Geographical Journal (Journal of the Royal Geographical Society): Volume VIII, No. 5 [‎33v] (69/154)

472 A JOURNEY IN THE VALLEY OF THE UPPER EUPHRATES—DISCUSSION.
thus oft'ered as a starting-place for expeditions into Armenia from the north. 1 he?
fact also that it lies on the natural road from Trapezus to the Euphrates, and
was thus open to communications with the sea, may have led to the choice of
the site for a legionary camp. The legion stationed here seems to have supplied
some of the garrisons in Armenia, for we know that in a.d. 185, a vexillatio of
the Legio XV., Apollinaris was stationed at Valarschapat (Caenepolis).*
Judging from our experience, it seems probable that very slight remains of the
Roman defensive posts along the frontier from Samosata to the Black Sea exist at
the present day, and, even though further evidence may be forthcoming in the
future, it seems certain that the lines on the eastern frontier were not laid out on
nearly the same scale as on the other Roman frontiers, on the Rhine and the
Danube, in Britain and in Africa. This is more easily to be explained on the more
southern reaches of the Euphrates above Samosata, where the great size and swift
ness of the stream, and the inaccessible nature of the mountains through which it
flows,f render the river, except at certain easily guarded spots, a formidable
barrier to the crossing of an army. But north of Melitene, and especially at and
above Keban-Maden, near which the two branches of the Euphrates meet, the
river is not so serious an obstacle to an invading force,! and it is natural to
expect to find remains of defensive works on a large scale near the river-bank.
Such remains do not appear to exist at the present day, and the inference may
with probability, but not with complete certainty, be made that no great defensive
works were ever carried out between the camps. In the same way little care
seems to have been bestowed on the connecting reads, and, with the exception of
the work of SeptimiusSeverus on the Cfesarea—Melitene road and the bridge which
he built at Kiakhta, there is no evidence which points to the roads having been
constructed differently to the generality of Roman roads in Asia Minor.§
If we are to seek a reason for this apparent neglect, it is doubtless to be found
in the different character of this frontier to that of the other frontiers of the
empire. This difference lies in the continual change of relations between Armenia
and Rome during the first two centuries of our era. During this period Armenia
was hardly ever regarded as a hostile country, and was often reduced nearly to
the condition of a Roman province with Roman troops quartered in it. Probably,
as Hogarth has pointed out, Septimius Severus was the first to depart from the
traditional views of the emperors. He is known to have laid out roads to Melitene
from the west and from the south; that he intended to extend these roads to the
north as well is not improbable, but the regative results of our journey tend to
show that, if ever conceived, this project was not carried out.ll
Before the reading of the paper, the President said: The paper we are to listen to
this evening is an account of an expedition under Mr. Hogarth by the upper waters * * * §
* ‘ C. I. L.,’ vol. iii. Suppl. 6052.
t This part of the river is best described by Von Moltke (‘ Briefe a. d. Turkei,’ edit
vi. pp. 308 ff.).
J It was easily forded by Taylor in September at a spot near Korpanik (Taylor,
loc. cit., p. 315).
§ Ramsay (op. cit., p. 46) describes the usual character of the roads in Asia Minor.
II S‘ nce this paper has been in the printer’s hands, a most valuable article on the
campaign of Basil I. against the Paulicians has been published by Mr. J. G. C.
Anderson in the Classical Review (April, 1896), in which he discusses manv of the
points of historical geography which are dealt with above. It is most satisfactory to
me to find that the conclusions at which he has arrived agree most remarkably with
the views expressed in this paper.