the Geological Survey memoir (Gunn, 1903) seems more satisfactory than that offered by the hanging valley hypothesis (Mort, 1914). According to the former, the waterfall is due to the presence of a band of schist indurated by contact-metamorphism. The argument that such a band should induce a similar fall in Glen Catacol, which has been advanced against this view, is scarcely valid, as the amount of metamorphism induced by the intrusion of the granite is very variable, a contact metamorphosed aureole 300 yards wide being found at some parts, while in others unaltered rocks are found within a few yards of the granite. Further, the nature of the schists, and consequently their resistance to erosion, is by no means uniform round the granite margin.

The general conclusion is that, while the valleys have been broadened by glacial erosion, the spurs being truncated and the walls steepened to some extent, there is very little evidence that there has been any great amount of over-deepening. Many of them must have developed their present courses and have been cut down to their present levels in pre-Glacial times, and it is possible that some of the through valleys which now form cols have, in the neighbourhood of the latter, been subjected to little downcutting since the 1000-foot uplift.

In the southern half of the island the topography at the time of the Pliocene uplift must have been very mature. The whole area had been worn down to base-level with the exception of a few low hills in the region of the great Cainozoic vent and the neighbourhood of Tighvein. The persistence of these must be ascribed to the superior resistance of the underlying rocks, the vent consisting mainly of granite and agglomerate with a mass of diorite on the eastern margin, and the Tighvein district of an igneous complex of granophyre and dolerite allied to the Marsco complex in Skye. These rocks would be better able to resist erosion than the soft Permian and Triassic sediments which surround them. Apart from the vent, the present drainage is moulded on the structure of the underlying rocks. The prevalent dip of the sediments throughout the whole area is approximately south-west, and the intercalated Cainozoic sills seem to have the same inclination. As the main streams flow south-west, they are therefore consequent, and flow down a dip-slope. The watershed runs north-west and southeast along a belt containing many igneous intrusions and stretching from Tighvein to Dippin Head, and a number of smaller streams flow down the scarp to the east. This escarpment seems to be migrating to the south-west, as there is a col at Sliddery waterhead which cannot have been lowered by the present streams which flow from it, and which was probably formed by the action of a stream whose source was further to the north-east.

The two chief south-westerly streams, the Sliddery Water and the Kilmory waters, are normal dip-streams, and the valleys are ordinary river-cut valleys. The Ashdale burn is mainly a strike stream in the upper (north-west and south-east) part of its course, but the position of the actual gorge is determined by a number of dykes of similar trend. The lower part of the course, where the flow is eastwards, is at right angles to the strike of rocks and against the dip, and a number of water

VOL. XXXIV.

H

falls are formed by the basaltic sills which are cut through. The final part of the course is a fairly deep V-shaped valley cut in the sediments.

Of the streams which debouch in Lamlash Bay, the Benlister burn is an obsequent stream which has cut through the original watershed and diverted part of the headwaters of the Clauchan Burn. A series of thin sills and dykes form a number of cascades in the stream, which falls rapidly from the 1000-foot level to the 200-foot level. The other valley, Monamore Glen, is of a different type, and seems to be along the line of an older valley. It is a deep, V-shaped valley, with great broad meanders for the main part of the course, but the upper portion is an immature gorge draining the Tighvein complex.

The Cainozoic Vent at the commencement of the present cycle of erosion seems to have been of the nature of a dome, and this has been dissected by an approximately radial drainage. The harder rocks of the eastern margin, however, seem to have constituted a barrier, as the streams to the east have not been able to cut back into the vent to any extent, while, on the other hand, Glen Craigag and Ballymichael Glen, draining north and west respectively, are typical deep river-cut valleys. These originally drained into a stream which rose in the northern granite area, occupied the middle part of the present course of the Machrie, and flowed southward past Blackwaterfoot. The consequent Clauchan Burn, draining the southern margin of the vent, was also a tributary of this stream. At a comparatively recent date the latter was intercepted by a small stream cutting back from Machrie Bay; the recent date is proved by the fact that the old river alluvium north of Shiskine is partly below the level of the 100-foot raised beach. Hence the capture must have occurred since the deposition of the 100-foot beach.

The valleys of the southern part of the island have undergone little modification by glacial action, with the notable exceptions of Glen Dubh and Glen Ormidale, draining into Glen Cloy, and to a less degree Benlister Glen. In the two former there are numerous moraines; in Glen Dubh a large crescentic series must have dammed up a lake at one time. The head of the glen forms a corrie, which is due partly, at least, to glacial action. These moraines, and the lateral ones which occur in Benlister Glen, are due to a local glaciation, proceeding from the vent. Although the whole area must have been covered by the main icesheet during the Glacial Period, the main function of the latter seems to have been protective, and there was probably no great amount of erosion during that period.

The area is chiefly interesting on account of the close dependence of the topography on the geological structure, and in this respect it presents many analogies with the basaltic plateaux of Skye (Harker, 1904). In both cases the intrusive sills dip gently in one direction, with a steep escarpment in the opposite; the thicker sills form sea-cliffs, as at Drumadoon, or inland escarpments, and the thinner sills, which give rise to such well marked terraces as in Glen Varrigill (Skye), have their counterpart in the prominent terraces on the south side of Glen Ashdale. In both cases the dykes give rise to gullies and ravines, and the present streams are often entrenched along the lines of these dykes. In Arran most of the waterfalls can be traced to the effect of the sills, the one notable exception being the lowest fall in Glen Ashdale, which is due to a band of hard sandstone (Tyrrell, 1915). Throughout the whole area there are practically no topographical features which cannot be accounted for on the basis of the geological structure.

REFERENCES.

BAILEY, E. B. 1910. In "The Geology of Ben Nevis and Glen Coe." Мет. Geol. Survey, Scotland, pp. 1-13.

GREGORY, J. W. 1915. “The Permian and Triassic Rocks of Arran." Trans.

Geol. Soc. Glasgow, vol. xv. pp. 174-87.

GUNN, W. 1903. In "The Geology of North Arran, etc." Mem. Geol. Survey, Scotland. 200 pp.

HARKER, A. 1904. "The Tertiary Igneous Rocks of Skye." Mem. Geol. Survey, Scotland. Pp. 434-51.

MORT, F. 1914. "The Sculpture of North Arran." Scot. Geog. Mag., vol. xxx. pp. 393-404.

PEACH, B. N., and HORNE, J. 1910. In The Bathymetrical Survey of the Scottish

Fresh-water Lochs, vol. i. pp. 439-513.

SMITH, J. 1896. "A New View of the Arran Granite Mountains." Trans. Geol. Soc. Glasgow, vol. x. pp. 216-56.

TYRRELL, G. W. 1915. “On...the Red Rock Series in Arran." Trans. Geol. Soc. Glasgow, vol. xv. pp. 188-99.

SOUTHERN PALESTINE: SOME NOTES ON HISTORICAL GEOGRAPHY.

By Captain W. R. KERMACK.

ONE result, at any rate, of the war, which is, perhaps, not wholly on the bad side is that, with its will or against its will, the great British public has been taught or retaught a certain amount of classical and Biblical geography. One British army nearly at the site of Troy, another in Macedonia, a third in Mesopotamia, yet a fourth in Egypt and Palestine, must have compelled many a paterfamilias to turn his mind painfully back to schoolboy days, and furbish up again his rusty armoury of classical information. Specially true, perhaps, is this in the case of the campaign in Palestine; and those who agree with Sir George Adam Smith, that "History never repeats, without explaining, itself," and who feel that the explanation of the history lies often in geography, will see with interest, but without surprise, our troops reaching in turn Beersheba, Ascalon, Ashdod, and Ekron, fighting in the footsteps of Joshua at Beth-horon, and facing Jerusalem, where Cœur-de-lion hid his face from the city he might not win on Nebi Samwil.

To the ancients Syria was first and foremost a bridge, between the two great river-civilisations of Mesopotamia and Egypt, with the desert on one hand, the Mediterranean on the other. Palestine, which is the southern portion of this bridge, consists of an inland strip of mountain land and a coastal plain, running roughly parallel to each other and to the sea coast, and each again divided lengthwise into two: the coastal plain into a belt of sand-dunes along the sea, and a broader inland plain of cultivation; the mountains divided by the Jordan valley. In the north of Syria the two great ranges of Lebanon and Anti-Lebanon are separated by the wide valley of Cœle-Syria. The highest summit of the Anti-Lebanon range is Hermon, on the slopes of which are the springs of the Jordan, which then flows southward through the high lake of Merom, descends 300 feet to the Sea of Galilee, and then in a series of twenty-seven rapids plunges down 1000 feet to its lowest depth (1292 feet below Mediterranean sea-level) in the great salt lake of the Dead Sea. Its deep valley, nearly 3000 feet below the level of the Judæan hills, and in most parts not more than eight miles wide, is part of a great "rift valley" which is continued southward to the Gulf of Akaba, which also forms a part of it, by the Wadi el Araba, and then, by the Red Sea into Africa. Western Palestine is thus cut off by this Jordan valley from the tableland of the Hauran and the Mountains of Moab.

But though there is access from north to south along the coastal plain, Palestine would never have taken the place it did in ancient history as a bridge between Egypt and Mesopotamia had the hill barrier west of Jordan remained unbroken. It is a capital point in the historical geography of Palestine that between the hills of Galilee, which are the roots of Lebanon, and the mountains of Ephraim, which are the northern end of the mountains of Judæa, there breaks the Plain of Esdraelon, an uneven stretch about twelve miles broad in the region between the Jordan and the Mediterranean Sea, reaching the Jordan through the narrower valley of Jezreel and, by the pass of the river Kishon, joining the coastal plain at the Bay of Acre, just north of the broad beach which rounds the promontory of Carmel. The western end of this Plain of Esdraelon is guarded by Megiddo, as the eastern is by Bethshan, while at its north-east and south-east corners are Mount Tabor and Mount Gilboa, four names ringing with historical association.

From Esdraelon southwards runs a long central spur, forming first the mountains of Ephraim, occasionally broken by wide plains and continuous tracts of vegetation, such as environ Shechem and Samaria, then the grey rounded hills of the Jebel El-Kuds, as it is known about Jerusalem, and then the still higher Jebel El-Khalil at Hebron. Thence it shades off gradually into desert. This plateau is, as a whole, well over 1500 feet above sea-level, while the hills on which Jerusalem stands rise to more than 2500 feet. Seen from the coastal plain, it seems to stand wall-like and aloof, its rock features bare and rounded, crowned here and there on their summits by a handful of trees or a mountain village. The passes of Southern and Central Palestine run east and west, climbing from the Jordan valley and the coastal plain to the central backbone ridge, on which are Jerusalem itself and Hebron, the earlier centre of the tribe of Judah, and, indeed, one of the earliest sites of civilised life in Palestine.1 From the plain routes of to-day reach the capital by the Wadi Ism'in, which is the course taken by the Jaffa-Jerusalem road by Latrun and Enab, by the Wadi es Surar (Valley of Sorek), the course taken by the railway, or by the Wadi es Sunt. But the usual route of ancient times was up the broad valley of Ajalon ("Valley of Stags," modern Wadi Selman), from which there is a gradual ascent by the Pass of Beth-horon from Beit Ur et Tahta (Lower Beth-horon) to Beit Ur el Foka (Upper Beth-horon), a descent to El Jib (the ancient Gibeon), and a climb again to the commanding height of Nebi Samwil, which is probably to be identified as the "Mountjoy" of the Crusaders, whence pilgrims had their first view of the Holy City, and more anciently as "the high place of Gibeon," where the Tabernacle stood before it was taken by Solomon to Jerusalem. It was this Pass of Beth-horon which was the scene of the great victory of Joshua over the confederacy of the five kings which gave Southern Palestine to the Israelites.1 Those indomitable road-makers, the Romans, in their turn drove a road straight -or as nearly straight as might be from Jaffa and Lydda (Ludd) to Jerusalem, over the most unpromising country by Beit Likia and Kubeibeh (Emmaus), a road which must have remained untouched since till taken in hand by our own sappers and pioneers in the present year of grace (1917). From Bethel, farther north on the main ridge, and on the north and south route along the ridge, there descends the pass which was the eastern counterpart of Beth-horon, the Wadi Harith, known to the Jews as the passage of Michmash or Ai, commanded at the Jordan valley end by the city of Jericho.2 Opposite Jericho are two of the four best known fords of the Jordan.3

Eastward from the mountains of Judea you look to the greater heights beyond Jordan; westward the view takes in the fertile coast plain, the dark expanse of fruit gardens round the white buildings of Ludd and Ramleh, Jaffa still farther off in a break in the sand-dunes, and the sea beyond. From some hill near his capital King Solomon may easily have watched the ships of Tyre bringing south the building material of the Temple-perhaps some

"Quinquireme of Nineveh from distant Ophir
Rowing home to haven in sunny Palestine,
With a cargo of ivory

And apes and peacocks,

Sandalwood, cedarwood, and sweet white wine".

though the principal port of Palestine was Ezion-Geber, on the Gulf of Akaba; and it is indeed small wonder that the inhabitants of these barren highlands should have gazed jealously on the richer Philistine plain that lay at the foot of their hills.

This coastal plain, the second great natural division of the land, stretches between the mountains and the Mediterranean, four miles wide at Acre, two hundred yards only where it rounds Mount 2 Ibid., pp. 158, 232.