Plate I.

NORTH SIDE OF THE BIRKET EL QURUN, LOOKING WEST.

SURVEY DEPARTMENT,
EGYPT.

THE TOPOGRAPHY AND GEOLOGY
OF THE
FAYUM PROVINCE
OF EGYPT

BY
H. J. L. BEADNELL, F.G.S., F.R.G.S.

CAIRO
National Printing Department
1905.

CONTENTS.

LIST OF ILLUSTRATIONS.

INTRODUCTION.

The geological survey of the desert surrounding the Fayûm was commenced in October 1898. At that time the area, although so near to Cairo, was little known; the Rohlfs Expedition maps marked the region as “unexplored,” and in fact with the exception of a publication by Schweinfurth, who had traversed the region from north to south, via Qasr el Sagha and Gar el Gehannem to Rayan, there was little information obtainable. The area being of considerable size (12,000 sq. kilom.) and almost unexplored, both geologically and topographically, the primary object was to construct as rapidly as possible a general map of the depression, at the same time laying down in broad outline the chief geological formations and trusting to future opportunity to examine in more detail places of special interest.

Commencing work at Sêla, on the eastern side of the depression, the survey was carried northwards along the east side of the cultivated lands and thence through the northern desert, up to the summit of the depression. After mapping westwards as far as the isolated hill-mass of Gar el Gehannem the work was temporarily suspended until, in the spring, the narrow defile of Wadi Muêla, and the Wadi Rayan, forming the southern part of the Fayûm depression, were provisionally examined.

In January 1901, samples of soil and water from the cultivated lands were collected as an experimental soil-survey, and the results have been published.[1]

During the winter’s work of 1902-03 a traverse was carried from Gar el Gehannem in a south-west direction through a hitherto unexplored part of the depression. On reaching a point midway between Cairo and the oasis of Baharia a connection was made eastwards to Wadi Rayan. In the winter of 1903-04 further exploration was carried out in the neighbourhood of Gar el Gehannem.

It will be convenient here to briefly relate the history of the discovery of the remarkable series of new and extinct animal forms, the recovery of which from the Fayûm deposits has created such widespread interest in the zoological world. When Schweinfurth crossed the region in 1879 he obtained fossil bones, which were examined and determined by Dames to be the remains of cetacea of the genus Zeuglodon, from certain beds of the escarpment west of Qasr el Sagha; these, it is believed, were the earliest vertebrate remains obtained from the Fayûm. During the early part of the survey of the district, remains of fish and crocodiles were frequently found in one of the beds of the Middle Eocene, probably on the same horizon as that from which Schweinfurth had collected. Fragments of bone were also commonly met with on a much higher horizon (i.e., near the base of the Fluvio-marine series) but nothing of particular interest was obtained, as no detailed search could be made at that time. In April 1901, during the survey of the western end of the Birket el Qurûn, some of the localities found to be bone-bearing in 1898 were re-visited in company with Dr. C. W. Andrews, who was in Egypt at the time and had accompanied the survey in order to obtain specimens of jackals, hares, etc., for the British Museum, in connection with the forthcoming work on Egyptian mammals. In one of these Dr. Andrews picked up several vertebrae which turned out to belong to a new species of Pterosphenus.

Further north, when descending the Middle Eocene escarpments at a place not previously examined, we crossed the outcrop of the bone-beds at a point where a considerable number of mammalian and reptilian bones lay exposed on the surface, many in an excellent state of preservation. The importance of the find was evident, and a short examination of the material on the spot enabled Dr. Andrews to pronounce the discovery to be of the highest importance from a palaeontological point of view.

Some three weeks’ work in the immediate neighbourhood resulted in a very good collection of vertebrates from the Middle Eocene beds, including several new genera afterwards described[2] under the names of Eosiren, Barytherium, Mœritherium, Gigantophis, etc. Moreover, a fossil tooth brought in by one of the camelmen from a point several kilometres to the north led to a careful examination of the lower beds of the overlying Upper Eocene formation, which resulted in obtaining well-preserved remains belonging to a new genus, since described as Palaeomastodon. All the material so far obtained was taken home to be worked up and determined at the British Museum and a preliminary description was published by Dr. Andrews in the Geological Magazine.

In the winter of 1901-02 the survey of the Fayûm was resumed with the special intention of following up the highest beds, those in which Palaeomastodon had been found. Continued search westwards eventually led to the discovery of the remains of a large and remarkable horned ungulate (Arsinoitherium), a preliminary notice[3] of which was published in the spring of 1902. Shortly after, the remains of several new smaller mammals and reptiles (Phiomia, Saghatherium), including the shell of a large land tortoise (Testudo Ammon), were obtained[4]. Further work in the winters of 1902-03-04 led to a great deal more material being obtained[5], mostly of course belonging to the same species, but including some new genera Geniohyus, Megalohyrax, Pterodon.

The amount of palaeontological material is now so large that the Egyptian Government has arranged with the Trustees of the British Museum for the publication of the whole in a monograph to be issued by the Trustees. The present report, therefore, deals only with the geology and topography of the district.

[1]A. Lucas, A preliminary investigation of the Soil and Water of the Fayum Province; Survey Dep., P.W.M. Cairo, 1902.

[2]Andrews, Extinct Vertebrates from Egypt. Parts I and II. Geol. Mag. N. 8. Dec. IV, Vol. VIII, Sept. and Oct. 1901, pp. 400-409 and 436-444.

[3]Beadnell, A Preliminary Note on Arsinoitherium Zitteli, Beadn. Survey Dept. P.W.M., Cairo, 1902. See also A New Egyptian Mammal (Arsinoitherium) from the Fayûm. Geol. Mag. N.S. Dec. IV, Vol. X. Dec. 1903, pp. 529-532.

[4]Andrews and Beadnell, A Preliminary Note on Some New Mammals from the Upper Eocene of Egypt. Survey Dept. P.W.M., Cairo, 1902.

[5]Andrews, Notes on an Expedition to the Fayûm, Egypt, with Description of some New Mammals. Geol. Mag. N.S. Dec. IV, Vol. X. Aug. 1903, pp. 337-343. Also Further Notes on the Mammals of the Eocene of Egypt (Parts I, II, III). Geol. Mag. N.S. Dec. V., Vol. I. March, April, May 1904.

Plate II.

BAHR YUSUF AT LAHUN BEFORE ENTERING THE FAYUM.

PART I.

TOPOGRAPHY AND STRUCTURAL GEOLOGY.

The Fayûm, a large circular depression in the Libyan Desert, is situated immediately west of that part of the Nile Valley lying between Kafr el Ayat and Feshn ([Plate XVII.])

The depression, which has an area, roughly speaking, of 12,000 square kilometres, is primarily divisible into three distinct parts—cultivated land, lake, and desert.

Section I.—CULTIVATED LAND.

The cultivated land has an area of about 1,800 square kilometres and, with the exception of the lake and part of the Wadi Rayan, occupies the lowest part of the depression. Cultivation is necessarily strictly limited to the area covered with alluvial soil. The latter, for the most part identical in origin and composition with the river-alluvium of the Nile Valley, covers a leaf-shaped tract between the bounding desert on the east side and the lake (the Birket el Qurûn) on the north-west. The easterly and central part of the cultivated area forms a more or less level table-land, from which the ground slopes gently away, especially on the north side, where the slope is towards the lake and very marked. The cultivated land of the Fayûm is directly connected with that of the Nile Valley by a narrow strip of low ground, a natural passage through the desert separating the Nile Valley and the depression of the Fayûm. Through this gap runs the natural canal known as the Bahr Yusef, which is practically the sole source of water in the Fayûm and irrigates the entire district.

The canal leaves the Nile Valley at Lahûn ([Plate II]), and follows a somewhat serpentine course through the desert for about 5 kilometres, irrigating a narrow strip of land on either side, which at Hawara rapidly broadens out into the wide cultivated area of the Fayûm. Once within the latter, the Bahr Yusef gives off numerous subsidiary canals which traverse the country in all directions, constantly splitting up into smaller branches until the water-supply is divided throughout the whole area. With the exception of the self-contained basin of Gharaq, on the south side of the Fayûm, the entire district drains into the Birket el Qurûn, which occupies the lowest part of the depression, to the north of the cultivation. The basin of Gharaq is irrigated by the Bahr el Gharaq, a canal which takes off from the Bahr Yusef soon after the latter enters the Fayûm[6].

The cultivated land of the Fayûm is traversed by two main ravines, cut down in many places to the Eocene limestone below the alluvium (Plates [III] and [V.]) At the present time these ravines carry canals for irrigating the lower parts of the district, and also act largely as drains to the higher lands. They were probably initiated by the escape of water through breaches in the Bahr Yusef during flood time, and have since been deepened to their present dimensions.

In addition to the main central cultivated area, the soil of which, as mentioned above, is essentially identical with that of the Nile Valley, large tracts of the surrounding country, more especially on the north, north-west, and west sides, are also covered with alluvial deposits. These latter, which include sands, sandy clays, and clays of a quite distinct type, represent the slowly formed accumulations of the quieter and more remote parts of the ancient Lake Moeris (and the earlier prehistoric lake). The material was mostly derived from the Eocene strata which formed the shores of the lake, augmented no doubt by a certain amount of very fine sediment drifted from the Bahr Yusef, and by sand blown in by wind.

It is noticeable that the thickest and most sandy deposits occur near the borders of the lake site; when close under the Eocene cliffs, as along the north side above the Birket el Qurûn, the deposits closely resemble those of the latter. The finer more calcareous beds occur further out and the true marls were accumulated only at some distance from the shores of the lake.

When in Ptolemaic times the lake became reduced to a fraction of its former size, large areas covered by these lacustrine clays were exposed and some portions were brought under cultivation. Subsequently, however, all these outlying districts were abandoned and became absorbed by the surrounding desert, until in modern times the cultivation was restricted to the central portion of the old lake bed, a portion almost identical with the area over which true “Nile Mud” had been deposited.

The construction during recent years of extensive irrigation works in the Nile Valley has made it possible to largely augment the water-supply of the Bahr Yusef to the Fayûm. High level canals are being cut in various parts of the district and already large areas of desert covered by these lacustrine deposits have been brought under cultivation, notably to the north of Tamia and in the neighbourhood of Qasr Qurûn. The approximate area covered with lacustrine deposits can be seen on the map and with a sufficiency of water probably the greater part of this area could be utilized, though the exact value of the soil compared with Nile deposit remains to be determined.

Section II.—THE BIRKET EL QURUN.

The lowest part of the depression, lying immediately to the north-west of the cultivation, is occupied by a sheet of water of considerable size, known as the Birket el Qurûn.[7] The lake, which has a length of 40 kilometres, and a maximum breadth under ten, covers at the present time an area of about 225 square kilometres. Sir Hanbury Brown obtained no sounding exceeding 5 metres in crossing the lake to Dimê, but according to the fishermen the depth increases towards the south-west.

Its long axis lies nearly east and west, and while on the north it is entirely[8] bordered by desert, along a large part of the south side the cultivated land approaches its shore, although even here a large area actually bordering the lake is waste salty land as yet unfit for cultivation. As already mentioned, with the exception of the Gharaq basin, the lake receives the whole drainage from the cultivated lands.

The Birket el Qurûn is the existing remnant of the ancient prehistoric lake which covered a large part of the floor of the Fayûm depression, and which in historic times was converted into an artificially controlled sheet of water—the celebrated Moeris—by Amenemhat I and his successors in the XII Dynasty.

Lake Moeris, being used as a regulator of excessively high and low Nile floods,[9] was of the greatest importance in connection with the irrigation of the Nile Valley. In more recent times, apparently under the Persians or Ptolemies according to Flinders Petrie,[10] Lake Moeris ceased to perform its function of regulator; since that time all water, except that required for irrigation of the reclaimed land, being carefully excluded, the surface of the lake has continually and gradually sunk to its modern dimensions.[11]

Lacustrine deposits, showing approximately the actual limits of the ancient Fayûm lake, can be traced over wide areas of now barren desert; these will be more fully dealt with later. The present lake-level is still continually sinking owing to an improved system of irrigation, by which a constantly decreasing amount of waste water drains into the lake. Its average annual fall has, during the last decade, been nearly half a metre,[12] and the slope of the land being very gradual, large areas have been reclaimed during the last few years, though whether the advantages derived from this constant lowering of the lake are not more than balanced by certain drawbacks is somewhat doubtful.[13]

With the new areas now being brought under cultivation the amount of drainage water finding its way into the lake will increase and the fall be checked. At the beginning of 1904 the level was markedly higher than in the previous winter, and a difference of even half a metre alters the shore line to a considerable extent, owing to the flatness of the land by which the lake is for the most part bounded.

Although under the present desert conditions practically no material from the surrounding desert is washed into the lake, doubtless a considerable amount of fine dust and sand is carried into it by the wind, especially during the violent sandstorms which occur frequently in the locality. The high cliffs which bound the northern shore of the lake throughout a portion of its length probably have the effect of checking the velocity of both north and south winds, thus causing a considerable amount of sand, which would otherwise be carried across, to be dropped on its surface. This material, together with the fine mud brought down by the canals on the cultivation sides, must have an appreciable effect in raising the level of the bed of the lake.

The phenomenon of the extraordinary freshness of the water of the Birket el Qurûn has been commented on by Schweinfurth, who shows that the degree of concentration of salt in a lake whose volume has been continually reduced, and to which salt has constantly been added, should be many times greater than the actual existing amount. An analysis[14] of the water at the west end of the lake (where the concentration is greatest, owing to the distance from the feeder canals) showed that the total salts amounted to only 1·34%, of which 0·92% was sodium chloride. Dr. Schweinfurth[15] concludes that the lake has a subterranean outlet, which alone would enable it to maintain its comparative freshness.[16] In this connection it is interesting to note the existence of distinct currents, which may possibly be caused by such outlets, in certain localities on the north side of the lake; and it is just possible that a careful survey of the lake itself would not only prove the existence, but show the exact position, of such underground outlets.

Most probably, however, the currents are simply local movements produced by temporary differences of level, which might conceivably be caused in such a large and comparatively shallow sheet of water, varying considerably in salinity in different localities, by wind and evaporation.

The comparative freshness of the lake and the possible presence of underground outlets are of the highest importance in their bearing directly on two of the most important questions in connection with the proposed utilization of the Wadi Rayan as a reservoir, i.e. what the leakage from such a reservoir would be and to what degree of salinity its water would attain.

Section III.—THE SURROUNDING DESERT REGION.

With the exception of the lake and the cultivated area the depression is practically entirely desert. The southern and south-western parts include the wadies Rayan and Muêla, where freshwater springs occur, surrounded by areas covered by a good deal of wild scrub. Apart from these, however, no springs occur outside the cultivated land.

The topography of the region is so intimately connected with its geological structure that an adequate description of the former is not possible without constant reference to the latter. Full geological details will, however, be reserved for later consideration.

Area and Limits.The part of the Libyan Desert dealt with in this report has, excluding the cultivated land and the lake, an area of some ten thousand square kilometres. While some portions have been examined and mapped in detail, others are still very imperfectly known, especially on the south and south-west sides. The irregular cliff-line forming the southern boundary of Rayan and the adjacent wadis may be taken as our limit in this direction, beyond lying an almost unbroken limestone plateau rising gradually and continually to the south. On the north and north-west the area under description is bounded by the southern limit of the great undulating high-lying gravelly desert-plateau which stretches with little change of character to the Mediterranean. On the east side the Nile Valley forms a convenient though not altogether natural boundary; while to the south-west our limit practically coincides with the boundary of the depression, where the floor of the latter insensibly merges into the general desert plateau.

Rocks forming the Area.The rocks forming the area within the above limits are almost entirely of sedimentary origin, the exception being a band of hard basalt intercalated at the very top of the series and exposed only on the extreme northernmost limit of the depression. The total thickness of sediments, from the lowest beds exposed in the bottom of the Wadi Rayan to the summit of the escarpments, a day’s march north of Tamia, is some 700 metres. These beds include every kind of sedimentary deposit—limestones, marls, clays, sandstones, sands and gravels, forming an ever-changing succession of rocks, varying considerably in hardness and capacity for withstanding the agents of denudation. It is not too much to say that the coming into existence of the Fayûm, with its plains, lowlying depressions, precipitous cliffs and escarpments, was largely dependent on the existence of this variable series of deposits.

Apart from the presence of sediments varying greatly in hardness and durability, the fact that the whole of the rocks have an almost constant northerly dip of two or three degrees is a point of prime importance. So small a dip may be scarcely noticeable in any one place, but over the large areas with which we have to deal its influence on the position and level of any individual bed is very marked and the topography of the region would have been essentially different if the strata had been quite horizontal.

Origin of the Fayûm.The unique character of the Fayûm is alone sufficient to show that special causes have acted in its production. Two main causes stand out:—(1) the presence of thick bands of comparatively soft arenaceous and argillaceous strata breaking up the usually continuous hard limestone of the Middle Eocene; (2) the effect of the Nile Valley fault in lowering the whole of the western desert (north of Assiut) relatively to the eastern. The former took place as the result of changed geographical conditions on the continent to the south at the time in question, with which however we need not deal here. On a homogeneous mass of rock weathering has little power to form depressions of any magnitude, and this is the cause of the continuous unbroken plateau which stretches southwards from the Fayûm, the underlying rocks being one continuous thick mass of hard limestone. Wherever softer intercalations are present differential weathering takes place, and all the great depressions of the Libyan desert owe their origin to the presence of soft easily denuded strata; if the great homogeneous mass of Nile Valley limestone had stretched unchanged westwards, the oases of Farafra and Baharia would never have existed. They owe their origin entirely to the presence of the underlying saddle of softer Cretaceous rocks. Similarly if changed conditions had not led to the deposition of soft beds of clay, marl, and sandstone, the western plateau would have continued unbroken northwards.

A comparison of the two sides of the Nile Valley between Cairo and Assiut shows that the tectonic movements, which largely determined the existence of the valley itself, resulted in a considerable lowering of the rocks forming the western side. This was brought about by differential movements along the north and south line or lines of fault, and by the presence of an east to west monoclinal fold which is especially well marked in the neighbourhood of Heluan. The depressions of the Fayûm would doubtless have existed irrespective of this general lowering of the western desert relative to the east, but denudation would have required an additional period of many thousands of years before the floor of the depression was low enough to allow of its actual connection with the Nile river.

As it has been maintained that the Fayûm is an area let down and enclosed by faults, it may be mentioned here that all available evidence points in an opposite direction; this question of faults will however be dealt with in detail later. The influence of the Nile Valley fault has been explained above and it must be remembered it is one affecting not the Fayûm alone but the northern part of the western desert as a whole.

For purposes of description it will be convenient to divide the whole region into three parts: first, the southern portion, including the wadis Muêla and Rayan; secondly, the central area, comprising the extensive plain forming the floor of the depression as a whole, and including the areas under cultivation and the Birket el Qurûn, as well as the desert separating the Fayûm from the Nile Valley. Thirdly, the northern portion, embracing all the rising ground between the floor and the northern rim of the region. These areas will now be taken in order.

Section IV.—WADI RAYAN AND NEIGHBOURHOOD.

This part of the Fayûm is of special interest in consequence of its possible future as a reservoir. Although the area has not yet been examined in detail by the Geological Survey it will be useful to bring together all the information that is at present available.

Colonel Western’s Survey.In 1882, as a counter-project to other irrigation schemes, Cope Whitehouse suggested[17] utilising as a reservoir the Wadi Rayan, a depression which had been referred to by Linant de Bellefonds.[18] At the request of Sir Colin Scott Moncrieff the Government deputed Colonel Western to make plans of the Wadi Rayan and surrounding country and to ascertain the capacity of the depression and its capability of being used as a reservoir. Liernur Bey under his direction prepared a contoured map, and Colonel Western’s report, plans, and estimates were published.[19] Some general details of the wadi and surrounding hills are given and the detailed survey showed that the 30 metre contour line (above sea-level) enclosed an area of 706 square kilometres (170,000 feddans). The lowest points of the depression were found at 42 metres below sea-level. The sand, scrub and springs are briefly referred to and the discharge of the latter is given as equal to that of a very slow-going four inch hand pipe, the water running out at about + 20 m. and disappearing in the sand. Wadi Muêla was found to be separated from the Rayan depression by sandhills and rock at a mean level of + 50 metres, the lowest point in Muêla being at + 25 metres. A line of levels was run from Rayan through Muêla to the Nile Valley, the highest point crossed being at + 105 metres; for fifteen kilometres the level was not below + 75 metres. In order to find the most suitable passage for a canal to connect the Nile with the Wadi Rayan two lines of level were made after a reconnaissance of the hills near Sidmant el Jebel: the southern, from Ezba Menesi Ali, near the Gharaq canal, to Mazana on the Bahr Yusef, being considered the best. Along this line the highest point was only at + 44·7 metres and the average + 35 metres along four kilometres. Borings were not made here but judging from the surface excavation would be mostly in soft limestone, sand, and conglomerate. A much shorter route is from Deshasleh on the Bahr Yusef over the hills about 5 kilometres to the south of Mazana or Sidmant into the Wadi Gharaq, a distance of 30 kilometres. This route was not however levelled but is fairly straight and apparently not much higher than the Mazana passage.

The survey of the + 30 metre contour line of the Wadi Rayan proved that there were only two outlets into the Fayûm, both on the northern side: these two openings are only from 400-500 metres wide and their lowest points are not below + 25 or + 26 metres.

Later Government Publications by Scott Moncrieff and Willcocks.In 1889 Sir C. Scott Moncrieff published[20] a further note, in which he briefly discussed the probable cost and benefits to be derived from the suggested reservoir, concluding that at least the project was one worthy of being thoroughly examined.

In 1894 the plans and designs in connection with the Wadi Rayan were published[21] and the possibility of utilizing the Wadi Rayan was examined by Sir William Willcocks, then Director General of Reservoirs, from an engineering point of view, and the questions of its probable cost and future utility were discussed. In this report it is stated that the routes proposed by Colonel Western in 1888 pass through salty marls and clays unsuitable for holding canals. Another route is suggested, which after leaving the Nile Valley crosses the high desert ridge in a straight line, passing through the so-called Wadi Liernur (Wadi Lulu of Cope Whitehouse); this depression is 12 kilometres long and has its bed some 24 metres below the general level of the desert. Plate 15 of the report shows the Wadi Rayan, the deserts between it and the Nile Valley and the cultivated land. The map was begun by Col. Western and completed by Willcocks. The lowest point of Wadi Rayan is shown as − 42 metres and the depression is separated from the Fayûm by a limestone ridge generally from + 34 to + 60 metres, except at two places where it falls to + 26 metres above sea level on a length of 600 metres. Within the + 27 metre contour line the wadi has an area of 673 square kilometres and a capacity of 18,743,000,000 cubic metres. Between it and the Nile Valley lie 30 kilometres of desert, of which 11 are occupied by a marked depression discovered by Liernur Bey in 1887. At the extreme western edge of the Nile Valley (here 20 kilometres wide) runs the Bahr Yusef. Comparing the proposed Wadi Rayan reservoir and the ancient Mœris and allowing for a difference of 4·5 metres between the levels of the Nile Valley in B.C. 2,000 and to-day, Willcocks assumes that the high water mark of Lake Mœris was at + 22·5 metres and its area 2,500 square kilometres, against 673 square kilometres of the Wadi Rayan at + 27 metres. It is pointed out that the ancient lake had the great advantage that in those days the Bahr Yusef was an important branch of the Nile, if not the main river itself, and the reservoir was connected with the Nile by a natural ravine of great length and short breadth, across which a massive embankment was thrown and destroyed annually, the surplus water of high floods being stored for the deficiency of low floods.

The published sections along the lines of borings put down show the different strata cut through by the proposed canal. The Nile Valley, along the line of the inlet canal, consists of hard clay 6 to 10 metres thick, lying on coarse sand. Along the outlet canal sandy clays and clays alternate to a depth of 10 metres. On entering the desert sands and sandy conglomerate, with gypsum and salt, are met with below the surface, then a yellow marl with salts, and finally a plastic black clay overlying the Parisian limestone. These beds are most extensive in the narrow neck of land between the Nile Valley and the Fayûm and to some 10 kilometres to the south of it. They rise to + 70 metres. There are some other marls inside the Wadi Rayan or in the adjacent depressions and as they have to be traversed by the canals form a serious factor, being easily dissolved in water; in consequence Willcocks chose the alignment of the inlet canal along the Bahr Belama where the extent of these beds would only be 2·5 kilometres against 9 kilometres on the alternative route marked on the plan. A narrow neck of land, some 15 kilometres in length, runs between the Fayûm and the depressions traversed by the proposed Wadi Rayan canal; this neck is the continuation of the salty marls and clays, but the limestone is near the surface and is overlain by a thin deposit of sand and pebbles, with freshwater shells on its northern slope at + 22·50 metres; the southern slope is entirely devoid of them. Willcocks points out that it is evident the ancient Mœris rose to + 22·50 metres but its water never penetrated into the Wadi Rayan. The report goes into details of inlet and outlet canals, discharge, necessary masonry works, cost, and compares the different reservoir schemes.

After a careful review of the whole question, the scheme, while considered perfectly feasible as far as available data went, was abandoned by Sir William Garstin[22] in favour of the less costly and more useful Nubian reservoir.

Plate III.

EL WADI, RAVINE NEAR QASR GEBALI.

Schweinfurth’s report on the probable salt-content in Wadi Rayan Reservoir.In an appendix[23] to the above report Schweinfurth discusses the question as to how salt the water of such a reservoir would become. He points out that the exact valuation of the salt which would be contained in this reservoir when the water had risen to + 27 metres cannot be accurately determined, owing to the absence of information on certain points. The maximum quantity of salt in the desert soil is estimated at 2% and this figure is used in his calculation, which includes the amount of salt which would be brought into the reservoir, (1) from the Nile during filling and in the extra water entering to replace that lost by evaporation in the lake and canals; (2) from the ground forming the bed of the lake (far the largest item); (3) from the bed and banks of the inlet canal, both in the desert and in the Nile Valley; and (4) from infiltration. The figure obtained is 7,500 million kilogrammes, equal to 0·04 per cent, or almost one twenty-fifth per cent of salt. This amount is only equivalent to half the salt existing in many of the well waters used in the country for irrigation. As Schweinfurth is careful to point out his calculation is based on maximum and assumed data.

Willcocks’ “Assouan Reservoir and Lake Mœris”.The question of the utilisation of the Wadi Rayan as a reservoir has recently been again brought to the front, notably by Sir William Willcocks in a paper[24] read before the Khedivial Geographical Society, Cairo. The author, after pointing out the value of such a lake, working in connection with the Assuan reservoir, discusses at length the position, dimensions, and functions of the ancient Lake Moeris. It is suggested that the main canal should be cut through the desert opposite Mazana and crossing the so-called wadis Liernur and Masaigega enter the Wadi Rayan at its easternmost point. These wadis would in time become covered with alluvium and be converted into valuable cultivated land. After examining the big ravines of the Fayûm, where similar beds are exposed, the author comes to the conclusion that the maintenance of canals in the saliferous marls, which form part of the desert through which the inlet canal would pass, would offer no particular difficulties.

With regard to the questions of leakage into the Fayûm and of the water of the lake eventually becoming salted, Sir William Willcocks says, “When the old Lake Moeris, or the present Fayûm, was full of water and 63 metres higher than the bottom of the Wadi Rayan and remained so for thousands of years, there was no question of the waters having become salted or having escaped into the Wadi. The Wadi was as dry as it is to-day and the great inland sea was always fresh.” As to the question of leakage into Gharaq the author considers that if water found its way into that depression it would be a distinct advantage, as such water could be pumped into the Nezleh canal and utilized elsewhere; he maintains at the same time that no leakage will take place. Incidentally it is mentioned that the Wadi Rayan is separated from the Fayûm by a limestone ridge, a statement which, as will be shown later, requires modification.

Wadi Rayan not yet examined in detail by the Geological Survey of Egypt.Until a detailed geological examination of the Wadi Rayan and neighbourhood has been carried out it will not be possible to form reliable opinions on many of the questions raised in connection with the prospective reservoir. The writer’s acquaintance with the area is limited to a traverse in 1899 from the Nile Valley through Wadi Muêla to Rayan and thence to Gharaq, and subsequently to a stay of a few days duration in the neighbourhood of the Rayan springs, after mapping the extreme south-west of the Fayûm depression. While the accompanying maps may be taken as representing fairly accurately the bolder topography of the region, they do not replace the older contoured maps of the floor of the depression and the country between it and the Nile Valley to the east, accompanying the report on “Perennial Irrigation and Flood Protection in Egypt.”

The following description of this part of the district is based on a traverse from the Nile Valley through the wadis Muêla and Rayan to Gharaq; the detailed geological sections measured and examined along the line of route will be given later.

Traverse from Nile Valley through Wadi Muêla to Rayan and Gharaq.Between the village of El Gayat and the mouth of the Wadi Muêla (16 kilometres to the north-west) stretches a gradually rising undulating gypseous plain, superficially covered with loose sand and rounded pebbles of quartz and flint. In occasional small hills the white limestone which forms the underlying rock is visible. Near the entrance to the wadi stands a somewhat prominent conical hill composed of hard whitish fossiliferous limestone passing down into more sandy and clayey beds. The bottom of the wadi is cut out in soft green and brown clays, its surface being covered with blown sand, fragments of limestone, flints and gypsum. From the mouth of the wadi the Nile Valley cliffs run north and south in a winding irregular manner. On entering the valley several outstanding flat-topped limestone capped hills are passed on the right hand; they are in part joined to the regular bounding cliff beyond; the eastern cliff is steep and well-marked, while that on the west only outcrops here and there, buried as it is in immense accumulations of blown sand, rising in places into definite dune-ridges. Wadi Muêla has a length of some 18 kilometres and lies nearly N.W. and S.E. The central part of its floor is a sandy scrub-covered area, the lowest points lying at about + 25 metres; just at the southern edge of the scrub stands a small hill composed of hard shaly clays capped by white limestone, surrounded by a saline, superficially dry. Holes dug in this are at once filled with excessively salt water, and by evaporation of the brine in shallow troughs supplies of white fairly pure salt can be obtained. The area is known as Warshat el Melh in Wadi Muêla.Warshat el Melh. Banks of reeds were found growing on the north side of the saline, the surface of the latter being here composed of a soft brown sandy salty deposit, caking here and there into a hard earthy impure salt.

In the lowest spots the saline frequently consists of soft wet sludge; its area is about half a square kilometre but the depth of the deposit is unknown. In the middle of the scrub-covered area to the north lies Ain Warshat el Melh, a pool of water, fairly fresh and drinkable, although ferruginous, measuring 10 by 5 metres in size and from 2 to 2½ metres deep. The water evidently rises from a spring on the west side, round which are fifty square metres of green rushes, with some larger bushes. The ground around and above is very saliferous; between the spring and the ruins to the north the ground is sandy, with many bushes and much scrub. This ground extends two kilometres to the west, whence it gradually passes up into great masses of drift sand; an occasional small outcrop of the top of the plateau above the sand is all that serves to locate the position of the buried cliff. On the east side the sandy ground with scrub extends about a kilometre, beyond which the plain gradually rises for another kilometre to the base of the cliff beyond, which is fairly steep and well-marked, though with an entire absence of indentations of any kind.

Der el Galamûn.Close to the north end of the valley, and about 33 kilometres from El Gayat, lie the ruins known as Der el Galamûn bil Muêla. At the time of our visit a new square stone building was in course of erection and five or six persons were inhabiting the place. There are several small palms scattered about to the south of the monastery and an excellent running spring of clear water five hundred paces to the south-west. A new well is being sunk within the premises. To the north of the monastery the eastern cliff takes a marked trend to the west for some three kilometres, whence it resumes a northerly direction, always maintaining its character of a steep well-marked escarpment rising some 100 metres above the floor of the wadi. At the corner of the cliffs the lowest bed exposed is a white limestone; this is overlain by gypseous clays passing up into sandy beds, the latter being surmounted by the white limestone capping the escarpment.

Wadi Rayan.We are here on the summit of the divide between Wadi Muêla and Wadi Rayan, the height of the floor being about + 105 metres; to the north stretches a gradually widening bay descending to the lowest ground of the Rayan depression. Immense accumulations of sand almost block the defile and stretch away to the east, and the hitherto well-marked cliff on that side bends back and is lost to view. On the other side however, the bounding wall gradually emerges from the dunes, getting more distinct as it is followed northwards until it becomes quite clear of the sand. The first glimpse of this cliff is seen a couple of kilometres west of the pass in an outcropping headland, the next point visible being some five kilometres further west. Between these portions of the cliff are one or two outliers, surrounded by quantities of blown sand. A depression known as Wadi Korif is reported to lie to the west, and much scrub and some water is said to exist there; such a wadi is marked on Schweinfurth’s map but apparently has not been examined.

Continuing in a N.N.W. direction high rather steep dunes occur on either flank, running N.N.W. and S.S.E. Between the dunes is a fairly hard undulating sand-flat affording an easy route; further on a narrow defile between the dunes leads down to the centre of the depression. The main areas occupied by blown sand are shown in the accompanying maps. The most interesting part of the depression is the bay lying to the south of the narrow well-marked promontory jutting out from the southern plateau, a huge pointer, as it were, in the direction of Gharaq; this is the Cape Rayan of Schweinfurth.

Springs in Wadi Rayan.The bay is on three sides completely enclosed by cliffs and its floor is thickly covered by a luxurious growth of wild scrub—chiefly tamarisk and ghardag; numerous isolated palm trees occur, especially in the neighbourhood of the water which exists at several points. There are three particularly good springs,[25] the positions of which are shown on the accompanying [maps.] According to Colonel Western’s survey the water emerges at about + 20 metres. In 1899 the water of the northern spring was found to have a temperature of 26°C. On our last visit we found an artificially constructed pool of two metres diameter and a depth of 30 centimetres; on the west side of this were two springs, marked by the motion of the grey sand rising and falling in the vents, down which a stick could be easily pushed to a depth of two metres. The output of these springs together amounted to six litres a minute; the water was quite clear and although soft and rather ferruginous not by any means unpalatable (see [analyses] below). The pool lies on an open bare sandy spot and is surrounded by scattered bushes, none of which however are within fifteen metres; a sand dune lies 150 metres to the south-west, with bushes and seven or eight young palms. The southerly spring has an output of 21 litres a minute, and its water does not differ essentially from that of the northern spring. Rising at the foot of a palm tree it forms pools on either side; thence it flows a distance of 20 metres into an artificially constructed shallow basin 2 to 3 metres across, from which it runs away down the slope and disappears after five or six metres. The east spring, which is situated on the east side of the dunes bounding the mouth of the bay, consists of a small hole cut out in soft sand. The water seemed good, although analysis shows the salts content to be high; this spring does not run, but if emptied the hole soon refills. The remains of old buildings occur near the well, in the shape of loose roughly squared limestone blocks, broken pottery, and remains of old walls; the latter are nearly level with the ground and very thickly and solidly built.

To the south of the promontory lies the so-called Little Rayan. Here there is a good deal of scrub, and water can be obtained on the lowest ground at a few metres depth, although there do not appear to be any surface springs.

Geology of Wadi Rayan in broad outline.The geological succession of beds exposed in the cliffs of the promontory is given later. Broadly speaking it consists of two thirty-metre bands of hard limestone separated by 68 metres of softer sandy and clayey beds. The lower of the limestone bands in places forms the floor of the depression but more frequently the latter is composed of the overlying sandy or clayey beds. The depression is bounded on the north side by the same succession, and, as far as could be judged from observations made on the traverse, the bed of limestone capping the ridge, and forming the plain stretching away to the Birket el Qurûn and to Gar el Gehannem, is identical with that capping the cliffs to the south, i.e. is the uppermost of the two thick limestone bands. At the two points more particularly noticed, namely, the spurs projecting southwards into the depression, 23 kilometres west and 18 kilometres W.S.W. of Gharaq basin, the sequence seemed to be the same as in the southern cliffs, although, owing to the northerly dip, the upper bed of limestone lies at a much lower level and the basal beds are not exposed at all. In both these localities, however, some of the underlying clays were exposed, as well as on the lowest spots crossed between the most easterly spur (18 kilom. W.S.W. of Gharaq) and the extensive dunes lying immediately west of Gharaq cultivation. These dunes, though of no height, have remarkably steep sides. In crossing Gharaq to the Fayûm cultivation occasional beds of yellow sandy limestone were noticed, but their horizon was not determined. Numerous bored blocks, probably belonging to the marine Pliocene, were observed scattered about. Apparently the uppermost thirty-metre band of limestone passes continuously northwards under the cultivated lands of Gharaq and the Fayûm; in the ravines of the latter this limestone is not observed, the soft limestones exposed below the alluvial deposits almost certainly belonging to the overlying Ravine beds. The country to the east of Gharaq has not been geologically examined and the exact locality in which the thick bed of limestone dips underground and is overlain by the succeeding beds is doubtful. Further north, in the desert ridge east of Qalamsha, we have observed the Birket el Qurûn beds and a section measured at this point is given later.

Character of Ridge separating Wadi Rayan from Gharaq and the Fayûm.As it appears to have been freely assumed that the ridge separating the Rayan depression from the cultivated lands of Gharaq and the Fayûm is formed throughout of solid limestone, it is important to point out that, on our assumption of the identity of the beds of limestone capping the cliffs to the south and the plain to the north of the Wadi Rayan, the dividing ridge would in part be formed of the underlying arenaceous and argillaceous beds.

Question of leakage through dividing ridge.The absence of Nile deposit and freshwater shells in the Wadi Rayan will, when confirmed after a thorough examination of the area, afford the strongest evidence that the depression was never directly flooded by Nile water. The fact that the dividing ridge is probably everywhere above the highest level attained by Lake Mœris, and by the still more ancient prehistoric lake, is almost sufficient in itself as a proof of this. It does not however follow that there was not leakage through the ridge into the Rayan basin, as such leakage might conceivably have taken place to a considerable extent without the water ever having collected in sufficient quantities to form even moderate sized pools within the depression. The bottom of the depression is for the most part covered with soft porous sandy deposits overlying the Eocene bed-rock below, and at the present time the water of the Rayan springs, though continually running, at once disappears from sight, drains down to the lowest parts of the depression and is then gradually lost by evaporation or underground leakage. In the lowest parts of the depression this water is, as already mentioned, met with on digging to a very moderate depth.

A careful examination of the flanks of the ridge separating the Fayûm and Gharaq cultivated areas from Rayan might prove if such leakage ever took place. If such was the case the seepage was probably along the line of junction of the limestone and underlying clayey or sandy beds. Even if it were proved that there never was leakage from Lake Mœris into Wadi Rayan, it would not be safe to assume that the converse would not happen, as the dip of the beds is from south to north and this fact is one to be reckoned with. Judging from the nature of the Eocene beds forming the Wadi Rayan, my opinion is that leakage on a large scale would not take place, and that owing to the northerly dip any water that escaped from the reservoir would pass indefinitely northwards and would not find its way through the overlying limestone to the surface either in Gharaq or the Fayûm cultivation. A detailed examination of the local geology would, however, be necessary to prove or disprove this. As to the question whether the Wadi Rayan as a whole would hold water, as far as is known there are no faults or other fissures of any magnitude through which the water could escape. No doubt a good deal of water would be lost before the smaller joints and passages, which exist in all rocks, were silted up. Schweinfurth supposes that the freshness of the Birket el Qurûn is due to the existence of subterranean outlets, and such might also be found to exist in the Wadi Rayan. In any case the argillaceous deposits from such a lake would very soon form a bed to all intents and purposes impermeable.

Degree of Salinity.With regard to the extent of salinity of such a lake Dr. Schweinfurth’s figures are of considerable interest and value, although based wholly on assumed data. The greater part of the salt would be derived from the rocks and soil forming the bed of the reservoir and only by extensive sample collecting and analysis can reliable figures be obtained. We believe that in the lowest parts of the basin the salt content of the ground would be found considerably in excess of the two per cent used by Schweinfurth in his calculation, although his total estimate would probably be found well within the mark.

Section V.—CENTRAL AREA OF THE REGION.

Central Plain at the Fayûm Depression.The great central plain, forming the floor of the depression as a whole, is composed of a hard bed of limestone some thirty metres thick. This limestone, forming the uppermost member of the Rayan series, is, as already mentioned, almost certainly identical with that capping the cliffs to the south of the depression, and in all probability in the eastern extension of the plain under description underlies the whole of the cultivated lands of Gharaq and the Fayûm. The feature of the plain as a whole is its marked and constant, though low, dip to the north; so that its surface, bared by denudation of the overlying soft limestones of the Ravine series, over a distance of some twenty kilometres, is a true dip-slope, at the base of which lies a strip of low-lying country extending from beyond Gar el Gehannem through the Birket el Qurûn to the Nile Valley ridge east of Tamia. The central and lowest portion of this low-lying area is occupied by the Birket el Qurûn, the bed of which lies fifty metres below sea level and is thus the lowest known spot in the whole of the Libyan desert. Thirty kilometres south-west of the western end of the lake, at the base of the dip-slope of the central plain and immediately under the southern scarps of the great outlying hill-mass west of Gar el Gehannem, lies another low lying basin, which receives the drainage from a considerable area of the plain to the south-west. The latter, consisting of the limestone above-mentioned, is here superficially covered by gravel, and its dark undulating surface is scored by numerous shallow winding water-courses marked by an abundant growth of scrubby vegetation; some of the principal of these drain into the basin just mentioned and after heavy rainfall the water collects and forms a pool 600 metres in length by 100 to 150 metres wide. The base of the basin, at about 80 metres above sea level, is marked by a level deposit of silt of considerable thickness, the east end of the site being surrounded by great numbers of luxuriantly growing tamarisks. Other similar basins exist on the plain to the south, and under an isolated hill five kilometres W.S.W. several full grown acacias were noticed. On the low ground to the north-west of Gar el Gehannem, and at several points between it and the head of the Birket el Qurûn, similar silt covered areas exist, some being only from 30 to 40 metres above sea level.

In the extreme south-west of the region the limestone forming the central plain is gradually overlain by the succeeding beds, so that the ground rises imperceptibly to the level of the plateau separating the depression from that of Baharia, distant some two days march. On the eastern side, if the superficial alluvial deposits could be stripped off, the underlying surface of limestone, sloping from south to north, would not differ materially from the plain further west, except that here, at any rate north of Gharaq, the Rayan limestone is overlain by the basal beds of the Ravine series.

Section VI.—RIDGE SEPARATING THE NILE VALLEY AND THE FAYUM.

The desert ridge separating the Nile Valley from the Fayûm has, to the north of the Bahr Yusef, an average width of some ten kilometres; further south it narrows, until due east of Gharaq the ridge is barely 2½ kilometres wide. The highest points are situated to the east of Sersena and Qalamsha respectively.

In both these localities the Eocene rocks, consisting of clays alternating with beds of calcareous sandstone and sandy limestone (pp. 39, 40) are overlain by thick deposits of conglomerate and gravel, attaining altitudes of over 100 metres above the cultivated land below. From these summits the slope is usually very gradual on the Nile Valley side but much more rapid towards the Fayûm.

The ridge is cut down, however, to a comparatively low level in four localities; to the north-east of Tamia; to the east of Sêla, where the railway crosses; between Lahûn and Hawara, where the Bahr Yusef canal enters; and to the south of Qalamsha, where along the site of the proposed Wadi Rayan canal the highest point is only some 40 metres above the Gharaq basin and 27 metres above the adjoining Nile Valley cultivation.

Outline of earliest connection of Nile with Fayûm.One of the most interesting problems connected with the Fayûm may be briefly alluded to here—When did the waters of the Nile first obtain access to the depression?

As will be shown later the Fayûm was occupied by the sea in Pliocene times, when the great gravel accumulations and gypseous deposits were formed. Later the area became dry and denudation of the land surface completed the work of erosion already begun in earlier times.

In Pleistocene times drainage down the Nile Valley appears to have become definitely established and probably the river in the lower part of its course eventually washed up against and broke down the separating barrier of gravel between the Fayûm and the Nile Valley, so that part of its waters obtained access to the depression, formed a lake on the lowest part, and gradually rose until the whole basin, up to the level of the channel connecting it with the Nile Valley, became filled. Every year thousands of tons of sediment were carried in by the floods and spread out on the floor in the shape of a fan. Probably later, as the Nile level fell, the valley and the depression again became disconnected, until the more modern river, with its gradually rising bed, again attained the requisite altitude. In early historic times the alluvial deposits had probably silted up the lake in its southern central part, and when in the XIIth dynasty the district was first taken in hand by Amenemhat I this part of it must have had the character of a huge marsh, nearly surrounded by open water, rapidly deepening towards the north.

Section VII.—THE NORTHERN DESERT REGION.

The Plateau bounding the Fayûm depression to the north.All along the north-west and north sides the ground rises rapidly from the base of the dip-slope of the plain in a series of escarpments to the summit of the rim of the depression, averaging 340 metres above sea level. Northwards from the summit stretches a rolling pebbly desert, the prevailing character of which is a dark brown, relieved by lighter brown grey and yellow patches, and especially flecked by the light sandy slopes of the undulations. Although the latter seldom rise to any considerable height above the general level of the plain, from the top of the most modest eminence an immense view in every direction can frequently be obtained. The monotony of this desert is only relieved by the occasional belts of sand, which although extremely narrow in width, run for immense distances in almost absolutely straight lines, and in a N.N.W.—S.S.E. direction. Although none of these dunes actually reach the rim of the escarpment we may mention here the beautiful Ghart el Khanashat, an almost straight and apparently unbroken ridge of sand, extremely narrow but of great length. Near its southern extremity the width does not exceed 100 metres; the slopes on both sides are frequently as much as 30°. The commencement of the Ghart el Khanashat was observed on a march from Wadi Natrûn to Mogara; it lay some way to the south of a line joining those two localities but could not be accurately fixed from the line of route. The belt dies out 24 kilometres from the rim of the Fayûm depression, its termination being particularly abrupt, although the height of the ridge diminishes gradually throughout the last kilometre or two. The line of the belt if continued would almost strike the western extremity of the Birket el Qurûn; near its termination the desert is almost flat, the surface being finely gravelly, with numerous groups of silicified trees; tufts of coarse grass grow in some profusion on the sandy ground at the base of the ridge on either side. A fairly well-marked road from the Birket el Qurûn to the Wadi Natrûn passes the end of the ridge and continues northwards at a distance of 200 metres from the east side of the dunes, although apparently gradually diverging eastwards.

Except to the north and north-west of Tamia, where a somewhat extensive and fairly level plain exists, the ground, as already mentioned, rises from the limits of the central plain in a series of escarpments to the summit of the rim of the depression. These cliff lines are broadly speaking three in number and represent the escarpments of the three great rock-stages which build up the northern part of the Fayûm, i.e., the Birket el Qurûn series, the Qasr el Sagha series, and the Fluvio-marine series. It would serve no useful purpose describing these different cliffs in detail; their positions and characters are apparent on the accompanying maps. The intervening plateaux are for the most part dip-slope plains formed of hard bands of rock, which resisting denudation, are left protecting the underlying strata while the softer beds above are cut back at a comparatively rapid rate.

Desert west and south-west of Gar el Gehannem.In December 1902 and March 1903 a traverse was made through the unexplored country west and south-west of Gar el Gehannem, finally connecting up with Wadi Rayan. The highest escarpment, i.e. that of the Fluvio-marine series, dies out about 20 kilometres west of Gar el Gehannem, gradually merging into the undulating gravel-covered plain. The lower escarpments, those of the Qasr el Sagha and Birket el Qurûn series, continue to a considerable distance in a south-westerly direction, although gradually losing the characters of well-marked cliffs. In fact westwards of this the depression gradually shallows, until at a point some 50 kilometres south-west of Gar el Gehannem the floor has attained the level of the ordinary desert plateau, on which the outcrops of the beds of successive rock-stages follow one another in regular order from south to north, but without forming well-marked topographical features, as in the depression.

Hills, capped with dark hard ferruginous silicified grits and puddingstone, were met with in the extreme south-west extension of the depression; these deposits, which will be referred to more fully later, considered in conjunction with the similar beds occurring within the oasis of Baharia, and in the hills of Gar el Hamra, on the plateau immediately to the north-east of that depression, are of considerable interest and importance, especially in connection with the question of the position of the early rivers which in Eocene and later times brought down quantities of trees and animals, the remains of which are so abundant throughout the later Fayûm deposits.

Jebel el Qatrani and escarpments north of the Birket el Qurûn.The boldest part of the region is the area lying between the Birket el Qurûn and the summit of the depression to the north. All three lines of cliff are here high and precipitous, and the uppermost escarpment, well known by the name of Jebel el Qatrani, formed of a highly coloured series of sandstones and clays and capped for a distance of many kilometres by a thick bed of hard black basalt, is of a most striking character. The eastern extremity of Jebel el Qatrani is perhaps the most conspicuous point in the whole region; here the two conical black basalt-capped cliff-outliers, known as Widan el Faras, stand side by side, and from their summits the eye commands the whole region from the pyramid of Lahûn on the one side, across Rayan to the south, up to the extreme limits of the depression to the south-west. The rim of Jebel el Qatrani has a fairly constant level of about 340 metres above the sea. From Widan el Faras the escarpment trends northwards for a few kilometres before again resuming an easterly direction, which is continued till the well-marked bluff of Elwat Hialla is reached. From this summit the pyramids of Dashûr, Saqâra and Giza are visible to the north, as well as Cairo and the Nile Valley southwards, backed by the bluffs on the Eastern desert limestone plateau.

To the south the isolated peaks of Garat el Gindi and Garat el Faras form conspicuous landmarks on the more or less open plain which stretches to Tamia and the limits of the Fayûm cultivated lands. Eastwards the escarpments continue in a broken irregular manner; the upper ones are gradually lost in an undulating plain, while the lower eventually join those forming the northern part of the ridge separating the Fayûm from the Nile Valley.

[6]For fuller details of the cultivated lands, water-supply, etc., of the Fayûm, the reader is referred to the excellent description by Sir Hanbury Brown in his work The Fayum and Lake Moeris, London, 1892.

[7]“The Lake of the Horns,” so called from the narrow horn-like promontories which jut out into the lake on the north side. Views of the lake are shown in Plates [I,] [IV,] [XVI.]

[8]This was the case until a year or two ago. At the present time a limited amount of freshwater finds its way to the area immediately north of the east end of the lake and small plots are cultivated by the arabs.

[9]Herodotus, Book II; Strabo, Book XVII; and Diodorus Siculus, Book I, Chap. LI. (See Brown op. cit. p. 19-22.)

[10]“Hawara, Biahmu and Arsinœ,” 1889.

[11]Brown, op. cit. p. 95. As mentioned above in some areas the cultivated land was formerly even more extensive than at present, notably near the modern villages of Roda, Tamia, etc.

[12]For details of evaporation and level-records of the lake, see Brown, op. cit. pp. 6-9, and P.W.M. annual reports.

[13]See Willcocks’ Egyptian Irrigation, 2nd edition, London, 1899.

[14]See A Preliminary Investigation of the Soil and Water of the Fayûm Province, by A. Lucas, Survey Department, Cairo, 1902.

[15]See Appendix II, A Note by Dr. Schweinfurth on the Salt in the Wadi Rayan, in Willcocks’ Egyptian Irrigation, pp. 460-465.

[16]The word “freshness” is used comparatively, as the amount of salt is sufficient to make the water unpalatable or unfit for drinking, except near the feeder canals. It is, however, quite good enough for most culinary purposes, and camels will usually drink from it, although it is not advisable to water the latter from the lake either before or after a fatiguing desert march, as in such cases the salinity of the water may have bad effects.

[17]“Bull. of the American Geographical Society, 1882, pp. 22 and 24.”

[18]Mémoires sur les travaux publics en Egypte, Paris, 1873, pp. 53, 54.

[19]G. A. Liernur, Col. Western and Col. Sir C. C. Scott Moncrieff, k.c.m.g. Notes on the Wadi Rayan, Cairo, 1888.

[20]Note on the Wadi Rayan Project, Cairo, 1889.

[21]Perennial Irrigation and Flood Protection in Egypt, by W. Willcocks, m.i.c.e. Dir. Gen. of Reservoirs, with A Note by W. E. Garstin, Under Secretary of State, P.W.M., Cairo, 1894.

[22]The engineering details of the Wadi Rayan reservoir project have since been more fully discussed by Sir William Garstin in his “Report on the Basin of the Upper Nile” Cairo, (pp. 6-9 Appendix I).

[23]A Note by Dr. Schweinfurth on the Salt in the Wadi Rayan: an appendix to Perennial Irrigation, etc.

[24]The Assuan Reservoir and Lake Moeris, London, 1904.

[25]The following analyses of the chief springs in the Wadi Rayan, made by Mr. Lucas, Chemist to the Survey Department, are of interest:—

The above figures are parts per 100,000 parts of water.

Although the above analyses prove the water to be of a very poor quality for drinking purposes, compared with many of the wells and springs of the oases, the water, which is quite clear, seemed good. Except for its softness and somewhat ferruginous taste, it is quite palatable, and on my last visit we used no other for five days. The south spring was found to yield 21 litres and the north 6 litres per minute. The water of the third spring does not run.

Plate IV.

WESTERN EXTREMITY OF THE BIRKET EL QURUN.

PART II.

TECTONICS.

Section VIII.—FAULTING AND FOLDING.

The Fayûm Depression formed by subaerial Erosion.More extended examination of the Fayûm region supports my original conclusion[26] that the depression owes its origin to the same causes as have given rise to the other oases-depressions of the Libyan Desert, namely Baharia, Farafra, Dakhla and Kharga[27]. No evidence has as yet been met with which would suggest that earth-movements have played any important part in the formation of the Fayûm depression. Local faults, for the most part of short length and slight throw, occur at certain points, but the influence of these is strictly limited to their immediate neighbourhood. In fact, an examination of the desert margin of the Fayûm conclusively proves that the depression has been cut out through the action of ordinary subaerial denuding agents. The somewhat prevalent idea that the central portion of the depression, that covered by alluvial soil and the water of the lake, is faulted down, also rests on no foundation, all available evidence pointing in an opposite direction. Throughout the margin of the alluvial covered area the Eocene beds forming the surface of the desert can be observed to pass regularly under the cultivated lands; moreover, the same strata are frequently exposed in the bottoms of canals, drains, etc., far within the cultivation. The big drainage ravines of El Bats and El Wadi are, through a large part of their courses, cut down to the underlying Eocene rocks (Ravine beds), and in every locality examined the strata were found in the position they would be expected to occupy if undisturbed by tectonic movements.

Deep boring at Medinet el Fayûm.The evidence yielded by the deep boring at Medinet el Fayûm is, as far as it goes, to the same effect. The ground level at the site was at 23·40 metres above sea-level and the following beds were passed through[28]:—

The method employed in this boring was such as to bring up the material as a ground-up paste, an examination of which does not afford absolutely conclusive evidence as to the age of the rock. The absence, after the first 18·5 metres, of sand or pebbles, common throughout the alluvial deposits of the Fayûm, suggests that the base of these beds was reached at that depth, but from this evidence alone it would perhaps hardly be satisfactory to conclude that the underlying 112·5 metres were entirely Eocene. Considering, however that in the two deep ravines of El Bats and El Wadi the underlying Eocene is very commonly exposed at an average depth of some 15 metres below cultivation level, it is highly improbable that in the centre of the area, at Medinet el Fayûm, the alluvial deposits greatly exceed the same thickness. To classify the 112·5 metres of marly clays as alluvium would give the latter a total thickness of 131 metres and would mean that over an extremely restricted area the Eocene rocks had been denuded to such an extent that the floor of the depression lay 108 metres below sea-level. The ground-up samples of rock closely resemble what might be expected from the clays and marls forming the Ravine beds and in all probability the greater part of the 112·5 metres belong to that series. The harder stone met with at 131 metres, which, with the exception of a band of soft clay, continued down to the bottom, must be regarded as belonging to the underlying Rayan series.

Dr. Blanckenhorn’s Fault theory.Dr. Blanckenhorn, in a paper published in 1901[29] dealing with the Pliocene and Pleistocene of Egypt, describes the Fayûm as a triangular depression bounded on all sides by faults. The position of these bounding faults, as well as of numerous others more or less parallel to the north shore of the Birket el Qurûn, is shown on an accompanying map[30] and in a section drawn from Abshawai to the summit of the plateau north of the lake.[31] Stratigraphical evidence, based on the supposed identity of certain strata in different localities, is brought forward in support of these faults, the author finally stating that the production of the Fayûm is clearly and distinctly to be referred to tectonic movements.

Our conclusion, formed after an examination of the region in some detail, is so diametrically opposed to the above, that it may be worth while to state here the evidence which we consider sufficient to refute the existence of the particular faults described by Blanckenhorn.

Blanckenhorn’s fault-lines lie for the most part within the area covered by the alluvial deposits and the water of the lake, so that for want of exposures it is in most cases impossible to directly disprove their existence, although strong presumptive evidence against them can be adduced. The fault along the east side, however, is shown as closely following the junction line of the desert and the cultivated land, but everywhere along this line we found the marls and limestones of the Ravine beds passing regularly from the desert under the cultivated lands, without any sort of break or dislocation. Moreover, an examination of the desert ridge to the east disproved the existence of any faulting on the desert side, while the appearance of the same beds in the ravine of El Bats, a few kilometres to the west, proved the continuity of the beds under the cultivated alluvium in this direction. Certainly no fault exists along this side of the Fayûm. Similarly with regard to the fault shown as running from the west end of the lake along the west side of the Fayûm cultivation into the basin of Gharaq; although the desert margin along this side of the Fayûm cultivation has not been so closely surveyed as that on the east side, no evidence in favour of the existence of faults was met with in the particular localities examined. With regard to the third main bounding fault, considered by Blanckenhorn to run throughout the length of the Birket el Qurûn and to be continued eastwards, possibly to the Nile Valley and at least to join the fault on the east side of the cultivation, we need only say that an examination of the desert near Tamia disproves its existence at that end; while it is difficult to imagine that a fault could traverse the lake from end to end without revealing its presence in the island Geziret el Qorn or in one or other of the promontories which jut out so far into the lake from its northern shore. Everywhere the strata are undisturbed and occupy their normal stratigraphical level and position.

Let us finally examine the series of more or less parallel faults stated to exist between the island and the northern shore of the lake, and on the mainland to the north and south of Dimê. Dr. Blanckenhorn publishes a detailed section (op. cit., fig. 2., taf. XIV) showing the positions of these step faults and their effect on the various strata through which they cut. Fortunately, in this neighbourhood the stratigraphical succession is well exposed and the presence or absence of faults become matters of easy determination. The sequence of beds from south to north is normal and uninterrupted and our interpretation of the area is shown in the accompanying sections (Plates [XIX,] [XXII,] and [fig. 4]). We have no hesitation in saying that such faults as those shown on Blanckenhorn’s section do not exist. Their insertion appears to be the outcome of an error in the correlation of strata at the three points Abshawai, Geziret el Qorn and Dimê. The bed capping the island is not identical with that forming the plain to the north of Dimê, although shown to be such on the section under discussion.

In a later publication[32] Blanckenhorn admits being in error in his correlation of the different beds in the localities in question and completely withdraws his former statements that the depression owes its existence to fracture and subsidence. The faults shown on his detailed section from Abshawai to Qasr el Sagha are admitted to be non-existent and in this retraction we may presumably include the remainder of the faults described by the same author, as the evidence for them is of a still less satisfactory nature.

In a wind-swept desert area like the Fayûm the slightest dislocations are as a rule markedly obvious, and faults of any magnitude could scarcely escape detection. Over the greater part of the region every bed is laid bare on the surface and can be minutely examined; while the marked irregularity of the escarpments afford sections cut through the different series in every direction. Some areas, however, are covered with superficial deposits, which more or less effectually obscure the underlying rocks; for instance, on the east side a large part of the central floor is hidden by the cultivated alluvium and by the water of the lake; in the south a considerable proportion of the floor of Wadi Rayan is buried under accumulations of blown sand; and again large areas on the north, west, and south-west sides are obscured by a superficial covering of loose gravel. But as already mentioned, there is no reason to suppose that faults of any importance exist within the areas thus partly obscured. The cultivated lands and the Birket el Qurûn do not occupy low areas produced by faulting but, as shown above, owe their positions entirely to the original northerly dip of the strata and to subsequent erosion.

Numerous small faults effects local.We have already stated that small local faults occur in various parts of the Fayûm and some of these may be specially mentioned. The most important is about 10 kilometres N.N.E. of Qasr el Sagha; the line of fault lies nearly north-west and south-east, has a length of six or seven kilometres, and affects both middle and upper Eocene beds; at its northern end the fault passes into a fold before finally dying out. Some of the Upper Eocene sandstones are hardened and silicified and form a succession of black knobs along the line of disturbance. To the south of these the axis of dislocation bends slightly eastwards and takes the form of a sharp fold; further south it again becomes a true fault, flanked by a line of highly tilted beds along its south-west side. The effects of this fault are very marked locally but entirely restricted to a limited area. The most important is the breaking of the continuity of the escarpment of the Qasr el Sagha series; the line of cliffs formed by those beds is a very marked topographical feature and the fault in question causes a lateral displacement of seven kilometres.

Fig. 1.—Fault near Qasr el Sagha.

In the neighbourhood of Qasr el Sagha and westwards for a considerable distance, small strike-faults are of common occurrence in the beds of the Qasr el Sagha series. As a rule these faults do not extend more than a few hundred metres in length, while the down-throw seldom exceeds two or three metres and in almost every case observed is to the north. The hade may be 65° or more. Fig. 1 shows an example near Qasr el Sagha. The most marked of these strike faults is seen to the east of Garat el Esh; commencing a little to the north-east of that hill it runs in a nearly due easterly direction till it cuts the cliffs of the Qasr el Sagha series after some five kilometres. Its down-throw is to the north and never exceeds a few metres; this small throw is however sufficient to cause a marked displacement of the highest bed of limestone forming the dip-slope surface of the plateau at the summit of the Middle Eocene beds.

At first sight it might be suspected that the very irregular trend of the different escarpments throughout the Fayûm was determined or influenced by fault lines; an extended examination of the cliffs however gave negative results, with one exception; the long narrow hill-mass to the north east of Gar el Gahannem is bounded by faults on both sides and that on the west can be easily traced for seven or eight kilometres northwards, and throughout its length its influence on the topography is very conspicuous.

[26]Beadnell. The Fayûm depression: A Preliminary Notice of the Geology of a District in Egypt containing a new Palaeogene Vertebrate Fauna. Geol. Mag. Dec. IV, Vol. VIII, No. 450, Dec. 1901, p. 540.

[27]See reports on Kharga Oasis (1900), Farafra Oasis (1901), Dakhla Oasis (1901), and Baharia Oasis (1903), issued by Survey Dept. P.W.M., Cairo.

[28]Public Works Ministry Report. Cairo, 1899.

[29]Blanckenhorn. Geologie Aegyptens, Berlin 1901, Pt. IV, pp. 339-344.

[30] „ p. 341, Fig. 10. Skizze der Strukturlinien des Fayûm.

[31] „ Taf. XIV. Querprofil durch den Fayûmgraben.

[32]Blanckenhorn, Neue geologisch-stratigraphische Beobachtungen in Aegypten, S.-Ber. d. math.-phys. Classe d. kgl. bayer. Ac. d. Wiss. Bd. XXXII 1902, Heft III, München 1902, pp. 428, 429.

PART III.

GEOLOGY.

Section IX.—GENERAL AND CLASSIFICATION OF STRATA.

The geology of the area[33] under consideration is almost entirely stratigraphical, the only igneous rocks being more or less local lava flows. The sedimentary rocks of the district have yielded an abundant fauna, both invertebrate and vertebrate; the latter is of unique interest, including as it does a number of highly interesting animal types quite new to science. An extended examination in the field, and comparisons with the stratigraphical succession in other parts of Egypt, checked by the determinations of the fossil molluscan fauna, make it possible to form a very fair estimate of the approximate age of the different rock-stages, although this may necessarily be subject to modification when the specific determinations of the entire collection of organic remains have been completed, and the development of vertebrate life has been correlated and compared with that in other parts of the world.

The depression is cut out in a great series of sedimentary rocks of Middle Eocene, Upper Eocene, and Oligocene age, and one of the features of the stratigraphy of the region is the constancy of many beds over wide areas. The dip of the beds throughout the area is nearly due north and at a very low angle, averaging 2° or 3°, but varying from 1° to 5°; this low dip is very constantly maintained except when locally affected by small faults. The structural geology and tectonics have already been discussed at some length in the previous sections.

The oldest beds found in the depression are the clays, marls, and limestones with Nummulites gizehensis, of Middle Eocene age. These are succeeded by a group of white marly limestones and gypseous clays, which largely underlie the cultivated alluvium of the Fayûm. They are followed by a series consisting of clays, sandstones, and calcareous grits, some beds of which are characterized by the abundance of small nummulites and Operculina. The latter series is followed by the uppermost truly marine Eocene beds, a group of alternating clays, sandstones and limestones, the “Qasr el Sagha Series” (or Carolia beds), characterized by an abundant invertebrate and vertebrate fauna, and equivalent to the Upper Mokattam beds of Cairo.

Above the Qasr el Sagha series, and well marked off from them both lithologically and palæontologically, is found a great thickness of variegated sands, sandstones, clays and marls, the “Fluvio-marine Series” (Jebel el Qatrani beds), divided near the summit by one or more thick intercalated lava sheets, the latter forming a convenient junction line. This series of variegated beds is of Upper Eocene—Lower Oligocene age.

No Miocene strata have been recognized within the area, but further north, as at Mogara, Lower Miocene deposits occur;[34] and it is probable that there is a continuous series of lithologically similar beds from the summit of the Fayûm escarpments (Lower Oligocene) to the Mogara Miocene.

The Pliocene is probably represented by the great terraces of gravel—raised beaches—which are such a marked feature in the geology of the district. Fossiliferous Pliocene deposits have also been recorded from the south part of the area by Schweinfurth.[35]

Pleistocene and Recent are abundantly represented by lacustrine clays, both ancient and modern, alluvial land and blown sand, the formation of which deposits is continuing at the present time.

The following table will show the sequence of strata and the classification adopted in the present memoir:—

Table showing Succession and Classification of Strata in the Fayum.

Section X.—MIDDLE EOCENE (PARISIAN).

A.—Wadi Rayan Series.—(Nummulites Gizehensis Beds).

(A.I.e. Schweinfurth, I.b. Mayer-Eymar,[36] Lower Mokattam of Cairo).

Beds of this group are chiefly found in the south of the depression. The wadis Rayan and Muêla, as already shown by Schweinfurth and Mayer-Eymar[37], are cut out in clays and limestones of Lower Mokattam age; the upper beds of limestone, containing among other fossil[38] numerous examples of the large Nummulites gizehensis, form the greater part of the floor of the depression west of the Fayûm cultivation, stretching from Jebel Rayan to the foot of Gar el Gehannem,[39] 28 kilometres west of the western end of the Birket el Qurûn ([Section XX]). Near the latter hill examples of N. gizehensis of inordinately large size occur.[40]

At the conical hill at the southern entrance to Wadi Muêla the following beds were noticed:—

At the corner of the cliff 7½ kilometres N.N.W. of the monastery of Der el Galamûn, in Wadi Muêla, occur about 80 metres of hard white nummulitic limestones, with beds of argillaceous sandstone and sandy clays. Fossils are numerous and include nummulites of several species (N. gizehensis, etc.), Carolia placunoides, different species of Ostrea, with gastropods (among others Terebellum sopitum), bryozoa, etc. It is very noticeable that the nummulites, especially the small species, occur in remarkable profusion not only in the limestones but often in the clays.

The following section will give a good idea of the general alternations found in this area; it was measured at Jebel Rayan,[41] 24 kilometres west of the western end of the cultivation of Gharaq basin.

Plate V.

ALLUVIAL DEPOSITS OVERLYING MARLY LIMESTONES (RAVINE BEDS) IN EL WADI, RAVINE NEAR QASR GEBALI.

The following is a section of the beds exposed in Wadi Muêla compiled from a paper by Mayer-Eymar on this oasis:—

There is a considerable difference in thicknesses between the above section and that of Jebel Rayan. Our heights agree closely with those of Schweinfurth, so that it is probable that Mayer-Eymar is in error, notwithstanding his challenge of Schweinfurth’s figures in the paper mentioned.

B.—Ravine Beds.

The beds of this series, consisting of gypseous clays, clayey marls, and white marly limestones, are met with bordering the cultivation on the east, west and north sides; they pass under the alluvial soil of the cultivated land and are frequently seen in the bottoms of canals, and especially in the deep ravines known as El Bats, and El Wadi (Plates [III] and [V]). The relation of these beds to the Rayan series below is well seen at the prominent outstanding hill Gar el Gehannem ([Fig. 2]); here the plain to the east and south is formed of the uppermost member of the Wadi Rayan series, a limestone full of Nummulites gizehensis. In the hill itself the latter is directly overlain by gypseous and glauconitic sandy clays and marls, with hard intervening beds of yellowish, often marly, limestone. The upper beds consist of alternating clays, sandy limestone and sandstone, at the top being a thick bed of the latter passing up gradually into the sandstones of the Birket el Qurûn series. The following is the detailed section:—

Fig. 2.—Section at Gar el Gehannem, showing the relation of the Wadi Rayan Series to the Ravine Beds.

The clays, marls, and limestones of the Ravine beds are generally found to contain fairly numerous shell-impressions, including Nucularia sp., Leda sp., Cardita sp., Corbula aff. pixidicula, Lucina sp., Oudardia ovalis, Desh., Tellina tenuistriata,[42] numerous small fish-scales, and occasional large teeth of sharks; while the skeletons of the toothed-whale Zeuglodon Isis are fairly common, although usually in poor preservation.

In the ravine of El Bats, about one kilometre west of Sêla, these beds (5-6 metres thick) are seen unconformably overlaid by 12 metres of false-bedded gypseous sands and clays passing up into the superficial cultivated loam. The junction of these alluvial deposits and the underlying Eocene is distinctly unconformable and an intervening pebble-bed is occasionally present ([Fig 3]).

In the large ravine known as El Wadi, which traverses the west side of the cultivation of the Fayûm, these beds are frequently well exposed; their lithological characters remain very constant. Here, as in El Bats, they are unconformably overlain by a varying thickness of Pleistocene and Recent clays. Their surface, a plain of subaerial denudation, represents the original floor of the depression before the entry of the sediment-carrying water from the Nile Valley through the Lahûn gap; its irregularity is seen in [Plate V.]

The plain bordering the cultivation to the east of Sêla and Rubiat likewise consists of these same white marls with fish-scales, etc.; they pass regularly under the cultivated land. Shaly marls, gypseous clays, and chalky limestones of the same age are seen in, and to the south of, the railway crossing the desert between Sêla and Medum. Eastwards they stretch into the Nile Valley, being found exposed along the desert-edge bordering the cultivation at Medum, Nawamis and Masaret-Abusia.

Fig. 3.—Sketch-Section across El Bats, 1 kilometre West of Sêla.

The same beds are exposed immediately to the east of the village of Sersena, midway between Sêla and Tamia. They are again well seen in the ravine below the last named village, and forming the narrow strip of the desert projecting into the cultivation as far as the northern end of the Tamia lake; they also occur on the shore of the latter at El Tuba, about 2 kilometres south of the village. At Tamia their exposure measures 25 metres in thickness.

At various points along the north side of the Birket el Qurûn exposures of this series occur, the beds forming the lower sloping part of the cliffs overlooking the lake, as well as the base of the island “Geziret el Qorn,” although only the upper beds are visible above the water of the lake. Both here and along the northern shore of the lake they are for the most part hidden by the high level recent lacustrine clays, but where occasionally exposed their identity is certain, the characteristic small brown fish-scales being abundant, besides occasional teeth, with shell-impressions of the different genera enumerated above.

Plate VI.

ESCARPMENT OF THE BIRKET EL QURUN SERIES NEAR THE WESTERN END OF THE LAKE.

At the western end of the lake the Ravine beds form the lower part of the cliff as well as the plain to the south; the underlying Nummulites gizehensis limestone not being exposed. The series consists of some 45 metres of white and grey shaly marls with harder bands of siliceous limestone intercalated throughout, one of which usually forms the uppermost bed. It is, in fact, the development in places of one or other of these hard beds of limestone near the top of the series that gives rise to the bold promontories, or horns, which occur on the north side of the Birket el Qurûn.

The greater part of the marls and clays met with from 18·5 to 112·5 metres below the surface in the boring at Medinet el Fayûm in all probability belong to the Ravine beds.

The maximum thickness of this series is 70 metres, measured at Gar el Gehannem.

C.—Birket el Qurun Series (Operculina-Nummulite Beds).

The above designation is convenient and applicable to these beds, which form the escarpment immediately overlooking the lake on the north side throughout its length.

The group includes all the beds between those last described and the well-marked Qasr el Sagha series, homotaxial with the Upper Mokattam (the brown beds) of Jebel Mokattam, near Cairo. It thus appears to be the equivalent of the upper part of the white beds (quarried limestones) of the Mokattam section, although the lithological characters are entirely different, the massive limestones of Jebel Mokattam being represented in the Fayûm by an arenaceous and argillaceous series, deposited probably in water of far less depth. Where the different members of this series are well exposed certain beds are found to be characterized by the abundance of two foraminifera, the one a small thin-shelled Operculina (O. discoidea)., and the other a small thick nummulite.[43] The tests of these foraminifera sometimes make up entire bands of rock. In addition, the series includes certain beds which at times become very fossiliferous, and contain a well-preserved molluscan fauna.

The series is well seen in the desert separating the Fayûm from the Nile Valley; on the south-east and east sides of the former; along the northern boundary of the cultivation and the Birket el Qurûn; and westwards in the cliffs to beyond the outlying hill-mass of Gar el Gehannem.

The following section was measured on the south-west of the Fayûm, from Ezba Qalamsha (on the confine of the cultivation) to the ridge summit 5 kilometres to the south-east.

To the north of the Lahûn pyramid the beds agree generally with the above. The following are the chief divisions here:—

The foraminiferal sandy limestones of this series are seen at points in the desert bounding the eastern margin of the cultivation, notably east of Sersena and at the top of the hill 15 kilometres north-east of Rubiat.

The following section was measured at the prominent hills 17 kilometres 28° N. of E. (magn.) of Tamia:—

In the north of the Fayûm the series is characterized by the presence of one or more very constant well-marked beds of hard calcareous sandstone, which almost invariably weather into huge globular masses. These masses should be regarded as huge weathered-out concretions, rather than as water-rounded blocks, although no doubt in many cases their roundness has been increased by the action of the waters of Lake Moeris as the level of the latter gradually fell, and possibly still earlier during the invasion of the Pliocene sea; from the latter time also may date the millions of parallel vertical borings with which these and other exposed rocks are often perforated. In the various places where one of these beds forms the present surface of the desert the concretions may be seen in different stages of exposure, from the initial, where only just the tops are laid bare, to the final stage where the globes are left completely weathered out, as seen in the illustration ([Plate VII]). The appearance of the desert when covered for many square kilometres with thousands of these blocks is more easily imagined than described.

The lower beds of the Birket el Qurûn series form the island Geziret el Qorn, and consist of clays and sandstones containing a considerable number of organic remains. These beds were collected from and examined by Schweinfurth[44] in 1879, the mollusca being subsequently described by Mayer-Eymar,[45] while the vertebrate remains, which included cetacean bones and numerous fish-teeth, were submitted to Dames.

The following species were determined by Mayer-Eymar, who indicated that the fauna as a whole had a Bartonian aspect[46]:—

Upper Bed.

• Ostrea plicata, Defr.
• Arca Edwardsi, Desh.
• Lucina pomum, Duj.[47]
• Lucina cfr. tabulata, Desh.
• Cardium Schweinfurthi, May.-Eym.
• Cytherea Newboldi, May.-Eym.
• Tellina pellucida, Desh.
• Mactra compressa, Desh.
• Corbula pyxidicula, Desh.
• Calyptræa trochiformis, Lam.
• Turritella angulata, Sow.
• Ficula tricarinata, Lam.

Lower Bed.

• Astrohelia similis, May.-Eym.
• Goniastræa cocchii, d’Achiardi.
• Heliastræa acervularia, May.-Eym.
• Heliastræa Ellisi, Defr. (Astræa).
• Heliastræa flattersi, May.-Eym.
• Ostrea digitalina, Dubois.
• Ostrea gigantea, Sol.
• Ostrea longirostris, Lam.
• Ostrea producta, Delb. et Raul.
• Isocardia cyprinoides, Braun.
• Turritella carinifera, Desh.
• Turritella transitoria, May.-Eym.
• Turritella turris, Bast.
• Turbo Parkinsoni, Defr.
• Pleurotoma, sp.

The cetacean remains, belonging to the genus Zeuglodon, were described by W. Dames,[48] who compared them with the American species Z. macrospondylus and Z. brachyspondylus, but did not then consider them to represent a new species; in a later publication,[49] however, the same author described similar but more complete remains, also collected by Schweinfurth (from beds belonging to our Qasr el Sagha series), as a new species, Z. Osiris. A considerable number of fish-remains from Geziret el Qorn are also described in the earlier publication. Although the difference in size of the bones of separate individuals was considered by Dames to be sexual, it seems probable that there are two distinct species of Zeuglodon, as the smaller type appears to have a much greater upward range than the larger[50]; both species, Z. Osiris, and Z. Isis occur in the Birket el Qurûn series, and a very fine mandible of the larger was obtained from these beds in the cliffs near the west end of the lake.[51] More recently a third species has been discovered by Stromer and described under the name of Z. Zitteli.[52]

Fig. 4.—Profile of beds of Geziret el Qorn.

1. Hard brown sand-rock with large concretions of weathered globular sandstone on the summit; ferruginous nodular bands containing shell-casts occur near top. 2. Soft gypseous clays with bands of sand-rock and sandstone with Ostrea, Cardium Schweinfurthi, Turritella, corals, Zeuglodon, chelonian and fish-remains. 3. Brown sand-rock. 4. Soft gypseous clays and harder brown sandstones. 5. White shaly marl with fish-scales; hard band at top and soft sandy shaly clays below.

The surface-slope is much less than shown in sketch and is generally covered by a deposit of lacustrine clays containing freshwater shells and fish-bones.

The accompanying profile (Fig. 4), measured during a hurried visit to the island for the purpose of correlating these beds with those of the mainland, shows the character of the lower beds of the Birket el Qurûn series at this point.[53]

Plate VII.

WEATHERED CONCRETIONARY SANDSTONE (BIRKET EL QURUN SERIES) ON NORTH SHORE OF LAKE NEAR GEZIRET EL QORN.

The upper beds of the Birket el Qurûn series in this part of the Fayûm are lithologically similar to those just described, consisting of alternating clays and sandstones, about 37 metres thick. They are, however, generally much richer in fossil remains, which are likewise usually better preserved than in the lower beds. Some of the brown sandstones of this series are literally crowded with perfect examples of many of the typical mollusca; and further west, near the end of the lake, foraminiferal bands again become noticeable. Near Dimê the escarpment of these and the lower beds is gentle and inconspicuous, but followed westwards it becomes a bold precipitous cliff, increasing in height towards the western end of the lake, where it is capped by the lower beds of the Qasr el Sagha series.

The following section was measured on the mainland[54] opposite the island Geziret el Qorn.

At the western end of the Birket el Qurûn the series is well marked, the sandstone beds forming the steep face of the bold precipitous cliffs which are so marked a feature at this end of the lake. The group has a thickness of some 50 metres and is overlain by the lower beds of the Qasr el Sagha series; it is more convenient here to give the entire section of the cliffs down to the base of the series under discussion:—

In the cliffs west of the end of the lake the upper bed No. 9 continues highly fossiliferous and yields the most perfectly preserved molluscan remains to be found in the Fayûm and probably in Egypt.

A few kilometres east of the end of the lake a band of large globular concretions occurs in the thick brown sandstone forming the vertical face of the cliff. In many places the effect of weathering of these rocks is of some interest, numerous “earth-pillars” having been formed; these are largely the result of the action of blown sand, assisted by rain, the concretions being left capping pillars of brown sandstone, the sides of which are sculptured by the wearing action of sand. The curious perforate or cellular appearance which the weathered surfaces of this sandstone assume after long exposure are particularly noticeable in this neighbourhood and in the Zeuglodon Valley further west.

In the well-marked hill distant 17 kilometres to the north-east of Gar el Gehannem, the soft fossiliferous sandstones of this series are crowded with Operculina, Nummulites, and many species of mollusca beautifully preserved.

At Gar el Gehannem the series is seen ([Fig. 2] and detailed section [page 36]) forming part of the slope of the hill, underlain by the Ravine beds, and capped by part of the Qasr el Sagha series. It here consists of yellow sandstones divided by a bed of clay; the sandstones are often crowded with nummulites (of two species); also Operculina (discoidea?), echinids, Balanus sp., Ostrea Reili, O. Fraasi, Carolia placunoides, and species of Pecten, Pinna, Cardita, Teredo, Turritella, and Cerithium.

Fig. 5.—Section of cliffs, western end of the Birket el Qurun.

Pleistocene.—(a) Lacustrine clays and sands with freshwater shells and fish-remains; Middle Eocene, 1. 14 Clays, sandstones and impure limestones; 15 White shaly clays and marly limestones.

In the Zeuglodon Valley, 12 kilometres W.S.W. of Gar el Gehannem, the brown sandstones of the Birket el Qurûn series are divided by a narrow band of fine-bedded grey clay. Most of the fantastically shaped hills on the south-west slope of the valley are carved out of the lower division of the sandstone. The concretionary beds of the Birket el-Qurûn series are not developed in this neighbourhood. Remains of Zeuglodon of both species (Z. Osiris and Z. Isis) are remarkably abundant and the skeletons of these cetaceans may be found in every stage of weathering. The larger species, Z. Isis, is the more common, and series of vertebrae, twelve to fifteen in number, can frequently be counted in situ. The remains are most abundant enclosed in the hard brown nodular bands of the series but in such cases it is almost impossible to extract specimens of any value. In one instance an almost complete skull of Z. Isis, measuring 116 cm. in length, was found enclosed in a large block of the nodular rock.[57] Bones are frequently to be observed protruding from the wind-worn sides of the small hills, while those portions of the skeleton already weathered out litter the ground below. Exposed they break up with rapidity, although where the enclosing rock is softer than the bone itself, parts of the skeleton beautifully preserved and perfectly free from matrix may sometimes be obtained.

The molluscan fauna is represented by very large numbers of pseudomorphs in sulphate of strontium (celestine) of the genera Lucina, Turritella, Fusus and Nautilus, the profusion of individuals of a species of the latter being very marked. In the case of lamellibranchs the radiating bundles of crystals of celestine are seen to originate from a point placed centrally on one of the valves, so that on this side (of a slightly weathered example) a radiating mass of crystals is seen, while on the other appear numerous contiguous circular areas, representing the terminal ends of the bundles of crystalline fibres or needles. Apart from the quantities of organic pseudomorphs, masses of crystalline celestine occur in the sandstones throughout the valley, and altogether the quantity of sulphate of strontium present must be very great. The gigantic oysters and other fossils which occur in some of the overlying higher beds, and the numerous individuals of nummulites in the sandstone itself, never seem to be replaced by celestine.

Nummulites of two species are very abundant in some bands and the presence in the Zeuglodon Valley of occasional individuals of the large N. gizehensis shows that in favourable localities this species persisted throughout the time represented by the deposition of the Ravine beds and ranged upward into the basal members of the Birket el Qurûn series.

In the higher hills within the valley, and in the hill-mass on the south side, the yellow sandstones of the Birket el Qurûn series pass up into the basal members of the Qasr el Sagha series. In their upper limits the sandstones become very nummulitic in places and at the top bands made up of Carolia and Ostrea occur. Above these, in the basal members of the Qasr el Sagha series, huge oysters and finely preserved specimens of Qerunia cornuta are conspicuous.

The dip in the valley is 2° north.

The southern face of the hill-mass lying immediately to the south of the Zeuglodon Valley is an almost sheer cliff of over 100 metres, descending to the silt covered basin below which has already been noticed ([page 23]). On this escarpment the hard nodular marly limestones of the Ravine beds are seen near the base, overlain by a mass of grey shaly gypseous beds; above, forming as a rule a vertical wall of rock, lies the hard massive brown sandstone of the Birket el Qurûn series, here undivided by clays; at the top, highly fossiliferous alternating clays and limestones are found forming the summit of the hills.

The exact junction between the Birket el Qurûn series and the overlying Qasr el Sagha beds is naturally perfectly arbitrary, many of the fossils being common to both groups. Carolia placunoides, which is perhaps the most abundant fossil in the Qasr el Sagha series, is sometimes very common in the upper beds of the underlying group, and, as shown before, is common enough in the still lower Nummulites gizehensis beds of Wadi Rayan. So that, though this fossil itself is no criterion, its relative abundance in the upper series justifies those beds being called the “Carolia beds,” the additional name of the Qasr el Sagha series being taken from the old ruin of that name where these beds are fully seen.

Plate VIII.

MIDDLE EOCENE ESCARPMENT (QASR EL SAGHA SERIES) 12 KILOM. WEST OF QASR EL SAGHA.

D.—Qasr el Sagha Series (Carolia Beds).

This division is strikingly developed in the north of the Fayûm, where it forms a bold escarpment of great length, consisting of an alternating series of very fossiliferous clays and limestones, with sands and sandstone in the upper beds, of a total thickness of 175 metres.

This series is the equivalent of the well known Upper Mokattam beds of Jebel Mokattam, immediately to the east of Cairo. The cliffs of this hill are among the best known in Egypt and have been studied by many geologists, including Zittel, Schweinfurth, Mayer-Eymar, etc.; these authors have classified the whole of the Upper Mokattam of Cairo as equivalent to the Upper Parisian (Middle Eocene) of Western Europe. The series is far better developed in the Fayûm than at Jebel Mokattam, where the total thickness is only some 70 to 80 metres.

In consequence of the discovery in these beds of a highly interesting vertebrate fauna, including land animals, the series becomes of the greatest importance. As already mentioned, as long ago as 1879, Schweinfurth, during a journey across the Fayûm, obtained remains of Zeuglodon in the underlying series from the island in the Birket el Qurûn. Subsequently[58] he obtained additional remains of the same cetacean in a violet marl belonging to the present series, from a locality 12½ kilometres west of Qasr el Sagha[59]; these remains, as already mentioned, were described by Dames as Z. Osiris. Since then important finds of land and marine mammals and reptiles have been made in different beds of this series; these will be referred to later.

The outcrop of the Qasr el Sagha series occupies a large part of the northern desert of the Fayûm. The beds are, however, best seen in the cliffs about 8 kilometres north of the Birket el Qurûn, where they form a steep double escarpment, running east and west, nearly parallel to the northern shore of the lake. The dip of the series being northward at a very low angle, and the upward slope of the ground being in the same direction, this cliff dies out a few kilometres north-east of Qasr el Sagha. A little further north, however, a N.W.-S.E. fold and fault again exposes nearly the whole of the beds of the series, forming prominent cliffs as before.

In the conspicuous hill 17½ kilometres 28° N. of E. (magnetic) of Tamia the series consists of innumerable alternations of clays and sandy limestone. The calcareous beds nearly always contain numerous examples of Carolia placunoides, Ostrea and Turritella of several species, but other well-preserved fossils are rare. The exposed beds here have a thickness of about 55 metres, and are underlain by the Birket el Qurûn beds with a well-marked band of concretionary sandstone, the thickness of the two series together being 127 metres. The upper beds of the former series are not here exposed, the top of the hill being formed of well-rounded flint and quartz pebbles embedded in a base of finely crystalline gypsum (2 metres thick), a deposit of Pleistocene times.

To the north of Tamia a large area of desert is occupied by the beds of this series; the district has the character of an undulating plain with occasional groups of hills and low irregular escarpments. At the groups of hills 12 kilometres N.N.E. of Tamia, and just to the east of Garat el Faras, the Qasr el Sagha beds are found to consist as usual of an alternating series of sands, sandstones, clays, marls and limestones, with numerous individuals of Ostrea, Carolia and Turritella, besides vertebræ, teeth and spines of large fish.

We may pass now to the locality where this series shows its best development and exposure, the beds being all concentrated in one bold escarpment, generally divisible into an upper and a lower cliff. These cliffs overlook the Birket el Qurûn, although distant usually about 8 kilometres, being separated from the lower escarpment of the Birket el Qurûn series (immediately above the lake shore) by a broad plain, the surface of which is usually the dip-slope of a hard bed of sandstone. From Qasr el Sagha (6½ kilometres N.N.E. of Dimê) these cliffs trend westward, keeping approximately the same distance from the north shore of the lake; they have been followed and mapped for a distance of 70 kilometres to a point 13 kilometres N.N.W. of Gar el Gehannem, whence they could be seen still trending in a direction slightly south of west (see [Plate XVII]).

Small faults are of frequent occurrence along these escarpments, but are not of other than local interest; they almost invariably have their downthrow to the north, and it seldom exceeds a few metres. [Fig. 5] shows a section through one of these faults near Qasr el Sagha.

The following detailed section ([Plate XXIII]) will show the character of the beds forming this division. As might be expected in such a series, although the calcareous bands are fairly constant, there is a continuous change of character among the sandy and clayey sediments from point to point; the false-bedding is in places very striking.

The main part of the section was measured 3½ kilometres north-east of Qasr el Sagha, but the lower beds not being exposed at that point, they were added from the cliffs at the ruin itself. The total thickness is 154 metres.

The chief divisions of the series remain fairly constant and can be recognized and followed for many kilometres westwards.[61] The lower beds form the summits of Gar el Gehannem and the neighbouring hills (see [Fig. 2] and section [p. 36]), the upper beds of the series being exposed in the higher escarpments to the north.

Although vertebrate remains are more common on some horizons[62] than on others, they are occasionally met with in most of the beds. The most prolific bone horizon is, however, about half-way down, i.e., those beds numbered 16 and 17 in the above section; bed 7 also yielded a number of remains. At the point where the upper part of the section was measured, 3½ kilometres north-east of Qasr el Sagha, the beds 16 and 17 yielded a considerable number of land-animal remains, all of which occurred within a fairly confined space, suggesting that they had been carried out from the land to this point by a strong river-current and deposited when the latter became too feeble to carry them further out to sea. The same beds were also examined in the faulted bay 8 kilometres to the north, but no bones, or at most a very occasional fragment or two, were obtained here. This is easily explained by the greater distance of this locality from the land-mass to the south. Westwards the same beds were always found more or less bone-bearing, isolated detached mandibles, limb-bones and vertebræ of Mœritherium, being of frequent occurrence, although no such complete remains were found as those from near Qasr el Sagha. Reptilian and fish bones are very widespread throughout the area. An extensive and detailed examination of these beds over a large area can hardly fail to yield important results, as other localities where skeleton-carrying currents came out from the land would very likely be discovered.

That the Qasr el Sagha series was deposited in fairly shallow water at no great distance from land seems certain, not only from the general lithological character of the beds but from the number of land-animal remains and the frequency of river and shore-frequenting whales, dugongs, crocodiles and turtles. The clays, moreover, are found to abound with impressions of plants, and in some cases are highly lignitic, being made up of compressed masses of vegetation including solid twigs, now found in a state more resembling charcoal than ordinary dense lignites; some bands approximate to an impure brown coal. In certain beds of the series further to the west, very thin seams of true coal occur; they were, however, never seen to exceed one or two millimetres. The intercalated bands of limestone are generally impure and do not indicate any great conditions of depth, but only rather a temporary cessation in the supply of sand and clay. Corals, moreover, abound along many horizons.

Plate IX.

UPPER BEDS OF FLUVIO-MARINE SERIES WITH BASALT CAP, LOOKING WEST FROM THE EASTERN EXTREMITY OF JEBEL EL QATRANI.

Section XI.—UPPER EOCENE (BARTONIAN)—LOWER OLIGOCENE.

E.—Fluvio-Marine Series (Jebel el Qatrani Beds).

Throughout the north of the Fayûm depression the Qasr el Sagha beds, forming the uppermost Middle Eocene, are followed by an unique series[63] of variegated[64] sands and sandstones, with alternating beds of clay and clayey marl. The ever-recurring bands of limestone, so common to the underlying marine beds, have now almost completely disappeared, being represented by only an occasional bed of calcareous grit, marl, or thin band of limestone. In the upper part of the series occurs a horizontal sheet of basalt,[65] in all probability contemporaneously interbedded; this forms a convenient datum line and may perhaps be provisionally taken as an arbitrary junction between the Eocene and Oligocene. Although as a rule remarkably barren of organic remains, certain bands, especially in the upper part, yield numerous individuals of a few species of mollusca, including Lucina, Arca, Mutela, Spatha, Unio, Lanistes, Turritella, Melania, Potamides, Cerithium and Pleurotoma. From such an assemblage we may without doubt conclude that the conditions under which the series was deposited were estuarine or fluvio-marine, and this is further proved by the non-marine lithological character of the beds. The enormous quantities of silicified wood which occur in certain beds, in the shape of hundreds of trees of great length and girth, together with the numerous remains of land-animals, crocodiles, tortoises and turtles, indicate that rivers of considerable size emerged from the land to the south, the coast-line of which was probably not far distant. In fact the retreat of the sea, which as already mentioned, had probably already begun in Middle Eocene times, was now still further continued, although the cause of this was apparently not so much due to elevation of the land as to the continued deposition of sediment from south to north beyond the land-shore. We may in fact regard the series as a huge delta deposit in an area of local depression, in which the great accumulation of sediment brought down from the land continually caused the gradual retreat of the sea to the north.

The same conditions would even appear to have continued on to Pliocene times, as from the Fayûm northwards stretches an immense plain of lithologically similar rocks, evidently accumulated under similar conditions, and which appear to contain newer and newer faunas from south to north. Thus, while in the Fayûm the remains are of Middle and Upper Eocene and Lower Oligocene age, when the latitude of Mogara is reached, some 70 kilometres further north, a fauna distinctly Lower Miocene in aspect occurs; further north again, as at Wadi Natrûn, Pliocene remains are abundant. We may hope therefore that this otherwise barren desert, when carefully and systematically explored, will yield us a continuous record of the vertebrate life of the northern part of the African continent from Eocene to Pleistocene times.

In the Fayûm, over a length of 80 to 90 kilometres, the basal beds of the Fluvio-marine series, at a height of only a few metres above the top of the Qasr el Sagha series, are frequently found to contain the remains of land-animals, often in sufficient quantities to form in places a true “bone-bed.” Besides land-mammals, remains of large tortoises, turtles and crocodiles, are very common, some of the latter being identical with those of the Qasr el Sagha series below. Chelonian and crocodilian remains are to be found on various horizons, but so far none but fragmentary mammal remains have been observed in the higher parts of the series. It is interesting to note that the bones in these beds appear to be most common near the accumulations of fossil trees, thus suggesting that they were floated out from the land at the same time and by the same river-currents. The porous character of the sands and sandstones of this group has resulted in the remains not being in nearly so hard or durable a condition, except when coated with ferruginous sand, as those in the series below, although the actual state of preservation is even more perfect.

Analysis shows that these bones, with the exception of the loss of all organic matter, have undergone very little change. A sample examined by Mr. Lucas was taken from a typically preserved pelvis of Arsinoitherium and gave the following result:—

The following composition of the bones of an ox, from an analysis by Berzelius, is appended for comparison:—

It is curious that these Eocene bones should have so completely preserved their original composition considering the almost universal silicification of the trees deposited in the same beds.

Most frequently the vertebrate remains are found in an unconsolidated false-bedded clean quartz sand, the grains of which are semi-rounded or angular; in some layers the sand is very coarse and polished, and mixed with fine gravel. These deposits of sand, apparently brought down by river floods, are not continuous along any particular horizon, but are intercalated here and there in the ordinary sandstones, clays and marls of the series; they occur chiefly, however, as local lenticular masses along a more or less constant horizon near the base of the series. The bone-remains are not absolutely confined to these deposits of river-sand, but like the silicified trees are far more common in them than elsewhere. Scattered mammal bones occur in the lower clays, marls, and hard concretionary sandstones, while the remains of aquatic animals, such as turtles and crocodiles, may be found almost anywhere.