Chapter IV

OF THE ANCIENT PLACES ON THE ROAD FROM ADALIA TO SHUGHUT, INCLUDING REMARKS ON THE COMPARATIVE GEOGRAPHY OF THE ADJACENT COUNTRY.

_Ancient Authorities—Cotyaeium—Termessus—Lake Ascania—Milyas—
Cibyra—Selge—Pednelissus—Cretopolis—Lyrbe—Sagalassus—Cremna—
Lysinoe—Sinda—Isionda—Tabæ, Tiaba—Mender-su at Sandukli the
ancient Obrimas—Ancient Sites on the four Roads of the Table,
which cross the modern Route from Adália to Shughut—Themisonium—
Cormasa—Celænæ or Apameia—Eumeneia—Apollonia—Euphorbium—Conni—
Eucarpia—Acmonia—Cadi—Azani—Synaus._

I shall now submit to the reader some observations on the ancient geography of the route of General Koehler and his party from Adália to Shughut.

This road traverses a part of Asia Minor upon which ancient history throws little light. The text of Strabo is almost contradictory in regard to some of the principal places which lay near the road; and the itineraries supply no routes in this direction, though there are five in the Peutinger Table which intersect it.

The march of Alexander from Pamphylia to Gordium in Phrygia, as related by Arrian; and the description by Livy of the progress of the Consul Cneius Manlius in his Expedition from Cibyra into Pamphylia and from thence by Sagalassus to Synnada and into Galatia, are the only historical documents. As the passage of Livy is very detailed and was borrowed from Polybius[149], its information deserves more confidence than is usually due to that of a Latin author in regard to Grecian geography; and it may hereafter be extremely useful, when the ancient ruins, with which Pisidia and the adjacent districts are known to abound, shall have been more explored. In the present state of our knowledge of the country, it supplies not much positive information.

The only point in General Koehler’s route which can be considered absolutely certain is Cotyaeium. The position of that city in Phrygia Epictetus, not far from Nacoleia, and Dorylæum[150], agrees perfectly with that of Kutáya, the resemblance of which name to the Greek Κοτυάειον is still more striking when we observe the identity of accent.

There are two other places also in General Koehler’s route, upon the ancient names of which we cannot entertain much doubt. These are Termessus and the lake Ascania. The latter corresponds with the salt lake of Burdur; for Arrian relates that Alexander, after having reduced Sagalassus and some other strong places in Pisidia, passed by the lake Ascania in his way to Celænæ (afterwards Apameia), and that the water of this lake was so salt, that the inhabitants had no need of sea salt for domestic purposes[151]. The same fact is mentioned by the anonymous geographer of Ravenna. Perhaps this is the lake Ascanius, of which Pliny remarks, that the upper surface of the water was fresh, while the lower was nitrous[152].

The great ruins which General Koehler passed through at the ascent of the mountains, on the second day of his departure from Adália, seem to be those of Termessus, which, next to Selge, was the largest of the Pisidian cities, and was situated at the passes of mount Solyma, leading from the maritime plains through Milyas to the lake Ascania[153], and from thence to Celænæ. Milyas was the country of the more ancient Solymi[154]; and being also described by Strabo as the mountainous district, which extended from the passes of Termessus to the district of Apameia, it answers exactly to the elevated region which General Koehler traversed after he had mounted the pass which I have supposed the Termessian.

Between Milyas and the valley of the Mæander were Cabalis and the Cibyratis[155]. The latter district, which long flourished under the monarchy of a family named Moagetes[156], was a tetrapolis; the four cities were, Cibyra, which had two votes in the general council, Œnoanda, Balbura, and Bubon. The Cibyratis is clearly indicated by Strabo to have been situated between Lycia and the parts of the valley of the Mæander about Nysa and Antiocheia[157]; in the height of its prosperity, its dependencies extended from Pisidia and Milyas to Lycia and Peræa of the Rhodii[158]. Balbura and Bubon having been given to Lycia by Murena, on the reduction of the last Moagetes, and Œnoanda having been included in the same province, in the arrangement of Constantine[159], while Cibyra was ascribed to Caria, it may be presumed that Cibyra lay to the northward of the three other cities. This in some measure agrees with Ptolemy, who places Bubon, Œnoanda, and Balbura in a district of Lycia called Carbalia; under this name, as a part of Pamphylia, he ranges also Termessus, Cretopolis, and six other towns; Cibyra he places in Phrygia. Such are the data afforded by ancient history, to assist the traveller in discovering the sites of the four cities of the Cibyratis.

Polybius[160], in his account of the proceedings of Achæus, king of the provinces _within_ Taurus, against Antiochus the Great[161], has furnished a few data as to the situation of some of the towns on the frontiers of Pisidia and Pamphylia. In relating the operations of Garsyeris, commander of the army of Achæus, whose ostensible object was to assist the people of Pednelissus against the Selgenses, Polybius appears to apply the name of Climax to all the ridge of the mountains Solyma, from the summit called Olympus on the shore of the Gulf of Attaleia, to the great heights of Taurus. Garsyeris was at first unable to penetrate through the passes of Mount Climax, leading to Pednelissus, because they were occupied by the Selgenses, and particularly the pass of Saporda—a place not mentioned by any other author. We know from Strabo[162], that Pednelissus was situated inland from Aspendus; and it has been seen that the principal pass of the Solyma was commanded by the city of Termessus: Saporda, therefore, may perhaps have stood at another pass which leads over the ridge of Solyma from Adália in a W.N.W. direction to Dauas and Denizli. Cretopolis in Milyas, where Garsyeris encamped before he attempted the passes, is shown from this circumstance to have been on the western side of Mount Climax: and the Etennenses, who, together with the Aspendii, joined the party of Achæus against Selge, are stated by the historian to have inhabited the mountains above that city,—being thus obviously the same people as the Catennenses of Strabo[163]; who describes them as bordering on Selge and the Homonadenses.

Lyrbe, which, as well as Etenna, was still a bishopric in the ninth century[164], under the metropolitan of Side, seems, from some verses of Dionysius of Charax[165], to have stood between Termessus and Selge, a little above the maritime plains, among hills once covered with olives, but now affording little but pasture.

There is great difficulty in reconciling the authority of Arrian with that of Strabo in regard to the site of Sagalassus, otherwise called Selgessus[166], one of the most important cities and most fertile districts in Pisidia[167]; and which could not have been far from the route of General Koehler. Arrian, in a passage already referred to, seems to place it to the south of Burdur[168]; thus far agreeing with Strabo, who, after describing the cities on the southern side of Mount Taurus, just noticed, remarks that Sagalassus was _within_, or on the northern side of Taurus, near Milyas[169], which district, as he tells us in another place, extended northward as far as those of Sagalassus and Apameia[170].

Strabo further informs us[171], that Sagalassus was one day’s journey from Apameia; whereas Arrian relates that Alexander was five days in marching from Sagalassus to Celænæ, passing by the lake Ascania.

Nothing but an examination of this country by an intelligent traveller can clear up this difficulty, or explain the passage of Strabo cited in the note below; and for this purpose the ruins seen by Paul Lucas in this country, and the others heard of by General Koehler, probably contain ample materials. The remarkable site which gave name to Cremna[172] could hardly elude research; and it is the more likely to preserve some remains of antiquity, as having been a Roman colony.

If by the _lake_, mentioned in the march of Manlius, Polybius, from whom Livy has taken all this part of his history, meant the lake of Burdur, Lysinoe may have occupied the site of Burdur; or more probably some situation near the opposite end of the lake, where the future traveller may perhaps find the river Lyses, from which Lysinoe seems to have taken its name. And this might also lead to the discovery of the lake Caralitis and Sinda[173].

It is evident from the passage of Livy just cited, that Sinda and Isionda were different places, and not the same place as has sometimes been supposed. Livy seems to agree with Strabo in placing Sinda to the northward of Cibyra at the extremity of Pisidia bordering on Caria and Phrygia; whereas Isionda appears clearly to have been on the Pamphylian side of Termessus[174].

Dombai seems to be a corruption of Tabæ: hardly, indeed, a corruption, as it is no more than the hard and rustic pronunciation of the Greek word Τάβαι. The situation of Dombai accords very well with that which Strabo assigns to Tabæ, for he places it in the part of Pisidia adjacent to Phrygia and Caria[175], and names it among the cities which lay around Apameia and Laodiceia, which is precisely the position of Dombai[176]. The fertile plain which has obtained the name of Dombai-ovasi, or Valley of Dombai, corresponds equally with the Ταβηνὸν πεδίον, which, according to another passage of Strabo, lay on the confines of Phrygia and Pisidia[177]. It can hardly be doubted that Livy has incorrectly described Tabæ as situated on the frontier of Pisidia towards the Pamphylian sea[178].

The river called the Mender-su, which General Koehler crossed at Sandukli, seems to be that branch of the Mæander anciently called Obrimas, the fountains of which were something more than two days’ march from Synnada, and not far from Metropolis on the side towards Apameia[179]. The modern application of the name Mæander (slightly corrupted) to a stream which was anciently considered a tributary of that river, is another instance of those natural changes of geographical nomenclature, of which a similar example has already been given in the case of the river Sangarius.

It has already been remarked, that General Koehler’s route was crossed by five of the Roman roads marked in the Peutinger Table. These are, beginning from the southward, 1. From Laodiceia ad Lycum to Perge; 2. From Apameia Cibotus to Antiocheia of Pisidia; 3. From Apameia to Synnada; 4. from Apameia to Dorylæum; 5. From Philadelphia to Dorylæum.—The real situations of all these cities, except Antioch, being known with sufficient exactitude, those of the intermediate places on the several roads would also have been determined, had the distances in the Table been accurate; but unfortunately, like some of those to which I have already had occasion to advert, they are either imperfect or they are obviously erroneous, when compared with the map.

1. From Laodiceia ad Lycum to Perge, passing through Themisonium and Cormasa.—Although the direct distance is upwards of 100 G. M. there are only 46 M. P. marked in the Table, namely, 34 between Themisonium and Cormasa, and 12 from Cormasa to Perge. If these two distances were correct, therefore, the omitted distance between Laodiceia and Themisonium ought to be supplied with about 100 M. P. It is impossible to believe however that Themisonium, which is named by Strabo among the smaller towns around Apameia and Laodiceia[180], could have been so far to the south-east. Cormasa, on the other hand, must have been much more than 12 M. P. from Perge; for it appears from Livy that Cormasa was at a considerable distance from the borders of Pamphylia towards Lysinoe and the lake of Burdur[181]; which agrees with Ptolemy, who names it among the cities of Pisidia and next to Lysinia. The suspicion of inaccuracy in this route of the Table is confirmed by the negligences which occur on its continuation to Side; where the distance between Perge and Syllium is wanting, and where Syllium and Aspendus occupy each other’s places. Upon the whole, therefore, this route serves only to give us the line of Themisonium and Cormasa, the distance between which two places (34 M. P.) may perhaps be correct. And so far it may be an useful approximation to the traveller.

2. From Apameia to Antiocheia of Pisidia.—There cannot be a stronger proof of the little progress yet made in geographical discovery in Asia Minor, than the fact, that the site of Apameia still remains unexplored. Under the name of Celænæ, it was the capital of Phrygia; and in Roman times, although not equal in political importance to Laodiceia, which was the residence of the proconsul of Asia, it was inferior only to Ephesus as a centre of commercial transactions[182]. It appears from Pococke to have been at a place called Dinglar (or some such name), situated, as well as we can discover amidst the negligence and want of precision which are the usual characteristics of Pococke’s narrative, at 8 or 10 miles on the right of the road leading from Khónos to Ishékle[183], and about 16 miles[184] to the southward of the latter place. Pococke himself had no doubt that some remains of antiquity which he observed at Ishekle were those of Apameia; thus overlooking, or failing to decypher, an inscription which he copied at that place, and which clearly proves it to be the site of Eumeneia or Eumenia[185].

Eumenia was situated on the river Glaucus, as appears from an existing coin[186]. Pliny names the Glaucus, but places Eumenia upon the river Cludrus. Possibly this may have been the name of the sources of the Glaucus, those fine fountains which Pococke observed at Ishékle, and which may perhaps join another stream in or near the town.

As Eumenia is marked in the Table on the road from Dorylæum to Apameia at 26 M. P. from the latter, we have a presumption in this datum alone that Apameia was not far from Dinglar, the site of which modern place, relatively to the other chief ancient cities of Phrygia, is in conformity with that of Apameia, as described by Strabo[187]. Our knowledge of the peculiarities of the place itself is derived from Pococke and some recent travellers, who were informed that at the place called Dinglar or Dizla there are many remains of antiquity under a high hill which has a lake on the summit and a river falling down the face of the hill; for this description of Dinglar accords precisely with that of Celænæ as given by several ancient authors. According to Xenophon[188] the Mæander rose in the palace of Cyrus, flowing from thence through his park and the city of Celænæ: and the sources of the Marsyas were at the palace of the king of Persia in a lofty situation under the acropolis of Celænæ. From Arrian and Q. Curtius[189] we learn that the citadel was upon a lofty precipitous hill, and that the Marsyas fell from its fountains over the rocks with a great noise: from Herodotus[190] it appears that the same river was from this circumstance called Catarrhactes; and from Strabo[191], that a lake on the mountain above Celænæ was the reputed source both of the Marsyas, which rose in the ancient city, and of the Mæander. Comparing these authorities with Livy[192], who probably copied his account from Polybius, with Pliny[193], with Maximus Tyrius[194], and with the existing coins of Apameia[195], it may be inferred that a lake or pool on the summit of a mountain which rose above Celænæ, and which was called Celænæ or Signia, was the reputed source of the Marsyas and Mæander; but that in fact the two rivers issued from different parts of the mountain below the lake: that the lake was named Aulocrene, as producing reeds well adapted for flutes, and that it gave the name of Aulocrenis to a valley extending for ten miles from the lake to the eastward: that the source of the Marsyas was in a cavern on the side of the mountain in the ancient agora of Celænæ: that the Marsyas and Mæander, both of which flowed through Celænæ, united a little below the ancient site: that to this junction the city was removed by Antiochus Soter, son of Seleucus Nicator, when he gave it a new name after his mother Apama; and that the united stream was soon afterwards joined by the Orgas and the Obrimas. Whether these inferences drawn from the ancient authors are correct, will be decided by the future traveller. He may also ascertain whether there are any volcanic rocks, the burnt appearance of which will justify the etymologist[196] who ascribed to that cause the origin of the word Celænæ; or he may discover the valley of Aulocrenis, the scene of the celebrated contest of Apollo with Marsyas, whose skin was still shown in the time of Herodotus, in the acropolis of Celænæ[197].

I have been thus particular in laying before the reader the ancient evidences on the site of Apameia, because it is a point of great importance to the ancient geography of the western part of Asia Minor,—not less so than Tyana is to the eastern: and because in regard to both these places, I have the misfortune to differ from the author in whose opinion the public is justly in the habit of placing the highest confidence[198].

The Roman road from Apameia to Antiocheia of Pisidia passed through Apollonia, otherwise called Mordiæum[199], which was 24 M. P. distant from the former, and 45 from the latter. Although on account of our ignorance of the site of Antiocheia, no exact comparison can be instituted between the amount of the two numbers just mentioned and the actual distance on the map, it is manifestly not very erroneous; and the position of Apollonia therefore was probably at no great distance from a town called Ketsibúrlu, which General Koehler passed through between Burdur and Dombai, and which according to Abubekr Ben Behren is a kadilik of Hamed, of which Isbárta is the chief city. Ptolemy places Apollonia near Antiocheia; and its situation, between that city and Apameia, which the Table gives, is in exact conformity with Strabo’s description of the conquests of Amyntas. Having taken Derbe, and received Isauria from the Romans, he made himself master of Antiocheia, and the country as far as the district of Apollonia, near Apameia Cibotus[200], together with Lycaonia and some part of Phrygia Paroreius. He took Cremna, but did not venture on attacking Sandalium: and after capturing the greater part of the places belonging to the Homonadenses, (whose tyrant he slew,) he was himself destroyed by a stratagem of the wife of the latter. Sulpicius Quirinius and the Romans afterwards reduced Homona:—all the late territories of Amyntas were then placed under the government of a præfect[201].

3. The ancient road from Apameia to Synnada must have crossed that of Gen. Koehler at or near Sandukli, on the river now called the Mendere (Mæander), but which anciently, I suppose to have been the Obrimas, a branch of the Mæander. The total distance of 73 Roman miles on this road agrees tolerably with the 66 geographical miles in direct distance, which the map gives between the assumed site of Synnada and that of Apameia at Dinglar. Euphorbium, the only place on the road mentioned in the Table, and which was midway between the two extremes, will fall at Sandukli. Euphorbium is noticed as a town in this part of Asia by Pliny only, who tells us that its people formed,—together with those of Metropolis, Peltæ, Acmonia and some other towns,—the _conventus_ held under the Romans at Apameia[202].

4. The fourth Roman road which crossed the modern route from Adália to Shughut, is that marked in the Table from Dorylæum to Apameia Cibotus, leading through Nacoleia, Conni, Eucarpia, and Eumenia[203]. Although the total distance of 148 M. P. on this road sufficiently agrees with the 100 G. M. in direct distance on the map, it must be confessed that the 26 Roman miles and the 15 geographical miles of direct distance, between Eumeneia at Ishékle and Apameia at Dinglar, do not bear the same proportion as the Roman and geographical numbers on the whole line; and that, if I am right in the position of Nacoleia, the 20 M. P. of the Table, between Dorylæum and Nacoleia, errs almost as much in defect, as the 26 M. P. between Eumeneia and Apameia does in excess. But it is in vain that we look for much accuracy of detail in the Table. The positions of Nacoleia and Eumeneia rest upon very satisfactory grounds. All that remains to be done, therefore, is to arrange Conni and Eucarpia between Doganlu and Ishékle, at the proportional distances of the numbers in the Table. This will place Conni not far to the southward of Altun Tash, near where the roads to Altun Tash, both from Karahissár and from Sandukli, cross the ancient road; a position which agrees with that of Conna in Ptolemy[204], according to whom it appears to have been not far from Cotyaeium, to the southward. Under the Byzantine emperors, Conna (then called Cone[205]) was a bishopric of the province of Phrygia Salutaris, of which Synnada was the metropolis.

Eucarpia was another bishopric of the same province. Its name was derived from the fertility of the soil[206], which by attaching the people to agriculture may have contrasted them with those of the neighbouring Euphorbium, celebrated probably for its flocks and pasture. The position of Eucarpia in the Table agrees with that which Ptolemy gives it to the southward of Conna.

5. The fifth and last of the ancient roads intersected by the modern road from Adália to Shughut was from Dorylæum to Philadelpheia: its two extremities are known points; its length in direct distance is equal to two degrees of latitude, or 120 G. M., which corresponds with as much accuracy as one can expect to the 155 M. P. of the Table. The _line_, as will be seen on referring to the map, leads directly through Kutáya. We cannot doubt therefore that _Cocleo_, the first name occurring on this road in the Table, is an error for Cotyaeio; especially as the distance of 30 M. P. answers very well to the real distance from Eski-shehr to Kutáya. The distance of 35 M. P. between Cotyaeium and Acmonia furnishes the traveller with a good approximation for discovering the site of the latter city, which is mentioned in one of the Orations of Cicero[207], and which was one of the towns of the conventus of Apameia, and afterwards a bishopric under the metropolitan of Laodiceia. It is difficult to reconcile the position of Aludda, 25 miles beyond Acmonia on the road to Philadelpheia, with that which may be inferred from Ptolemy, who names Alydda among the towns of the greater Mysia, together with Pergamum and Apollonia on the Rhyndacus. Clanudda I suspect to be an erroneous writing; but its correction I am unable to discover.

It is in the unexplored part of Phrygia Epictetus[208], lying between the Thymbres and the branches of the Rhyndacus on the southern side of the Olympene mountains, that the future traveller will seek for the Phrygian cities of Cadi, Azani, and Synaus. One is much disposed at first sight to consider the remarkable position of In-óghi, which General Koehler passed through in his way from Kutáya to Shughut, to have been the site of one of these cities of Phrygia Epictetus; but upon further examination, they all appear to have been situated considerably to the westward of this position. The Azanitis, or district of Azani, contained the sources of the river Rhyndacus, which, after passing through the lake of Apollonia, joined the Propontis opposite the island of Besbicus, having first received the united waters of several streams from Mysia Abrettena, particularly the Mecistus, which flowed from Ancyra Abassitis, a Phrygian town on the frontier of Lydia[209]. Synaus appears to have been near this Ancyra; for in the acts of one of the Councils, a bishop of the Phrygian Ancyra signs himself Αγκύρας Συννάου, no doubt in order to distinguish this Ancyra from the Galatian. Cadi also may be presumed to have been to the westward of the meridian of In-óghi and Kutáya; for we find that Cadi is assigned by some authors to Mysia[210]. It is precisely in the situation, which may be inferred from this circumstance, combined with what has been said of the position of Synaus and Azani,—that is to say, between the Thymbres and the sources of the Rhyndacus,—that we find a town of the name of Kodús, which has not been visited by any modern traveller, but which is briefly described by Hadji Khalfa—as situated on the banks of a river, in a plain surrounded by mountains. He adds that the river, which bears the same name as the town, descends from Mount Morad, and passes by Magnesia into the Gulf of Smyrna. We know from modern travellers, that this river, which is the ancient Hermus, is still called Kodús or Ghedís in all the lower part of its course; and Kodús, it can hardly be doubted, is the same place as Καδοί, the name of which the Turks received from the Greeks, in the usual Romaic form of the accusative case Καδούς.

In exploring the equally unknown country which extends to the southward of this part of Phrygia Epictetus, towards the mountains Messogis and Tmolus, and which formed the frontier of Lydia and Great Phrygia, the traveller may derive assistance from a passage in Strabo[211], where he enumerates the principal plains in their order from west to east. Adjacent to the Caystrian, which lay between Tmolus and Messogis, was the Cilbian, then the Hyrcanian, the plain of Cyrus, the Peltene, the Cillanian, and the Tabene. It cannot be doubted that a journey through these plains would lead to a knowledge of the general distribution of the geography of the country, as well as to that of the sites of some of the towns which gave name to the several plains. Peltæ, Lysias, and Silbium appear to have been in the country northward of the upper Mæander, which is traversed by the caravan route from Smyrna to Tokát: but the few names and distances which Tavernier and Seetzen have left us between Alláh-Shehr and Karahissár, throw no light whatever upon ancient geography.