CHAPTER XVII

TO ANI AND TO KARS

October 14.--We left the cloister at half-past eight, our little party of five persons including the Armenian cook. We had hired in the district ten miserable ponies, of which five carried our effects. The most direct way to Ani crosses the basal slopes of Alagöz, from the southern to the most westerly extremities of the shield-shaped mass. You proceed from Edgmiatsin in a north-westerly direction, the ground rising at every step of your advance. On the point of course, beyond oases of verdure in the foreground, lie the stony and arid declivities of the mountain--contours of immense length and low vaulting, joining the plain to the horizontal outline in the sky. The belt of verdure consists of fields of the cotton and the castor-oil plants, of patches of orchard and vineyard, and sparse groves of poplar, rising from the dusty and boulder-strewn waste. It is sustained by runnels which exhaust the waters of the Kasagh or Abaran Su, the stream which collects the scanty drainage of the volcano upon its eastern flank. The boulders are worn by water and have been dispersed by the swollen river, during the season of spring floods. Where we crossed the Kasagh itself, or principal channel, it was a languid and soil-charged body of water, threading these stony tracts. We passed several villages within the irrigated area, some inhabited by Armenians, others by Tartars, and a few by both races alike. Hiznavuz, or Kiznaus, an Armenian settlement, containing the State-school of the district, was the last of these hamlets of the fertile zone. We stayed a few minutes before the open windows of the schoolhouse, listening to a lesson given in Russian to Armenian boys. Behind the village, a sterile eminence leads over into the barren highlands which compose the pedestal of Alagöz.

The moderate elevation of these highlands above the plain of the Araxes and their long extension from east to west are conditions favourable to the full appreciation of the landscape, and of each new feature in the slowly-changing scene. Their free position contributes to invest them with the character of a natural gallery, which commands unbroken prospects over some of the grandest works of Nature in her most inspired moods. The European, whose conception of mountain scenery is founded upon the arbitrary peaks and scattered valleys characteristic of his Alps, who has looked with emotion upon the doubtful features of his lowlands from the summit of some famous pass, can scarcely fail to be deeply impressed by the attributes of a panorama in which reliefs and depressions of stupendous scale are disposed as members of a great design, and are seen in the pure atmosphere of an Eastern climate with all the clearness of a model in clay. At his feet lies a plain which is level as water, which in no very remote geological period was covered by an inland sea. It is a distance of some thirty miles to its opposite confines; yet the towns and the plantations are pencilled upon its surface as though they had been traced by a draughtsman's pen. The plain is bordered by the volcanic range which we have come to know as the Ararat system--a chain of which the jagged and fantastic outline is already familiar from many a rich sunset effect. The summits rise to nearly 8000 feet above the campagna; but how humble they appear behind the train of the fabric of Ararat, gathering immediately from the floor of the plain! The bold snow bastions of the north-western slope are seen in face from these highlands; and it is difficult to realise that the pronounced lineaments which compose that airy figure are removed by a space of nearly forty miles. We had not yet lost sight of the line of poplars which screens the cloister when the distinctive features of this magnificent landscape were unfolded to our view. The several ranges and mountain masses were disposed in the form of an amphitheatre, of which we seemed to occupy one of the middle tiers. In the east, along the Araxes, the crinkled buttresses of the northern border were still visible, projecting in a southerly direction beyond the cock-combed hill of Karniarch. In the west, at an interval of sixty miles from those eminences, the level ground extended to a double-peaked mountain which juts out into the valley from the Ararat system, and is known under the name of Takjaltu. Face to face with one another stood Alagöz and Ararat. In the plain we could discern an isolated hummock, north of the Araxes and bearing about south-west. It marks the site of Armavir.

That this scene--in itself a world, and a world which fills the mind with wonder--has of necessity been the theatre of momentous events in the life of humanity, the traveller realises at a single glance. His pious predecessors were surely justified in accepting the ancient belief of the Armenians, that our first father and mother loved and suffered in this plain. [255] If we are to seek the site of Paradise within the limits of Armenia, neither the Euphrates nor the Tigris crosses a country equally appropriate to have been the earliest and fairest home of man. It looks the land of hope which Noah tilled and planted with vineyards, the second nursery of the human race. The Armenians, whose mythical history connects them closely with Babylonia and Assyria, who from the earliest times have been accustomed to receive Jewish immigrants and to see Jewish colonies established in their midst, must at a remote date have localised the events of the Biblical narrative in this the most favoured of all their valleys and at the foot of the loftiest of their mountains. [256] If the Jewish writings which they inherited were believed to have reference to their native surroundings, it was only natural that they should identify with the same districts the primeval setting of the later creations of the Jewish mind; the whole countryside became hallowed by religious tradition; nor need we feel surprise when we read that a tree in the neighbourhood of Karakala on the Araxes was believed to have sheltered Job and his three friends. [257] When the horizon narrows and embraces the particular history of the Armenians, we find that some of the first beginnings of their history are placed within this fertile and spacious plain; it was the chosen seat of Armenak, the son or grandson of their progenitor, Hayk, to which he descended from the mountains about the head waters of the Euphrates, accompanied by his whole race. Here were situated their most ancient cities, of some of which the relics still stand above ground and invite discussion of which city they denote the site. Armavir, the contemporary of Nineveh, with the grove of plane trees which worked the magic of the oaks of Dodona, has been identified with the ruins that are found on the little hillock which we distinguish from the detail of the landscape at our feet. [258] Further west, on the southern bank of the river, where it is enclosed by rocky cliffs of basaltic lava, due to the passage of a lava stream, modern travellers have discovered considerable remains of ancient masonry, which have been utilised to build the castle of Karakala, and which are still, I believe, in want of their older name. [259] Traces of the fortress of Ervandakert, and of Ervandashat, its companion city, which were built in the first century of our era by an Armenian monarch of Arsakid descent, have been remarked on either bank of the Arpa river, the ancient Akhurean, where it issues from the elevated country on the north of the Araxes and effects its confluence at the head of this plain. [260] In the days when those cities flourished, the haughty Araxes was spanned by bridges of which, here and there, a pier or a buttress still survives. [261] Below the lofty rock of Takjaltu lie the famous salt mines of Kulpi, which have been exploited from immemorial times.

After leaving the Armenian village we continued in the same direction over the barren highlands, in possession of the landscape which I have endeavoured to describe. We were riding at walking pace; our immediate surroundings were indifferent to us; nor for the space of three hours did we meet a single settlement, except here and there a group of Kurdish tents. When at midday the clouds cleared above the summit of Alagöz, we remarked that the fangs of its rocky core were invisible behind the bulging contours of the outer sheath. Above us, upon those slopes, we could discern some small green patches, which mark the site of hamlets, peopled by Tartars and Armenians who eke out a scanty subsistence on the mountain side. When we had reached a point some thirteen miles in direct distance from Edgmiatsin, we crossed a close succession of deep ravines. The first of these was the most considerable of the three, and contained the broad bed of a dry watercourse, which descends from the central mountain mass. On the further side of the last among them we came upon the remains of a large church, of great simplicity but of much beauty of form. It was built of hewn stone, in the style of the best Armenian architecture; and the ancient frescos still stained the walls of the apse. But the lofty dome had fallen in, leaving nothing but a yawning circle, with fragments of cloud crossing the blue above our heads. An inscription in the interior bears the date 876 (Armenian era), which corresponds to the year A.D. 1426. Just beyond this ruin is situated the little Armenian village of Talysh, on the southern confines of which we visited the remains of some towers which are probably of the same period as the church, and which overlook the ravine upon the west. Both the starshina and the priest of Talysh were absent from the settlement; the inhabitants professed complete ignorance of the history of their antiquities, which, since they could neither read nor write, was perhaps not feigned. The afternoon was well advanced when we left this pleasant site; a mist arose, and developed into rain. In less than two hours we were glad to find shelter in the Tartar village of Akhja Kala, a refreshing oasis of green willows on these sterile slopes.

The essential majesty of the Armenian landscapes derives enhanced value from the presence at all seasons of clouds. In this respect Armenia is more favoured than Persia, where month after month you long for a cloud to temper the glare. To the radiance of her pellucid atmosphere is added the charm of effects of vapour; but the vapour has already been tamed in the passage of the border ranges, and floats in quiet masses over the central regions of the tableland. We awoke on the following morning to a scene which is characteristic of the season and of this plain. The whole valley of the Araxes was covered by a sheet of white mist, and had the appearance of a vast sea. From invisible limits in the west to the foot of the Ararat fabric the deceptive substance followed the base of the mountains, as though we had suddenly been introduced to that geological period when the waters washed these rocky shores. In the east several islands rose above the shining surface, eminences of the plain. The high ground upon which we stood was bathed in pure sunlight, and all Nature was intensely still.

As the morning advanced the vapours lifted or were dissolved; films of white cloud were wafted across the blue. We continued our march over highlands of the same stony character as those which we had traversed during the preceding day. But beyond the village the land had been cleared in places, and wheat planted, which was showing green above the ground. It is protected by the snows which cover these slopes during winter, and it is reaped in spring or early summer. The rocky heart of Alagöz was still concealed behind the declivities which swept towards us, on our right hand. In the great plain, which still lay beneath us, we missed the stretches of pleasant verdure which in that direction had become familiar to our eyes; desert tracts, seared by gullies, had taken the place of the gardens; while further west the valley was broken into hummock waves. A ground of ochre, washed in places with rose madder--such were the colours which clothed this naked expanse; the delicate tints were continued up the sides of the mountains which border the plain upon the south. These lower slopes of the Ararat system receive the light at sunrise; and, being composed of a marly substance, which is modelled into soft convexities, display a variety of tender hues. Bold peaks, of which the summits had been strewn with snow during the night, rise along the spine of the range; but they are dwarfed, even at this distance, by the fabric of Ararat. We could discern on the west of the mountain the pass which leads to Bayazid, and we had not yet lost sight of the mound of Armavir. But it was evident that the even ground in the valley of the Araxes was coming to an end. The western limits of the level plain may be placed in the neighbourhood of Karakala; and, according to Dubois, the last canal which derives from the Araxes waters the fields on the west of the village of Shagriar. [262]

Villages became less rare as we rounded the mass of the mountain and opened a view over the country in the direction of the Arpa Chai. An hour from Akhja Kala our attention was attracted by a still distant eminence, rising above the shelving land upon that side. It was the crag of Bugutu, which is probably due to a later eruption on the flank of Alagöz. We passed two Tartar settlements, and crossed a couple of ravines, the first of which must have had a depth of nearly a hundred feet. It contained a pleasant growth of lofty poplars and other trees, and it was threaded by a babbling brook. When the prospect extended to the upper slopes of the mountain, we observed that they were sprinkled with fresh snow. A stage of two and a half hours brought us to the village of Talin, a prosperous and picturesque little township at the foot of Bugutu (Fig. 61).

Both the Pristav and the priest were quickly forthcoming; we were by them conducted to a house which contained two storeys, and which was the residence of the priest. While food was being prepared, we were accompanied by our hosts in a walk round the place. We were informed that it contained some thousand inhabitants, all of whom were Armenians. It possesses a church, but is still without a school. The old prejudices survive, and it was impossible to persuade the young women to submit to the camera. But Talin is distinguished by the close proximity of a piece of architecture which appears to date from the golden period of the Bagratid dynasty and which ranks among the most charming examples of the Armenian style. It is a church--they call it cloister (vank), and it perhaps belonged to a monastery--which, although in ruins, is fairly well preserved. The roof has fallen in; the walls display wide breaches; but the masonry is still sharp and fresh, as when first put together, and the traceries might just have undergone the finishing touch. With its bold windows--no mere apertures--and bands of elegant sculpture, I thought it the most beautiful building I had yet seen in Armenia. I reproduce some of these chiselled mouldings of the exterior. The first, a vine pattern (Fig. 62), belongs to the southern transept; and the second (Fig. 63), representing a pear or apple, is taken from that upon the north. On the south side of the ruin we observed a sun-dial, carved in stone; and we were shown a square block, which had been found among the débris, and upon which was sculptured a relief, representing the Virgin and Child, attended by two angels. A graveyard surrounds the building; some of the old crosses have been built into the walls of the village church. A little on the east we noticed the remains of a small chapel. The ground was strewn with fallen stones, some red, others grey--the two colours which are so skilfully blended or placed in contrast by Armenian architects upon the broad, undecorated spaces of their walls. We enquired the history of the ruin, and were referred to a partially defaced inscription on one of the piers which once supported the dome. It mentions the name of King Sembat, a member of the Bagratid dynasty, which reigned from the ninth to the eleventh century. [263] The grandfather of the priest informed us that both the monastery and the church had been maintained up to a comparatively recent period. He said that the priests had fled during the campaign of Paskevich, since which date the buildings had been allowed to fall into decay.

Ker Porter, who crossed the district on his way from Ani to Edgmiatsin, mentions the existence in this neighbourhood of extensive ruins--the deserted relics of two churches, of walls and houses, which he saw at a distance, but did not stay to examine. He calls the place Talys, and Ritter hazards the conjecture that these may have been the remains of Bagaran. [264] That city, which was founded by the same monarch who gave his name to Ervandakert and Ervandashat, became a royal residence of the Bagratid dynasty, and at the end of the fourteenth century of our era still continued to exist. We did not hear of further antiquities in the vicinity of Talin; but the correspondence of name suggests that Ker Porter's account may have been called forth by the former condition of the site which we visited. It was evident that these highlands had been the seat of a flourishing civilisation, later in date than that which produced the vanished cities of the plain. First at Talysh and next at Talin we discovered traces of this mediæval culture, of which the evidence was lavished upon us when we had reached the banks of the Arpa, at Ani and at Khosha Vank.

The upper chamber of the priest's house and the company therein assembled recalled the simplicity of the early Christian times. Our host was still a young man, and his natural capacities had not been blunted by indigence and ill-treatment. His villagers were well off, and appeared to live on terms of friendship with their neighbours of Tartar race. A Tartar khan, a grandee of the district, happened to be visiting the place on business (Fig. 64); and we were glad to see that his intercourse with the principal people was marked by tokens of mutual respect. His grave face and dignified figure contrasted with the vivacity of the Armenians; his presence added to the interest of the group which I photographed, and which included the Pristav (Fig. 65) and the priest (Fig. 66). Neither the official head of the village nor our clerical acquaintance possessed any education, except what had been provided by an Armenian primary school. But both, and especially the former, were men of great intelligence, and did honour to the peasant class from which they had sprung.

We were in want of another pony, which we were able to hire at Talin; his owner, a Tartar belonging to Akhja Kala, accompanied or followed us on foot (Fig. 67). Measured on the map, it is a distance of sixteen miles from the village to the point at which we struck the Arpa Chai. We owed it to the nature of the ground and to the sorry condition of our horses that we were four and a half hours in performing the stage. It seemed an interminable ride; the landscape was monotonous; and we soon lost any glimpse of the valley of the Araxes, as we continued our north-westerly course. We crossed the neck of the ridge which culminates at its western extremity in the crag of Bugutu; and, on its further side, descended into the little Tartar settlement of Birmalek, where a stream trickles down from Alagöz. A dam had been constructed which, aided by the nature of the ground, had forced the waters to collect into a small lake. Beyond Birmalek a second ridge was placed athwart our way, and constrained us to deviate towards the west. In the hollow we passed a small settlement of Kurds, called Sapunji, of which the inhabitants were the wildest people we had yet met. It speaks well for the Russian officials that they did not dare to lay hands upon us, travelling, as we were, alone and unarmed. This second ridge was succeeded by another, similar in character, which was followed by several more. They are the outworks or spurs of the central mass of the mountain, from which they radiate outwards in a westerly direction towards the trough of the Arpa Chai. Although their relative elevation above the valleys is not considerable, our guide preferred to turn them than to take them in face. Their sides were clothed with burnt grass, or were sterile and strewn with stones, like the depressions which they confined. For more than two hours we continued among such dreary surroundings, crossing the western basal slopes of Alagöz. These decline, by an almost imperceptible transition, into a tract of open and undulating ground. We were refreshed by the sight of a village, which stood alone and without neighbours on the bare surface of the more even land.

It belonged to a colony of Armenians from the plain of Alexandropol. Let us hope that they will be followed by further migrations of their countrymen into the valley of the Arpa Chai. That classical river of their ancestors crosses a region which was long famous for its salubrious climate and productive soil. It has not yet recovered from the state of abject desolation to which it was reduced when it formed the borderland between the Turkish and Persian empires. During a ride of nearly two hours from this settlement to the bank of the river, we were not aware of any sign of the presence of man.

Yet the features of this more level zone reminded us of the plain of Alexandropol, of which in some sense it forms an outlying part. We stood in face of the western declivities of Alagöz, with the rocky core of the volcano again disclosed. The contours of the mountain were composed of a number of ridges, which in perspective appeared to belong to two principal groups. One group declined away into invisible limits on our left hand; the other into an uncertain distance on our right. We were placed in the fork between these two diverging branches. It was evident that the last group separated us from the valley of the Araxes; nor could we doubt that the principal and humble ridge in the reverse direction was the only barrier between us and the plains on the north (Fig. 68).

In the west, to the far horizon stretched the loamy tracts about us, bare of surface, like the sea. Above the outline of this high land rose the peaks of the Ararat system, fretting the sky from south-west to a bold mountain in the south, which we recognised as the familiar Takjaltu. We knew that we were overlooking the trough of the Arpa; but the river was hidden from sight. The light was failing as we entered the Armenian village of Khosha Vank, on the left bank of the stream.

It is a picturesque little settlement of some 120 tenements, grouped around a stately church. I have referred to it under the name which I received from the priest and the Pristav, but which more properly belongs to the neighbouring monastery. It is called Kizilkilisa (red church) on the Russian maps. It was our intention to sleep in Ani, after fording the river at this village; and we were surprised to learn that the ruins were four hours distant, and that it would be almost impossible to reach them that night. Since the baggage was behind us, we listened to the counsel of our informants, who conducted us to a stone house, containing a single room--the only decent building in the whole place. Although without a school, the inhabitants are no dullards; they seemed extremely ready to make a little money, and pleased to be able to exchange ideas. In fact we discovered on the following day that they had deceived us about Ani, with the express purpose of retaining us for the night. We waited some time in vain for the luggage to overtake us, and then composed ourselves to sleep.

When morning came our effects had not yet arrived; we reflected that we had given the rendezvous at Ani, and, although we felt sure that the laggards would cross the river at our village, decided to push on. The Arpa flows between high banks, a deeply eroded and sinuous bed, hidden by precipitous cliffs of black rock. You form the conception of a trough or fissure in the surface of the tableland, which undulates away into the distance on every side. After fording the stream, we proceeded along the right bank, and, at no great distance, opened out a romantic valley on our left hand, similar in character to that which adjoins the site of the Armenian village. In both places the river describes a complete S, and is lost in the gloom of overhanging walls. The disposition of these rocky sides assumes the appearance of a glen, in which are situated the remains of an extensive monastery, bearing the name of Khosha Vank. Just beyond this standpoint we gained the high land above the river; and there before us, on the plain, lay the ruins which we had been seeking, at the distance of an hour's canter from the cloister, or of a couple of hours' ride from Kizilkilisa.

Descrying horses in the direction of Ani, we galloped forward and overtook them; they proved to be our missing cavalcade. They had passed the river at a place lower down than where we had crossed it, and were pursuing their way in a most leisurely manner. After opening one of the cases in order to replenish the slides of the camera, we returned to the glen, and again forded the stream. We spent a considerable time at the cloister and in its neighbourhood; it was certainly the most remarkable building which we had yet seen. Reserving a description of its ancient church and halls of audience, I shall only refer to a couple of illustrations in this place. The one (Fig. 93, p. 386) shows the ensemble of the monastery; but, having been taken from the east, where the ground is open and the landscape tame, misses the peculiar characteristics of the site. The other (Fig. 94, p. 387) may convey some conception of the appearance of the glen, when seen from the river-bed below the cloister. From the flat and water-worn bottom rises a little tongue of higher land, upon which stand the remains of two little chapels. On the cliff above the ravine you see the pier of a ruined gateway, outlined against the sky. The track to Ani leads up the cliffside and passes that ruin, which stands on the plain in which the still-distant city lies.

It was late afternoon when we reached the walls of the ancient capital (Fig. 70, p. 369), and passed within the great gateway. No massive doors creaked upon their hinges; we rode through empty archways into a deserted town. From among the débris of the public and private buildings rose the well-preserved remains of a number of handsome edifices--here an elegant church, there a polygonal chapel. An old priest with a few attendants were the sole inhabitants--they and the owls. We had only to follow the track to be brought to the humble tenement in which the priest lived. He stepped forth to meet us, a grey head, a feeble figure; he walked with difficulty, and with the demeanour of a man who is awaiting death. He told us that he had dwelt here since 1880, the only custodian of these priceless architectural treasures, and the only exponent of the topography of the site. He had been attacked in his house by a band of Kurds in 1886; they had inflicted knife wounds, and stripped him of everything he possessed. We remained two whole days within the walls of Ani, examining the creations of a vanished civilisation, and collecting material with which I propose to deal in a separate chapter. At nine o'clock on the morning of the 19th of October we took leave of our aged host; and, leaving the city by the same gate through which we had entered it, pursued a track which leads in the direction of Kars.

Clouds were clinging to the hill slopes upon our point of course and concealing the shield-shaped mass of Alagöz. Lost fragments of opaline vapour lay on the surface of the grassy plain. Here and there we perceived the ruins of little chapels and other buildings, or the scattered débris of masonry. From these suburbs we looked back upon the bold line of the city walls, with their double girdle and towers at regular intervals. It seemed as though the stream of life had wandered off into other channels, leaving behind this eloquent evidence of its former course. We could not descry the form of man or of animal in the landscape; even the sky was without a wing. We rode in silence and at ease along a beaten path, where the burnt herbage had been worn away from the rich brown soil. West of Ani, at a distance which leaves the site of the city open, rises a hill of irregular shape and moderate elevation, known as the Alaja Dagh. It is due to volcanic action, and covers a respectable area; its sides and summits are overgrown with grass. It is placed across the direct line between Kars and the ancient capital, and compels you to deviate a little to the north. As we rose along the north-eastern slopes of the mass, we were lifted at a convenient altitude above the plains.

Outspread before us lay a vast extent of undulating ground, on the south, on the east, towards the north. After we had passed the small Armenian village of Jala, we could just discern in the lap of the expanse the city of Alexandropol, at a distance of over twenty miles. We had again opened out the northern slopes of Alagöz; and we could even see the meridional range which intersects it upon the east, and the gap through which we had journeyed to Erivan. When one reflects upon the significance of this panorama, it must be recognised that our standpoint on the skirts of the Alaja deserves a high rank among those apposite and commanding positions which Armenia appears to lavish upon her admirers, and which imprint her features indelibly upon the mind.

We might be said to have been standing on the dividing line between two landscapes and even of two climates. On the north lay the immense plains around Kars and Alexandropol, vague and grey in spite of the clear atmosphere, and with their distant limits shrouded in haze. These pass over, along the course of the deeply-bedded Arpa, into the ever-widening valley of the Araxes, bathed at all seasons in sun. Had it not been for the projecting spurs of the hill which we were skirting, the prospect would have embraced the peaks of the Ararat system, bounding the expanse upon the south. Snow had fallen upon the upper slopes of the mountains--Alagöz, no longer a shield but a towering parapet; the Chaldir system, the border range in the far east.

As we proceeded towards the west, the instructive lesson was developed--no ridge to cross, but continuous tracts of level land. The plain rises with gentle gradation from the right bank of the Arpa to the labyrinth of hills on the west of Kars. Its surface is slightly vaulted, and the configuration of the ground is such that you lose the outlook towards the east. We passed through Subotan, a prosperous village of Turks and Greeks. The gay dresses of the Greek girls formed a brilliant patch of colour, and their trinkets sparkled in the sun, which was already high (Fig. 69). Education is provided in a little schoolhouse, built and maintained at the charges of the Christian inhabitants, but supplied with a teacher by the State. A little further on we entered a second and smaller settlement, and again found ourselves among Greeks. I am under the impression that these scattered colonies date from the campaigns of Paskevich, when Christians in considerable numbers accompanied his armies across the frontier after their evacuation of Turkish territory.

On and on we rode over the spacious plain, beating the brown and idle soil, with nothing to divert us from the simple pleasure of cantering along. Vague tracks came converging towards us from the distance, the arteries along which the supplies of the fortress flow. It was evident that there was a pronounced slope of the ground upwards; and, at length, on the western horizon we opened out a long, low ridge, against which we could just discern without the aid of glasses the yellow masonry of the castle of Kars (Fig. 98, p. 406). As we neared the site, we were impressed by its strange and romantic character. From the hills upon the west a mass of gloomy basalt projects towards the east into the level and loamy land. Concave towards the plain, to which it presents a line of cliffs, it forms an extensive bay and terminates on the east in a commanding promontory, called the Karadagh. The answering horn of this sinuous line is composed or accentuated by the cluster of modern buildings which the Russians have erected, and which jut out from the ancient city on the side of the cliff into the even ground. Their white faces and iron roofing, coloured a quiet red or green, present a contrast to the black masonry which mounts the slope behind them--groups of houses, a few minarets, a large church. Above these towers the well-preserved pile of the old castle--an object which is rendered the more conspicuous by the yellow stone of which it is composed. Further eastwards along the summit of the ridge you see the ruins of the old Armenian fortress, with the remains of a wall rising towards it from the foot of the cliff. In the bay itself you will always find a confused medley of sheep and cattle, of bullock-carts threading the piles of hay and stores. We were met and challenged by a gendarme upon our arrival, but were allowed to proceed to a modest inn.

I am conscious of having hazarded to tire my reader with the continuous narrative of a journey of four days' duration and of more than the usual variety of interest. Anxious to avoid diverting his attention from the features of the country, I have not suffered him to rest, as we rested, at Ani; but have taken him without a break from the sunny depressions at the foot of Ararat to the wintry highlands about Kars. He has almost traversed from east to west one of the central regions of Armenia; and I would ask him to reflect that he has not crossed a single mountain barrier, but has throughout been riding upon the margin or over the surface of immense plains. In so far as it may be possible to parcel out this level surface, a triple division is suggested to the mind. In the north the basin-like area of the plain of Alexandropol (5000 feet) declines along the banks of the Arpa Chai; on the western side of the river the ground again rises and develops into the spacious plain of Kars (5700 feet). In the south lies the sheltered valley of the Araxes, commencing on the west with an elevation, in the neighbourhood of the confluence of the Arpa, which is rather less than 3000 feet above the sea.