82 HERODOTUS.

[CHAP.

to the other two combined, and much broader than they were
*" Though he was well aware that the Arabian Tnlf

i Boundaries * i , vjuii

between them, was a long and narrow sea to the eastward of
Africa, he maintained that the western frontier of
Egypt formed the true boundary of that continent on the side
towards Asia1. He admits that the Nile was the limit usually
accepted by the Greeks, but objects to it on the ground that
on this supposition Egypt, through which that river fl0wS) would
be partly in Asia and partly in Libya. Again, whilc he men-
tions the view that the Cimmerian Bosporus and the Tanais
were to be regarded as dividing Asia from Europe, he himself
would draw the line at the Phasis, the Caspian Sea, and the
Araxes, in a direction running due east3, ' The mention of the
last-named river introduces one of the most puzzling questions
^Hisconfu- of Herodotean geography. ^He seems indeed to
sion about the have confused together two streams called
Araxes— one the modem Aras, which discharges
its waters into the western side of the Caspian, for he says that
the Araxes rose in the land of the Matieni, i.e. ia tjjC nQT^
of Armenia, where the sources of the Aras lie ; the other—
which Cyrus had to cross in order to attack the Massagetae
and which is described as a great river, not much inferior to
the Danube— the Jaxartes, which flows into the sea of Aral-
but as that piece of water was unknown to the ancients it is
not surprising that Herodotus represents it as reaching the
Caspian3. But the difficulty does not end here, for £ his
description of the boundary line between Europe and Asia he
says definitely that the Araxes, which lies beyond the Caspian
flows towards the rising sun4; that is to say, in a direction
exactly opposite to the course of the Jaxartes, and away from
the Caspian. Unless we accept Canon Rawlinson's somewhat
drastic suggestion that the writer here made a lapsus and
described the river as running east, when he meaat to say that
it ran west, the confusion appears to be inextricable

In his account of the geography of Asia Herodotus introduces

.
4. 40 ; 6 Apdftj jrora/ify ftw vpbs iJAiov di^ovra.