rovince of Cordyene, inhabited
by a brave and hardy population, and afterwards the seat of fifteen
fortresses which brought the Roman dominion to the very edge of
Adiabene, made them masters of the passes into Media, and laid the whole
of Southern Mesopotamia open to their incursions. It is probable that
the hold of Persia on the territory had never been strong; and in
relinquishing it she may have imagined that she gave up no very great
advantage; but in the hands of Rome Kurdistan became a standing menace
to the Persian power, and we shall find that on the first opportunity
the false step now taken was retrieved, Cordyene with its adjoining
districts was pertinaciously demanded of the Romans, was grudgingly
surrendered, and was then firmly re-attached to the Sassanian dominions.

(ii.) The Tigris is said by Patricius and Festus to have been made the
boundary of the two empires. Gibbon here boldly substitutes the Western
Khabour and maintains that "the Roman frontier traversed, but never
followed, the course of the Tigris." He appears not to be able to
understand how the Tigris could be the frontier, when five provinces
across the Tigris were Roman. But the intention of the article probably
was, first, to mark the complete cession to Rome of Eastern as well as
Western Mesopotamia, and, secondly, to establish the Tigris as the line
separating the empires below the point down to which the Romans held
both banks. Cordyene may not have touch the Tigris at all, or may have
touched it only about the 37th parallel. From this point southwards,
as far as Mosul, or Nimrud, or possibly Kileh Sherghat, the Tigris was
probably now recognized as the dividing line between the empires. By the
letter of the treaty the whole Euphrates valley might indeed have been
claimed by Rome; but practically she did not push her occupation of
Mesopotamia below Circeshim. The real frontier from this point was the
Mesopotamian desert, which extends from Kerkesiyeh to Nimrud, a
distance of 150 miles. Above this it was the Tigris, as far probably
as Feshapoor; after which it followed the line, whatever it was, which
divided Oordyene from Assyria and Media.

(iii.) The extension of Armenia to the fortress of Zintha, in Media,
seems to have imported much more than would at first sight appear from
the words. Gibbon interprets it as implying the cession of all Media
Atropatene, which certainly appears a little later to be in the
possession of the Armenian