Southport : Original Sources in Exploration

Troy and Ilium: Results of the Excavations at Troy 1870-1894

Wilhelm Dorpfeld


Chapter 2 (part 21)

10. The excavations in the lower town.  (p.234)

The mound of ruins, the nine different strata of which we have hitherto discussed, is the extreme north-western spur of a ridge which forms the southern edge of the Simoeis valley. The hill rises only slightly above the average height of the next part of the plateau, even lower than its highest point. But as it is separated from it by a depression in the ground, it was very useful as an isolated hill for a castle.

[p.235] The general shape of the ridge can be seen from the sketch map of north-western Troas (Plate I) and from the plan of the lower town drawn in a larger scale (Plate II). However, the latter plan only includes that part of the plateau which belonged to the Roman lower town and which was again separated from the ridge extending further to the east by a depression.

The former extent of the Roman lower town can still be clearly recognized today by the numerous roof tiles, marble fragments, building stones of all kinds and other remains of the old town that cover the ground of the fields. In addition, the course of the ring wall that once surrounded the city is secured by the terrain design and often also by a low earth wall originating from the remains of the wall. Only in a few places one can be in doubt about the exact boundary of the city. In the plan of the lower town, which Herr Ritter Wolfif took up in 1883 and from which our plate II is drawn, the Roman wall line could therefore be indicated, although the wall itself has not been preserved. According to this plan, the length of the ring wall was around 3300m and the area of the city was more than 60 hectares.

But did the older settlements, the remains of which are preserved on the Acropolis hill, also have a lower town? And can its extent still be determined?

In the years 1879 and 1882 Schliemann had already created a large number of shafts and ditches within the Roman lower town, while in the first year he found only Graeco-Roman buildings, a few graves and pottery from the historical period (cf. Ilios, p.683 and plan II), many sherds of "the first two prehistoric cities", i.e. layers I and II, were later excavated on the plateau near the castle hill (cf. Troy (1882), p.28 and 68 ). Schliemann initially concluded from this that there was only a Greco-Roman lower town, but later, after the discovery of the prehistoric shards, he also assumed a lower town of the II layer and also drew the presumed extent in Plan VIII of Troy (1882) of this prehistoric lower city.

In order to check his statements and at the same time to come to a clear conclusion about the questions connected to the various lower towns, we had decided for the excavations of 1893 and 1894 to uncover a larger part of the lower town and to thoroughly explore its entire area as (p.236) a set task. To our keen regret, however, we were not able to complete this work in either of the two years. The finding and partial uncovering of the stately ruins of Layer VI, the real Homeric castle, so occupied our attention and manpower that little time was left for exploring the lower town. But we dug a few trenches and several shafts. A. Götze took over the special direction of this work. It is not possible to print his detailed report on this excavation here because of its size. However, some passages from it may be shared here, which relate to three different, particularly important deposits:

I. "A ditch excavated south-east of the castle hill (marked A in Plane II) contained the remains of a Roman settlement in its upper layers: a stylobate with a column base, two water pipes, parts of a mortar base for a pavement, of which a red marble slab still survives was, also larger and smaller pieces of marble columns and other pieces of architecture, all of which lay south of the stylobate, as well as small implements, of which a surgical knife may be mentioned. The layer with these Roman inclusions, which was mixed down with some older sherds, consists of rubble; its thickness is on average 2 m. Among the many Roman sherds only three Greek sherds with varnish painting were found".

"Below this layer follows another layer of fine brown earth, which contains clay sherds exclusively of the type from the VI city, also double-conical and spherical spindle whorls without ornament, a conical clay weight with a round base, half of an oval hand mill and a piece of flint, also running in this one level two low wall foundations of rough stone diagonally on top of each other. The natural soil is at a depth of 3.25m except for one spot".

“Here the brown layer descends deeper. After a small and poorly constructed wall had broken away, a ring of double layers of stone, 2.20 m in width, appeared below it at a depth of 3.40 m. It was the edge of a round well, 9.50 m deep from here, which had been worked through the rock into a water-bearing layer. It was filled with soil, stones, artefacts of the VI layer, and frequent cavities under larger stones suggest that it was not filled up gradually, but was filled up in a relatively short time. Among its contents are several large pieces of architecture, including a large stone, apparently coming from a sloped edge of the VI castle wall, and a base for a double column (?), as well as clay sherds from the VI city, fragments of a brazier and a bone spindle".

2. "On the western slope of the castle hill, outside the ring wall, two trenches (E and F on Plate II) were dug, which show that the (p.237) earth on the higher, more heavily sloped part of this slope was only very thin ( 1/4 - 1/2 m) and only at the foot of the hill up to 3 m thick on the rock. In the upper part of the longer ditch, three of those peculiar well-like depressions were excavated, a number of which were found here in 1893. In one shaft, 1.25 m wide, the bottom had not yet been reached at a depth of 6 m. It contained bad late pottery and a piece of terra sigillata with relief palmettes. Two small shafts were only 0.50 and 0.80 m deep. In the lower part of the ditches there was a water pipe, some small wall foundations and 6 simple ditches (see Section VII). Outside the tombs a number of Hellenistic, Roman and Byzantine objects have been found, which may have come from destroyed tombs. In particular, the fragment of a gray clay tablet with a relief depiction of the front part of a horse, similar to the tablets discussed in Troja 1893, p. 72 ff., should be mentioned".

3. "A ditch laid out on the plateau of the lower town about 200 m south of the castle ring (B on Plate II) revealed that the stratification conditions are quite analogous to those in the first ditch (A): down to the depth of the Roman floor, from which in 0.85 m deep a larger area of white screed was preserved, only Roman inclusions, below that there is a layer reaching down to a depth of 1.80 m, in which Roman and older sherds are mixed, and from 1.8 m to 2 m, where the rock begins, exclusively potted goods of the VI layer in a brown soil layer. Of distinctly Greek sherds, only three pieces with black varnish and the sherds of a Hellenistic red clay vessel with reliefs of a skeleton, vases, palm branches and other things were found here. In the Roman layer, remains of the walls of two buildings were also uncovered, which lie on either side of a space 1/2 a meter wide, probably a street going from N. to S., since five parallel water pipes were found here. In the vicinity of some of the larger stones protruding from the foundations of a transverse wall lay the sherds of the Hellenistic vessel just mentioned, with depictions in relief. In the lowest layer was a skeleton in a stretched position, below it was a fairly circular pit (upper diam. 2.65m, lower diam. 2.42m) with carefully smoothed bottom and walls, another 0.50m in the rock incorporated, it contained only monochrome sherds of the VI layer".

Finally, A. Götze sums up his judgment on the results of the excavations in the lower town as follows : " These excavations yield the following with regard to the question of the lower towns of the various strata : First of all, it must be emphasized that not a single object was found, which should be dated to the pre-Mycenaean period. Schliemann claims the existence of a lower town for the second settlement (p.238) of the castle hill and justifies this with the occurrence of ceramic finds, allegedly from the first and second town, in the lowest layers of the plateau (cf. Troja p 96) However, his observations are limited to the immediate vicinity of the castle hill, about which there are no new observations to check. If there really was a small lower town here for the second stratum, it certainly did not have the extent that Plan VIII in Troy (1882) gives it. The oldest finds in the ditches made in 1894 belong to the VI town, namely they are found in large numbers in the western half of the plateau and in a continuous layer lying directly on the rock. In Mycenaean times, the plateau was settled, the circumference of which seems to be roughly circumscribed by the red line that is supposed to designate the Homeric lower city on Plan VIII in the book Troy, with the only difference that the southern border is probably even further to advance to the south until close to the Roman enclosing wall".

"Almost none were found from the older Greek period, and so few remains from the Hellenistic period that it seems questionable whether they should be regarded as evidence of a lower town or only as pieces scattered from the castle hill. Apart from earlier finds, the ditches we have excavated have provided sufficient examples of the existence of a large Roman lower town with large, beautiful buildings".

From the results of all previous excavations, namely those conducted by A. Götze, the following results for the lower towns of the various layers:

While the first settlement was limited to the castle hill, the adjoining plateau may have been partially settled at the time of the second layer. Outside the castle there may have been a few houses because Schliemann found sherds of this layer in some of his ditches. However, a larger lower town does not seem to have existed in this epoch.

That the villages of the III, IV and V strata occupied not only the upper surface of the hill but also its slopes, is stated above (p.102). These settlements, however, hardly extended over the plateau of the lower town.

With regard to the VI layer, the excavations of 1893 and 1894, as A. Götze explains above, prove that a large part of the lower town was settled. In the western half of what later became the city district, the characteristic sherds of the VI layer were found in the lowest layer and in some places also the remains of walls of a simple type made of rubble and clay. Given the small extent of the uncovered areas, no judgment could be made about the floor plans of these buildings. It was also not possible to determine whether these were isolated houses or a closed (p.239) settlement. No trace of a ring wall of this lower town has yet been found, at least none of the wall remains found can claim to be a ring wall of the lower town from Mycenaean times.

It would have been very important for the comparison of the local facts with the Homeric poems if the existence of a walled lower town could have been proved with certainty next to the strong and stately castle. So far, however, this has not been permitted, and it must be left to later excavations to bring full clarity about this. However, the possibility of the existence of a fortified lower town seems to me to have been proven by the discovery of the remains of a VI settlement on the plateau south and south-east of the castle. Because this settlement could be temporarily or permanently surrounded by a ring wall.

No walls or other remains have been found in the area of the lower town from the time of the VII stratum. On the other hand, isolated ancient Greek and Hellenistic pottery, i.e. remains of layer VIII, came to light. If the city of Ilion had occupied the entire area of the Roman lower city since the time of Lysimachus, i.e. since 300 BC, as was almost universally believed in the past, then the walls and remains of their houses would have had to be found to a greater extent below the Roman houses. Since this has not hitherto been the case, the Greek city which we assume next to the sanctuary of the Ilian Athena on the castle hill can at most have extended over part of the plateau. I consider it certain that it was not limited to the acropolis, but that at least in the Hellenistic period there were also houses outside the fortified castle. But whether there were very many and where they were, I cannot say. Hopefully later excavations will provide us with information about this. Incidentally, several of the inscriptions cited by A. Brückner in Chapters VI and VIII speak for a considerable expansion of the lower town in the Hellenistic period.

From a Hellenistic building outside the acropolis we know of a theater from an inscription (see Chapter VI, No. 2). This will certainly have been situated on the same spot where the ruins of a large Roman theater are still preserved, namely north-east of the castle on the northern slope of the city plateau (cf. Plate II). A part of the skene building was uncovered by Schliemann in 1882 and its design and plan turned out to be a Roman building. Unfortunately, in 1894 we were not able to examine it more thoroughly and uncover it. We therefore do not know whether there are still remains of a Greek skene under the Roman walls. The walls uncovered by Schliemann were partly broken down by Turkish soldiers and partly buried again; which they didn't even bother to measure. We do not know any other (p.240) buildings in the Greek lower town. Only from the inscriptions (see Chapter VI, Nos. 14, 16, 19, 25, 35) can we infer the existence of a prytaneion, dicasterion, stadium, basileia and several sanctuaries. Whether they were all in the lower town is unknown.

We have already shown that nothing was found of a Greek city wall, and that there was hardly one. We believed that Strabo's report (XIII, S93) about the building of a 40-stadia long ring wall by Lysimachus had to be related to Alexandria Troas (p. 207). The circular wall of Ilion, the remains of which are still in existence, the course of which is indicated by stippling in Plate II, is assigned to the IX layer, i.e. to the Roman city. Wherever we have found the remains of its foundations, it consists of soft porous blocks and corresponds in its design to the foundations of the columned halls of the Athena district. On the south side of the city (at G on Table II) we uncovered it in two places and measured its thickness to be 8 feet. On the north side, the castle wall and boundary wall of the Athena district could have served as the wall of the lower town at the same time. It is possible, however, that a special city wall existed deeper on the slope, because in 1882 I saw a wall corner built from right-angled blocks on the northern slope during excavations, which Schliemann describes in the book Troja (1882), p.20 may belong to a ring wall of the lower town. We could not find them again in 1894.

We are just as little informed about the time when the great city wall was built as we are about the time when it fell. In any case, since it must have been built after the destruction by Fimbria, it seems to me most probable that it was built in the time of Julius Caesar or Augustus. From the former we know that he enlarged the territory of Ilion and gave the city complete freedom, he is said to have even had the plan to transfer the seat of government from Rome to Ilion (cf. E. Meyer, History of Troas, p. 95 and Haubold, De rebus Iliensium, p. 41). Not only is it certain that Augustus did good to the city, but he is even called "ktistes", i.e. founder of the city, on a coin. The people here also stamped his head on their coins alongside those of his successors up to Claudius (cf. Haubold, De rebus Iliensium, p. 45).

None of the buildings inside the Roman lower town have been completely excavated, but we have found many in the excavated shafts and ditches. Walls of ashlar and brick, floors of lime screed, marble slabs or tesserae, columns of marble and syenite, pieces of entablature of marble and porous limestone have been found in large numbers, proving that the whole area within the encircling wall was occupied by buildings of various kinds. It is to be wished that soon a larger part of the city area would be freed from the masses of rubble. The place south-east of the Acropolis (p.241) should be recommended as such, because according to finds to date the agora of the city may have been located there. Such an excavation would also soon make it possible to determine whether the public buildings in Hellenistic times were in the lower city or were still on the Acropolis, as they had been before.

Finally, the water supply of the lower town has to be discussed. From the discovery of a few in the dug ditches we can conclude that there were numerous wells in the houses of the city. Wells could easily be dug in the tertiary limestone of the soil, reaching water-bearing strata at a depth of about 100 m. Even on the Acropolis several wells have been brought down to these strata (cf. p.180 above). The same water that was drawn from these wells also appeared in springs on the slopes of the plateau.

Three such springs are still to be found on the northern slope, that is, in the Simoeis valley. One is drawn northeast of the theater on Plate II, where it bears the letter L, a second is in the depression just east of the town, and a third a few hundred yards further east. All three still provide good and plentiful drinking water (cf. Ilios, p. I28). A fourth was dug up and cleaned by us in 1882 on the west side of the city, that is in the Skamander valley (cf. Troja 1882, p. 71). A previously known tunnel (K on panel II) leads into the rock here and divides into three arms, all of which supply good water. In Roman times, this water was fed through clay pipes to a basin in front of the tunnel and then certainly to a running well. A canal made of small stones, which we found under the clay pipe on the bottom of the tunnel, dates back to earlier times. It proves that the spring was already in use and tapped in pre-Roman times. Even today it still gives water. From the course of the horizontal curves in Plate II it can also be inferred that the water from this spring ran westwards through a fold in the terrain to the old Skamander from ancient times.

Although these springs and the numerous fountains provided good water for the old castle and also for the city, the people of Roman times were no longer content with this. They wanted fresh, running water both inside the city and on the Acropolis, so they built a large aqueduct that brought good spring water from far away from the mountains to Ilion. Clay tubes fed by this line have been found in many places in the city and on the Acropolis. The clay pipes in which the water was carried up to the Acropolis under pressure have also been preserved. Given the location of the city, this water could only come from the south-east, from the foothills of the Ida Mountains. In fact, it was brought to Ilion either from the upper Thymbrios (now Kemar-Su), or from a still higher point. For a stately arch of an aqueduct still exists about 6 kilometers above the confluence (p.242) of this river into the Skamander, and it is this that has given the Thymbrios its current name "arched river". Measured as the crow flies, the site is about 9 km south-east of Ilion.



photo 32: Arch of the Roman aqueduct from Ilion in the Thymbraios valley.

We see the stately arch, which once led the water from the southern bank to the northern one, on photo 32. The considerable size of the span can be seen in the people depicted on the bottom left and top right. The arch itself and its abutment walls are made of cut stone, while the upper part of the structure is made of small rubble stones with lime. Noteworthy is the relief head located on the keystone of the arch. Perhaps it represents the Roman emperor who built the large complex for Ilion, the ancestral seat of the Julian family. It is not known which emperor it was. However, there can be no doubt that the aqueduct is certainly a Roman and not a Hellenistic work, given the design of the arch. A line that contained such arches must have been a magnificent work, and so this one arch still bears loud testimony to the great heyday of Ilion in Roman times.

Wilhelm Dorpfeld




[Continue to Chapter 3]

[Return to Table of Contents]


Southport main page         Main index of Athena Review

Copyright  ©  2023    Rust Family Foundation.  (All Rights Reserved).

.