Geographical Journal (Journal of the Royal Geographical Society): Volume VIII, No. 5 [‎27v] (57/154)

460 A JOURNEY IN THE VALLEY OF THE UPPER EUPHRATES.
From hero we followed the carriage-road, which keeps to the left bank
!, a small stream flowing into the Kelkid Irmak (Lyons), and ,n five
and a half hours reached Sadagh, which lies a short distance to the left
!f the road Sadagh is a Turkish village of about a hundred and fifty
houses, built for the most part of old squared stones. It is situated in a
depres ion in the hills on the left bank of one of the main sources of the
Kelkid Irmak, and has a fine water-supply which flows from a reservoir,
which is artificially banked up above the village. The site ,es close to
the route up the Kelkid Irmak from Enderes „„ Ashkaleh to Eraerum.
This route, inasmuch as it formed part of the old route by land from
Constantinople to Erzerum,* was of great importance before the half
sea route via Trebizond to Erzerum was opened up, and in consequence
the village has been mentioned in the writings of some of the early
travellers to the East.f „ , . 1QftQ i.
In more recent years, it has been visited by Taylor in 1868, who
copied a Latin inscription, and reported the existence of some Byzantine
inscriptions in the village; + and subsequently by Biliotti, who sent in a
description of the site, accompanied by a plan of the rums, to the British
Government A Latin inscription was also copied on the site by Colonel
Everett, and has been published in the Corpus.§ On the strength of the
reports Sadagh was identified by Kiepert in his maps with Satala, the
campofLegioXV.Apollinaris.il
There was nothing, however, in these reports and inscriptions to
make this identification certain, and we devoted two days to an exami
nation of the site, and a search for further epigraphical evidence. Our
efforts were rewarded by the finding of several legionary titles inscribed
with the words “ Leg. XV. Apoh,” the discovery of which must remove
all doubt as to the identity of the site with Satala. Considerable remains
of an ancient fortified town exist near the modern village. The walls,
which consist of a rubble core faced with well-squared stones, seem to
have formed a square enclosure, the sides of which faced the lour points
of the compass. They are best preserved on the north side, where the
towers are still standing along the whole line and at the two corners.
On the south and along part of the west side the ancient walls have
been almost obliterated by the modern village, which is built over the
south-west corner of the enclosure. As far as can be judged from the
style of construction, the walls would seem to be of Byzantine rather than
of Roman date, and are probably in their present form not older than the
time of Justinian, who is known f to have restored the fortifications of * * * §
* It is described in Ritter, vol. x. p. 751.
f Tournefort, Let. 21; Morier, ‘ Persia,’ p. 331.
X Taylor, loc. cit., pp. 287 ff. The inscription which he copied can be partially
restored, and is undoubtedly one of the Emperor Aurelian. It has not yet found its
way into the Corpus.
§ ‘C.I.L.,’ vol. iii. Suppl. No. 671-1. || ‘Notit. Dignit. Orient ,’ cap. xxxv.
Procopius, ‘ De iEd.,’ iii. 1.