CHAPTER XX

BACK TO THE CENTRAL TABLELAND

We were received with the greatest kindness by Shakir Effendi upon our return to Uran Gazi. The vigorous old man came to sit with us in our tent, and gave us some account both of himself and of his people. It appears that he has held the office of Kaimakam of Adeljivas, and that he occupied that dignity for four years. He is the Reis or supreme chief of all the Circassians in these districts; and he gave me a list, which should prove of some interest, of their villages. [263] He added that the population was increasing. The founders of the settlement, of whom Shakir was one, came to these seats after the last Russo-Turkish war. They were emigrants from the district of Kars, a home which they had adopted after the Russians came into possession of their native mountains. When the Russians captured Kars they received notice to quit, or, as Shakir put it, they were told to get out (Aideh!). They took ship, and landed upon the shores of the Black Sea within Turkish territory. But no arrangements had been made to settle them anew. They were starving and being decimated by sickness, when the Queen of England came to their aid. Her Majesty told the Turks that they must either find the land without delay or she herself would provide land within her dominions. This speech spurred the Turks on. In this way they became established in Uran Gazi. This kind action on the part of our Queen would always live in their memory. They are on good terms with the Turks, but they are preparing to move on again. That inexorable Russian advance!

As for the Kurds, they regard them as scarcely human beings and do not fear them at all. But they are held in great awe by the Kurds. Unlike the wretched Armenians, they are allowed to carry arms, which they know how to use with effect. And you hear the laughter of children in their villages.

While I was engaged in writing, the indefatigable Oswald scoured the plain in all directions. He found the limestones on its margin highly marmorised, full of corals; they must belong to the Eocene period. The efflorescence on the border of the Jil Göl is not in fact an incrustation, but is due to a bleached felting of confervæ. Rushes abound, but they had already been cut. The waters find an egress through two funnel-shaped basins, near a large crack in a boss of lava. They disappear beneath the ground in little whirlpools, and are believed to come to the surface at Adeljivas. Shakir assured us that, if anything, the lake is now on the increase; it has been increasing since the earthquake which was so destructive at Adeljivas about five years ago. This earthquake did little damage at Uran Gazi. Although the village is now about a mile distant from the lake, good water may be found at any point in the vicinity by digging a short distance down.

August 8.--Nimrud and Sipan having now yielded up their secrets, it was our next object to explore Bingöl. But, on the way, we were anxious to follow the course of the Murad, the reaches of that river between Gop and Charbahur being practically unknown. We were also desirous of climbing Khamur. Our first day's stage was to be the village of Gop. [264] We therefore crossed the plain in a north-westerly direction, the way being indicated by a Circassian guide. After riding about three miles we reached the foot of some low hills, confining the plain on that side. Against their first slopes lay a Kurdish hamlet--Karaghun. Issuing into a valley, we rose above it to the crest of the hills, which, as we expected, were down-like in character. For some little distance our way led over these downs.

But I need not tire my reader by taking him over old ground; he will readily recognise that our surroundings were much the same in character as during our journey across this region from Melazkert. For the second time we were crossing the block of stratified rocks on the west of Sipan; but on this occasion we were already within their northerly and less elevated zone, and we might, no doubt, have descended to the plain, and followed along their base. We struck our former route above the village of Demian, after passing through the same valley to which we had then come down at Akhviran, and which we now entered above the large Kurdish village of Shebu. It is an inlet of the great plain at the foot of Sipan. But our guide preferred to take us along the slope of the mass, all the way from Demian to the village of Leter. For that is the direction which these heights pursue.

I have little doubt that these stratified rocks come up again on the east of Sipan, and the view from the eastern summit disclosed in that direction very similar block-like heights. They probably sink beneath the volcanic system of the Ala Dagh. In this northern zone the downs consist of lacustrine deposits, sandstone, and a limestone full of the little shells known as mytilus. [265] The sandstone underlies this mytilus limestone, indicating, as Oswald observed, that the great lake, which once covered this region, grew deeper before the latest earth-movements set in. Throughout the ride from Demian to Leter, a distance of 5 3/4 miles, we overlooked the flat region which is due to the action of those former waters, and through which the Murad flows. But a gale of wind was in our face; the plain was shrouded in haze--a treeless and little-inhabited district, which might, no doubt, be made fertile and prosperous.

Leter, a large village, partly Kurd and in part Armenian, is situated on the confines of the plain. It is built upon lava, black and slabby in character, which has broken through the lacustrine deposits. Similar bosses, resembling those near Uran Gazi, but larger, rise up from the level expanse beyond. The direction in which the inhabitants pointed towards the village of Gop was plainly not that of the lake of the same name. We had already obtained a glimpse of its waters, lying almost west of where we now stood. Anxious to visit the lake, we shaped a course which we thought would find it beyond a screen of low hills. The plain in that direction was very rudely cultivated, white hollyhocks and a large mauve thistle crowding out the ragged corn. At about 6 1/2 miles from Leter we passed through the first village we had since seen, the large Armenian settlement of Kekeli. And in another ten minutes we stood on the summit of the eminence which had concealed the lake for so long.

It was nothing more than a low hill, an isolated mass of lava rising up from the plain. It was crowned by a little chapel, put together with stone and mud, and provided with a wicker door. Looking through, we discovered a large stone, engraved with a cross, which was, no doubt, the object or symbol of worship. Before it, three little lamps reposed on a horizontal slab. From this standpoint we overlooked the extent of the waters, of which the nearest shore was still some two miles off. The lake is bordered by level ground upon the east and south, and by considerable heights on the west and north. On the west it is Bilejan, sending outwards radial buttresses with deep valleys from a central, meridional ridge; lesser heights in connection with the mountain, and of volcanic origin, descend to the waters along the northern shore. A slight depression separates this series from Bilejan, and they, in turn, sink somewhat steeply into the plain. At that point, and at their foot, lies the large village of Sheikh Yakub, beside which flows the stream giving issue to the lake.

In the opposite direction, beyond the plain on its southern confines, it is overlooked by an extension of the heights on the west of Sipan, which are continued up to the mass of Bilejan. It would appear that volcanic action has been busy throughout this region; and we thought we saw a grassy crater among those heights. Beyond the outline of the barrier emerges a little conical peak, which we recognised as the cone on the west of Nazik. Bilejan itself does not look as if it ever could have possessed a crater, and it is probably due to upwellings of lava along a meridional fissure. The highest points along the central ridge may have an elevation of some 9000 feet.

A brisk breeze was blowing as we made our way to the brink of the water, churning up its muddy depths. Indeed this lake is thickly charged with dark sedimentary matter--a characteristic which has given rise to a name under which some know it, Lake Bulama, or the muddy lake. [266] Another and not more savoury feature is its odour, which is fetid and nauseating. An abundance of fresh-water mussels were strewn on the shore, and several pelicans were floating on the waves. In shape the lake appeared to be almost circular. Its elevation, as one would expect, is much less than that of Nazik, being only 5550 feet. In point of verdure its surroundings are quite as mournful as those of its neighbour, while the lake itself does little to relieve their monotony.

From the north-eastern extremity of this unattractive sheet of water we followed the course of the foul stream by which it is drained. It took us to the foot of the heights already mentioned, and through the village of Sheikh Yakub. It is a very large Armenian village, which has probably been prosperous, but which is now in a state of extreme destitution. All the inhabitants were in rags. Boys up to the age of puberty were quite naked, and girls to their fifth or sixth year. The village was full of soldiers, who were standing on the roofs. I summoned their officer, and enquired what their business might be. He answered that Ibrahim Pasha, adjutant of Kurd Hamidiyeh, was about to visit the place. In Gop I ascertained that the object of his visit was to restore some property which had been carried off by Kurds. Such at least was the explanation which I received. It was certainly not a bad idea to quarter all these troops upon starving people; they would think twice before claiming redress a second time. But I suspect that it was a rather clumsy lie.

Gop is situated in the plain, some miles distant from the lake, at the foot of the extreme slopes of the heights which border the northerly shore (alt. 5150 feet). Although the place is the capital of the caza of Bulanik, it is a large village rather than a town. The Kaimakam informed me that there were 400 houses, all but 50 inhabited by Armenians. The district of Bulanik comprises some of the most fertile land in all Armenia, and is of considerable area. Towards the east it includes a large portion of the plain of the Murad below the town of Melazkert; while, on the west, it reaches across the mass of Bilejan and its outliers to a second extensive stretch of fairly level ground. That region slopes away from the northern border heights of Mush plain to the Murad and the opposite heights of Khamur; it sends a tributary to the left bank of the great river, and one of its principal and central villages is that of Liz. The fecundity of the soil is probably due to a happy combination of calcareous marls with the detritus of eruptive rocks. The grain which it produces is of excellent quality, in spite of the fact that the fields will be full of thistles. The peasants are miserably poor. The Kaimakam explained that their rags, and squalor were matter of custom (tabiat); and, in fact, they had plenty of money, hoarded away. It is possible that such an hypothesis may indeed govern some of his actions; but I doubt whether he put it forward in good faith. The main cause of their destitution is plainly the want of security, coupled with the impossibility of exporting their crops. But usury is also a factor of considerable importance, the husbandman having generally borrowed to buy his seed at rates which rob him of most of the earnings of his toil.

From Gop we made a second excursion to the lake, riding to one of the most conspicuous of the volcanic eminences which rise from its northern margin. It is a distance of about four miles. The ascent commences on the outskirts of the village; but it is at first very gradual, the slope consisting of marly clays. These beds were full of mytilus in perfect preservation, and were seen to have been overlaid with tuffs. About halfway, we came to the walled monastery of Surb Daniel, containing the relics of a saint of that name. The ancient chapel has been restored. Over the altar was conspicuous a picture of the Virgin and Child. The one or two resident priests were sunk in abject ignorance, but they were in possession of some good farm buildings within their enclosure. We remained for some time upon the peak which we had selected, and from which we obtained a fine view of the lake and its surroundings. While I was mapping, Oswald sketched. We could see two villages on the level ground south of the lake--Khashlu and Piran. In the plain towards the Murad several settlements were visible upon a line between Leter and Gop.

August 10.--It was ten o'clock in the morning of a fine summer's day when we resumed our journey, and set out in a north-westerly direction across this spacious plain. Travelling at this season is most agreeable in Armenia; it scarcely ever rains, yet one is never overpowered by the heat of the brilliant sun. Pleasant breezes float across the expanse. The harvest was being gathered in. Our landmarks were in full view--Sipan, Khamur, Bilejan. A little river meanders through the deep soil, on a course towards the Murad. It receives the waters which irrigate the village of Gop, and, among them, those of the stream from the lake. It has its origin some distance east of Gop. It is called the Kör Su. At first our track took us about parallel with its banks; then we crossed it at the large Armenian village of Yungali. Anxious to visit the point of confluence of the Bingöl Su with the Murad, we now diverged towards the north. The nature of the ground compelled us to cross the latter river a little above the junction. It was flowing in a very broad, alluvial bed. In width it may have been about a hundred yards; nor in any place did the water reach much above our horses' knees. Except for the great islands of mountain about us, we might have been standing upon the Mesopotamian plains. Our approach disturbed a group of large eagles, so heavy that they were obliged to run before taking wing. The Bingöl Su came in through a deep channel, which washed the girths of our horses. It did not seem to be more than forty yards wide. But, although sluggish, it must bring a very considerable volume of water; for its contribution extended to about half the width of the joint river, being clearly distinguished by the quantity of sediment which it sustained. From this confluence we followed the course of the Murad, riding over the plain on the right bank, with the stream. A flock of wild geese were resting in the pebbly bed, nor did the shapely birds move as we passed them by. One of our escort was successful in securing a fine specimen with a bullet, which provided us with an excellent meal next day.

But the features of the landscape soon underwent a change; for the river was approaching the foot of the Khamur heights. At first it was low hills, consisting of lake deposits, which we skirted on our right hand. But near the Armenian village of Karaogli a bold ridge comes into prominence, and it extends all the way to Shakhberat. It is of eruptive volcanic origin. It is an important member of the series of heights of which Khamur forms the dominant mass. East of Khamur that series rises to a considerable elevation before declining to the valley of the Bingöl Su. The highest ridge, as seen from this district, lies some distance towards the north, and is called the Zirnek Dagh. On the other hand, this volcanic parapet comes right up to the river, which follows along its base. At the same time hills started up from the plain upon the left bank. It was evident that they were volcanic and in connection with Bilejan, of which we were opening out the more westerly and less deeply carved side.

These features transformed the scene with startling rapidity; the idle river was no longer able to flow where it pleased. Some two miles below Karaogli it enters a deep gorge, and throughout its course to the plain of Mush it is, with little intermission, confined in a narrow bed. Except during the passage of the block of heights on the north of that plain, the Murad performs no considerable feat. It follows the general trend of the lines of elevation, and one would expect its course to be fairly tranquil through this region. But the lavas tease the river; they have welled up along fissures, and have converted the wide valley into as inhospitable a district as any through which it passes on its long journey to the Persian Gulf.

Our mid-day halt was spent beneath the shade of a grove of willows, on the margin of some fields of hemp and cabbage, which softened the site of Karaogli. But, the village left behind, we soon entered the narrows, the track being taken along the cliff-side, at some considerable height above the hissing, silvery water. The Murad pierces a mass of lava belonging to the ridge on its right bank. It seemed a wayward thing to do; for the ground is lower just south of the gorge, and appeared to invite the river. While still within the cleft, it was spanned by a wooden bridge resting on several piers of solid masonry. This is probably the first bridge over the Murad below Tutakh. Issuing from the cliffs, the tortuous reaches opened out into an easier country, and a wider prospect was unfolded on either side. For the first time we obtained a view over the plain on the west of Bilejan, bordered on the south by the still distant heights, on this side of the depression of Mush. But the volcanic hills on the left bank were not long without a successor; the outline was taken up by a second block of similar origin; and the scene again became restricted to the immediate surroundings of the river, which were stony and bare and bleak. We passed only a single Kurdish hamlet during our ride to the cirque or caldron of Shakhberat. There the river makes an S-shaped bend through a fairly wide valley, enclosed on all sides by volcanic heights. The ridge and peak of Kolibaba is seen to full advantage, confining the valley on the west. Two little Kurdish villages, Arenjik and Shakhberat, lie on the slopes and in the lap of this spacious cirque.

August 11.--The level of the Murad at Shakhberat was tested by two readings of the boiling-point apparatus on successive days. It was found to be 4900 feet. The village commands a view of the summit of Khamur, the highest point of the amphitheatre in which the hamlet is placed. But that lofty ridge is in part screened by the slopes of Kolibaba, and by the parapet which has skirted the right bank of the river all the way from Karaogli. That parapet joins the mass about opposite the summit; and it is only at the head of the valley between Khamur and Kolibaba that the outline and slopes of the principal ridge are fully exposed. The shortest way to the summit, but certainly the steepest, would lead up that valley by a fairly direct course. But our guide preferred to take us by a more easterly approach, up the face of the parapet. He was a very pleasant fellow, a khoja or village priest; and he looked well with his clean white cottons, astride upon his mare. But the notion that we really intended to mount to the actual peak was repugnant to his good sense. Climbing Khamur meant to him proceeding to an adjacent eminence and thence contemplating the airy heights above your head. Such a spot was provided under circumstances of luxury by the site of a hamlet high up on the ridge. A rustling stream flows through it, which has been dammed and made into a lake; and round the pool trees have been planted to shade the flocks. It was indeed a charming foreground to the immense landscape which already extended to Nimrud. But the khoja's dallying was soon cut short; the track ceased at this village, which bore the unworthy name of Ganibuk. He was forced to lead us across a beach of large boulders, and through some thickets of oak scrub. The ascent became pronounced, and, when at length the flat top was reached, the main mass was still distant, and looked very high.

The composition of the ridge, which we had now surmounted, is at once interesting and typical of the whole region. It consists of a series of deep beds of lake deposits, separated one from the other by bands of lava. At first the lava was seen to be basaltic in character and compact; but towards the summit it became scoriaceous. The fact would seem to indicate that, while the earlier issues were submarine, the latest flows were outpoured when the land had risen above the water, and the present configuration was being attained. The platform upon which we stood was composed of a sheet of lava, and so was the summit of the opposite ridge of Khamur. But as we rode into the shallow trough which separates the two eminences, the greyish-white marls again came to view. We could see them on the escarpments of both ridges, which, further east, became gradually separated by a deep latitudinal valley. We could observe the soft material where it was baked into a yellow porcelain by contact with the cap of lava on the Khamur ridge. Far and wide, towards east and south, the landscape wore the same hues and appearance; the same character appeared to belong to the heights on the north of Mush plain. Descending into the depression and rising again on its further side, we reached the actual peak or highest part of the Khamur ridge after a zigzag climb up a slope overgrown with fennel. We had attained an altitude of 9850 feet.

Although the outline of Khamur assumes a somewhat pointed shape when seen from the south, as from the cirque of Shakhberat, yet the summit is nothing more than a fairly flat and narrow platform, which slopes away with some abruptness on the north and south. The lava upon this platform is slabby in character and may be described as an augite-andesite. There is no crater on the summit and one cannot speak of Khamur as a volcano in the proper sense. In fact it is a considerable block of elevated land, in the western portion of which volcanic action has played a great part. The foundation of the block is probably composed of Eocene limestone, which has been overlaid by later lake deposits. This limestone comes to view in a remarkable manner as you survey the eastern half of the mass. The ridge upon which you stand extends for a mile or two in that direction, and presently sinks to a somewhat narrow upland valley. This depression can be clearly seen from the adjacent region, whence it has the appearance of a notch in the outline of the mountain. Its eastern slope leads over into a very broad block of mountain, of which the central region is hollow and basin-like in shape, and the outer sides steep and high. They are perhaps steepest and most lofty on the south. It is in fact one grand synclinal, described by beds of hard limestone, which, from a distance, groups with Khamur in a single mass. The axis of the mass in that direction is about east-north-east.

The prospect towards the west is not of lesser interest, and is certainly even more strange. Time was wanting to examine this extraordinary region, which, indeed, it would require several days to explore. The ridge of Khamur is joined on to the northern portion of another great block of elevated land. The eastern wall of this plateau projects some distance towards the south, almost up to the right bank of the Murad. But the deep valley, which is formed by this projection and by the ridge of Khamur, is filled up by the lofty pile of Kolibaba--a peninsular mountain, only connected beyond a considerable depression with the slopes of the main ridge. It is plainly of eruptive volcanic origin, and it is somewhat circular in form. Its sides are strewn with talus and clothed with oak scrub. Our guide and the people of the district knew it under the name which I have given, and which they averred to have been that of a holy man who had been buried there. Though who this prophet might have been, or what he wrought, or when he lived, not a soul among them knew. Throughout this district, as far as Bingöl, the tops of mountains will be often crowned by the rude enclosure of some sage's grave. Such a monument was a conspicuous object on the very peak of Khamur; and with its headstone, a huge slab grimly resembling a human bone, might have been disposed to receive a giant's remains (Fig. 189).

But to return to the scene before us--this adjacent plateau on the west extends all the way to Bingöl. Indeed it is connected with the southern margin of the Bingöl pedestal by a bold saddle, due to a flow of lava. This feature was, of course, scarcely visible from Khamur; but the continuation of the Khamur ridge might be traced throughout the region, being distinguished by a succession of bold bosses, rising along its northern margin. These peaks are especially pronounced at their inception, and appeared to rise almost immediately from the northern shore of a large lake which was irregular in form. Near its south-east corner lay a second, much smaller and circular lake. Neither figure on any map which I have seen. They are evidently rather deep, for their colour is an intense blue. Such are some of the characteristics of this curious region, which may be included among the Khamur heights. It rises above the Murad with cliff-like sides, which scarcely decline at all to the elevated level of these lonely azure lakes.

The view from the summit of Khamur may be divined by my reader; nor need I attempt to describe it in any detail. It embraces Palandöken and Bingöl on the north; Nimrud, Bilejan and Sipan on the south. On one side lies the plain of Khinis, bordered by the peaks and ridges of the Akh Dagh, which are seen to their fullest advantage. On another it is the basin of the Murad, with Sipan rising in all his majesty from an expanse of level and cultivated ground. In yet another you overlook the plain and village of Liz and the course of a winding river. From no standpoint does the character of the country, as a succession of sea-like plains, become imprinted with greater clearness on the mind. Nor is any district in the nearer Asia better adapted to become the granary of a prosperous and highly civilised land. We descended in failing light by the valley on the side of Kolibaba, and reached our camp with some difficulty in three hours.

August 12.--On the following morning we set out to follow the Murad, as far as its egress from this region through the heights which border the depression of Mush. Much the same features were continued throughout the stage. The cirque of Shakhberat is enclosed upon the west by a stream of lava from Kolibaba. This emission has flowed in an almost meridional direction, and has forced the river to bend away to the south. After passing through the Armenian village of Akrag, which is placed on a higher level of the basin-like area, we breasted the bold ridge which has been formed by this lava, and crossed it just south of a little parasitic cone, emerging from the side of the mountain. Some low oaks flourish among the boulders; the rock is a glassy augite-andesite. The ridge leads over into a little plain, bordered upon the north by the wall of the Khamur plateau, and with a high rim upon the side of the river, which cannot be seen. The plain has evidently been covered by a lake in fairly recent times. A crack in the rim of the basin displays the channel through which it was drained. On the side of Kolibaba an old beach line was visible, some fifty feet above the level of this plain. The soil consists of a black clay which is not cultivated, but which must be very rich. It took us nearly an hour to cross to the opposite side, where it is confined by an outwork of the Khamur heights.

Our further journey, which occupied the better part of a whole long day, need not be followed step by step. We had arrived at a spot where the dominant lineaments of the landscape had already become pronounced. These prevailed with little variety all the way. On the right bank, at an interval of two to three miles or more, rose the wall of the Khamur plateau. The further west we proceeded the more irregular it became, the less distinguishable from the massive spurs which it put out. These outworks descend into the river valley, which is flooded and choked by the lavas. Both in the valley and on these slopes the lavas have the upper hand; but the grey lacustrine marls are seldom absent or for long. They provide favoured stretches, covered with luscious herbage, where a little stream may trickle down from the barren heights. Still the scene remains wild, bleak rather than of impressive ruggedness; there is space along the margin of the river, which flows in a deep cañon through the sombre eruptive rock. Some stunted oak springs from the crevices among the boulders, but it rather enhances than relieves the mournful aspect of the surroundings.

On the left bank a new feature came into prominence: a long and fairly lofty ridge, with perfectly horizontal outline, many miles away on the south. But the slope of this broad mass was continuous to the brink of the river, where it was broken by the stream into cliffs. Its gentle gradient and almost level surface somewhat softened the rigour of the landscape. It was seen to consist of a sheet of lava, which had covered up the marls, and which must have issued in a very liquid condition. The heights upon which it is built are the northern border heights of Mush plain; and this block of heights approaches closer and closer to the Murad, as it eats its way through the district on a westerly course. Such was the character of the country beyond the winding, hissing river throughout the whole stage. Villages there were none, and hamlets few. A single oasis of any importance was observed high-seated upon the slope in the south, near the break in the outline where the Murad pierces through the block. We should have been pleased to spend the night in that extensive and leafy grove, which belongs to a village called Ali Gedik. But we were assured that the river was not passable, and we were obliged to push on to Charbahur. After following a romantic gorge where the Murad has again been wayward, and has preferred to saw a passage through a towering parapet of lava rather than to follow easier ground upon the south, we rode for some distance along a wide stretch of alluvial soil in which the river at length reposes from its arduous labours. The Circassian village of Charbahur is placed at some distance from the waters, on the northern margin of the broad strip of willow-grown land. [267]

Charbahur is backed by a barren slope of the Khamur heights, and is screened from all freshness on the side of the north. On the other hand, it is exposed to the sultry southern breezes, which find their way through the passage of the Murad, acting like a funnel to the furnace of Mush plain. There were said to be some sixty houses in the village; but I should say that there were more. Some of the tenements are well built, resembling neat cottages; but unfortunately they swarm with fleas. The standard of living is far higher than among the Armenians; but one feels that there is little or nothing in the race. Our impression of the Circassians did not improve upon longer acquaintance; although they are by no means the worthless and predatory people which they are sometimes represented to be. Their conspicuous characteristic is an inordinate love of swagger; and their handsome figures encourage the tendency of their disposition. One afternoon, as we were busy at work, a bugle sounded; and immediately a band of horsemen galloped into the village. One by one they passed our tent at the utmost speed of their horses, jumping to the ground and vaulting back into the saddle, while still at full pace. Those Cossack manoeuvres heralded the approach of their chief, Suleyman Pasha, who, it appeared, was riding over from the neighbouring capital of the caza in order to honour us with a visit. When he arrived the place became full of irregular troops, with whom were combined a small detachment of regular cavalry. Dismounting from a well-bred horse, he came towards us with hands outstretched, tall and supple, with a rhythm of movement which at once revealed his Circassian blood. His large and animated eyes, the thin, aquiline nose, the high forehead and the black hair, waving on brow and chin, were set off by the contrast of a very correct uniform--a deep-blue tunic with a pale crimson collar. The voice suited the man; it was resonant and was meant to be so, and his words were accompanied by a profusion of gestures. He was followed by two valuable English pointers, which, however, he did not treat with proper respect. To him the world was a gallery; yet he lacked the mind of the actor; and, while his principal occupation was the giving of orders, his directions were not less empty than his words. But these defects were in the nature of inherited failings; personally he was extremely kind, and, I believe, a staunch friend. He spoke with gratitude, which was sincere, of the service which had been rendered to his countrymen by England and England's Queen. It has sunk deeply into the hearts of Circassians. At home we are too much imbued with excellent business principles; and few of us realise the value in politics of sentimental considerations, especially when we are dealing with the untrained peoples whose destiny happens to link with ours.

The most interesting occupants of such a village are, no doubt, the girls and young women. They retain their fair complexion even in this climate, as well as their roundness of face and form. Several among them would come to the margin of an adjacent stream, in order to wash their grain. Their bare feet were as shapely as their hands. From Charbahur we made an excursion to the passage of the Murad, riding first to the confluence of the important stream which collects the drainage of the southern slopes of the Bingöl plateau. A ridge from the Khamur heights extends across the wide valley, choking it up and checking the drainage of its considerable extension towards the west. The stream cuts through this obstacle a little west of Charbahur, issuing into the alluvial plain at the Circassian village of Charbahur Tepe. It joins the Murad at the egress of the river from the valley. It comes in beneath the shade of willows and silver poplars. It brings a large addition to the waters of the Murad, and is by far its most important tributary since it received the Bingöl Su. Unhappily this affluent bears the same name as that river; but I need not fear that my reader will confuse the two. This Bingöl Su had a width of about 30 yards; its depth was fairly uniform, and it reached above our horses' knees. [268] The Murad now becomes a stately river, recalling, both by its volume and the manner in which it flows, the course of the Danube in Upper Austria. We forded it at a point some 3 miles down the passage, where it was over 100 yards wide and reached above our horses' girths. It had descended to a level of 4570 feet.

The cutting through the broad block of mountain which is interposed between the plain of Mush and the long valley through which the river has been flowing for so many miles--a valley which is continued as far west as the little plain of Dodan--is perhaps too broad to be described as a gorge. Yet the heights on either side descend to the margin of the Murad, which has turned at right angles to its former course. It pursues this southerly direction until it has gained the floor of the Mush depression. From the ford we mounted the slopes on the eastern side of the valley, and, after a sharp climb, reached the summit of the block. Our position was a little south of that pleasant grove which has been mentioned, belonging to the village of Ali Gedik. We stood on a sheet of lava; but the limestone was all about us, on the face of the cliff, in the bed of the river, where it formed long ridges, fretting the current into rapids. It was seen to contain fossils of the cretaceous period, and its strike or axis of elevation was towards east-north-east. The heights on the opposite bank appeared to be of similar nature. The view extended over the plain of Mush. Mush itself was seen nestling in a recess of the border range. We could see the village of Sikava, well in the plain, and the almost imperceptible break in the wall of mountain where the Murad issues from the plain. In the north, the line of cliffs belonging to the Bingöl plateau dominated the scene. Bingöl itself was either hidden behind their lofty edge, or could not be distinguished from the mass. We returned to Charbahur not along the valley, but down the gentle southern slope of these heights. Its even nature is due to a flow of basaltic lava. We found the Murad above the junction with the Bingöl Su to be flowing in two separate channels, which we forded and so returned to our camp.

August 15.--To reach Gumgum and the westerly extension of the long valley, it is necessary to cross the ridge from the Khamur heights which I have mentioned; such was our purpose and our next task. We found it to consist of grey lacustrine clays and marls with interbedded lavas. A thick layer of tuff occurs high up on the ridge; and the summit of the whole formation displays a cap of basaltic lava, sloping northwards in the direction of Gumgum. The parapet lessens in height as it stretches obliquely into the valley towards the block on its southern verge; yet even at the lofty col, over which our track lay, it was less elevated than the corresponding ridge which joins the Khamur heights to the Bingöl plateau, and which is surmounted by the road from Gumgum to Khinis. As we descended, a pleasant stretch of fairly even ground lay beneath us, in the lap of which we could see the capital of the caza. It was watered by several streams, which issue from the slopes of the wide amphitheatre described by the Khamur heights and the bold outline of the Bingöl cliffs. One river alone was seen to proceed from the very heart of the Bingöl system, coming into the plain through a tremendous chasm in the cliffs. Above that abyss we obtained a glimpse of the western summit of Bingöl. Further west the great valley was choked up with minor heights, rising up from its floor. On the south it is bounded by the commanding block of mountain which continues, across the passage of the Murad, the long wall of the northern border of Mush plain. Limestones, buff and white, could be seen high up on that flat-topped mass, with the same axis of elevation as those further east. The scene was bleak, without a tree and scarcely a bush.

Gumgum had evidently blossomed since my last visit, for it possessed at least two stone houses above ground, besides several little shops. We found it in a state of extraordinary commotion, owing to the presence of Suleyman Pasha. A troop of regular cavalry, mounted on white horses, had met us on the road. They had been sent as a guard of honour to escort us into the place. The scene before the Government building was extremely picturesque; and what was our astonishment when we beheld, among the medley of Circassian cavalry, a ragged band of horsemen whom we at once identified as Kurds, and in whom we recognised the much-talked-of Hamidiyeh! Here indeed was food for the note-book and the camera! On the steps of the building stood the Kaimakam, not my friend of the first journey; and beside him the Hakim in a black robe. Behind these were gathered the notables, and among them a giant who enhanced the imposing nature of the show. When we had received and returned the greetings of this distinguished company, we were ushered into the presence of the Pasha, seated in an inner room. He overwhelmed us with every token of kindness; and, when the Kaimakam read me a telegram relating to a supply of money, he waved him aside with a gesture of magnificent contempt, and drew from his pocket a reel of gold which he begged me accept. A little speech, modelled on his own, seemed to allay the sting of my refusal; but he insisted upon our taking with us to our camp on Bingöl a detachment of cavalry. This offer was gratefully accepted. Orders were at once given to prepare a repast. The servants left the presence with a deep obeisance; but, alas! it transpired, after a considerable interval, that there were no viands in the house and none to be found. All this time the audience chamber was filled full of as strange a company as it had ever been our privilege to see. Suleyman Pasha appeared to hold a roving commission in connection with the Hamidiyeh. But the men of his own race, settlers in the country, had come in from all directions to do honour to a countryman in his high position, and to a nobleman in whose veins their bluest blood flowed. The Circassians furnish recruits to the regular army, differing in this respect from the tribal Kurds. But, jealous of their ancestral customs, they maintain the irregular cavalry, of which a strong contingent was gathered together in Gumgum. The principal men, one by one, were introduced into the apartment; each bowed low and kissed the Pasha's hand. To each was assigned a seat on the divan. Most had passed the middle age; their wizened and wrinkled faces harmonised with the drab hues of the Cossack dress. The Pasha was resplendent in his blue and crimson uniform; several swords, in richly engraved and valuable scabbards, rested by his side. Near him sat a grave and gloomy personage in European uniform. His cruel face displayed the true Tartar lineaments and expression; yet he was a Kurd, and the colonel of one of the four Hamidiyeh regiments recruited among the Jibranli tribe. The Pasha treated him with great courtesy, if with a little condescension; but, although he received the many orders which were addressed to him with military obedience, his manner scarcely concealed the irritation which they produced. There was mischief in the man's face. He is seen on the left of my illustration (Fig. 190); his bugler, a young Kurd, richly attired, is placed on his left hand. Behind him are some of his horsemen, of which in all there were mustered a hundred, after extraordinary exertions on the part of the Pasha. Yet the nominal strength of the regiment is six hundred. The whole force--regulars and irregulars, Kurds and Circassians--were drawn up in a half-circle for our benefit. The regulars were, as usual, a fine body of men; of the rest the very refuse were the Kurds.

We did not regret to leave a scene which was pathetic as well as humorous, and to set forth on an expedition to one of the most remarkable of those works of Nature with which Asia--past mistress of violent contrasts--appears to mock the contemporary littleness of her sons. We had experienced the greatest difficulties in obtaining supplies; for the wretched shopmen, alarmed at the inundation of undisciplined soldiery, had absconded after barring up their humble booths. The promise of some cavalry had proved empty; none came or intended coming. We had said good-bye to our excellent escort from Akhlat, of whom the officer, a handsome man with charming manners, had suffered in health owing to the hardships of the journey. But we had been met by our tried and trusted zabet from Erzerum; and to him was attached a fellow-officer from Gumgum with several men. We might have proceeded on a fairly direct course to our mountain, which indeed is situated almost north of the little town. But I was anxious to retrace my former journey as far as Dodan, in order to complete my rough survey of this interesting region, interrupted on that occasion by failing light. Our course was therefore directed up the long valley, with the outline of the stupendous Bingöl cliffs on the one side, and, on the other, that of the border heights of Mush plain. At the hamlet of Alagöz we forded the stream which comes down through the great chasm, and which, perhaps, for want of a better name, we may call the Gumgum Su. It unites at this point with the combined streams which water the plain, and the joint river flows off through a gorge in some minor heights to effect a confluence with the Bingöl Su. I have already mentioned that the valley is choked up with insignificant hills; on its southern margin flows the river last named. Eruptive volcanic action has played a great part in its configuration; and the axis of the masses of lava which rise up from its floor is about the same as that of the plain of Mush. These eruptive hills are varied by heights composed of limestone, or of marls and clays, interbedded with lava and tuff. After a long ride through this wild scene we at length emerged upon the plain of Dodan, level as the lake which it must have supported in fairly recent times. Dodan lay beneath us; but we pushed on to a further village, the picturesque and pleasant settlement of Gundemir.