unknown before in Asia Minor, was first introduced there by
the Persian army. The site of the battle is not precisely
known. According to Herodotus, the fight took place in the
great plain before Sardes, which is crossed by several small
tributaries of the Hermus, amongst others the Hyllus. Radet
recognises that the Hyllus of Herodotus is the whole or part
of the stream now called the Kusu-tchaî, and he places the
scene of action near the township of Adala, which would
correspond with Xenophon's Thymbrara. This continues to be
the most likely hypothesis. After the battle Croesus would
have fled along the Hermus towards Sardes. Xenophon's story
is a pure romance.

Croesus was again worsted on the confines of the plain of the Hermus,
and taking refuge in the citadel of Sardes, he despatched couriers to
his allies in Greece and Egypt to beg for succour without delay. The
Lacedaemonians hurried on the mobilisation of their troops, and their
vessels were on the point of weighing anchor, when the news arrived
that Sardes had fallen in the early days of December, and that Croesus
himself was a prisoner.* How the town came to be taken, the Greeks
themselves never knew, and their chroniclers have given several
different accounts of the event.**

* Radet gives the date of the capture of Sardes as about
November 15, 546; but the number and importance of the
events occurring between the retreat of Croesus and the
decisive catastrophe--the negotiations with Babylon, the
settling into winter quarters, the march of Cyrus across
Phrygia--must have required a longer time than Radet allots
to them in his hypothesis, and I make the date a month