But in one respect the geological structure in this part of Garhwal is unique, so far as I know. The appearance of the Outer formations underlying the schistose series is not confined to one line of country, but is equally noticeable at nearly every point round the margin of the Inner formation, whilst up the Huil and Rausan rivers offshoots of the Outer formations appear among the schistose series, fitting in with them more like a piece of gigantic inlaid work rather than lying as unconformable outliers upon them.

When I commenced field-work last season, I was in a complete state of uncertainty as to which way to interpret the sequence of the rocks. On the one hand, I was exceedingly loth to believe that an enormous thickness of phyllites, schistose slates, schists, and even garnetiferous schists could normally belong to an era subsequent to the deposition of the ordinary slates, limestones, and sandstones of the Outer formations; whilst, on the other hand, I was equally averse to straining what then seemed to be the plain facts of the case in order to draw up a more plausible stratigraphical table. Besides, Mr. R. D. Oldham's work in Jaunsar Bawar1 and Mr. H. B. Medlicott's at Simla had already shown the extreme probability of an upper and comparatively younger schistose series normally overlying slaty and calcareous strata; and I could not of course neglect telling evidence of this kind, although so great a span of country lay between their working grounds and mine.

I was also in a state of uncertainty with regard to the Outer formations themselves; for they by no means preserved a uniform relation to one another; so much so that my first statement that the massive limestone overlaid the Tal beds has had to be discarded and reverse positions assigned to them.

The most pressing difficulty then was that of stratigraphical succession. Although the superficial relations of certain rock-series to one another had been made tolerably clear, it was not manifest which of them was really the newer, and which the older; inasmuch as sometimes they appeared in one order of superposition, and sometimes in the inverse order. In other words, the problem before me was to unravel their order of deposition in time, from conflicting appearances, due to disturbance of the strata ; for, in a region of true mountains, it is not enough to see one set of beds dipping beneath another set; but in every case the question must be put—is this the normal order, or is it an inverted order?

The presence, or absence of fossils, makes all the difference in the ease with which such a question is answered; though a single fossiliferous series is not sufficient by itself. But when two or more definite fossil horizons are fixed, among a set of formations roughly coinciding in dip; the sequence in time, evinced by those horizons, will necessarily proclaim the true time sequence of the whole.

Applying this principle to the comprehension of the problems before us, it was imperative to find out how the presence of the two distinct fossil horizons of the nummulitics, and the mesozoic Tal beds, in a region where the other formations are unfossiliferous, would help in settling the true order of those associated unfossiliferous rocks whilst it was extremely probable that if the Outer formations were once chronologically arranged, a clue would be obtained which would fix the age of the schistose series. Until recently, the unfossiliferous formations had not been met with in such fortuitous conjunction with the fossil-bearing series as to enable me to Mem. G. S. I., Vol. III, 1863.

1 Rec. G. S. I., Vol. XVI, 1883.

rightly deduce their true stratigraphical order, but the time has now come when I can speak with certainty on this point, and I can throw the whole of the strata of Western British Garhwal into comprehensive groups, arranged in true historical succession (see table of formations).

Within the confines of this paper I shall merely endeavour to show how I have arrived at my conclusions, by reference to the geology represented by the accompanying maps and sections.

Referring to the 1-inch map, it will be seen that within the general curve made by the Ganges there are a set of boundaries, marked as faults, each roughly parallel to one of the reaches of the river, and, in result, giving a compound boundary somewhat resembling the course of the river. Within this boundary, except for a narrow band up the Huil river, the rocks composing the Inner formation agree in being very compact purplish quartzites, without much granular structure visible; glossysurfaced slates, generally slightly purplish, or considerably metamorphosed into schistose slate, and schists. On the other hand, outside the boundary there are fairly regular groupings of the Outer formations, viz., the nummulitics, the Tal (mesozoic), the unfossiliferous, massive, blue-grey limestone and the purple slates and volcanic breccia.

The first point to which I would call attention is that the order of superposition (whether a true order or not) of the Outer formations, up the Ganges valley, is from the purple slates and ashes below, up through the massive limestone and the Tal to the nummulitics, which are brought to a check by the boundary (see sections AA, BB). Now, in order to discover whether this order is a normal or inverted order, we must reason in this way. The nummulitics being by their fossil contents of later age than the Tal beds, which are mesozoic in age, it follows that the apparent position of the latter, dipping underneath the former, represents the true original sequence in which they were deposited. That being so, the dip of the massive limestone beneath the Tal beds must also be an original true dip; and likewise that of the purple slates and ashes beneath the limestone.

Thus, the present arrangement of the group of Outer formations, as seen near the bend of the Ganges at Lachman-jula, from the purple slates and ashes up to the nummulitics, must be the original historical one; and so part of the previous difficulty is solved once for all.

Coming now to the nature of the boundary between the Outer and Inner formations, and the question of their relative positions: it is certainly, at first sight, a most astounding coincidence that the Outer formations in nearly every locality should persistently dip towards the presumably older schistose series; especially when we have just learnt that the Outer formations, among themselves, are in a natural order. This occurrence, so marked in many places, and the lie of the Outer formations completely encircling the schistose area, make it difficult to get rid of the first impression already alluded to that the whole is a synclinal trough, with the Outer formations below, and the Inner above. One seems almost driven to conclude that if a boring were sunk through the centre of the schistose area, we should inevitably strike the Tal beds below. The very often curved direction taken by the boundary between the Outer and Inner formations, whereby it wanders irregularly, sometimes even V-ing inwards with the inequalities of hill and valley after the man

ner of a moderate dip-plane outcrop, further enhances this belief; whilst the apparent synclinal, into which the schistose series itself is thrown, seems to clinch the matter. Indeed, if it were not for certain facts, which are at the same time pure stratigraphical accidents, and some elementary reasonings, which might be passed over, I venture to think that the above would be the natural and most plausible interpretation of the features. But nevertheless these facts, not only render the above interpretation unacceptable, but emphatically negative it; whilst I hope I shall be able to show that, owing to the exigencies of mountain structure, apparent anomalies of the above kind are certain to present themselves.

First, as to the facts: in numerous instances, as the map testifies, the Tal beds, dipping down against the schistose series, are not in direct contact with them; but there is an intermediate deposit of shales, clays, and earthy limestones of nummulitic age, which also, in like manner with the Tal, dips down against, and apparently under the schistose series from several sides. And I wish it to be understood that this statement is no general one; but the two widely distinct rocks, namely, soft shales, and highly metamorphic schists and quartzites are in actual contact, without any semblance of what could be called a transition rock. Thus, if the Tal beds in reality continue beneath the schistose series, the nummulitics, where present, also do the same: that is to say, a soft, shaly, tertiary rock, not only must lie as a foundation on which the schists are piled, but also must be beneath them in direct contact. To satisfy a condition of this kind the most glaring case of selective metamorphism would be totally inadequate; it results then that the schistose series must be older than, and therefore normally below the whole of the Outer series, including nummulitics, Tal, massive limestone, and the purple slates and ashes. The preservation of the thin deposits of nummulitic age may be called a stratigraphical accident, but the key it gives to the chronological order of the Outer formations, renders any other interpretation of the above impossible.

But if this is so, the false position of the Tal beds and the other Outer formations, where present, cannot be looked upon as nothing but a coincidence; there should be some inherent necessity for such a steadily anomalous position, a position of being, so to speak, tucked in all round under the margin of the inner and older formation, a necessity due, probably, to the exigencies of mountain structure. I think the following considerations will make clear what this necessity is. During the deposition of the nummulitics, the whole of that portion of the Himalaya where they now exist, must have been beneath the sea. Between then and the present time they must have been raised into a mountainous tract; and from the inclined position of these same tertiary beds we conclude that the cause which tended to raise the hills into existence was a lateral compression, acting chiefly south-west and north-east, though complicated to a certain extent by compression in other and opposite directions. Whatever be the immediate cause of this lateral compression, into a discussion of which there is no call now to enter, it is sufficient for present purposes to notice that its effect was to urge the crust of the earth to take up less horizontal space than before. This could only be brought about in two possible ways; namely, by corrugation and by faulting. That the former actually took place, we have abundant evidence; and that the latter, viz., the snapping and tearing of the strata, whereby faults would be produced, also happened, we can well believe.

But if we now enquire as to what sort of faults would be produced, we shall be obliged to own that only such as could enable the strata to take up less horizontal space, in compliance with the lateral pressure, would be of any service in relieving the state of strain. From this we see that a vertical fault, having no effect of this kind, would be of rare occurrence. On the other hand, a reversed strike-fault, inclined at some angle with the vertical, would not only relieve the constant strain and bending of the rocks, but would directly contribute to the horizontal compression of the region by allowing the rocks on one side to work up over those on the other side; thus increasing the vertical thickness of the earth's crust at the expense of its horizontal extension. There being such a manifest relief accruing from such faults as these, it is but natural to suppose there would be a strong tendency to their formation. But their direct consequence is to bring about stratigraphical complexities of precisely the kind we have to deal with, by forcing older beds over the top of younger formations. On the other hand, a fault with the down throw towards the hade would obviously increase the horizontal extension of the rocks, and there could be no predominating tendency to its formation in a tract of country subject to great lateral compression. It seems, indeed, impossible to imagine them occurring in any mountainous tract, other than some few mountains due to great vertical elevation; except as dip-faults, or as secondary results due to local and intermittent relaxations.

There is thus a good à priori reason for expecting to find reversed strike faults, and them alone on a large scale, in the district under discussion. But that the boundary of the Tal beds with the schistose series should be so persistently a reversed fault on the dip side, an uninjured synclinal of the younger rocks scarcely ever (in only one instance that I know of) being preserved, does seem a little remarkable, until we remember that the Tal beds, save for the slight capping of nummulitics, must have been the uppermost rocks at the time when the elevation consequent on the lateral pressure began, and therefore it would only be in very favourable conditions that any trace of them would survive from the rapid denudation consequent on emergence from the sea and subsequent atmospheric waste. Simple folding would no doubt tend to bring this preservation about by allowing the anticlinal folds to be swept away whilst the synclinals were saved by being depressed below the action of the denudation agents; but it would not give results so decisive as that ultimate phase of a synclinal1 (in combination with an anticlinal) developing a thrust plane, or reversed strike fault; whereby the Tal beds, and in some cases the nummulitics, would be actually thrust and buried under a capping of older beds; and so protected by them from all subsequent wear and tear of denudation. We are obliged to admit, from the evidence before us, that the synclinal, except in one instance, was found insufficient, and the sigma-flexure and reversed thrust plane sufficient, for their preservation; and in believing this we shall do no violence to thought, and the apparent remarkable coincidences become reasonable necessities.

A certain additional weight is given to these conclusions, when we remember that the position of the Siwalik and other upper-tertiary rocks with regard to the Outer formations, where they front the plains, is exactly what that of the Outer formations is

1 Called a sigma-flexure, folded flexure, reflexed fold or overfold. See Heim's Atlas "Untersuchungen über den Mechanismus der Gebirgsbildung," and Lapworth's "Secret of the Highlands." Geological Magazine, 1883.

with regard to the schistose series. It has long since been shown by Mr. Medlicott and others that the later tertiary rocks (about the age of which there is no doubt) dip down against older Himalayan rocks, from which they are separated by a reversed fault; and there seems every reason to believe that the causes which produced this condition were simply a repetition in more recent times of a similar type of earth-movement to that which I have advocated above. In ages to come, when the Siwaliks have been worn away at their outer edge more completely by denudation, there will be left a narrow band of rocks analogous to the present reduced state of the Tal; and presenting the same phenomena of persistently dipping beneath much older beds.

It has now been proved: first, that the apparent order, among themselves, of the Outer formations near the Ganges, is their natural historical order; and, secondly, that the Inner schistose series is of older formation than the whole of the encircling Outer series.

It is satisfactory to note as a favourable issue that the complete proof now established shows that the chronological succession of the rocks as here interpreted, is a natural, and not a strained one: for, first, the more highly metamorphosed rocks are found to be the oldest; whilst decreasing age is similarly marked by a commensurate decreasing metamorphism. Secondly, the non-fossiliferous formations pose naturally as older than those which display fossils, whilst the latter, as we should expect, are together and the youngest; and, thirdly, the difficulty of the Kalogarhi gneissose mass is overcome; for since the schistose series, among which it appears, are the oldest beds, its presence whether it prove to be an archæan gneiss, which under pressure and high temperature, at extreme depths, has at some period been liquified, and so merited the name of granite; or whether it be an intrusion of subsequent date, i.e., a granite in which contemporaneous or subsequent foliation has been superinduced by pressure-in either of these cases there is nothing so abnormal as to be improbable.

In this paper I have only touched on the main difficulty, ignoring for the present all less prominent questions, which a close scrutiny of the maps will suggest, e.g., the unconformability of the members of the outer series amongst themselves and the marked unconformability indicated near the Sour and Kotedwar glens between the Tal beds and the schistose series. The map there shows the great thrust plane absent, and the Tal beds and nummulitics in an undisturbed synclinal. Hence the Tal beds are disposed indiscriminately on the schistose series, the purple slates, &c., and the massive limestone; a condition implying contortion and denudation such as to expose the schistose series at the time when the Tal beds were deposited. I have also refrained from going into more descriptive detail than was absolutely necessary to bring out the points of chief significance.

Above all, it seems to me that the foregoing results afford a useful lesson in enjoining the utmost caution and accuracy of mapping; in order that, amongst such intensely folded rocks, a right distinction may be drawn between ordinary dip planes and thrust planes; especially when the latter have a perfect parallelism in strike with the strike of the rocks, and hence cannot readily be recognized as differing in any way from ordinary dip planes. Extreme circumspection in this matter is a sine quá non if good work is to be done in mountainous regions.