Ramsay, William Mitchell: The cities and bishoprics of Phrygia: being an essay of the local history of Phrygia from the earliest time to the Turkish conquest (Band 1,2): West and West-Central Phrygia (Oxford, 1897)

586 XIII. THE BANAZ-OVA.

in the earlier Notitiae between Bria and Trajanopolis, and in the later
Notitiae next to Alia1. The comparison of the order in these cases
suggests that this bishopric was situated in the middle part of the
Banaz-Ova but towards E., so as to come naturally after Sebaste
and next to Trajanopolis, and the site of Hadjim suits excellently.
A bishopric is needed for the great stretch of country between
Senaros and Hippourios; and it is a strong proof of the inadequacy
of M. Radet's theory, that he leaves that country without an episcopal
centre2. Moreover, it also tells against his theory that he takes no
account of Hadjimlar: the fact that is most nearly certain about this
stretch of country is that the ancient and the modern centre is at
Hadjimlar. He places ancient cities or villages at Kalin-Kilisa and
Yapaklar, where I could see no sign of ancient life, and leaves
Hadjimlar a blank.

In Nov. 1881 I spent a night at Hadjimlar3, and saw at once that
it has been an ancient site, but found no inscriptions. In 1883 my
travelling companion Sterrett visited it on his way from Ushak to
Sebaste. He saw digging going on, from which building-stones were
being taken. Several of these, bearing inscriptions, had already been
defaced, and he was only in time to copy the fragmentary decree of
a city. He also pronounced that Hadjimlar was an important ancient
site. Further it is the busiest village of this district that I have
seen4; and must be regarded as the market town. Its importance
lies in the fact that it is the furthest south place which is able to
take advantage of the ridge acting as a watershed between Senaros
and Hippourios. It has a natural water supply: the villages in the
centre of the plain are so deficient in this respect that development is

1 The order in the Notitiae is very Sterrett was in the same condition; but
irregular in Phrygia, but retains some- we used the name Hadjimlar and were
thing of geographical arrangement understood.

broken in fragments. The geographical i I have visited many of the villages;

character of Hierocles's arrangement is but there are still some to be examined,

well marked in this province, as a glance Banaz-Ova is difficult and unpleasant to

at the map shows. explore, because a wide view cannot be

2 He places Agathe-Kome further got. The plain is not a dead level, like
north at Yapaklar; but we have seen the great flat plains E. and NE. from
reason to place it elsewhere (p. 261); Iconium ; it is undulating : most of the
moreover Agathe-Kome appears only in villages (Hadjim and Geubek are excep-
the late Notitiae. tions) lie in the hollows: the roads

3 The form in singular Hadjim (i.e. keep as a rule to the hollows also.
' my hadji') is also used. The Cassaba Often one maybe unaware of the exist-
Railway Survey calls it Adjama. I ence of a village only 150 yds. distant,
cannot vouch for the true foi-m. In as the eye looks over it, owing to some

I!

5i I knew no Turkish, and in 1883 very slight intervening ridge.