Col. William Leake, Travels in Northern Greece (v. 1-4)

§ 3.234 At the entrance some remains of a wall constructed of mortar and small stones, are seen on the slope of either hill; the pass, however, of which these works formed the defence, although remarkable, is not very important, as the passage over the hills on either side is easy, particularly to the north. Towards the middle of the pass, on a small rock by the side of the paved road, the word ΟΛΠΑΙ is engraved in large letters on the rock. Olpae may perhaps have been the name of the pass, derived from ελπίς, Aeolice όλπίς, in allusion to the expectation which the traveller feels of being quickly gratified by a view of the maritime plain and sea, and by the speedy termination of his journey. At the issue of the glen stand Khaivit on the right and Laina on the left. The latter is very small, but Khaivat contains a large church and 300 cottages, inhabited by Bulgarian Christians, a people which occupies, with the exception of two or three large Greek villages, all the great maritime plain of Lower Macedonia. Few of the women in the Bulgarian villages can epeak Greek. The houses of Khaivat, like those of the Bulgarians in general, are neat and comfortable, with plastered walls and floors, covered with a yellow wash which borders also the outside of the door. Our baggage, which I quitted to visit the baths, arrives at Khaivat at 11.40.

§ 3.235 Nov. 15.—The late χειμώνας (so the Greeks call a day or two of stormy weather) has covered the mountains to the north and west with snow, end this morning a. strong gale from that direction brings frost with it. At a well and large plane tree, a little below the village, lies a marble inscribed with characters of a good time, but containing only names. In half an hour we descend into the plain of Saloniki, and winding to the left along the foot of the range of Khortiatzi, enter at the end of another hour the Turkish cemetery which surrounds the city, and which contains many fragments of columns and sori dispersed among the tomb-stones. The city walls towards their foundations, are in part composed of ancient marbles, and there is every appearance of their having followed the ancient line. At the end of an hour and three quarters from Khaivat, we enter the Vardar-kapesi, or gate of the Vardhari. In a tree before it hangs the body of a robber. Just within the gate the street is crossed by an ancient arch about 14 feet wide, supported by pilasters, which are buried apparently to half their original height. Below the capital of each pilaster, on the western side, a Roman togatus is represented in relief, standing before a horse. The frize above the arch is decorated with the caput bovis united by festoons. The whole construction consists of large masses of stone, but the monument could

§ 3.236 never have been very magnificent, and appears hardly worthy of the time of Antony and Octavius, to which it is attributed by Beaujour, who supposes it to have been a triumphal memorial of the victory of Philippi. Nor does an inscription below the arch which contains the names of the eight archons in whose magistracy the monument was erected seem to favour his opinion, as the names are chiefly Roman, which they would hardly have been at so early a period. They are styled Politarches, as when St. Paul visited Thessalonica, 93 years after the battle of Philippi. Two of these magistrates were the gymnasiarch and the tamias.

§ 3.237 Nov. 17.—In the evening (being the proper time during the Ramazan) I visit Musa (Moses) Pasha. This is the same gentleman whom I saw in exile at Epakto, cooking his pilaf with oil for want of butter, and stealing our consul’s wood. Since that time he has been in Egypt, whither he was sent to supersede Mehmet Aly, who was ordered by the Porte, on the plea of his being a Macedonian, to exchange the government of Egypt for that of Saloniki. Mehmet Aly, however, was not to be displaced so easily. Musa Pasha had chiefly founded his hopes of success on the dehlis in Mehmet’s guard, the chiefs of whom were his friends and formerlv in his service, and attributes his failure to the Kapitan Pasha, whom he accuses of having been bribed by Mehmet Aly to delay a march to Cairo, which had been concerted with Elfi and four other Mamluk beys, until it was rendered impracticable by the rising of the Nile. Musa’s troops had a skirmish with Mehmet Aly’s, but without any advantage on either side. The Porte, convinced that their project had failed, ordered Musa to assume the government of Saloniki, and the Kapitan Pasha to return to Constantinople with his fleet. Musa came with the fleet as far as Cos. He affirms that Mehmet’s forces amount only to 4000 Albanians and 6000 others, that he is detested for his oppressions, and for having ruined commerce, and that no Red Sea goods can pass the desert, as the merchants are afraid of being plundered by the Pasha at Cairo. On the event of the battle of Austerlitz, the Turkish government assumed a certain degree of insolence, and supported by the French, immediately set about attempting two objects upon which they had long fixed their wishes, though until that moment without much prospect of attaining them:

§ 3.238 1. The extending of the Nizai-djedid, ite imposte and military discipline over Rumili; 2. The withdrawing from all rayas the protections of the European courts, and particularly the Russian flags from the Greek ships. To effect the former of these objects a very large force was raised in Asia, and sent into Rumili, and for the latter a firmahn had already been issued last March. The Janissaries of Constantinople, however, and particularly all the Turks of the country extending from Adrianople to the capital, having united against the Nizami, the Asiatics were entirely defeated and dispersed before they got beyond Selivria, where the remainder were surtounded and in danger of being cut off, while their adversaries threatened to march to Constantinople and depose the Sultan as a ghiaour. The project of the Sultan was immediately renounced and the Turkish ministry changed. Salonica, as the Italians and English name this city, is by the Turks called Selanik, by the Greeks Σαλονίκη, and by all the educated among them Θεσσαλονίκη. Being situated in great part upon the declivity of a hill rising from the extremity of that noble basin at the head of the Thermaic gulf, which is included within the Capes Vardar and Karaburnu, and being surrounded by lofty whitened walls, of which the whole extent, as well as that of the city itself, is displayed-to view from the sea, it presents a most imposing appearance in approaching on that side. The form of the city approaches to a half circle, of which the diameter is described by a lofty wall, flanked with towers,

§ 3.239 extending a mile in length along the sea shore, and defended by three great towers, one at each extremity, the third overlooking the skala or landing place, where stands a small suburb, between the tower and the sea shore. Since the invention of gunpowder, batteries on a level with the water have been added to the maritime defences in the most important points, and a fortress, or fortified inclosure, has been constructed at the western angle of the city. The eastern and western walls follow the edges of the height, where it falls on either side towards a small valley watered by a rivulet, and terminate above in the walls of the citadel, which has a double inclosure towards the town flanked with square towers. The heads of the valleys on the east and west are separated only by a ridge connecting the citadel with the falls of Mount Khortiatzi, which command it at a short range. The citadel, like that of Constantinople, is called Ένταπΰργιον, which the Turks have translated into Yeddi Kulelar, the Seven Towers; for doubtless at both places the name is older than the Turkish conquest. Saloniki bears the usual characteristics of a Turkish town; no attention is paid to cleanliness or convenience in the streets, the exterior of the houses is designed to conceal all indications of wealth, nor can any correct opinion be formed of the population from the central part of the town, or a visit to the bazar, where crowds are collected during the greater part of the day, while the rest of the city is a solitude. The houses in the lower part of the

§ 3.240 town are shut out from all external view by the narrow streets and the high town walls, but in rising higher, a noble prospect opens of the grand outlines of Olympus, Ossa, and Pelium, seen above the promontory of Karaburuu, together with a part of the Chalcidic peninsula to the southward, and to the westward the immense level which extends for 50 miles to Verria and Vodhena. All the principal mosques were formerly Greek churches, and two of them were Pagan temples, which had been converted into churches. The most remarkable is that which is still known to the Greeks by the name of ιταλεα MqrpowoXtc, or more vulgarly Eski Metropoli, an appellation employed also by the Turks. Hence it seems to have been, in the time of the Byzantine Empire, the cathedral church of the metropolitan bishop. It is a rotunda built of Roman bricks, with two doors, one to the south, the other to the west. The thickness of the walls below is 18 feet, their height about 50 feet, the diameter within, 80 feet: above these walls was a superstructure of slighter dimensions, the greater part of which, as well as the dome which crowns it, may perhaps have been added when the building was converted to the service of Christianity. It is lighted by windows in the middle height of the building, which in all is about 80 feet.

§ 3.241 Possibly these windows also are a Christian repair, the ancient temple having perhaps been lighted from the dome. The inside of the dome is adorned with the representation of buildings and saints, in mosaic, interspersed with inscriptions which, as usual in Greek churches, explained the subjects, but are now too much injured to be decypherable, though the Turks have not destroyed any of these ornaments, nor even a figure of the Almighty which occupied a niche opposite to the door where once stood the Pagan idol. In one place they have supplied a fallen mosaic with a painting in imitation of it. Eski Djuma, or Old Friday, is the name of another mosque, the masonry and form of a great part of which shows that it was once a building of the same age as the Eski Mitropoli, or perhaps still older, but such have been the repairs and alterations which it has undergone in its conversion first into a church and then a mosque, that the ancient plan cannot easily he traced. It is supposed by the learned to have been a temple of Venue. Ai Sofia is a mosque, so called by the Turks, and which like the celebrated temple at Constantinople, was formerly a church dedicated to the Divine Wisdom. The Greeks assert it to have been built by the architect of St. Sophia, of Constantinople: its form at least is similar, being that of a Greek cross with an octastyle portico before the door, and a dome in the centre, which is lined with mosaic, representing various objects much defaced; among these I can distinguish saints and palm trees. The Turks, contrary to their usual custom of destroying, or at least of hiding with a coat of plaster, the

§ 3.242 figures in the Greek churches which they have converted into mosques, have allowed all the figures of St. Sophia to remain, with the exception of a piece in the centre, which they have replaced by an Arabic inscription, having been justlyshocked, perhaps, by a huge human face, looking down, as I have frequently seen in Greek churches, and which is generally inscribed with the word Παντοκράτνρ. St. Demetrius is a long church with a triple aisle, supported by a double order of columns of several kinds of variegated marble, and very much resembling an old Latin church, such as are seen in Italy, Sicily and the Holy Land. It may possibly have been built by the Latins when in possession of Thessalonica in the 13th century. Within this temple a sepulchral marble is inserted in the wall, which very much resembles many similar monuments in Christendom, being in that common form which represents the end of a sorus crowned with a pediment It is ornamented with flowers well executed, within which is an inscription in twenty-two Greek Iambic verses, in honour of one Luke Spanduni, who is described as a scion of Byzantium and the Hellenes, and who died in the year 6989, or A.D. 1481, whence it would seem that the Turks did not deprive the Greeks of their church of St. Demetrius immediately after the conquest. As the verses on this monument are rather creditable to the learning of that time, and have been published only by Paul Lucas, who, among other inaccuracies, has omitted two lines, I subjoin a copy of them.

§ 3.243 The modern poet, to make his Hellenic attempt the more complete, has imitated the ancient character, and avoided any division of the words. The word οια shows that it is a woman who grieves for the loss of Spanduni. Among the ecclesiastical antiquities, in which Saloniki exceeds any place in Greece, as the churches just mentioned show, are two of the most ancient pulpits in existence; they are single blocks of variegated marble, with small steps cut in them. One of these βήματα, as they are still called by the Greeks, is in the mosque of Eski Mitropoli: the other is lying in the yard of a church of St. Minas, which is still appropriated to the Greek worship. Και τήν πατρίδα αποβεβληκώς, οίμοι, Tης βαρβαρικής ου μετέσχες κηλίδος...

§ 3.244 Among the remains of Pagan times, may be mentioned some small portions of the walls, which there is every reason to believe, follow the line and foundations of the inclosure of Cassander, and which being in their general structure much higher and more solid than such as the Ottomans build, seem to consist for the most part of successive repairs of the Macedonian work, before the Turkish conquest. Therme we can hardly suppose to have been so large as Thessalonica, and as it could not have left the citadel unoccupied, probably did not extend as far as the sea. That the main street, add two principal gates, and consequently the whole inclosure, of the Roman Thessalonica, corresponded with those of the modern town, we have an infallible proof, in two ancient arches which still cross that street; one already mentioned near the Vardar gate, the other not far from the corresponding gate at the eastern end of the same street. The latter, which had two smaller lateral arches annexed to it, now destroyed, consists of two piers 14 feet square, faced with stone, which were covered on all sides with a double range of figures in low relief, representing the sieges, battles, and triumphs of a Roman Emperor. A great part of the piers are concealed by shops of the bazar, which cover all the lower parts of the figures on one side, and the whole of them on the other. Entering a bakehouse in the latter situation, I found the sculpture still more defaced than in other parts, but in none is it in good preservation, and the whole appears to have been of a very declining period of art.

§ 3.245 The arch which rests upon the piers is still more deprived of its facing, and is now a mere mass of Roman tile and mortar. Zosimus seems to give some support to the tradition which attributes this monument to Constantine, by his remarking, that when Constantine had subdued the Sarmatians, he went to Thessalonica, and there constructed a port. But the execution of the sculpture is perhaps better suited to the age of Theodosius, whose victories over the Goths were a common subject on the monuments of his age. To the westward of this arch, near the main street, are the ruins of a portico with a double order of architecture, consisting of four Corinthian columns, not of the best design or execution, and the shafts of which are now half buried in the ground. On their architrave stands an upper order, consisting of four plain pilasters, on the opposite faces of which are Caryatides, eight in all: the figures are of the human size, or near it, and each of them represents a different subject. On one of the pilasters the two opposite figures are Leda and Ganymede; the former embraces the swan, whose head reposes upon her breast: Ganymede is held by the eagle, whose wings are spread over his back, and whose talons rest on his hips, while the head of the eagle reaches over the left shoulder of the youth, looking in his face. This is a very good piece of sculpture, and not much injured by lime. The other figures seem inferior in merit as they are in preservation; nor can the subjects be easily understood.

§ 3.246 The next to Ganymede, on the same side, is a man with a Phrygian bonnet, at whose feet is a bull’s head; the third and fourth are females in light drapery, the latter with wings. On the opposite side, or that of the Leda, the figures are so much ruined that Ι cannot distinguish the subjects. This monument is in the house of a Jew, and is known in the Spanish dialect of the Jews by the name of Incantada, “the Enchanted,” on the supposition that the figures are human beings petrified by the effect of magic. Its central position, and the nature of the construction, support the idea that it was connected with the ancient agora. The space which lies between the sea and that part of the main street where the Incantada and arch of Constantine are situated, is said to have been occupied by the hippodrome, noted for having been the scene of a promiscuous massacre of the assembled people of Thessalonica by order of Theodosius. In many parts of the town, particularly at the fountains, sepulchral stones and inscribed sori are to be found. Wherever figures occur upon the latter, their heads have, as usual, been destroyed by the Turks, nor is it easy to find an inscription that is perfect. The most interesting that I have observed are, 1. A simple mnema, valuable only

§ 3.247 its having a double date, by which it appears that the year 302 in one epoch corresponded to 186 in the other: as the difference 116 is the exact interval between the destruction of Corinth and the battle of Actium, there can be no doubt that these were the two events from which the dates were taken. 2. An epitaph in verse, wanting one or two lines at the beginning, where the name of a woman occurred, whose husband Eutropus constructed the tomb for her and himself. In two prose lines in smaller characters, which follow the verses, he declares that whoever shall place any other corpse in the tomb, except those of his children, shall pay a fine to the public chest of 10,200 denaria. 3. Another inscription contains the names of those who contended for the prize in a certain funereal contest, in which there were trials in the pancratium and in wrestling by boys, by young men, and by adults. It is to be supposed that the prior name of each pair was the victor.

§ 3.248 The population of Saloniki is reckoned at 80,000, but probably does not exceed 65,000, of whom 35,000 are Turks, 15,000 Greeks, and 13,000 Jews, the remainder Franks and Gypsies. I was unable to obtain permission to enter the citadel. It appears from Beaujour, and other travellers, that there are some columns of verd antique, and an arch erected by the city in honour of Antoninus Pius, his wife Faustina, styled Σεβαστή, and his adopted sons, Marcus Aurelius and Lucius Commodus, the former of whom is entituled Caesar.

§ 3.249 All the Turks of Macedonia who bear arms are Spahls, Yuruks, or Janissaries. The Spahis are the cavalry found by the holders of the zaims and timaria, when called upon by the government. The Yuruks cultivate their own lands chiefly in the mountainous districts. The Janissaries are the garrisons of the fortified places, among whom are generally enrolled the greater part of the heads of families engaged in trade or manufactures, or who have landed property in the neighbouring plain. A thousand pounds sterling a year in land is considered a large estate. Hadji Mustafa, the Bash Tjaus of the Janissaries, has seven tjiftliks worth 20,000 piastres a year (or 1200L), though he lives at the rate of not more than eight or ten thousand. Under a government which makes every one feel danger in displaying his wealth, and renders property and life insecure even to its most favoured subjects, the extremes of parsimony and extravagance are naturally to be found. Turks as well as Jews often carry the former to excess, and the latter is by no means uncommon among the young Osmanlis. An under-employe in the Mekheme is pointed out to me, who in a few years dissipated 2000 purses and seven tjiftliks. These Turkish landed proprietors, however, are the persons of the greatest stability in Turkey; and the Frank merchants who bargain for their corn, cotton, and tobacco, can, without much risk, make advances upon their crops. The Jews of Saloniki are descended from the largest of those colonies, which settled in Greece at the time of their expulsion from Spain at the

§ 3.250 end of the fifteenth century; but a considerable portion of them have become Musulmans since that time, though without being altogether acknowledged by the Osmanlis, and forming a separate class under the denomination of Mamins. Inheriting the Jewish spirit of parsimony and industry, they are generally rich, and among them are some of the wealthiest Turks at Saloniki. Hassan Adjik, one of the ministry at Constantinople, and his brother, who is Gumrukji, or collector of the customs at Saloniki, are Mamins. They are naturally objects of extreme dislike to the idle, poor, and profligate Janissaries of the lower class. They go to mosque regularly, and conform to the Mahometan religion in externals, but are reproached by the other Turks with having secret meetings and ceremonies, with other peculiarities of which the best attested is their knowledge of the Spanish language. They are said to be divided into three tribes, two of whom will not intermarry with the third, nor will the latter give their daughters in marriage to the Osmanlis. The πολιτεία, or Greek community, is presided over by the metropolitan bishop, who with the archons arranges all civil disputes in which Turks are not concerned, unless when the Christiane think fit to resort to the Mekheme. By a strange distortion of ancient geography, Thessalonica and Berrhoea are ecclesiastically επαρχίαi, or provinces of Thessaly; thus the

§ 3.251 bishop of Thessalonica is styled υπέρτιμος και έξαρχος Πάσης Θετταλίας; he claims the privilege of the epithet παναγιώτατος in his own province, but elsewhere is intitled only, like other metropolitans, to the πανιερωτατος. The bishoprics of his province are Kitro, Kampania, Platamona together with Lykostomo, Servia, Petra, Ardhameri, of which the residence is Galatista, and Ierisso which includes the Aion Oros. There are some opulent Greek merchants at Saloniki, most of whom are indebted for the undisturbed possession and increase of their wealth to the protection which they have enjoyed as dragomans or barataires of the European missions. Now that these protections are about to be abolished, their situation will be much more precarious. There are three sorts of kharatj paid by the rayahs; the first, called edina, is of 3 piastres, to which boys under 14 are subject, but which is generally exacted from all under 11; the second, the efsat, of 6 piastres, is paid by artisans, servants, and all the poor, even beggars; the third, alia, taken from all the classes above the last, amounts at Salonica to 12 piastres a head. Mr. N—, the principal Greek merchant, who is procurator for Mount Athos, informs me that he pays only 3600 kharatjes for the whole population of the peninsula, though there are 4000 monks alone, besides laics.

§ 3.252 It is almost the only place where the kharatj is underrated. Those who farm it having generally the means of making good their claims for an increase in the rayah population, it most frequently happens that individuals pay more than the regulated sum, and scarcely ever the reverse. Sometimes they are called upon for the double or triple. The Turks are probably aware that Mount Athos is rated below its numbers, but being the abode of persons devoted to religion, it is intitled to favour by the Turkish usages, for custom is a powerful argument among them, though seldom employed, as in the instance just mentioned, for the benefit of any but themselves. A Pasha of Saloniki having received orders to join the Grand Vezir’s army, was waited upon by a merchant acting as English consul, to whom he was indebted about 30L. My friend, said he, where am I to find a para? I have not money to pay the bread I have been eating here; the Porte indeed has sent me 500 purses, but it will not discharge one fourth of my debts. At least, says the consul, you will give me an acknowledgment in writing. Adet deil: it is not the custom; was the only reply. It is the custom to admit Christians to see the mosques of Saloniki, which have been once churches, probably because the imam gets a fee by it. The menials of a Turkish family at Saloniki, such as the kahuedji [coffee-man], tutunji [smoke-man], akhdji [cook], receive about 10 piastres (12 shillings sterling) a month. A yazji, or scribe, 30 piastres.

§ 3.253 Greek women servants in the Frank families have about 50 piastres a year, with some articles of clothing; in all cases with board. The finest bread is now 15 paras the oke of 2 3/4 lbs., and mutton 18 or 20 paras an oke; beef only 8 or 10, as it is consumed only by Jews and Franks. The ordinary price of silk is 50 piastres the oke; and almost every family raises silkworms. Ordinary cotton and woollen stuffs for the clothing of the common people are also woven in the private houses as well as in the surrounding villages. A considerable quantity of cotton towels are made here, sometimes with a border of gold threads, for the νίψιμον, or washing of the upper classes before and after meals, which in every part of Greece is practised as in the time of Homer. Silken gauze for shirts and mosquito curtains, are another fabric of the city, but the chief manufacture is the tanning and dyeing of leather, which is entirely in the hands of the Janissaries. The commerce of Saloniki has very much declined during the war, and even since Beaujour described it in 1797. Tobacco sent from hence in imperial ships is now the only considerable export. No English ship has loaded here for 12 years. The beys have their magazines full of corn, which by a firmahn of the Porte, issued last year and renewed this year, they are forbidden from sending to Christendom.

§ 3.254 Meantime the Porte demands a certain proportion from all the most productive corn countries of the empire, Macedonia among the rest, at a low price, on the pretence of fitting out fleets and armies. The consequence has been, that last year, when the price of corn at Athens was very high, it was sold by the government at Constantinople to foreigners, at a much lower price than they might have received for it in Greece, including the expence of sending it there. Three or four hundred thousand Stambul kila of wheat might be procured here in a month, and cattle in any number that could be required. The Beys of Saloniki suffer more than the more distant landlords, because the smuggling of corn can be more easily carried on from any other part of the coast. In general the orders of the Porte against the exportation of corn are converted into a source of profit to the local governor; but in a fortified place, under the eyes of a Pasha, and in time of war, more attention to the imperial orders is necessary. In reading descriptions of China one is struck by the similarity of the customs of that country with those of Turkey, arising from the same Tartar origin. Their dress and architecture, their custom of interchanging presents, their habit of smoking, and the amusements at their festivals, are almost identical. Public employments are generally venal, in spite of the Sovereign. The quantity of escort when a man goes out, is the measure of his grandeur. It is unpolite to speak of any but agreeable subjects at visits, and even to use certain words conveying hateful ideas.

§ 3.255 The Emperor gives only two audiences to ambassadors, one at coming, the other at departing. When a great man passes through the streets, his approach is indicated by a small drum. A drum marks the watches of the night. Provincial governors are changed very frequently. Tjay, of which word tea is the softened English form, preserves its original sound from Japan to the Adriatic. From the Lettres Edifiantes, we learn that the Mongol Tartars distinguish black tea by the name Kara Tjay, like the Turks. The latter, however, now make very little use of tea, except medicinally, nor is any brought to them overland as formerly, their supply being entirely, as well as that of the greater part of their coffee, from Europe. In Barbary the custom of drinking tea, particularly green tea, still prevails. There are many words in Turkish, which having been borrowed from the Greek, seem to show that the Turks had not in their own country the objects expressed by them; for example, lelek -stork, liman -port, keremid -tile. The borrowing of titles is more easily accounted for, as Effendi from αυθέντης. Effendem in Turkish, and αυθεντήμου or more vulgarly αφεντήμου in Greek, is the common mode of addressing a gentleman among both people. The Turks have a certain manly politeness, which is the most powerful of all modes of deceit, and which seldom fails in giving strangers an erroneous impression of their real character. It covers a rooted aversion to all European nations, as well as to the individuals who have the misfortune to have any dealings with these plausible barbarians.

§ 3.256 Though in the most splendid aera of their history their feelings may have been those of contempt, founded upon ignorance, fanaticism and the pride of conquest, it has been changed by their weakness and their dread of the Christians of Europe, into a mixture of fear and hatred. Thus there are two things which the European who has any political dealings with the Turk, should never lose sight of: 1, that he hates us: 2, that he fears us. By the latter only can we counteract the effects of the former, added as it is, to the most profound dissimulation, a keen sense of self-interest, and an obstinate perseverance in defending it. The Turks have so long experienced the advantages of conduct founded on this basis, and that of the mutual jealousy of the several European powers, that we may rely upon their adhering to it, as long as they have a foot of land on the continent of Europe. To say that the Turks have more honour and honesty than their Christian subjects, is a poor commendation: they have not the same necessity for the practice of fraud and falsehood. What other arms against their tyrants, are left to the unfortunate rayahs! It is not in the materials, but in the machinery of war, that the Turks are defective, and have hence become contemptible as a military power: they possess great numbers of armed men, strong, courageous, and enduring, and who, if properly managed, might oppose the most formidable resistance to the march of a numerous regular army through Turkey, where supplies are so scanty.

§ 3.257 Their very irregularity would in some respects render them more destructive to the formal tactics of an European power. But this powerful engine is rendered inefficient by the impotence of the government: repeated firmahns, which have lately arrived at Saloniki for the movement of the Macedonian troops to the northward, have produced only the march of a few Janissaries from this city. All the Yuruks and Janissaries of the subordinate towns have pleaded the insufficiency of their force for their own defence, and yet Macedonia is considered one of the most military provinces in the empire. The Albanians justly hold both Janissaries and Yuruks cheap in comparison of themselves; but they have a considerable respect for the Turkish cavalry.

§ 3.258 CHAPTER 27 MACEDONIA Nov. 26.—From Saloniki to Alaklisi in five hours and fifty minutes, with menzil horses and baggage, and deducting halts. The road lies all the way through the plain. At an hour and a half from the city a rivulet named Galliko crosses the road and flows directly to the gulf; half an hour beyond it is Tekeli, a small village, where the horses are changed; and an hour and a half farther a bridge over the river Axius, now called Vardhari, by which name it was known before the twelfth century, as appears from Anna Comnena. To the right, between Tekeli and the bridge, two pointed tumuli are very conspicuous objects; one in particular is of uncommon magnitude. The bridge of the Vardhari is about 1800 feet long, and crosses an island lying in the middle

§ 3.259 of the river, which occupies about a third of the whole breadth between the banks. The stream is now rapid, deep, and swollen with rain, though not so high as it usually is in winter. Below the bridge, about midway to the mouth, the river leaves Kulakia, a large Greek village, at no great distance on the left, and widens so much before it meets the sea, as to be near two miles in breadth. Kulakia, which is in the road from Saloniki to Katerina as well as to Verria, is the residence of the bishop της Καμπανίας, one of the subordinates of the metropolitan despot of Thessalonica. The bishop of Campania formerly resided at Kapsokhori, another Greek village, situated between the Karasmak, or Μαυρονέρι, and the Injekara, or Βιστρίτζα, in a well-wooded part. of the plains, around which are some other Greek villages. All the rest of the population of these great plains of Lower Macedonia consists of Bulgarian cultivators of the Turkish tjiftliks which are dispersed over it. One hour and ten minutes beyond the bridge, a small flat-topped height is on the left of the road, on the summit of which are some ancient foundations, and around it a Turkish burying ground, in which are many fluted and plain shafts, and other fragments of architecture, together with a pedestal bearing an imperfect inscription. This place is about a mile distant from the south-eastern extremity of a high mountain, which stretches from the right bank of the Vardhari in the direction of Vodhena.

§ 3.260 The valley of that river is seen to our right branching to a considerable distance among the mountains. Midway between the artificial height and. Alaklisi, which is 1 hour and 10 minutes beyond it, a tumulus rises close to the road od the right, then five more, nearly in a line, the last of which is at a musquet shot from Alaklisi. These tumuli stand on the last slope of the mountain, where a mile on the left begins an immense marsh, which extends as far as can be seen southward towards the sea, and westward towards the Olympene range of mountains which border the plains on the west. The tumulus nearest to Alaklisi is a great heap of earth based upon the rock, which all around is covered only with a thin layer of mould. An opening cut in the rock, covered above with a semicircular arched roof of masonry, and having a small chamber on either side of it, leads on a descent 33 feet long, to two chambers, which are excavated in the rock, under the centre of the tumulus, and are now nearly filled with the earth washed into them through the entrance. Of these, the first chamber is 56 feet long and 10.9 broad, the inner 13 1/2 by 11 1/2. The plain between Saloniki and Alaklisi is by no means so well cultivated or peopled.as that of Serres, on the road we met only some small caravans of camels; but it feeds a great number of herds and flocks, and abounds in hares, plovers, and woodcocks.

§ 3.261 On the lake there are myriads of the duck tribe in the winter; and partridges of the red-legged species on the slopes of the hills. The English breed has been introduced by some of the merchants of Saloniki, but has not propagated far from the neighbourhood of the city. Alaklisi, meaning in Turkish Godchurch, is by the Greeks named στους Αποστόλους, and by the Bulgarians Postol. It contains 40 or 50 poor cottages, and belongs to Selim Bey, of Saloniki, who maintains here an Albanian Subashi, with a small guard. The village is not in the direct road to Yenidje, but half a mile to the right of it. Nov. 27.—On the descent from Alaklisi into the main route, the fields are covered with fragments of former buildings, and of ancient pottery, such as are generally observable on the sites of Hellenic cities. The foundations of a wall of the construetion of those times is seen at right angles to the road, and terminating apparently at the marsh, the edge of which is parallel to the road at the distance of half a mile. A little beyond these foundations, following the road towards Yenidje, occurs a fountain, below which, on the edge of the marsh, is a small village, named Neokhori or Yenikiuy, where a low mound of considerable extent, and apparently artificial, seems to have been intended as a defence against the encroachment of the marsh. At 20 minutes from Alaklisi, and 10 beyond the first fountain, is another much more copious source, which is received into a square reservoir of masonry, and flows out of it in a stream to the marsh. This source is called by the Bulgarians Pel, and by the Greeks Πέλλη.

§ 3.262 As the ancient cities of Greece often derived their names from a river or fountain, the same may have occurred in the instance of the celebrated capital of Philip and his successors, which the description of Livy, compared with the tumuli and other ancient remains, clearly show to have stood in this situation. It would seem as if the name of Pella had survived even the ruins of the city, and had reverted to the fountain to which it was originally attached. The word was appropriate to a fountain, whether derived from the same etymon as πέλλη mulctrum, or from πελός black, an epithet which has been very generally applied by the Greeks to a source of water, from the μέλαν ύδορ of Homer to the mavromati of the present day. Below the fountain are some remains of buildings, said to have been baths, and still called τα Λουτρά. The baths of Pella are alluded to by a comic poet cited by Athenaeus. There is nothing remarkable in the taste of the water, but it has a slight degree of warmth, which perhaps might not be perceptible in summer. The reservoir stands upon the foundations of a Hellenic wall, above which, in a corn-field, is a large piece of masonry, constructed with mortar: all the cultivated land around is covered with pottery and stones, and hereabout the coins which the labourers of Alfiklisi collect in great abundance, are chiefly found. Eight minutes beyond the baths begins a second line ot tumuli, of which there are three parallel to the road, at a short distance to the right of it. The westernmost, or last towards Yenidje, is the largest

§ 3.263 of all, and has either been excavated, or has fallen In by natural decay, for it now exhibits the appearance of a double summit, with a hollow in the middle. It might naturally be supposed, that some of these tumuli, were royal sepulchres, especially the last mentioned, as well as that nearest to Alakliei, which contains chambers in the rock; but as we are informed upon good authority that Aegae continued to be the. burial place of the royal family, even after the seat of government was transferred to Pella, that the body of Alexander was destined to be sent to the same place, had not Ptolemy caused it to be carried to Egypt, and that Philip Aridaeus, his wife Eurydice, and her mother Gynna, were buried at Aegae by Cassander; it is more probable that the tumuli of Pella are the tombs of some of the noble families of Macedonia. That which I examined near Alaklisi might have been the receptacle of a family during a long succession of ages, and from the arched entrance it seems to have been used for this purpose, as late as the Roman Empire. Although so little remains of Pella, a tolerable idea may be formed of its extent and general plan by means of the description of Livy, compared with the existing traces. The interval between the westernmost of the eastern tumuli and the easternmost of the western was probably something more than the maximum of the diameter of the city, as we cannot but suppose these monuments to have

§ 3.264 stood on the outside of the walls. Its circumference, therefore, was about three miles. The two sources were probably about the centre of the site, and the modern road may possibly be in the exact line of a main street which, traversed it from east to west. The temple of Minerva Alcidemus is the only public building mentioned in history’, but of its exact situation we must remain in ignorance, unless some excavation or accidental discovery should hereafter reveal it. Of the construction of the city towards the lake, the historian has left us the following description, derived undoubtedly from Polybius: “Pella stands upon a height sloping to the southwest, and is bounded by marshes, which are impassable both in winter and summer, and are caused by the overflowing of a lake! The citadel rises like an island from the part of the marsh nearest to the city, being built upon an immense embankment which defies all injury from the waters; though appearing at a distance to be united to the wall of the city, it is in reality separated from it by a wet ditch, over which there is a bridge, so that no access whatever is afforded to an enemy, nor can any prisoner whom the king may confine in the castle escape but by the easily-guarded bridge. In this fortress was the royal treasure.”

§ 3.265 The mound near Neokhori marks perhaps the line where the wall was separated by the wet ditch from the citadel, but no vestiges of the island are to be perceived, which is not surprising as the citadel of Pella has now for not less probably than fifteen centuries been abandoned to the incroachments of the lake and the effects of the seasons. Beaujour asserte that he saw the remains of a port, and of a nicely-levelled canal communicating from the port to the sea. I am informed, that in summer when the marsh recedes from its present limits, some remains of a canal may be traced from the heights above Alaklisi, but as to the port, I can neither perceive the least traces of it, nor can I discover where M. Beaujour found any mention of it in ancient history. Nothing seems to have been wanted for a water communication between the city and the sea but to clear a passage through the marshes, which in all the deeper parts are capable of receiving vessels of a considerable draught of water.

§ 3.266 Scylax seems to have been sensible of this fact, for he merely states that there was a navigation from the sea by the Lydias to the royal residence of Macedonia, which was 120 stades in length exclusive of the Lydias. The lake was named Borborus, as appears from an epigram, in which Aristotle was reproached for preferring a residence near the Borbarus to that of the Academy. From the baths of Pella to Yenidje is a ride of 50 minutes. Two miles to the right of the last tumulus of Pella is the village of Alatjaushluk, standing on the slope of the mountain. Iinnitza, or Ghiinitza, more commonly known to the inhabitants, being chiefly Turks, by the corrupted Turkish form of Yenidje, appears, to have declined considerably of late years, as the number of houses is now by no means proportioned, to the eight minarets which the town still exhibits. There are however several good Turkish dwellings, and in the middle of the town that of Abdurrahman Bey, an Osmanli of an ancient family, and possessor of a large proportion of the

§ 3.267 neighbouring lands, which produce grain, cotton, and tobacco. The last of these, which occupies most of the land in the immediate vicinity of Yenidje, is renowned in every part of Turkey for its aromatic tutun, which, together with coffee, supplies the Turks with a stimulant at least as agreeable as the meagre ill-made wines of modern Greece. The leaves have been lately gathered, strung together, and hung up to dry, which operations are chiefly.performed by the women: every wall in. the town is now festooned with tobacco leaves, but particularly the open galleries which surround all the houses, and into which the inner chambers open. As the apartments in general have hearths only, without chimneys, the smoke of the wood which is burnt upon them circulates amidst the tobacco leaves, and gives the tobacco a peculiar flavour, which Italians object to, but Turks admire. The herb of Yenidje is of the species called garden tobacco, and has a small yellowish leaf. The territory yields in good years 2000 bales of 80 okes. The late harvest of corn has been abundant, and the Bey has his granaries overflowing for want of a market. Yenidje is commonly known among the Turks in distant parts of the country by the name of Vardar Yenidje, to distinguish it from the Karasu Yenidje, still more renowned for its tobacco, and which is situated about as far from the Nestus or Karasu eastward, as the Vardar Yenidjέ is to the westward of the Axius.

§ 3.268 The lofty and conspicuous mountain which rises behind Pella and Yenidje, is named by the Bulgarians Paik, and by the Greeks the mountain of Iannitza. The ancient name I know not where to look for. On the southern side it is for the most part bare and rocky, but on the summit and northern face it contains foreets chiefly of chestnut trees. Beyond it is the district named by the Christians Moglena, and by the Turks Karadjovasi, into which there is a direct road across the mountain from Yenidje, but the more frequented route makes a circuit of the western end of the mountain. Nov. 29.—Many remains of Hellenic antiquity, such as squared blocks of stone and fragments of architecture, are to be seen in the streets and burying-grounds of Yenidje, which has been built and repaired with the spoils of Pella. In quitting the town this morning for Vodhena I diverge to the right of the direct road, for the purpose of visiting Balakastra, as the Turks call Paleokastro, a tjiftlik of Abdurrahman Bey, which he recommended to my notice as a place containing antiquities, and arrive there in forty minutes.

§ 3.269 Just above the tjiftlik a copious source issues from the foot of the mountain, turns several mills, and waters some gardens belonging to the farm which is on its right bank. On the opposite side of the stream are many ancient wrought blocks in and around a rained chapel; others are observable in different parts of the tjiftlik, as well as at the mills near the source; so that there can be little doubt that Paleokastro was an ancient site. The position is very agreeable, being well furnished with wood and water, and commanding a prospect over an extensive level bounded by the mountain of Iannitza, the lake of Pella, and the heights near Vodhena. This plain is much better cultivated than any part of that towards Saloniki, being now almost a continued field of nascent corn, without a single fence. Leaving Paleokastro exactly at noon, we follow a carriage-road through the plain, and pass several small Turkish villages with burying-grounds, in which the tombstones are for the most part ancient wrought blocks or fragments of architecture. Many of these have probably been brought from Paleo-kastro, or even from Pella, for the Turks often resort to a considerable distance for the stones, which they convert into sepulchral monuments. At 1.40 we cross a large river by a bridge which derives its name of Koluden Kiupresi from a small village a little below it on the left bank. The river flows from the valley of Karadjdvasi, or Moglena, which is separated from the plain by a range of small hills, admitting only a narrow vale for the passage of the river, and connecting the mountain of Iannitza with the great range which is a continuation of Olympus.

§ 3.270 A lofty summit to the northward of Vodhena, called Nitje, bounds Karadjdvasi on the west; and is the highest point of the range except Olympus itself. Mogleni is a Greek bishopric, under the name of Moglena and Moleskha. The former name, as well as Vodhena, is older than the twelfth century, as we learn from Anna Comnena. They are both to be traced to the language of the Sclavonic tribes, who occupied the Macedonian plains about the ninth century, and drove the Greeks -into the Chalcidic peninsula, or into the low grounds near the sea, where the marshes and rivers which intersect them offered means of resistance. To these two parts of Lower Macedonia the Greeks are now chiefly confined, and there the names of places are of Greek form and derivation. The Turks of Karadjovasi are supposed, for the most part, to be Bulgarian apostates from Christianity. A high snowy mountain makes its appearance to the northward of Mount Paik, which is said to be not far from Istib and the plains of the Upper Axius. The river of Moglena is called Karadjk by the Turks, Meglesnitj by the Bulgarians, and by the Greeks Moglenitiko. The ancient name is not certain, possibly it was Lydias, or Ludias, for it is the largest of the rivers which fall into the lake of Pella, and its course before it enters the lake is in the eame direction in which the Karasmak, or Mavroneri, which we know to have been the Lydias, pursues its course to the sea, after emerging from the lower end of the lake.

§ 3.271 At 2:10 we arrive at the extremity of the plain, which is not less than fifty miles long, in a direct line from its opposite end near Saloniki. Turning a point of the heights which branch from Mount Nitje, and bound the valley of Vodhena on the north, we enter that valley, which is about a mile broad, and is included on the southern side by the lowest falls of Mount Turla, a. summit of the Olympene range, which rises above Niausta. Nitjk is a link in the same chain, and is separated from it only by the pass of Vladova behind. Vodheni. The valley of Vodhena, at the end of four miles, is closed by precipices over which the river falls in one principal and several smaller cascades. On the edge of the cliffs stands the town of Vodhena. Ascending the valley we soon reach the left bank of the river formed by the reunion of the torrents which fall over the cliffs; it is a small, but deep and rapid stream, confined by high banks. At 3.15 we cross it by a bridge, and immediately afterwards a smaller branch by another bridge, then enter the vineyards and mulberry grounds which extend to the foot of the precipices of Vodhenk; pass soon afterwards some foundations of Hellenic walls on the road side, and at 3.40 arrive at the clifis. Leaving these to the right, we mount the heights by a circuitous stony road, which in one place is cut through the rock, and enter the town through a wall of sun-baked bricks. Vodhena, in the grandeur of its situation, in the magnificence of the surrounding objects, and the extent of the rich prospect which it commands, is not inferior to any situation in Greece. As Horace

§ 3.272 said of Tibur and the precipitous Anio, neither Sparta nor Larissa, although both combining sublimity and beauty of scenery in the highest degree, appear to me so striking as the rocks, cascades, and smiling valleys of Vodhena, encased in lofty mountains which expand into an immense semicircle, and embrace the great plains at the head of the Thermaic Gulf. There cannot be a doubt that this is the site of Aegae, or Edessa, the ancient capital of Macedonia, to which it was well adapted by its lofty, salubrious, and strong position, at the entrance of a pass which was the most important in the kingdom, as leading from the maritime provinces into Upper Macedonia, and by another branch of the same pass into Lyncestis and Pelagonia. Such a situation would have been ill exchanged for the marshes of Pella, had not the increasing power and civilization of the Macedonians rendered maritime communication of more importance to their capital than strength of position, while in the winter Pella had the recommendation of a much milder climate. Vodhena, so called from the Bulgarian Voda with a Greek termination, in allusion to its plentiful waters, is a metropolitan bishopric, comprehending about one hundred villages of Bulgarian Christians, who in general are ignorant of the

§ 3.273 Greek language. The bishopric is still known by the name of Edessa as well as Vodhena; ecclesiastically it is considered subordinate, together with several other metropolitan and episcopal sees, to the archbishop of Achris, or Bulgaria, who received this authority from the emperor Justinian, when he founded at Achris the town which he named Justiniana Prima. Hence the archbishop of Akhridha is still in the Greek church αυτοκέφαλος, and independent of the three patriarchs; though the Turkish government not acknowledging his independence of the Patriarch of Constantinople, and the duties and influence of the hierarchy being almost entirely local, his authority is little more than nominal. Numerous ruins of churches on the skirts of Vodhena show its former importance under the Greek Empire. At present it contains 1500 Turkish and 500 Greek houses, but many of the Turkish houses are let to Greeks. The bazar is extensive and well-furnished. There are five or six mosques, and a high tower containing a clock, but the most striking building, more however from situation than magnitude or structure, is the bishop's palace adjoining the metropolitan church. Standing on the edge of a projecting rock in the middle of the cliffs, it commands a

§ 3.274 prospect of the plains as far as the Bay of Saloniki and Mount Khortiatzi, and itself furnishes a most picturesque object, especially when viewed in profile, crowning the cliffs which overhang a beautiful concave slope terminating in the valley which consists of gardens, vineyards, and orchards. The chief produce of Vodhena is silk and fruit; the yearly amount of the former varies from 2000 to 4000 okes, with a price equally variable, being sometimes 15 and sometimes 40 piastres the oke: this year it is 17. Every market day, which in Greece is commonly on a Sunday, is attended by men from Sarighioli, Ostrovo, Filtirina, and other surrounding districts, for the sale of their agricultural productions, or to furnish themselves with manufactures from the bazar, or with the fruit grown in the gardens of Vodhena, consisting of jujubes, apricots, apples, plums, and grapes: the latter are raised in large quantities, and are chiefly used for making a sweetmeat common in Turkey, by boiling the juice of the fruit into a thick hard syrup, which is mixed with almonds and walnuts. Nov. 30.—At a distance of 50 minutes above the town there is an upper cascade, where the river falls over the rocks in a single body. The road thither leads through gardens watered by numerous derivations from the main stream, and affords many beautiful views of the town seen through the trees, with the great mountain of Niausta in the background. At a superb grove of plane-trees a fair is held on the 15th of August.

§ 3.275 Beyond the gardens the plain narrows, and is occupied by meadows and vineyards on the bank of the river as far as the cascade, which is not large but extremely picturesque, falling into the meadow over a rocky steep covered with bushes. The perpen-dicular fall is not more than 50 feet, but above it there is a rapid descent at an angle of about 45°, more than equal in perpendicular height to the former. Above the cataract stands the little village of Vladova, so named from the fall, at the entrance of a green valley which terminates at the end of two miles in a small lake, from which the river issues. The vale is about half a mile in width, and is bordered by the woody summits of two parallel ridges which meet at a pass at the further end of the lake: through the opening appears the great snowy peak northward of Kastoria called Vitzi. The valley leads, at the end of two hours more, to the town and lake of Ostrovo, near which the road branches to the left into Saiighioli, and to the right by a precipitous ascent over the ridges which unite Mount Vitzi with the summits on the northern side of the pass of Vladova and with Mount Nitje. The latter route leads into the plains and valleys watered by the tributaries of the Erigon, or great western branch of the Axius, called Tjerna by the Bulgarians, and by the Turks the Little Karasu. The pass of Vladova being the opening made by nature for the passage of the river of Vodhena, which rises in Sarighioli and Mount Vitzi, is the easiest of all the communications which lead across the Olympene range from Lower into Upper Macedonia. The two others most remarkable are

§ 3.276 those behind Niausta and Verria, both which descend into the plain of Sarighioli, but are rendered less important than the pass of Vodhena, as well by their difficulty and steepness as by their conducting into a part of the country more distant from the passes which lead into the basin of the Erigon. Having crossed the river near Vladova, I return to Vodhena along the right bank, and in descending the hill of the cascade pass through a deep passage which has been cut through the rocks for a road, and is probably a work of the ancient Macedonians. The rivulets diverted from the main stream for the sake of watering the gardens behind the town, are conducted through every street, and even through many of the houses, until approaching the cliffs they reunite, and fall over the precipices in four principal cascades, which, after watering the gardens below the cliffs, they again constitute the single stream which flows through the lower valley to the Moglenitiko. The largest fall of water over the cliffs is towards the northern end of the hill, where it forms the main river which we first crossed in arriving; this branch receives a tributary from Mount Nitje before it unites with the streams from the other cascades. Notwithstanding the importance of the ancient city which stood at Vodhena, the Hellenic remains are few; the advantageous position has doubtless been always occupied by a considerable town, and new constructions have been continually operating the destruction of the more ancient. The only vestige I can discover of the Hellenic fortifications is a

§ 3.277 piece of wall which supports one of the modern Houses on the edge of the cliff; but there are many scattered remains in the town, and among them some inscriptions of the time of the Roman empire. A stele, surmounted by a pediment, which has been placed over the gate of the Bishop’s palace, preserves a catalogue of young men who had passed through their ephebia under an ephebarch named Lysimachus, son of Abydianus. It is curious for two particulars: 1. Some of the ephebi are distinguished by the mother’s name without any mention of the father’s, as, Αλέξανδρος και Ειουλιος οι Μαρκιας, ‘Έσπερος Σεμέλης, Είουλιος Καλλίστης. I have already given an example of this Macedonian custom from the Vardar gate of Saloniki. 2. The inscription has the date 328, which, calculated from the capture of Corinth, is the year a.d. 182, in the reign of Commodus, but from the battle of Actium, is a.d. 298, in the reign of Diocletian. The latter epoch is to be preferred, not so much from the style of the monument as from the certainty afforded by a coin of the emperor Philip bearing the date 275, and which was struck probably at Berrhoea, that the latter epoch was then employed in Macedonia. In the metropolitan church are two fragments, which appear to have belonged to one and the same inscription.

§ 3.278 The epsilon and sigma are of a singular form but of which there are other examples in Macedonia. A third inscription might be ascribed to a late period of the Roman empire, from the angular form of the omicron and theta, thus, [ ]; but this also may have been a Macedonian peculiarity, for the composition shows no decline of taste among the Edessaei, being an elegant epitaph in three elegiac couplets in memory of one Graphicus, whose wife survived him The poet in saying that “God had placed the divine soul of Graphicus in the plain of the blessed,” may be thought, perhaps, to have written in Christian times, but the words are not inconsistent with the Platonic doctrines. The epitaph is inscribed on a sarcophagus standing at a fountain (now dry) which is called by the Turks the fountain of the Mirror, because one of the lacunaria of a Corinthian ceiling has been placed over it, with the stone set on its edge over the pipe. The sculpture thus placed the Turks have likened to a mirror. Aly Pasha was not slow in discovering the advantages of the position of Vodhena, and having introduced himself into it ten years ago as Dervent Aga, he has now the power of descending at pleasure into the plains of Lower Macedonia, or the means of defending this approach to his dominions from the side of Constantinople.

§ 3.279 The Ayan who now governs is a native, but is entirely under the influence of Aly, who maintains here a guard of Albanians. The military importance of Edessa was still greater under the Romans, in consequence of its lying in the great road from Dyrrhachium to Thessalonica, the establishment of which was one of their first cares after the conquest of Macedonia Although this road was furnished through its whole extent of 267 miles with milestones, and the distances of the several stations are given in all the three itineraries, the Antonine, Jerusalem, and Tabular, and some parts of it twice over in the first, there are not many points on the road which can be accurately fixed until the whole shall be submitted to a careful examination, so as to ascertain some of the ancient sites. Nor until then can any safe criticism be exercised upon the itineraries themselves, which as usual differ from one another in many of the distances. A few remarks on this important route may nevertheless be acceptable to future travellers. In proceeding westward from the pass of Vodhena, the road crossed two great valleys and three remarkable ridges before it arrived at Clodiana, from which there was a bifurcation to Dyrrhachium and Apollonia. From the Tabular Itinerary we learn that at 19 M.P., beyond Lychnidus, the road crossed a bridge named Pons Servilii, which could have been no other than a bridge over the Drin, anciently Drilo, at ite issue from the lake Lychnitis. We thus obtain the point

§ 3.280 from whence the road crossed Mount Candavia to Clodiana, which appears to have been situated on the Genusus, for the name Clodiana is probably derived from Appius Claudius, whose camp was upon that river when he was employed against Gentius, at the same time that the Consul Aemilius was carrying on the war against Perseus in Macedonia, in the year B.C. 1681. And hence it becomes evident that the Genusus was the river now called Skumbi, or Tjerma, consequently that the mountain which lies between the sources of that river and the northern end of the lake Lychmtas was the proper Candavia. It is the same mountain of which I observed the bearing from Kozytzk to be N. 23 W. by compass. Although the distance of Clodiana from Apollonia is no less than 8 M.P. greater in the Jerusalem than in the Tabular, Itinerary, yet as both these authorities place the Apsus about midway, we have thus an approximation which may assist in ascertaining the exact site of Clodiana. Skumbi is obviously a corruption of Scampis, a name found in all the Itineraries at about 21 M.P. eastward of Clodiana, consequently on or near the Genutus, perhaps at the modern Elbasan. The branch of the Genusus upon which that town is situated may have been named Scampis as well as the town, and by a common kind of change may have superseded the name of Genusus, as that of the entire course of the stream below the junction.

§ 3.281 As there was a distance of about 17 M.P. from the bridge of Servilius to Lychnidus, this chief city of the Dassaretii was near the southern extremity of the lake, on the eastern shore, where the road, after having been diverted by the lake to the northward of its general direction, recovered that line by following the eastern shore from the bridge of Servilius to Lychnidus. From thence it crossed the mountains which rise from the eastern side of the lake into the plains watered by the Erigon and its branches. These mountains, which have a north and south direction, are divided into two parallel ridges by a longitudinal valley, where are situated Peupli and Prespa, and, if I am rightly informed, three lakes, of which the southern, called that of Ventrok, sends forth, as I have before observed, the river which flows through the pass of Tzangon, and forms the principal, or at least the longest branch of the Apsus, and which I suppose to be the Eordaicus of Arrian The disagreement of numbers in the several Itineraries renders it difficult to deduce from them the exact position of any of the places on the road between Lychnidus and Edessa; the only one of any importance was Heracleia, the chief town of the province of Upper Macedonia, called Lyncus, or Lyncestis. Heracleia was distant from Lychnidus about 46 M.P., from Edessa 64,—total from Lychnidus to Edessa 110; which, compared with the 56 o. m. of direct distance on the map, gives a rate of 2 M.P. to the horizontal g. m., not

§ 3.282 an unreasonable rate in itself, as the road is in great part mountainous, nor as compared with the rate on the level road from Edessa to Thessalonica, which is 1.4 M.P. to the o. m. According to the proportional distances, Heracleia stood not far from the modern town of Filurina, at about 10 o. m. direct to the southward of Bitolia, which is now the principal town in that part of the country, and occupies the site of the ancient Pelagonia, thus agreeing in reference to ihe supposed position of Heracleia of Lyncestis, inasmuch as the ancient authorities show that the Lyncestae were situated to the southward of the Pelagones, and between them and the Eordaei, who appear to have occupied the country of Ostfovo and Sarighioli. But I shall have occasion to revert to the geography of Lyncestis, in reference to the military operations at the beginning of the contest between Philip, son of Demetrius, and the Romans. Dec. 1.—Among the vineyards at the foot of the precipices of Vodhena are many fragments and foundations of ancient buildings, together with remains of barbarous times, probably those of Greek or Turkish houses, which were once dispersed among these gardens. It is said that several marbles sculptured in relief were once to be seen here, and among them some broken statues, in particular part of a horse of very large dimensions. Lower down the stream there are some other fragments of antiquity; from all which, as well as the foundations of Hellenic walls, both above and below, it is evident that Edessa occupied both sites.

§ 3.283 With the decline of Macedonia after the Roman conquest, the lower town may have gradually been abandoned, and the upper, which was anciently the acropolis, and probably the royal residence, may have become the part principally inhabited, as indeed the inscriptions, being all of that date, tend to show. At 10 we leave the point where we crossed the two bridges in approaching the town, and following the foot of the heights on the southern side of the valley, arrive at 10.30 at a projecting point where a copious source of water issues from under the hill; then pass along the plain at a short distance from the foot of the mountain, and at 11.25 join the direct road from Vodhena to Niausta, which descends from the southern extremity of the former town into a small circular plain lying at the foot of the hill on that side, and then crosses over the heights of Mount Turla, which enclose that plain to the southward. At 12.5 we halt, till 12.34, to dine at a brook, and then after having crossed a small stream which descends to the lake of Iannitza from the mountain on the right, arrive in sight of the singular topography of Niausta, to which we soon begin to ascend, and arrive in the town at 1.45. At the upper end of a deep rocky glen, between two of the highest summits of the mountain, three tabular elevations rising one above the other, look from the plain like enormous steps; they present a front of cliffs not so high as those of Vodhena, but which terminate laterally also in cliffs separated on each side by ravines from the great heights of the mountain. Niausta occupies the middle and widest terrace, and, like Vodhena, is watered by numerous branches of a

§ 3.284 stream which, flowing from a ravine behind the upper tabular summit, passes through the middle of the town in a deep rocky bed, over which there is a bridge. As at Vodhena derivations from this stream pass through every house in the town, and fall over the cliffs, after which they turn some mills, and are again united into one river in the low grounds. Niausta is a Greek town, the Bulgarians not having obtained possession of the Olympene range to the southward of Vodhena. The name is properly Νιάγονστα, perhaps a corruption of Νέα Αίγονατα. Although now in the power of Aly Pasha, it is still governed by its own magistrates, whose authority, the place being an imperial appanage, and the inhabitants well armed, has been generally respected by all the neighbouring Pashas and other men in authority, including the robbers, though Niausta has occasionally been at war with them all. By an effect of the republican system of the place, I am detained two hours in an empty house, while the powers are consulting as to the konak in which Ι am to be lodged; at length I am conducted to the house of Thomas, who is married to the widow of Lusa Papafilippo, a name of some note in Macedonia, and formerly proestos of Niausta. The decline of the place, and its subjection to Aly, which will be followed by the usual consequences of his insatiable extortion, is to be attributed to that spirit of dissension which seldom fails to ruin the Greeks when they have the power of indulging in it. Not many years ago Niausta was one of the most commercial places in Northern

§ 3.285 Greece, and like Verria, Siatista, and Kastoria, had merchants who traded to Christendom as well as Turkey, but not one of whom now remains here. Papafilippo, who is spoken of in terms of high respect by his own adherents as a benefactor of his native town, was poisoned with several others, about 20 years ago, by the adverse party, at the head of which was one Zafiraki, son of Theodosius, who afterwards became proestos, and enjoyed all the authority until last year, when the party of Papafilippo, by applying to Aly Pasha, gave him the long-desired excuse for introducing his myrmidons into the town. But he met with a stout resistance from Zafiraki and his brother Konstantino Musa assisted by a party of Albanians, under two Albanian brothers Vrakho and Litjo. Those whom the Pasha first sent having been fired upon from an inclosure of mud bricks, which is the only artificial defence of the place, he found it necessary to increase their numbers to 2000, who quickly destroyed every thing on the outside of the town, but not having cannon, could not ruin the fortifications, slight as they are. They proceeded therefore in the manner of an ancient πολιορκία, building towers on a level with the walls, from which they could fire into the town. Their loss was very great, according to the people of Niausta, of whom about fifty were slain. At length the besieged, after having lived for some time upon wild herbs, branches of trees, and bread made of the refuse of their rice mills, were obliged to surrender, but not until the four chiefs above

§ 3.286 mentioned had fought their way one night through the besiegers with 50 palikaria, and had arrived safe at Saioniki, where I saw them, and where they still remain. All the persons found in Zafiraki’s house have been carried to Ioannina, where they are now in prison, and the house is occupied by the Albanian commandant, and by a Stambuli Bostanji residing here as agent of the Sultana, who enjoys the revenue of the town and its district. In one year Aly has exacted 500 purses from the people, and no longer apprehending any resistance, has reduced his Albanian guard to 20, which, united with those stationed at Verria and Vodhena, are sufficient both to maintain his interests and to protect the passes against the robbers, to whom he has been indebted for his justification with the Porte for introducing his troops here. These kleftes during the last summer blockaded Varia as well as Niausta, and advancing to the walls of the latter, carried away children, cattle, and sheep. At length Aly sent his trusty Tepeleniote Mutjobon, or Μετζομπόνος, as the Greeks write his name, who has dispersed or taken them all, except a few men under a Musulman Albanian named Sulu Proshova, who not long before was at the head of 700 men, for the most part Christians.

§ 3.287 He still haunts these mountains which as far as Bitolia, Prillapo, and Velesa, furnish so many impenetrable retreats, that it is almost impossible to eradicate the thieves from them. Not long since, Sulu took a boy of Niausta going to Verria, who was to have been ransomed by the village for 16 parses, when, two days before the money was to be paid, the boy escaped, and arrived here a day or two ago. The principal church, dedicated to St. George, has a monastery attached to it, and is surrounded by a quadrangle of cells or small apartments for the monks, which they generally let to strangers. The people of Niiusta were formerly noted for working in gold and silver, and still carry on the manufacture in a smaller degree. The productions of the territory are wheat, barley and maize in the plain; rice in the immediate neighbourhood of the marshes adjacent to the lake of Iannitza; on the heights vines, supplying one of the best wines in Macedonia, in sufficient quantity for a large exportation, and in the valley mulberry plantations, which yield about 300 okes of silk per annum. The town is well supplied with fish, particularly with large pike from the lake of Iannitza, and with trout from their own river, the principal source of which is at a short distance above the town. Many persons suppose it to be the discharge of a katavothra in the lake of Akridha, but can give no better reason for this opinion, than that the lake is the only one in Macedonia which produces trout. The sheep which feed on the mountains behind the town, furnish a fine wool, and mutton of the best quality.

§ 3.288 Niausta, as might be expected from its natural advantages, stands on the site of an ancient city, of this the coins which are found in the fields below the hill, and some vestiges of ancient buildings in the same situation, leave no doubt. But these are the only remains I can discover, except a Doric shaft, of a soft kind of stone, in the gallery of the church of St. George, and at one of the fountains in the town a sepulchral marble, with figures in low relief. The natives suppose that the Macedonian city stood higher in the mountain; it occupied, perhaps, all the three terraces, the upper having been the citadel. I am inclined to think that Citium was the ancient name. Livy states that in the plain before Citium Perseus reviewed his army before he marched into Thessaly, when after a peace of twenty-three years, he began that celebrated war with Rome, which in four campaigns put an end to the Macedonian kingdom. That Citium was between Pella and

§ 3.289 Berrhoea, may be inferred from the king having sacrificed to Minerva Alcidemus at Pella, just before he joined his army at Citium, and from his having marched from thence in one day to the lake Begorrites in Eordaea, and on the succeeding day into Elimeia, where he encamped on the bank of the Haliacmon, and thence proceeded to cross the Cambunian mountains into Perrhaebia. Hence also we may infer that the lake Begorrites was the Kitrini of Sarighioli, for the lake of Ostrovo would not have been in the direction from Pella to the Haliacmon, unless Citium had been at Vodhena, nor could the king have marched in one day from that lake to the Haliacmon. In the epitome of the 7th book of Strabo, it is stated that the lake of Pella is formed by a certain Λχόβτασμα, or stream diverging from the Axius, which can only be reconciled with the reality, by supposing the sources of Pella and Paleokastro to be derived from the Axius through the mountain. But this would be so unusual a phenomenon, that it cannot even be considered probable, until a derivation from the Axius is found flowing into the opposite side of the mountain; nor if it were true, would the quantity of water be any thing approaching to a sufficiency for the lake of Pella, which is evidently fed, not only by the springs of Pella and Paleokastro, but also by the Moglenitiko, the rivers of Vodhena and Niausta, and many smaller torrents, assisted perhaps by some subterraneous springs; the excess of all these over the water carried off by the Ludias, is the cause of this extensive tract of lakes and marshes.

§ 3.290 Dec. 2.—Setting out from Niausta for Verria at 12.30, we descend the hills obliquely, and having reached the plain follow its margin, pass two small villages beautifully situated among the rich elopes of the mountain, while to the left is the plain, equally well cultivated, and extending to the marshes of the Pellaean lake. At 3, turning a projecting point of the mountain, we arrive in sight of Verria, and at 3.30 cross a deep rivulet, which issues from a gorge in the mountain to the right. Here are some foundations of an ancient bridge, consisting of loose materials cemented with mortar, but faced with large quadrangular stones, accurately laid in the best Hellenic style. An ascent from thence of ten minutes conducts to the modern gate of Verria, after passing through a Turkish cemetery, which contains many fragments of ancient architecture, and a little beyond it a large piece of the wall of the ancient Berrhoea, founded on the rocky bank of the rivulet, and apparently one of the lower angles of the inclosure of the city. Verria, as the name is pronounced, or Βέρροια, as it is still written, stands on the eastern slope of the Olympene range of mountains, about five miles from the left bank of the Vistritza or Injekara, just where that river, after having made its way in an immense rocky ravine through the range, enters the great maritime plain.

§ 3.291 The territory produces corn and maize in the lower plain, and at the foot of the mountain hemp and flax, which are supplied with the necessary irrigation from the rivulet on the northern side of the town. This stream, which has its origin in the mountains to the westward, emerges from a rocky gorge in them, falls in cascades over some heights which rise abruptly above the town, and after turning several mills, rushes down the mountain between steep rocky banks to the bridge, over which we crossed it, and from thence into the plain. The town contains about 2000 families, of which 1200 are Greek: the houses are lofty, and for Turkey well built. Water flows through every street, supplied either from springs or from the rivulet; which advantage, together with the lofty and salubrious situation, the surrounding gardens, many fine plane-trees interspersed among the houses, the vicinity of the mountains, and a commanding view over the great level to the eastward, renders Varia one of the most agreeable towns in Rumili. The manufacturing part of the population spin the hemp and flax grown at the foot of the mountain, and make shirts and towels, particularly the makrama, or large towel used in the public baths, and of which there is a great consumption in all Turkish towns, four of them being required for each bather, besides two more for sheets to the bed on which he reposes after the bath. Many of the water-mills around the town are for fulling coarse woollens and carpets, which are made in the surrounding villages or by the Jews of Saloniki.

§ 3.292 The remains of the ancient Berrhoea are very inconsiderable. I have already noticed that which appears to be the north-western angle of the walls or perhaps of the acropolis; these walls are traceable from that point southward to two high towers towards the upper part of the modern town, which appear to have been repaired or rebuilt in Roman or Byzantine times, as the large quadrangular stones of which the work is partly constructed are mixed with mortar, tiles, and fragments of ancient monuments. I can discover only three inscriptions at Verria: in one, Popillius Summus the younger is honoured by the council and people; the other two are sepulchral monuments, one of which was erected by Annia Epigone, in memory of her son Flavianus, and her grandfather, who is not named; the other by Porus, son of Ammia, to Caius Scirtius Agathocles, his son, and Scirtia Zosime, his wife, who are styled heroes of virtuous life. In this inscription we have another instance of the Macedonian custom of recording in some cases the mother’s name instead of the father’s; and it is remarkable that one of the Politarchons of Thessalonica was also the son of an Ammia. In the plain below Verria, at no great distance, are two barrows, or tumbe, as the Turks call them. The name Vistritza, which is applied by the Greeks to the Haliacmon, although betraying a Sclavonic modification in its termination, may possibly be a corruption of Astraeus, for we learn

§ 3.293 from Aelian that there was a river called Astraeus, flowing between Thessalonica and Berrhoea, which although not a very correct description of the Vistritza, inasmuch as this river is not crossed on the road from Saloniki to Verria, would be still less suitable to the Moglenitiko, or to the river of Vodhena, as lying so far to the right of that line, or indeed to any but the two great streams which we know to have been anciently named Axius and Lydias. Perhaps Haliacmon was the ordinary appellation of the river above the gorges of Berrhoea, and Astraeus below them: in the same manner as Injekara and Vistritza are used in the present day. The river is noted at Verria for guliani of immense size. I before remarked that the same fish grows to enormous dimensions in the lake at Kastoria, which is one of the sources of the Vistritza. The district of Verria contains about 300 villages, extending eastward nearly to the Lydias, or Karasmak, and to the west to Sarighiul. To the south the village of Kulindros, standing on the heights which terminate the plains at their southern extremity, not far from the gulf, formerly belonged to Verria, but is now enumerated among the villages of Elassona. The voivoda of Verria is Halil Bey of Grevena, who lived here many years as kharatji, or farmer of the Christian capitation tax, and upon the death of Osman Aga, a short time ago, obtained the government, having first secured

§ 3.294 the approbation of Aly Pashi, whose influence is thus established in Verria. Though the Verriotes suspect Aly to have been sometimes instrumental to their having been annoyed by the thieves in order to make the necessity of his own services manifest to the Porte, they are so far satisfied with the result as to agree in commendation of the police of Metjobon, and to admit that all this part of Macedonia now enjoys great security: nor has Aly yet ventured to lay any heavy contributions on a place which is at the farthest extremity of the country under his influence, and the revenues of which are attached to the imperial family. His encroachments in this quarter have, however, created a panic, and there are now several large houses in the town of which the building has suddenly been suspended. Dec. 3.—In the afternoon I receive a visit from Metjobon, who here assumes the Turkish name of Mehmet Bey: he is a little spare man, of simple Albanian manners and mild address, and is said to be gifted with a remarkable share of promptitude, coolness, and sagacity. He showed great ability lately in his proceedings against the robbers, most of whom he made prisoners. In this part of Macedonia it is customary for the keepers of wine-houses to suspend an evergreen bush before them, being the same as the old English custom, from whence the proverb, “Good wine needs no bush.” In the southern parts of Greece, it is generally a long stick with shreds of painted paper on a string.

§ 3.295 I have frequently had occasion to notice the extraordinary celerity of some of the pezodhromi, or foot-messengers in Greece. A celebrated one of Verria may compete with any of them. He carried letters on foot to Saloniki in seven hoars, remained there one hour, and returned to Verria at the end of the fifteenth hoar. After having performed this feat more than once, he was commonly known to the day of his death by the name of Anemos, an adjunct as honourable to a courier as Africanus to a Scipio. Dec. 4.—The weather, which has been fine, with a northerly wind, ever since the day of my arrival at Saloniki, as well as on the road from thence, is said to have been the reverse at Verria for several days, and last night the rain fell heavily. At 6.30, Turkish time, Ι set out for Kozani, accompanied by one of Aly Pasha’s tatars, a guard of six Albanians supplied by Metjobon, and Musa Pasha’s tatar, who has accompanied me from Saloniki. We begin immediately to ascend the hills at the back of the town, and soon enter a narrow vale watered by the stream which descends to the town. At the upper end! of this valley, at 8.4, stands the derveni, a straw hut for lodging the Albanian guard, from whence we begin to ascend Mount Bermium, in defiance of the assertion of Herodotus, that it is impassable, and although the historian has every possible advantage in the season, and weather, that of last night having covered

§ 3.296 the mountain with snow to a great depth. Very soon after entering a forest of large chestnut trees, we arrive, at 9.40, at Kastania, a small village, of which all the houses, except two or three, are now deserted, in consequence of the demands for provisions, which were alternately made upon them by the robbers and their Albanian opponents. Aly Pasha endeavours to encourage their return, and declares his intention of building here a large village, with kules on the mountain for his soldiers, and thus to secure to himself this important pass between Lower and Upper Macedonia. The mountain abounds with wolves, wild boars, fallow deer, and roes. The swine are killed for the sake of their skins, which are in request for making shoes. A peasant informs me that not long since he shot one of these animals in the woods, which weighed 90 okes. The flesh of the roe is esteemed by these people, but not that of the deer. Dec. 5.—We leave Kastania at 3.5, Turkish time. The snow continued to fall during the night, but the weather has now become bright and calm, with a hard frost. As we advance the woods are of birch, in the highest parts of beech, and amidst them numerous traces of the wild animals are observable. On the summit, which is not more than three miles in a straight line from the Vistritza, we leave the highest point of the mountain

§ 3.297 now called Dhoxa, or more commonly Xerolivadho, from a village of that name which once stood near it, six or eight ibilea on our right, and descend to Khodova, a village of about 50 Turkish families, from whence there is a further descent of about three miles to the Vistritza, which is seen from our road. There is no passage to the same point from Verria along the river, as both banks are here bordered by impracticable precipices. Above those on the right bank are the villages of Kokova, Katafyghi, and some others, from which the mountain rises to a lofty summit, one of the Olympene chain, and separated only from Olympus itself by the elevated pass of Petra. To the northwestward of the mountains the Vistritza is again seen flowing in a valley which extends to Servia. Katafyghi is on the shortest route from Verria to Servia, which crosses the Vistritza near Verria, but in some parts is so difficult that the pass of Kastania is often preferred. Having passed Kokova at 5.10, we descend from thence along a narrow valley, which at the end of an hour conducts into the plain of Budja. To the left this plain is separated from those of Tjersemba and Servia on the banks of the Injekara, by a low root of Mount Bermium, which is connected at the other end of the plain of Budja with the mountain of Kozani, which is a branch of Mount Burino. The highest and middle point of these lower heights is called by the Turks Ghioztepe, a name analogous to the Greek Skopo, and meaning a point which commands an extensive view.

§ 3.298 The plain of Budja widens as we advance, and contains many small Yuruk villages, situated at the foot of the mountains on either side. To our right a root of Mount Dhoxa, advancing to the westward, leaves only a space of two miles between it and a similar projection of the mountain of Siatista; but beyond the opening the level again widens into the more extensive plain of Sarighiul. A little on this side of the opening stands the small Turkish town of Djuma, which contains a bazar, and is the market town of a district of small Turkish villages. The plains of Ostrovo, Sarighiul, Djuma, and Budja, seem, with the enclosing mountains, to have formed the ancient Eordaea. At 6.50 we halt to dine at a rising ground in the plain, spreading carpets and capote on the snow, which still lies here though the sun is now hot; then proceeding at 7.35, leave soon afterwards Djuma two or three miles on the right, and at length arrive in the lowest part of the plain, in which there is no longer any snow. The plain is fertile, and well cultivated with corn. The entrance of the Boghaz of Siatista appears at a distance of seven or eight miles on the right. At 9.20, having arrived at the end of the plain of Djuma, and passed a little to the right of several small Turkish villages situated at the foot of the hills of Ghioztepe, we turn to the left of our former course, through a narrow passage between the Ghioztepe range and some other small hills connected with the mountains near Kozani. At the entrance of the opening stands a khan and a small Turkish village called Sulinaria: half an hour further begins an undulated country, which

§ 3.299 extends on the right to Kozani, and the mountains behind it, and descends to the left to the Vistritza; at 10.30 we arrive at Kozani, vulgarly pronounced Kodjani. This is a town of six or seven hundred houses, with a good bazar and a market on Saturday for the neighbouring country: formerly it had a considerable commerce with Hungary and Germany, and several opulent merchants resided here. My lodging, which belonged to one of them, is constructed like the houses at Siatista, with thick walls, and apartments, which, though smaller, are more commodious than those in ordinary Greek and Turkish houses. There is a cellar below the house for the wine, which is here made from an extensive tract of vineyards surrounding the town. The greater part of the Kozanite merchants, whom Turkish oppression, particularly that of Aly Pasha, has driven from hence, have settled in Hungary. Dec. 6.—The market this morning is much frequented by both Turks and Greeks from the neighbouring country. Kozani and Servia form one episcopal diocese in the province of Thessalonica; the bishop has a house in both places, and is now at Servia, but his ordinary residence is Kozani. At the foot of the steps of his house, is a square stone of the annexed form, which serves the bishop for a mounting-block when he rides out.

§ 3.300 It is an επιτυμβιος στηλη, erected in honour of one Cleopatra, by her husband Crispus, in union with his daughter Crispina: a square excavation in the upper surface may perhaps have supported a vase of stone. On two opposite sides of the stele, is a repetition of words, intended probably for an Iambic verse, and signifying “Farewell ye heroes: and fare thee well also traveller, and good journey to thee.” The plural form of ήροες appears to indicate that these two inscriptions were added after the death of Crispus and Crispina, and when they had been buried in the same sepulchre with Cleopatra. The sigma is rectangular, and there are several siglae or conjoined letters, a mode of engraving which seems to have been more common in Macedonia than in the southern provinces of Greece, but was probably seldom or ever employed even here, before the end of the first century of the Roman Empire, to which date the inscription may with probability be attributed. The monument having been discovered in one of the corn-fields above the village, where several small sepulchral marbles, with figures in relief, or other remains of antiquity, have also been brought to light, it is evident that Kozani occupies the position of an ancient town, though I search in vain for any other indications of it, such as town walls, or remains of architecture.

§ 3.301 Kozani is the native place of Dr. George Sakellario, translator of a part of the Voyage D’Anacharsis and some other works, which he undertook for the benefit of his countrymen. The comfortable residence in which I find his family, shews the sacrifice he makes, or rather is forced to make, in residing at Berat as physician to Ibrahim Pasha. His brother-in-law, Papa Kharismio, who is now residing at Kozani, is an author also, and has written a Pantheon for the use of the schools of Greece.

§ 3.302 CHAPTER 28 MACEDONIA, PERRHAEBIA THE plain or rather low undulated country included between the Vistiitza, the mountain of Kozani, Mount Burino and Ghioztepe, is called Tjersemba, a Turkish word, written by the Greeks Τζεσιμπάς. Its inhabitants are chiefly Turks, occupying small villages. The soil produces good corn, but it is more particularly noted for saffron, which is sent by land to Germany, by the merchants of Kozani and Tzaritzena. When the trade of Egypt was closed by the consequences of the French invasion, the saffron of this country was worth 80 piastres the oke, but it has now fallen to 50 and 40.

§ 3.303 The only other district which produces it, is that of Venja, on the opposite side of Mount Burino, and lying between Tjersemba and Grevena. The name Βύτίηο appears to belong, like Vistritza, to the ancient language of Macedonia, and may have been derived from the same root as Bora, Bermius, Bertiscus. Beyond Burino to the southward, is seen a ridge of nearly equal height, which takes a southerly direction towards Trikkala, and separates the waters of the Haliacmon from those of the Peneius. They are the mountains anciently called Cambunii, a word of which βοννος is obviously the root They form a continuation of the heights above Katafyghi, and at their foot, a few degrees to the right of the summit of Olympus, is seen the town of Servia, called Selfidje by the Turks, a name which they attach also to the entire district stretching along the right bank of the Injekara, opposite to Tjersemba. In Tjersemba there are said to be remains of antiquity in four places, but in none of them are they described as being formed of that beautiful masonry which is so distinguishing a mark of Hellenic works. This the Kozanites very justly account for, by the nature of the stone of the surrounding

§ 3.304 mountains, which being brittle and incapable of being hewn into large blocks, apparently obliged the inhabitants of this part of Upper Macedonia, who moreover were semi-barbarous before the time of Philip son of Amyntas, to build in a manner different from that of the Southern Greeks. The four ruins are: 1. At Ktinia, on the side of Mount Burino, where a height is crowned by a castle having a double inclosure, and thin walls. 2. At Kaliani, a small Greek village, three hours from Kozani, near the left bank of the Injekara, a little on this side of a boghaz leading from the valley of Tjersemba into that of Venja. Here are the remains of a building, of which my informant gave me a rude drawing. It was constructed with a double row of arches, of which the larger were supported by white marble columns, with Corinthian capitals, in bad taste. The building is in ruins on three sides, but the fourth still preserves the place where the statue is supposed to have stood. The arches have been walled to form it into a Greek church. The neighbouring fields are said to be strewed with broken pottery; coins also are often found, and sometimes small idols. 3. At Kesaria, about half way between Kozani and Servia, half an hour to the right of the direct road, are similar appearances, with fragments of marble and sepulchral monuments; and there are remains of the same kind also between Kesaria and Kaliani. So deficient are the ancient details of Macedonian geography, that no opinion can be given of these places, further than that one of them

§ 3.305 bore the common name of Κβισαρία, and that they were all subordinate towns of the Elimeia, for that Elimeia extended thus far to the eastward, and here bordered upon Eordaea and Pieria, seems evident from Livy, in a passage already referred to, where he relates that Perseus marched from Citium to the lake Begorrites in Eordaea, from thence to the Haliacmon in Elimeia, and on the following day into Perrhaebia, which lies immediately to the southward of Tjersemba on the western side of Mount Olympus, whence it is evident that the encampment of Perseus, previously to his entering Perrhaebia, was exactly on this part of the river. As it is equally manifest from other authorities that Elimeia extended westward to the range of Pindus, it may be defined as comprehending the modern districts of Grevena, Venja and Tjersemba. Of the three other subdivisions of Upper Macedonia, namely, Eordaea, Orestis and Lyncestis, Eordaea comprehended probably, as I have before remarked, the modern districts of Budja, Sarighiul and Ostrovo— Orestisthose of Gramista, Anaselitza and Kastoria— and Lyncestis, Filurina and all the southern part of the basin of the Erigon. These seem to have been all the districts which properly belonged to Upper Macedonia, the country to the northward, as far as Illyria westward, and Thrace eastward, constituting Paeonia, a part of which (probably, on the Upper Axius ) was a separate kingdom as late as the reign of Cassander, but which in its widest

§ 3.306 sense enveloped on the north and north-east both Upper and Lower Macedonia, the latter containing the maritime and central provinces, which were the earliest acquisition of the kings, namely, Pieria, Bottieeis, Emathia, and Mygdonia. Even a part of these was occupied by Paeonians before the establishment of the Macedonian monarchy. Paeonia extended to the Dentheletae and Maedi of Thrace and to the Dardani, Penestae and Dassaretii of Illyria, comprehending the various tribes who occupied the upper valleys of the Erigon, Axius, Strymon, and Angitas, as far southward as Sirrhae inclusive. Its principal tribes to the eastward were the Odomanti, Aestraei and Agrianes, parts of whose country were known by the names of Parstrymonia and Paroreia, the former containing probably the valleys of the Upper Strymon and of its great tributary the river of Strumitza (Aestraeus?) the latter the adjacent mountains. On. the western frontier of Poeonia, its subdivisions bordering on the part of Illyria inhabited by the Penestae and Dassaretii were Deuriopus and Pelagonia, which together with Lyncestis comprehended the entire country watered by the Erigon and its branches. The respective limits of these subdivisione were not well defined, nor in all ages the same. Strabo considered Pelagonia, as well as Lyncestis, a division of Upper Macedonia, but as Stobi is described by other authors sometimes as a city of Paeonia, and Sometimes of Pelagonia, as Stymbara, another important place on this frontier of regal Macedonia is stated by some as belonging to Deuriopus, and by others

§ 3.307 to Pelagonia, and as Bryanium, placed by Strabo in Deuriopus, was near the passes leading into Eordaea, and consequently in Lyncestis, it is evident that no exact definition of these districts prevailed, at least among the ancient writers whose works have reached us. Lyncestis, although originally a part of Paeonia, having become a separate kingdom, which was annexed to Macedonia as early as the reign of Philip, son of Amyntas, may, with reference to a later period, be ascribed to Upper Macedonia; at the same time that all beyond it, to the sources of the Erigon, was still a portion of Paeonia, the whole of which, however, was united to regal Macedonia before the Macedonic wars of Rome. There is no occurrence in ancient history which better illustrates the ancient geography of that part of the cotintry than the operations of the consul Sulpicius against Philip, in the campaign of the year B.C. 200 Philip, who flattered himself that he should be able to deprive the Romans of the assistance of the Aetolians and Dardani, had for the purpose of preventing the entrance of the latter people into Macedonia, stationed his son Perseus in the passes of Pelagonia, when the consul having marched from Apollonia of Illyria through Dassaretia into Lyncestis, there encamped on the banks of the Bevus, and from thence sent foraging parties into Dassaretia, where the corn of the open country had already enabled him, on passing through that district, to save the supplies which he brought with him from his winter quarters.

§ 3.308 One of his parties having suddenly encountered a body of Philip’s cavalry who were in quest of information, an action ensued, with a loss nearly equal on both sides. Upon learning the force and position of the enemy, Philip found it prudent to recal Perseus from the passes of Pelagonia, and having thus brought together 20,000 men, he occupied a height distant only 200 paces from the Roman camp, and which he fortified with a ditch and rampart. On the third ensuing day, the consul having drawn forth his line at a distance of 500 paces from the enemy, Philip ordered out 700 of his cavalry, attended by the same number of light infantry; these the enemy met with an equal body of horse and foot, and obtained an advantage, the Greeks having shown themselves, in both kinds of force, inferior in firmness to the Romans, and the velites of the latter being much better armed than the Illyrians and Cretans who accompanied the Macedonian cavalry. Two days afterwards, Philip equally failed in drawing the enemy into an ambuscade of peltastae, whom he had stationed during the previous night in a position between the two camps. On the following day Sulpicius drew out his whole army, with elephants in front, and offered battle to the king, when the latter, not accepting the defiance, the consul moved his camp 8 miles to Octolophus, for the sake of being able to forage in greater security than could be done while the enemy’s camp was so near.

§ 3.309 The armies remained inactive in their respective positions until the Roman foragers had become negligent of their security, when the king advancing suddenly with all his cavalry, and some Cretan infantry, cut off the Roman foragers from their camp, and slew many of them. The consul, upon being made acquainted with the occurrence, advanced his legions in a close column and sent forward his cavalry, who came to action with the king. At first Philip had the superiority, but at length he was defeated, and lost 300 horsemen, of whom a third were made prisoners, and the rest were killed or perished in some neighbouring marshes. The king himself -was nearly taken, having wandered for some time in the marshes before he recovered hie camp. He now resolved upon a retreat, being partly actuated by the report that the Dardani, under Pleuratus, were approaching. He concealed this intention from his adversary by a proposal for a truce to bury the dead, and by lighting fires in his camp at night, while he was retiring towards the mountains. The consul remained several days in the same position, ignorant of the enemy’s movements, when, having exhausted the supplies of the neighbouring country, he removed to Stymbara, and from thence, after having collected the corn from the fields of Pelagonia, to Pluvina, still ignorant of the motions of Philip, who had in the meantime encamped at Bryanium, and having better information of his opponent’s proceedings, alarmed the Romans by suddenly approaching them, but did not venture to bring on an action.

§ 3.310 The Romans then proceeded to encamp on the river Osphagus, while Philip entrenched himself at no great distance on the bank of the Erigon, when, perceiving that the Romans intended to cross the mountains into Eordaea, he retired, and fortified the passes with trees, stones, ditches, and ramparts. But from these works he derived little benefit. The Romans forced or turned them without difficulty, chiefly because the Macedonian phalanx was useless and unmanageable in such a narrow and rugged field of action. Philip having retired, the Romans ravaged the fields of Eordaea, entered Elimeia, and from thence moved into Orestis. Here the consul received the submission of Celetrum, and from thence, proceeding into Dassaretia, took Pelium, “a town conveniently placed for making incursions into Macedonia,” and having placed a garrison in this place, he returned with his captives and plunder to Apollonia. This narrative, extracted undoubtedly from Polybius, seems so clear, that a traveller commanding sufficient leisure and security might hope to determine the position of the first encampment of Sulpicius as well as that of Octolophus, to identify the branches of the Erigon, named Bevus and Osphagus, and perhaps to ascertain the sites of Pluvina, Bryanium, and Stymbara. In this he would be greatly assisted by the evidence which the Itineraries have left us of the position of Heracleia, the chief town of Lyncestis. As the historian states the first encampment of the Romans

§ 3.311 to have been at Lyncus, on the river Bevus, and as Lyncns is described as a town by Stephanus, it might be supposed that Heracleia was sometimes called Lyncus, and that the camp of Sulpicius was at Heracleia itself. But notwithstanding the words “ad Lyncum ” seem to favour this opinion, it is more likely that Polybius employed Lyncus on this occasion in the same sense which we find attached to it in two other passages of Livy, as well as in Thucydides and Plutarch; that is to say, as synonymous with Lyncestis, or the country of the Lyncestae, once a small independent kingdom, and afterwards a province of the Macedonian monarchy. Lychnidus and Heracleia lying nearly in the line between Dyrrhachium, or Apollonia, and Thessalonica, were the principal places in the centre of the Candavian or Egnatian way—the great line of communication by land between Italy and the East, between Rome, Constantinople, and Jerusalem. A road of such importance, and on which the distance had been marked with milestones soon after the Roman conquest of Macedonia, we may believe to have been kept in the best order, as long as Rome was the centre of a vigorous authority; but it pro bably shared the fate of many other great establishments in the decline of the empire, and especially

§ 3.312 when it became as much the concern of the Byzantine as of the Roman government. Of this we discover some strong symptoms in the Itineraries; for although Lychnidus, Heracleia, and Edessa, still continued, as on the Candavian way described by Polybius, to be the three principal points between Dyrrhachium and Thessalonica (nature in fact having strongly drawn that line in the valley of the Genusus, branching from the maritime country of Illyria, and penetrating Mount Candavia in the same easterly direction in which the vale of the river of Edessa issues into the plains of Lower Macedonia) there appears to have been a choice of routes over the ridges which contained the boundaries of Illyria and Macedonia, and which separate the lake of Lychnidus from the valleys watered by the Erigon and its branches: a strong

§ 3.313 indication that the great Roman work was out of repair. In the original road described by Polybius, the portion between Lychnidus and Heracleia led through Pylon, which received that name from its being the limit of the two provinces. The Antonine Itinerary gives two routes in this part; one passing through Scirtiana (Scirtonia?) and Castra, the other through Nicia (Nicaea?), which is the same as that in the Tabular Itinerary. In the Jerusalem the road passes through Brucida (Brygiada, i. e. Brygias?) and Parembole. Now there seems little doubt that these names Castra, Parembole, and Nicaea, have reference to the military transactions of the Romans in Lyncestis, who not many years after, those events constructed a road, which happened to pass exactly over the

§ 3.314 scene of the former exploits of their army. Castra or Parembole, therefore, indicates the first encampment of Sulpicius on the Bevus; and Nicaea the place where he obtained the advantage over Philip’s cavalry, near Octolophus, which was eight miles distant from the first encampment: consequently, Nicaea was about eight Roman miles from Parembole or Castra—and probably to the northward of it, because after the battle near Octolophus, the consul proceeded in a northerly direction to Stymbara, in search of provisions, having already exhausted the country around Heracleia. It appears, therefore, that Nicaea, Parembole, and Heracleia, formed a triangle, of which the sides were 8, 11, and 12 M.P. in length; that the northern route from Lychnidus descended upon Nicaea, or Octolophus, and the two southern upon Parembole, or Castra, on the river Bevus: this was evidently the southern branch of the Erigon, near the issue of which into the plains Heracleia might be sought for, and nearer to its sources the town of Beve. As to the route described by Polybius through Pylon, the names which he mentions being of much earlier times than those in the Itineraries, it is very possible that the former route may have coincided with one of the latter, notwithstanding the difference of names. The pass over the mountains which separated Lyncestis from Eordaea, where Philip made his unsuccessful stand against the Romans, is described by Polybius as αι εις την’Εορδαίαν υπερβολαι, and Thucydides terms a defile in the same mountains

§ 3.315 η εισβολη της Λύγκον, in relating the attempt of Perdiccas against Lyncestis, in the eighth year of the Peloponnesian War, which ended in a separate negotiation between his ally Brasidas and Arrhibaeus king of the Lyncestae. It was by the same pass that Brasidas, in the following year, effected a skilful retreat from the Lyncestae and Illyrians, when, having descended into the plains of Lyncus with Perdiccas and a joint force, composed of 3000 hoplitae, 1000 cavalry, and a large body of barbarians of Thrace, they were obliged to retreat in consequence of the Illyrians, who had promised to join Perdiccas, having suddenly ranged themselves on the side of Arrhibaeus. The Macedonians of Perdiccas, and the undisciplined barbarians, having taken the alarm, moved tumultuously in the night, and rendered it necessary for the king himself to accompany them without communicating with Brasidas, who was stationed with his forces at some distance. Thus abandoned, the Spartan general began his retreat on the following morning towards the pass, forming his hoplitae in a square, placing his light-armed within it, and covering the retreat of this body with 300 chosen men under his own command. He thus not only resisted the attacks of the enemy, but having seized upon one of the heights which bordered the entrance of the pass, prevented them from intercepting him in it. He was then allowed to retreat without farther molestation, and arrived the same day at Arnissa, the first town in the territory of Perdiccas. Arnissa, therefore, seems to have been in the vale of

§ 3.316 Ostrovo, and possibly it may have been the same place as the Barnus of Polybius, B being a common Macedonian prefix; for the words of Strabo are not imperative in placing Barnus between Lychnidus and Heracleia, although bearing undoubtedly that interpretation. It is from the remark of Polybius that the Candavian way passed through the country of the Eordaei, in proceeding from that of the Lyncestae to Edessa, together with the historical authorities just referred to, and that other passage in the Latin historian, wherein he describes the march of Perseus from Citium in Lower Macedonia, through Eordaea into Elimeia, and to the Haliacmon, that we obtain a knowledge of the exact situation of Eordaea, which thus appears to have extended along the western side of Mount Bermius, comprehending Ostrovo and Katranitza to the north, Sarighioli in the middle, and to the southward the plains of Djuma, Budja, and Karaianni, as far as the ridges near Kozani and the Klisura of Siatista, which seem to be the natural boundaries of the province. The only Eordaean town noticed in history is Physcus, of which Thucydides remarks, that near it there still remained some of the descendants of the Eordaei, who had been expelled from all other parts of Eordaea by the Temenidae. But there is some reason to add to this name those of Begorra and Galadrae as Eordaean towns, the Begorrites locus, to which Perseus marched from Citium, having probably been so called from a town of Begorra; which stood

§ 3.317 perhaps at Kaliari, by the Turks called Sarighiul, the central and otherwise advantageous position of which leads also to the conjecture that it may have been the city Eordaea of later times. As Lycophron couples Galadrae with the land of the Eordaei, and as Stephanus attributes that town to Pieria, it might best be sought for at the southern extremity of Eordaea, towards the Haliacmon and the frontiers of Pieria, its territory having consisted chiefly perhaps of the plains of Budja and Djuma. If Galadrae was in the southern part of the province, and Begorra in the middle, Physcus was probably to the northward, about Katranitza, towards the mountains of the Berman range, such a situation being the most likely to have preserved the ancient race. The modern routes over the mountains which separated Lyncus from Eordaea, are, from Tilbeli to Oslova, to the eastward, and from Banitza to Ostrovo to the westward: the former is in the ordinary route from Bitolia to Vodhena; the latter from Filurina to the same place. Although Filurina is nearer than Bitolia to the site of Heracleia, I should conceive the Egnatian Way to have crossed by the former route, as it descends into

§ 3.318 the Eordaean valleys nearer to the situation of Edessa. The only place which the three Itineraries agree in placing between Heracleia and Edessa, is Cellae, but the distances given are too conflicting to lead to any certainty as to its position. At or near Banitza are. the mineral acidulous waters of Lyncestis, much renowned among the ancients, who imagined that they possessed intoxicating qualities; they were noticed by Dr. Browne in the year 1669. Although Livy employs the name Pelagonia in his narrative of the campaign of Sulpicius only as that of a large district containing Stymbara, it is evident from his account of the division of Macedonia into four provinces after the Roman conquest, that if not at the former period of time, thirty-three years later at least, Pelagonia was the appellation of the chief town of the Pelagones, which then, became the capital of the Fourth Macedonia. It was perhaps not specifically employed as the name of a town until the two other cities of Pelagonia were ruined: for that Pelagonia or a portion of it once contained three we may infer from the adjunct Tripolitis given to it by Strabo, who also shows, if I rightly apprehend his meaning, that one of the

§ 3.319 three towns bore the same name as the Azorus of Perrhaebia Tripolitis. The name Pelagonia still exists as the designation of the Greek metropolitan bishopric, of which the see is Bitolia, or Monastiri, which latter Greek name the Turks have adopted. Bitolia is now the chief place of the surrounding country, and the ordinary residence of the governor-general of Rumili. At or near the town are many vestiges of ancient buildings of Roman times. These the natives suppose to have belonged to a city named Tripolis: a tradition

§ 3.320 which accords with the existence of a Pelagonia Tripolitis as attested by Strabo, and which is not adverse to the identity of Tripolis with the city Pelagonia of Livy, since it is easy to conceive that after the redaction of the two other towns of the Tripolitis (and Strabo asserts that all the towns on the Erigon, Stymbara included, were ruins in his time), the surviving city may have been known by the name of Tripolis, as formed from the three former towns, and that it may also have been often known by the name of the district, Pelagonia. Bitolia being a word of Greek origin, may possibly be a corruption of a third name of the same place, or that which the city bore when the three towns of Pelagonia still existed: the Hellenic name most resembling it is Epitalia. The passes of Pelagonia, in which Perseus was stationed by his father Philip, Ι take to have been the passage over the mountains in the modern

§ 3.321 route from Akhridha to Bitolia, which now forms the main communication instead of the old line or lines of the Via Egnatia, that change having probably been caused by the circumstance that A khridha and Bitolia being now the chief places instead of Lychnidm and Heracleia, and lying respectively to the northward of the. two ancient places, have caused the road to assume a more northerly line in this part, and which has occurred the more easily, as anciently the Egnatia was here diverted from its direct line by the necessity of passing round either the northern or southern end of the lake Lychnidus, and had no advantage therefore in shortness over the present line. The pass of Pelagonia was of great importance as one of the direct entrances from Illyria into Macedonia by the course of the river Drilon, now called Drin. Hence it was necessary for the kings of Macedonia to maintain strong garrisons in Lychnidus and some other positions on the lake, as well as in Stymbara and Heracleia. By means of these garrisons. and the strength of the frontier, the kingdom was not so liable to invasion here as on the side of Scupi, which commanded the entrance from Dardania into the plains of the Upper Axius, and which place having been generally held by the Dardani, gave them great facilities of offence against Macedonia. Stymbara or Stubera appears from Polybius and Livy to have stood in the most fertile part of the country, to the northward of Bitolia;

§ 3.322 a situation which accords with its having been the place from whence Perseus marched in three days to Uscana, the chief town of the Penestiana, situated probably on the Drilon, at or near the modern Dibre. Stymbara would seem to have been near Prfllapo, by the Turks called Pyrlepe, and Pluvina, between Stymbara and Bryanium which was not far from the passes leading into Eordaea. If Strabo is correct in naming Alcomenae as a town on the Erigon, its situation appears to have been above Bryanium, for below that town, or between it and the junction of the Erigon with the Axius, the Tabular Itinerary shows that we ought to place Euristus (the orthography is not quite certain) and Stobi. By Ptolemy both these towns are ascribed to Pelagonia, and by other authorities Stobi is designated a city of Paeonia; but these, and some other conflicting testimonies of the same kind, are reconciled, if we admit that Deuriopus was sometimes considered a subdivision of Pelagonia, smd the latter sometimes a subdivision of Paeonia. I have already remarked how exactly Livy’s description of Celetrum, as well in relative position as in its situation on a peninsula in a lake, agrees with Kastoria. By means of this datum we have the exact course of the march of Sulpicius on his return from Pelagonia into Dassaretia. From Eordaea or Sarighioli he crossed a part of the plain

§ 3.323 of Grevena, and through Anaselitza to Kastoria, from whence his route to Pelium in Dassaretia could have been no other than through the pass of Tzangon, which, being the only interruption in the great dorsal ridge of Northern Greece, was undoubtedly one of the most frequented of the communications between the two sides of the country, and particularly from Orestis into Dassaretia. It was precisely near Pelium that Arrian describes a remarkable pass, through which flowed the Eordaicus, leaving in one part space only for four shields abreast; a description which corresponds so exactly with the pass of Tzangon, both as to the river and the breadth of one part of the pass, that the identity can hardly be questioned. Pelium was situated at the foot of a woody mountain, near the pass; a description which may be applied either to Pliassa or to Poyani, but the former has the preference by its name, which seems to be a vulgar sounding of Πηλιάσσα. The march of Alexander in approaching Pelium, as well as his subsequent progress to Pelinnaeum in Thessaly, may furnish some further illustrations of the relative chorography. He was returning from an expedition against the Gets, who dwelt beyond the Danube, and had arrived in the country of the Agrianes and Paeones, when he received intelligence that Clitus and Glaucias, who shared between them all maritime Illyria, had declared against him, and had prevailed upon the Autariats to attack him on the route.

§ 3.324 But Langarus, king of the Agrianes, having frustrated the latter design by invading the country of the Autariatae, Alexander was enabled to march without interruption along the Erigon, and from thence to Pelium, near which the Illyrians were encamped. After some operations which are not very clearly described, he surprised the Illyrian camp in the night, when Glaucias fled, pursued by Alexander as far as the mountains of the Taulantii, while Clitus retired into Pelium, from whence, after having burnt the city, he proceeded to join Glaucias in Taulantia. Soon after this event Alexander received advice of the revolt of Thebes, when, crossing Eordaea and Elimeia, and passing the mountains of the Tymphaei and Paravaei, he arrived in seven days at Pelinnaeum in Thessaly. Without the comparison afforded by Livy’s account of the proceedings of Sulpicius, it might be supposed from the circumstances stated by Arrian, that Pelium was not far from the Erigon, or the name Eordaicus might lead to the impression that Pelium was in Eordaea, instead of having been upon a river which flows to the western coast. It is clear, however, that Pelium was not far from the mountains of the Taulantii, a people who occupied the plains extending to the western coast. Again, it might be thought that Alexander marched from Pelium to Pelinnaeum by the most direct route; but se in that case he would not have passed through any part of Eordaea,

§ 3.325 the historian has probably omitted to mention that Alexander returned home to Pella before he received intelligence of the revolt of Thebes: on which supposition the road to Pelinnaeum would have led through the centre, first of Eordaea and then of Elimeia, as Arrian relates. If the situation of Pelium as deduced from the combined evidence of Arrian and Livy be correct, it will follow that Dassaretia comprehended not only the great valley which contains the lake of Lychnidus, but also the plain of Korytza: and that plain being an extensive corn country, the inference accords with that abundance of grain in Dassaretia which enabled Sulpicius to save his own stock while he passed. through that district, and which induced him afterwards to send back his foragers thither, though he was encamped in an equally fertile plain, but of which he had not the same military possession. The western part of Dassaretia was a contrast to the eastern, consisting entirely of lofty and rugged mountains intersected by branches of the river Apsus: its extent was very great. If Berat be the site of Antipatria, as I have shown some reason for supposing, it will follow that the Dassaretae possessed all the mountainous country lying between Korytza and Berat, beyond which latter the frontiers of the Dassaretae met those of the Taulantii, Bylitmes, and Chaones of Epirus. On the north they bordered on the Eordeti and Penestae, and partly on the Taulantii, while to the eastward the crest of the great central ridge very naturally formed the line of demarcation between them and

§ 3.326 the Pelagones, Brygi, and Orestae, or in other words, between Illyria and Macedonia. It results from these boundaries that Dassaretia was not less than 60 miles in length, and as much in breadth, an extent such as we are in some measure led to expect from Polybius, who in addition to the towns on the lake of Lychnidus, represents the Phebatae, Pissantini, Calicaeni, and Pirustae, all as tribes of Dassaretia. The situation of some of these tribes may be deduced from the testimony of the same author, as preserved in the Latin text of Livy. When Sulpicius was encamped on the Apsus between Dyrrhachium and Apollonia, before he advanced into Lyncestis, he sent Apustius against the neighbouring possessions of Philip. Corragum, Gerrunium, and Orgessus, were captured, not without resistance; after which, Apustius laid siege to Antipatria, a large city in a narrow pass remarkable for the strength of its position and walls. Having taken this place he slew the men, destroyed the walls, burnt the town, and gave up

§ 3.327 the plunder to his soldiers, which so intimidated the people of Codrion, that they surrendered to him, although their city was well garrisoned and fortified. Ilion,. another town, was taken by force, after which the Romans, in returning to Sulpicius loaded with plunder, were attacked at the passage of the river by Athenagoras, one of the most distinguished of Philip’s officers, but without suffering much damage. Gerrunium (Gertunium?) and Codrion seem to be the same places which in the text of Polybius are written Gertus and Chrysondion, for he names them together with Antipatria as frontier places which Scerdilaidas had taken from Philip, and which the latter retook in the second year of the Social war, B.C. 221. As Gerrunium and Antipatria were in Phoebatis, and Orgessus was a town of the Pissantini, it seems probable, assuming Antipatria to have been at Berat, that the Phoebatae chiefly inhabited the valley of the Uzumi, and the Pissantini that of the Devol; and that as Gertunium was attacked by Apustius before Antipatria, it was lower on the Uzumi than Berat, perhaps, near the junction of the two rivers.

§ 3.328 To the eastward of it on the Devol, may be placed Orgessus, and somewhat nearer than either to the camp of Sulpicius, Corragum the first named of the three. Codrion and Ilium, seem to have been in the valley of the Uzumi above Berat on the slopes of Tomor. This great mountain still bears probably its ancient name, of which the Greek form was Tomarus. It is easy to conceive that, like the names of mountains and rivers in general, Tomor was a generic word belonging to the aboriginal language of Epirus, and that hence it became attached also to the more celebrated mountain near Dodona. The ancient fortress near the modern village of Tomor may, like that village, have borne the same name as the mountain itself, according to a custom which seems to have been prevalent in Greece in every age. In the same chapter of Polybius just referred to, the historian proceeds to relate that Philip, after having recovered the three towns of Phoebatis abovementioned, proceeded to capture other places in Dassaretia, namely, Creonium and Gerions, (not the same place as Gertus,) and four towns on the lake Lychnitis, namely, Enchelariae, Cerax, Sation, and Boei, then Bantia of the Calicaeni, and Orgessus of the Pissantini. That the four towns on the lake were on its western shore, may be inferred from the Itineraries, but especially from the Tabular, which evidently followed the eastern side of the lake from the bridge of the Drilon to Lychnidus, and which makes no mention of any of the places named by Polybius. The same silence as to those towns may perhaps be considered as an argument to prove that all the three routes in the Itineraries led along the eastern shore, but it is very possible that one of them at

§ 3.329 least may have approached the southern end of the lake obliquely from the pass of Candavia, so as entirely to avoid the western shore. Ι am inclined to believe that the road in the Jerusalem itinerary passed round the southern end of the lake, and that Patrae was situated at that extremity. The Pirustae would seem to have been on the northern frontier of Dassaretia, as they joined the Taulantii and some other more northerly Illyrians, to assist the Romans in the reduction of Gentius. They probably occupied an intermediate tract between the Pissantini, on the lower part of the Devol, and the southern extremity of the lake Lychnitis, in which case there seems to remain only the plain of Korytza to the left of the Eordaicus for the situation of the Calicaeni. Possibly Korytza may be the site of Bantia. Dec. 6.—Quitting Kozani for Servia at 7.45, Turkish time, we leave Akbunar, by the Greeks called Nizvoro, or Izvoro, not far to the left, at the extremity of the vineyards of Kozani, then descend over downs covered with corn-fields, and inter-spersed with small villages, until at 8.45, Hadjiran, about the same size as Akbunar, is 1-1/2 mile distant on the left of the road at the foot of the Ghioz-tepe: all these places are Turkish. At 10.6 we arrive at the river Injekara, or Vistritza, which is bordered by white clifis along the left bank, and on the opposite side by low level ground: follow the sands on the bank of the river for nine minutes, then cross it in a broad flat-bottomed boat,

§ 3.330 capable of containing ten or twelve horses, and in an hour and 8 minutes from the river reach Servia, having passed over rich meadows and a fertile plain, beyond which is an ascent of 20 minutes to the town. Servia contains about 500 Turkish houses, and a few Greek. It is situated on the northern side of an opening, in the ridge which commences at the gorges of the Vistritza, near Verria, and terminates in the mountains of Khassik, to the north of Trikkala. The most valuable produce of the fields of Servia is a small species of tobacco, bearing a yellow leaf like that of Yenidje. The streets of the town are bordered with the herb which is hung to dry along the sides and galleries of the houses, as well as round the yards attached to all the better class of houses. Dec. 7.—The episcopal church of Servia, which stands on a height, rising from the lowest part of the mountain behind the town, is now in ruins, and the bishop’s house, which is in the town, is not in much better condition, though he still occupies it. The bishop, whom I visit this morning, supposes Servia to be a κτίσμα, or colony of Servians, whose descendants were driven out by the Turks, which is not improbable. Another opinion of hie holiness seems more questionable, though he advances it as a fact not to be disputed, and the honour of his see being concerned I do not contest it with him. He asserts that St. Paul passed through Servia on his way from, Berrhoea to Athens.

§ 3.331 Undoubtedly, if the apostle crossed Mount Bermius, Servia was in his way to Athens by Larissa, but it does not appear whether he went to Athens by sea or by land; and even if we suppose the words ώς την θάλασσαν to mean, that in order to elude his enemies he departed from Berrhoea to the coast “as if he intended to embark,” but that in reality he travelled by land, it is much more probable that he should have continued his way through Pieria and by the direct and level road of Tempe, or even by the pass of Petra, than that he should have made a circuitous journey over two ranges of mountains. Having dismissed the guards who were furnished to me by Metjobon at Verria, Ι take six others from Aly Pasha’s derventji at Servia, who is an Albanian Mussulman of Kolonia, and set out for Livadhi, first visiting a ruined castle on the summit of the hill above the episcopal church, and accompanied so far by the Albanian commandant, who when he finds that I have some knowledge of the distant objects in view from the castle, shows great satisfaction in answering all my geographical questions, for Which he is well qualified by his extensive knowledge of Macedonia, acquired in the course of his military services. AH Tjersemba is seen from hence, inclosed by Mount Burino and the βΐήοζ-ίβρέ; between which summits the mountain of Siatista shows itself nearly in a line with Kozani, and beyond it to the

§ 3.332 left Siniatziko; a little to the right of the latter Peristori is also seen, which looks down on the plains of the Erigon and Bitolia. To the north-eastward rises the great Dhoxi, or Bermius, and to the right of it is seen Velvedhos, or Velvendosa town of 300 houses, which, though conspicuous by its minaret, is chiefly inhabited by Greeks. Velvedhο is 3 hours distant from Servia, and similarly situated on the same mountain; it lies in a line with the great ravine of the Haliacmon, through the opening of which appears the mountain above Pella. The castle of Servia was so placed as to command the ascent to the Portes, as the highest point of the pass is called, which here conducts from the banks of the Haliacmon into the valleys watered by tributaries of the Peneiue. Being the most direct and easy passage across the Cambunian ridge, it is the natural gate between Macedonia and Perrheebia, and the position could not have been neglected by the ancients, though Ι have been unable to discover any Hellenic remains, either in the castle or town. It is now the most important station of the dervent Aga’s troops on the beylik, or post road from Larissa and Trikkala to Bitolia, the first post on which from hence is Kaliari, and the second Filurina. The road from the castle to the Portes is wide and level, and occupies the whole of a natural opening in the mountain.

§ 3.333 At the farther end of the Portes are vestiges of a fortification apparently of the same date as the castle, and once forming part of the same system of defence. The road to Trikkala follows the eastern foot of the mountain as far as another opening between it and a round hill on the left, where it enters the valley of one of the branches of the Titareskis. This round hill, which is visible through the pass of Servia from Kozani, is called Vigla, a modern word equivalent to Phyle, and is said to retain some vestiges of an ancient fortress. Instead of passing through the Portes, I pursue a higher track along the southern face of the mountain, which stretches northward to Katafyghi and the gorges of the Vistritza above Verria. As we ascend, the peak of Samaiina appears to the northwestward through the upper straits of the same river, or those which at the southern extremity of Mount Burino, near Kaliani, separate the plains or valleys of Grevena and Venja from those of Tjersemba and Servia. Our route all the way to Livadhi follows the side of the mountain, gradually ascending and crossing many deep ravines and rocky slopes of dangerous footing. At about half way we begin to look down to the right upon a plain which extends five or six miles from the foot of this mountain to another called Amarbes, in the direction of Dheminiko. Amarbes is the principal summit of the Cambunii montes: westward it is connected with another named Bunasa, which rises from the left bank of the Vistritza, opposite to Burino.