argentiferous galena occurs, accompanied by green fluorspar. Traces of copper ores are found in the clay-slate of the Ura-Júke. According to Pander, gold and turquoises are found in the granite and porphyrite of the Bukantán hills.

Mesozoic strata containing land-plants rest on the paleozoic and metamorphic rocks. According to D. Romanofsky, many of the plants found in the mesozoic series of Turkistán are identical with jurassic, others with rhætic, and some even with triassic species. The entire mesozoic series however is of fresh water origin, marine forms being absolutely wanting. Romanofsky believes that Turán formed part of a continent from early triassic to cretaceous times, which extended a considerable distance eastwards. Neumayr looks upon it as the Turánian island in a jurassic ocean.

Jurassic deposits are only met with at very few localities in Turán; but they are of great practical importance since they contain coal-seams. Coal-basins of importance have so far only been found along the borders of Turán, in the valleys of the rivers Bodam, Seiram, and in Ferghána. In the latter area they are well developed. Others may possibly also be found along the western borders; at present they have only been shown to exist in Mangishlak. The jurassic deposits rest invariably discordant on the paleozoic rocks; whereas the cretaceous strata and tertiaries are usually conformable to them. Only locally is there an unconformity visible between the jurassic and the overlying cretaceous formations.

Cretaceous and terti

The cretaceous and tertiary deposits are the principal rocks of Turán where they reach an enormous thickness; altogether about 2,000 feet in Ferghána and 5,000 in Hissara. They are always closely connected and are conformable to each other. The strati graphical features are most complicated in the hill-ranges, especially in Keless and in Ferghana; but the further away from the hills of Eastern Turán, the simpler becomes the structure until, in Central Turán, the strata are perfectly horizontal; an exception forms the Sultanis-Dagh and Bukantan (in Kizilkum), where a northwest strike prevails.

The sedimentary strata of Turkistán. Plant-series.

aries.

The lithological character of the cretaceous deposits varies considerably. Where they are well developed, they consist of a great thickness of bright-coloured beds of gypsum, clay, marls, and shell-limestone, with micaceous and ferrugenous sandstone. The sandstone, limestone, and marls are remarkable for their regular development, whereas the clay-beds and gypsum deposits often thin out at short distances. The first group of rocks is found all over Turán, the other only in the Tian-Shán, Pamir, and Alai; in Ferghána, for instance, the complete series of rocks is found; whereas only the first group of beds occurs near Kilef on the Oxus,

Away from the eastern hill ranges the lithological character of the rocks becomes modified-the shell limestone passing into oolitic and dense limestones, and the bright coloured (red and green) micaceous sandstones becoming more uniformly light grey and yellowish coloured. The fossils are mostly so badly preserved that a division of the Turánian cretaceous group into horizons is hardly possible. Romanofsky divides the cretaceous group into two horizons, namely, an upper or Ferghána formation, containing numerous species of ostrea (near the Oxus mouth some Ammonites of senonian types also occur), and a lower horizon characterized by some ostrea, rudistes, echinderms and a few brachiopods. In Ferghána, the creta

ceous formation yields naphta, especially at points where there is a change in the strike of the beds. According to Konshin, the naphta and ozokerite of the Western Turán belong to a younger geological horizon than the Ferghána deposits.

Tertiaries.

The tertiary deposits are closely connected with the cretaceous group in central Turán; they are unfossiliferous in the spurs of the TianShán, casts alone occurring and then only in a very few beds. Nummulitic limestones occur on the Aral, but are not found elsewhere in TurkisThey are overlaid by lower and middle oligocene fossiliferous deposits (sandstones, clays, and limestone), followed by miocene limestone and sarmatic clays.

The tertiary series of the Tian-Shán is poor in fossils and can scarcely be divided into separate horizons. Romanofsky believes that eocene, oligocene, and miocene formations are represented, overlaid by a largely developed pliocene formation.

As eocene, he considers strata with Sphenia rostrata, La m., Modiola subcarinata, Lam., Modiola jeremejewi, Rom., and Avicuta trigonata, Lam. From the oligocene he cites Alligator darwini, Ludw., Osirea raincurt, Desh., Ostrea longirostris, Lam., &c.

tán.

The miocene and pliocene chiefly consist of conglomerates and sandstone beds, and are closely connected with the younger Aralo-Caspian formations. A characteristic Valvata is found in the largely developed pliocene deposits.

It was observed that the tertiary formations also change in lithological character further away from the hills; the shell-limestones disappearing and being replaced by sandstones. The green gypsum and rocksalt-bearing clays which are so largely developed in the east gradually thin out westwards, and the great salt deposits disappear altogether. The coarse conglomerates pass into fine-grained varieties.

According to Mushketoff, it appears that the tertiaries of the Tian-Shán have been deposited near a coast line; whereas those of the Aral neighbourhood are of purely pelagic nature. He believes that the nummulitic sea had a considerable depth which gradually decreased in the following periods: for instance, the UstUrt became dry land during the sarmatic epoch, whereas the lower Sir-Dariyá and Oxus regions remained a marine area till pliocene times.

The pliocene marine area was the forerunner of the present Aralo-Caspian basin, which gradually broke up into several smaller and isolated lake-basins, many of which exist to the present day, although constantly lessening in size and depth.

The entire series of formations, from the cretaceous to the younger tertiaries, forms one connected whole in which the boundaries between the several groups are exceedingly obscure, owing to the gradual changes which took place during the deposition of the series. During the whole period of time, from the cretaceous to the pliocene, Turán was covered by a sea, which spreading over the jurassic continent existed at first as an open ocean. This gradually changed to a mediterranean sea; and after deposition of the sarmatic detritus, it became a completely landlocked basin, which steadily lessened in extent until it is now represented by several isolated lake areas.

The stratigraphical features prove that contemporary with this change in the extent of the ocean area, an uninterrupted process of folding took place in tke surrounding land area; the outlines of the present hill ranges having begun to form before the deposition of the cretaceous formations.

The Aralo-Caspian deposits are of very varying thickness. Generally they are yellow or greyish-blue sandy clays with fine, often false bedding, completely corresponding lithologically with the upper deposits of the Kalmück desert. The brown and dark blue clays and white quartz sandstones of the Caspian deposits are entirely wanting in Turán.

The fauna of the Aralo-Caspian formations agrees with the living fauna of the Aral and Caspian seas. Cardium edule, L., Dreyssena polymorpha, Van Ben., Neritina liturata, Eichw., Adacana vitrea, Eichw., Hydrobia stagnalis, L., Anodonta ponderosa, Pfr., and a sponge, Metshnikowia tuberculata, Grimm, were found in Karakum, north-east of the Aral. These are forms, which in the Caspian live close in shore, in depths at most of 8 fathoms: Cardium edule being alone found at greater depths. Hydrobia stagnalis, Metshnikowia tuberculata, and Anodonta ponderosa live at still lesser depths. Nearer the Aral was found Lithoglyphus caspius, a species which occurs in the Caspian at the present day at depths varying from 7 to 108 fathoms. The Aralo-Caspian deposits in Turán are to be considered therefore as shallow-water formations.

Grimm has shown that the distribution of the living fauna of the Caspian is closely connected with the existence of great deposits of drift-sand. Where the coast-line is influenced by waves of drift-sand, he found animal life almost entirely wanting, a fact which, as Mushketoff believes, explains the sporadic occurrence of fossil remains in the Aralo-Caspian deposits.

The boundaries of these deposits can only be observed here and there. According to Mushketoff (contrary to the opinion of Barbot de Marni), the Ergeni hills west of the Caspian formed the old shores, and near the parallel of Manytsh a narrow channel formed a connection with the Pontus. Ust-Urt and the Mugojár range partly interrupted the connection, forming a long and narrow promontory in the Aralo-Caspian sea, and thus divided it into two separate basins. Sewerzoff thinks that the eastern basin may have reached as far as the Balkash lake; whilst to the southward the parallel of the Sultanis-Dagh may have partly formed the boundary.

The western and larger " Caspian" basin was the deeper. The channel which connected both runs between the two Bálkhan ranges; therefore along the Usboij to eastwards, into the so-called Sarykamish basin which was again connected with the Aral-basin through the strait of Aybugir between the Ust-Urt and Sultanis-Dagh.

Mushketoff expresses himself decidedly against Grimm's opinion that the Usboij was the former course of the Oxus.

Climatic changes brought about, and are still active in bringing about, the narrowing of both sea tracts of the Aralo-Caspian area; the disappearance of the channels of communication between both basins having been the first step in the direction of this shrinkage.

Mushketoff divides the drift-sand into two kinds, which form the littoral and the continental areal deposits.

Dry and cold north and north-east winds are the prevailing air currents during the dry seasons (summer and autumn); they absorb moisture, even such moisture as the soil may receive from rain and snow during the remainder of the year. This accounts for the rapid destruction of the cretaceous and tertiary sandstone, the decomposed material of which is carried away by the air currents of the dry seasons and re-deposited in the deserts of the Oxus and Sir-Dariya.

Crystalline and Metamorphic Rocks of the Lower Himalaya, Garhwal, and Kumaun. Section I, by C. S. MIDDLEMISS, B.A., Geological Survey of India (with map and plate).

INTRODUCTION.

Although much has been done recently, by Colonel McMahon, to bring to light the structures of the gneissose and other crystalline rocks of the N. W. Himalaya, and to frame reasonable hypotheses as to their cause, it must be admitted by every one who has looked into the subject that much yet remains unaccomplished. Microscopical work of high order, and detailed sections carried along one or two lines of country, are of the greatest service to geology; and the value of the papers which Colonel McMahon has from time to time contributed to this journal cannot be over-estimated. But for the rational account of an expanse of crystalline rocks something more is wanted which the above cannot supply; I refer to the actual tracing, step by step, of their relations in the field, or in other words, a patient mapping of the district on an adequate scale. Petrography at the present day is in such a state of transition with regard to many of its most vital principles, that the reading of any one district by the petrographical alphabet made out in another should be supplemented by deliberate endeavour to ascertain how far that alphabet can apply to the district, and by how much it falls short. With this end in view, it seems to me that fairly detailed descriptions of the arrangement, internal structure, lie, and associations of the crytalline rocks of this part of the Himalaya, drawn from exhaustive field observations, will constitute the safest and most impartial evidence attainable in the great questions always pressing for solution among crystalline and metamorphic rocks. I therefore propose to describe here and subsequently a series of localities where these rocks appear to show to the best advantage, the descriptions in

all cases being based on a thorough mapping of the area on the one inch scale. There will then be no need for elaborate argument and careful balancing of probabilities; each locality will tell its own tale, so far as structure, lie, and behaviour of the rocks can tell it; each district will depend entirely on its own internal evidence alone, and I shall endeavour as much as possible to banish the risky proceeding of generalizing over large areas. Detailed field exploration is the basis of geology, and sooner or later will extract the secrets from the rocks; and that without recourse to interpretations borrowed from other countries, which may be, but are equally likely not to be, applicable, and without appeal to principles and theories supposed to be universally accepted, but which may be after all only partially true.

DUDATOLI MOUNTAIN, BRITISH GARHWAL.

I will begin these descriptive notes by an account of Dudatoli Mountain and its surrounding spurs. I have already mentioned it in a previous number of the "Records," but without giving any details. This mountain, which rises to the height of 10,188 feet, constitutes the source of the Ramganga and the east and west Nyar rivers, all of which are tributaries of the Ganges. It is by far the highest mountain in the neighbourhood. Most of its upper slopes are covered with dense forest of pine, with an undergrowth of Ringall reeds; snow lying for a great part of the year among its sheltered and rocky fastnesses. Without implying more than is directly stated, I may say that it bears a very strong resemblance to Kalogarhi Mountain1 (Kálandanda) situated further to the south-west and near the plains and also to the Chor Mountain near Simla; both in the way it rises above the surrounding country and in its geological structure.

Speaking roughly, Dudatoli and its neighbouring hill spurs may be said to be composed of two sets of rocks, namely, mica schist and Geological structure. gneissose granite. These vary among themselves and shade off into different varieties which will be mentioned in due course. Their general surface arrangement will be seen from the accompanying map. The structural features are those of a quaquaversal synclinal elongated in the direction northwest-south-east. The uppermost and lowermost beds are schists of varying character; whilst the gneissose granite appears at numerous horizons insinuated among the schists, seemingly bedded with them, but in lenticular beds which expand or thin out with extreme rapidity. On the east, south, and west sides these lenticular beds are very prominently developed, whilst on the north there is only one thin but continuous band. The Dudatoli ridge itself is very nearly uniformly composed of the gneissose granite, the one or two thin bands of mica-schist which intervene being very unimportant. East of a line between the two Dudatoli peaks, the gneissose rock rapidly breaks up into a fringe of thin beds whose foliation planes dip roughly towards the south-west. They ultimately thin out altogether. West of the two peaks there is one blunted protruding mass in the direction of Gulek, whilst the long thin band alluded to above passes from Dobri trigonometrical station by Tarakakand, Banjkot, and Kotkhandia trigonometrical stations, and the villages Sountee and Hartur. Near the last mentioned place it is interrupted by a fault, but apparently was originally continuous into re-connection with the Dudatoli massif by means of the band

'Records, XX. pt. 1, p. 40.