does it now lie concealed under newer formations? There is no doubt that great denudation has taken place repeatedly in the Himalaya and subordinate hills; yet basins nicely protected by eruptive or metamorphic rocks, bottoms of valleys or down-thrown beds might have escaped removal. Not a trace of true coal has yet been found in the Himalayas, the Punjab or the Afghan mountains, excepting (geologically speal.ing) the few grains of coal which fill in the cellular tissue of the lepidodendron-like plants described in para. 43, as having been found in one of the layers of the Wean group. This is not very encouraging; but any person who has observed what a thick mantle the Miocene sandstones and the old and new alluvia form over the older formations, would not expect to find coal cropping out in a conspicuous manner. If coal does exist, it will be one day discovered, no doubt; but the discovery will be made by patient and careful study, and not by digging at random with a pickaxe wherever something black is observed. It may be said with truth that the means hitherto employed, by Government or persons interested in the search for coal, have been such that not the smallest reasonable chance of success could be anticipated. But all this is foreign to our subject.

91. The end of the Palæozoic epoch or beginning of the Secondary period was marked by new volcanic action, trifling indeed, if we compare it to the intensity of volcanic power displayed during the Silurian time, but yet highly interesting. I allude to these local outbursts of hot vapours, gases and waters, charged with several minerals, which have taken place in many distant places of the Himalayas and their dependencies. The action is geyserian rather than volcanic, as no true volcanic rocks, that is, no lava, no scoriæ and no ash appear to have been discharged by these vents. The existence of this force is mostly manifested by the metamorphism it has caused in some of the upper beds of the Carboniferous limestone, and by the peculiar way it twisted rocks, then soft, in a manner which appears now incomprehensible, and totally abnormal to the surrounding layers. In some localities, however, it seems that the waters, erupting through the calcareous mud, were so rich in felspars, that this crystallised in

*This remark applies only to the Punjab and the mountainous districts studied in this paper.

minute crystals which now form a sort of intrusive band of a friable incoherent rock.

When this geyserian action subsided, the Palæozoic animals had died out.

92. I now enter upon debatable ground. I have said before, that the salt, gypsum and red marl of the Salt Range-and I need hardly say the gypsum and red marl of Spiti, the gypsum of Rukshu (and that of Rodok ?), and most probably the salt of the Yarkandkash valley, and also that of the Lataband mountains in Badakshan, all belong to the same epoch and have probably a common origin. I have said before that, this Saliferian formation has been placed by Dr. A. Fleming in the Devonian. Dr. Jameson makes it superior to the Carboniferous; Major Vicary and M. Marcadieu believed it to be Miocene or Pliocene; some will have it volcanic, others sedimentary; but nobody gives a good and well defined section of the relations of this formation to the rocks above and below it.* This is much to be regretted, and I will not increase the confusion by discussing here the reasons which make me believe that the salt and gypsum of the Himalayas belong to the Trias or the Permian. My opportunities of observing the Saliferian formation have been few and of short duration, and I have no good section to give in support of my opinion. I shall therefore refer the reader to the note to para. 64, and proceed with the next formation.

93. Whatever had taken place beeween the end of the Carboniferous epoch and the beginning of the Jurassic, it appears tolerably evident that the Jurassic sea bathed the shores of a long strip of land or succession of large islands, very similar to those which the Carboniferous sea had bounded. The Jurassic sea does not appear to have been much deeper than the Carboniferous one had been; the same impurity of the limestone is noticed, the same admixture of sand and clay with the calcareous matter, the same rarity of clean drifted sands, the same prevalence of thin-bedding, false-bedding and continual

* Dr. A. Fleming gives some sections in his Report on the structure of the Salt Range; but only two of these show the relations of the salt marl to the Carboniferous limestone, and in one, sect. No. VIII., a number of more or less theoretical faults are introduced which, if placed at the base of the mountain limestone escapements, would then make this rock inferior to the salt. Another section, No. VII, shows an anticlinal across a ravine, and then the salt marl appears indeed to be placed under the Carboniferous limestone.

change of the nature and weight of materials. All these conditions, and the frequency of ripple marks, indicate a shallow sea easily influenced by heavy out pours of muddy waters from the land. The thickness of the Jurassic rocks vary veries much, and the extent of the beds is limited to very small areas, compared to those of the Carboniferous. This is probably due to the deposition taking place in creeks of a deeply indented coast, and in great part to the oscillations of the land and sea bottom, causing in some localities repeated denudation of materials newly deposited, and in others a steady sinking and consequent thickness of formation. The fossils being frequently much deformed, is a good evidence of these oscillations having taken place.

The Jurassic beds have always been considered conformable to the Carboniferous. I am inclined to believe that this conformity is only apparent. The dip of both formations is generally great, seldom under an average of 45°. In such highly up-tilted beds, a difference of a few degrees is not easily appreciated, unless a careful measurement is taken, and I fancy that most writers have been satisfied with an approximation. However this may be, there is no doubt that the Jurassic limestone presents, in very many places, indeed in most, the appearance of having sustained very sharp local upheavals, soon after the end of the Secondary period, but of little extent; and here again we find the salt, gypsum and red marl always underlying these sharp and dome-like anticlinals. We remember how Sheikh Bodeen is thrown into a succession of short, gothic, arch-like anticlinals; and that under the Jurassic beds the Saliferian are to be seen, perfectly conformable to the limestone and following it in all its oscillations. At Maree on the Indus, a similar appearance occurs: thick masses of salt, gypsum with bi-pyramidal crystals, quartz, red marl and magnesian mud stone more or less cellular, support a very sharp anticlinal of Jurassic limestone; and the Saliferian and Jurassic are conformable not only in general dip, but in all the details of the fold. Moreover, both the Silurian and Jurassic dip S. (2 or 3 degrees E.) and N. (2 or 3 degrees W.) on both sides of the anticlinal dip, which are not the usual ones of the other rocks of that portion of the Salt Range, the Nummulitic and the Miocene dipping N. E.

Whether these local upheavals are merely due to the swelling of

the gypseous beds from the change of anhydrite into common gypsum by absorption of water, is more than I can say. The Saliferian beds would naturally break, dislocate and lift up the superincumbent Jurassic when swelling itself into undulations. We should thus obtain undulated beds of Saliferian and Jurassic. Let such undulated layers be submitted to the lateral pressure which must have accompanied the great upheaval of the Afghan-Himalayan system, and we have the undulations folded into arches and sharp bends.

The Saliferian and Jurassic have been very much denuded, their debris being extremely abundant in some beds of conglomerate and sandstone of the Miocene, especially on the western side of the Indus, in the districts of Kohat and Bunnoo.

94. There are but few traces of the deposits which may have taken place between the Oolite and the Nummulitic, and I have never myself seen any cretaceous rocks in the western Himalaya* or the Afghan mountains, neither have I found any pebbles with cretaceous fossils in the conglomerates of the Miocene. From the development of considerable vegetation in the shales near the base of the Nummulitic formation, it is evident that a steady rising of the land went on during the time of the upper Jurassic and Cretaceous periods, and with such a rising we would naturally associate the great denudation of the Jurassic beds, soon after their deposition. Little doubt can be entertained that during the Cretaceous period, the Himalayan and Afghan islands had become united into a continent of considerable extent, traversed by chains of extinct volcanic ridges, and therefore receiving an abundant rain-fall which caused great denudation. We know how quickly volcanic mountains decay, when once they have ceased to receive fresh supply of ejecta. I believe that the cretaceous beds which have been found in and near the Himalaya are very limited in extent, even more so than the Jurassic beds. The small horizontal area of these Secondary beds contrasts widely with the great superficial extent of the Carboniferous, the Nummulitic and Miocene formations; and yet when they do occur, the Jurassic beds at least have considerable power. A continent with a deeply indented coast appears to be indicated by these peculiarities of the Secondary beds.

* Dr. Stoliczka has found Cretaceous rocks in the mountains of Spiti. Editor's note.

95. The Nummulitic epoch must have been a long one, if we can judge by the thickness of its deposits. There does not appear to have been any violent volcanic action, nor any great and sudden movement during the period, but there was a great deal of very slow and probably imperceptible oscillation. Thus we first find the base of the Nummulitic to be generally a sandstone without fossils,* this is gradually impregnated with calcareous matter, becoming a sandy, very impure limestone, full of shallow water fossils and containing only a few very small species of Nummulites. This has been therefore a period of slow and trifling sinking of the land, and it is probable that the sea never covered it by more than a few feet. Then the oscillation went the other way, and the land appeared again, and was covered by forests. Another slow sinking brought on a fresh incursion of the sea, which soon covered the forests (lignite) with a layer of limestone, full of large Nummulites and other shells. The depth of the sea was greater than before the growth of the forests, but it probably did not much exceed 20 fathoms. Another movement upwards again exposed the land, and again forests grew and formed thin seams of lignite. Again the land sank and the sea covered in the lignite-beds with calcareous mud. At first the depth was trifling, little exceeding 20 fathoms, but the sinking continued to the end of the Nummulitic period, and the limestone assumes more and more the appearance of a deep-sea formation as we get higher up the series. It is, however improbable that the volcanic mountains of the great bars of the Himalaya and Afghan mountains were ever covered by the Nummulitic sea, as no nummulite has ever been found amongst the central chains;† but that sea filled up the whole of the space between the arms of the great everted V formed by the Himalayan and the Afghan chains, and probably also bathed the outside shores of the arms of the V. This slow, gradual and long continued sinking of the land, during the deposition of the Upper Nummulitic formation, accounts for the appearance of no great depth in rocks which have

* Sometimes a fragile limestone with Planorbis, and probably fresh-water. See note to para. 66, chap. iii.

† Dr. T. Thomson reported having observed Nummulitic Limestone in Little Thibet at an elevation of 16,500 feet. But I much doubt the accuracy of the observation, and cannot help imagining that the Thibet nummulites are, like those of Manus Bal, weathered encrinite rings. See "Introduction,” page ii.