thicker, is, however, found pretty abundantly, though not in swarms like the N. variolaria. A ribbed cardita is the only bivalve which appears tolerably abundant.

67. Resting on the nummulitic formation of the Salt Range are thick beds of Miocene sandstone, clay and conglomerate. I have described in chapter I. how these sandstones form a great plateau between the Salt Range and the foot of the Maree Hill, and indicated that thay may be considered as the upper Miocene Bed, whilst the Maree Hills and the whole of the mountains between the Jheelum and the Pir Punjal chain are to be regarded as lower Miocene. The upper bed is rich in mammalian fossils, and is identical to the Sewalik formation. The lower bed is devoid of fossils,* containing only a few debris of plants, rootlets, small stems aud occasionally small niduses of lignite. The upper Miocene has probably been a great deal denuded; remains of the bed are, however, to be seen in the valley of Poonch; they are there rich in very well preserved fossils, teeth of elephants being common and very perfect.

68. The sandstones and conglomerates just mentioned form a great belt from the E. N. E. to the W. S. W. (see Map) and to the north of it appears another belt, having a similar direction and composed of nummulitic limestone and slate. It begins in Hazara in Lat. 34°, and forms all the superficial covering of the Hazara mountains as far as the Sirun river and as high north as Mausera, being about thirty miles in width as the crow flies. It proceeds from N. E. to S. W. towards Attock, keeping the same width and extending in that district from the Indus to Janika Serai. Crossing the Indus, it forms the whole of the Akora Kuttuck and Afreedee hills between Peshawur and Kohat, extending about sixteen miles south of Kohat. It has been followed as far as longitude E. 70°. The beds of this nummulitic formation have a general dip to the N. W. A similar

*It is said that one or two bones have been found in the lower Miocene, but this is doubtful; if they exist, they are at any rate very rare. Mr. Medlicott has pointed out a non-conformity between the lower and upper Miocene; he makes three beds of the formation.

nummulitic tract follows the foot of the Himalayan ranges along the southern versant of the Pir Punjal chain and its continuation to the S. E. It begins in the valley of Poonch; it is seen north of Rajaori, and the pebbles of the streams near Rajaori are often nummulitic limestone, though the parent-beds have not yet been discovered. I cannot say whether nummulitic beds are to be seen to the north of Tummoo, Basaoli, and Noorpoor or in Kangra, but they appear near Subathoo in long. 77° lat. 31°, and have further been just discovered by Captain G. Austen on the east of the Ganges in Kumaoon. But this nummulitic along the foot of the Himalaya is either much denuded or much covered up by Miocene, and does not make such a show on the surface as the other belt which follows the direction of the Afghan mountains.

To the north of these zones of nummulite we meet the volcanic hills, which I have described in the first chapter.

69. The stratum of nummulite in Hazara, occasionally broken through, or faulted or denuded sufficiently to allow of older rocks making their appearance.

At the northern end of Mount Sirbun near Abbottabad, carboniferous limestone resting on volcanic rocks is quarried for building purposes. The limestone belongs to the Weean and Kothair groups and is thin-bedded, arenaceous, marly and occasionally conglomeratic. It is of considerable thickness and immediately covered in by limestone, the lower beds of which are so poor in fossils that it is impossible to identify them, the upper being nummulitic.

The following is a section near the small village of Sheikh Wandie, from E. to W.