I had resided for nearly two years at Erzerum, in Armenia, when I quitted that city on my way to Persia - not taking, however, the direct road, but one which should lead me into it by Kars, Gumri, Anni, and Kurghesman. The following is an itinerary of that part of my journey only.

20th May - We quitted the village at about seven in the morning, our road leading us southward along the banks of the Arpa Chai. On the opposite side of the river, akabas traversing the plain in all directions, and labourers busy at the plough, presented a cheerful sight: numerous villages scattered about the plain were visible in the distance. The Russians have guards stationed at certain distances along the bank of the stream, for the double purpose of preventing the importation of contraband articles and of hindering their soldiery from deserting. Both objects, however, are but imperfectly attained. The contraband traffic is carried on with the assistance of the villagers whose habitations lie near to the stream, which is crossed during the night, and the goods transported by bye-paths from village to village to the place of disposal. I was told, however, that the smugglers have of late been much more closely watched, and that they have been obliged in a great measure to suspend their operations.

Our road led us by the Persian villages Karaklissia, and Aralik, which latter has a tower; and, fording the Kara Khan river twice (the same which we yesterday passed at Kizil Charchak), came to another village named Bash Surijeh, which has also a tower, a large and handsome building in the style of a church, which it probably was at some distant period. Here our road turned off to the west; we ascended to higher ground and proceeded south-west. At the third hour we forded the Kars river, and immediately passed the village Arghuzun. A little south of this the Kars river forms a junction with the Arpa Chai (the latter having already been augmented by the Kara Khan), and the three united streams then flow towards the Arras. Our road from Arghuzun led in a general direction south past the village Illanli which is inhabited by Persians, near another village, and soon after to the banks of the Arpa Chai, after its junction with the Kars river.

Continuing in the same direction, we reached the plain of Anni, at the commencement of which we passed on our left a church with a village near it: the plain is of no great extent, and is an arid, stony, and almost desert tract. Thus far we had traversed since the morning a fine grazing country, which is cultivated only in the vicinity of the villages; few flocks or herds were to be seen, and the country seemed deserted except on the immediate line of our road. We crossed the plain in a south-westerly direction, and reached Anni at about half-past 1 p.m.

Outside this ruined and deserted town we found a few inhabited houses, in one of which we took up our quarters; and I then proceeded to visit the ruins. They are situated on the south-west corner of the plain on a plateau, which is bounded on all sides but one by a deep ravine. The Arpa Chai flows at the foot of it but on one side only, and continues its course south-west; a bridge of ancient construction affords a communication with the opposite hank of the stream. The sides of the ravine exhibited a deep stratum of very soft sandstone, hollowed into innumerable holes and caves, probably in most cases for sepulture; many of them, however, were shaped like apartments, and may have served as habitations for the living: at present the wandering Kurds sometimes make these caves their abodes during winter.

Ascending to the town from the ravine, I was struck with the solid and beautiful masonry of what remains of the walls and palace - the former are double on the side unprotected by the ravine, the outer one bearing evident marks of the strife that must once have raged beneath them. They are marked with innumerable small indentations, produced very probably by the points of arrows or other sharp weapons with which the defenders of the walls were assailed. The remains of the palace are extensive, and several of the apartments are entire, with the exception of the roofs. The quantity of materials which has been carried away for buildings in various parts of the country is probably the chief cause, more than anything else, of the ruined state of this building and that of the city walls, both which, as I have before observed, being of remarkably solid construction, seem capable of resisting the wear of time for some centuries to come if undisturbed by earthquake and the destroying hand of the Turk. In many places, however, the walls have been actually undermined to get at the materials. In some places I observed great masses of stone supported merely by the adhesion of the mortar, which would prove the latter to have been of a very superior quality.

There are many long Armenian inscriptions on the walls in excellent preservation; I also observed on them the figures of a tiger, a horse, and of a man holding something globular in one hand. The area within the walls is covered with the prostrate remains of the houses, the forms of some of which may still be traced among the ruins, as may the outlines of some of the streets. Several spots unencumbered by ruins would probably mark the former positions of public squares.

Four or five churches, still in tolerable repair and of neat construction, remain as proofs of the taste of their Armenian founders; one of these, indeed, may justly be termed an elegant and handsome building: all have long inscriptions in Armenian, which it would require some days to copy; and the interior of the walls are painted like other Christian churches in Turkey with the figures of saints and other subjects from sacred history. Besides these edifices are two towers resembling Turkish minarets; one of them has an Armenian inscription with a few Turkish characters above it, which may perhaps be symbolical of the vassalage under which the government of this petty Armenian kingdom was held. On the highest part of the plateau overlooking the ravine was situated the citadel, now apparently little more than a mass of ruins, though I did not ascend to it, owing to my indisposition at that time.

Anni, the capital of the Pakiadian kings, was captured with great slaughter of the inhabitants by Alp Aslan, one of the Seljukian princes, who succeeded Togrul Beg, the Turkish or Tatar leader in 1063, and subsequently filled the Persian throne. Kaghig, the king of Anni, was banished to Caesarea, and the town after its capture was given by the conqueror to a tribe of Kurds, who in their turn yielded it, after a violent struggle, to one of the kings of Georgia. This place is situated at 9 hours from Gumri, and nearly the same distance by the direct road from Kars.

21st May - We took our departure this morning at about eight o clock, our road leading us for a short space in a direction west, and presently south-west. Soon afterwards we proceeded south over a flat country, in taking which direction Ararat bore south-southeast. of us. Passing on our right a village called Cosujah, situated at the foot of low hills, we presently altered our direction to the south-west by south, and ascended to higher ground, where we found a small encampment of Kurds. During this much of our journey today I have observed no appearance of cultivation except near Cosujah, but the country affords fine grazing. Our route altered to the south-west, and we traversed some flat ground, passing a village named Alem, which has a church: the inhabitants are Armenians and Persians. Ararat bore S. by E. of this village, and S. and by W. is an elevated peak, with Keorogli Kalissi, a castle, now in ruins on its summit. To our left, part of the plain of Erivan was visible - still no symptoms of cultivation except immediately around the villages. We passed those of Digiole and Zebni towards the end of the fourth hour, the latter having a church in it. From hence the direction of our route was S. by W., which took us across some hills, and we descended into a valley to a large Armenian village called Zechchi, possessing also a church of neat construction but falling to decay.

Our road from this led us south-west, through a well-cultivated country and near two or three villages, but we were overtaken with such a fierce storm of rain, wind and thunder, that whilst it lasted, and that was for a considerable time, it was impossible to take any notes; it ceased as we approached the Kurghesman valley, into which we descended on foot by a long and difficult path alone its side until we reached at about 6 p.m. the pretty Armenian village of Kess, situated in a recess of the valley amidst gardens and groves of fruit-trees, chiefly the apple, pear, mulberry, and walnut. I wished to push on to Kurghesman that evening, but was dissuaded from making the attempt by the wretched condition of our cattle (the same which had brought us from Kars), and by the peasants assuring us that it would be impossible to get across the Arras during the night, as the ferry by which it is passed is not continued after sunset, and the river was too much swollen to admit of our fording it.

On the opposite side of the valley we could observe a large encampment of the Jennukki Kurds, amounting, I was told, to about 300 tents. These people frequently pass the river, and plunder the unfortunate villagers of whatever comes in their way; and so reckless are they of the injury they may be doing these poor people that they not infrequently, I was assured, in the wantonness of their mischief, destroy the vines and trees, on the produce of which the villagers mainly depend for many of their little comforts.

22nd May - We left Kess at about 7 am. and continued our road up the valley, which runs nearly east-west. Its general feature is a bold and rugged outline; its length, breadth and depth, are very considerable; but it is not otherwise remarkable, being devoid of trees and cultivation, except in some few parts. Mountains of great height form its southern side, in the recesses of which I was told the oak is found. We gradually approached the Arras, on the opposite bank of which Kurghesman is situated; the ferry being higher up the stream, we passed Kurghesman, and on our way met the mutzellim of that place, who appointed one of his attendants to accompany me to it and provide us with horses. In about half an hour we reached the ferry, which is effected by means of a raft of wicker-work supported on inflated skins - it is so unwieldy an affair that to track it up against the current to the proper place for a start is a work of some difficulty and time. We, however, got across on it very well, the current bearing us along with great rapidity till we reached the opposite bank. A man with a long wooden shovel, which he used both as an oar and rudder, sometimes on one and sometimes on the other side of the raft, steered us with much dexterity across the stream. Our horses were then driven over, the water occasionally reaching above their shoulders. At this part I should think the river must be about 80 yards in breadth. Having replaced our baggage on the cattle we turned back to Kurghesman, which proved to be a large cassaba, containing about 200 families of Turks and Christians, and is embosomed in gardens and groves of fruit-trees; the pear, apple, plum, cherry and walnut trees, appeared to be in great abundance; and it is from hence that the surrounding country is supplied with these productions. I was unable to procure horses to proceed with at once, and was detained here to the following day.