s
Halfoon and Guardafui the country is fine and well watered with small
streams, and the climate delicious, as is the coast from Cape Guardafui
westward to Berbera.

Harrar stands in a beautiful, fertile, and well-watered valley,
surrounded with hills, the soil rich, and producing fine coffee
abundantly. It is strictly Mahommedan, and, comparatively speaking, a
considerable place, though much shorn of its dominion and power from
those days when it had become the capital of that portion of eastern
Africa ruled by the Mahommedans; and when under Mahommed _Gragne_,
(left-handed,) it overran and desolated the whole Abyssinian empire,
then under that unfortunate sovereign King David. In the county south of
Berbera there is abundance of fine wells of excellent water. Waggadeyn
is a very beautiful country, and produces abundance of myrrh and
frankincense, as in fact every portion of the eastern horn, from Enarea
inclusive, also does. It is the great myrrh and frankincense country,
from which Arabia, Egypt, Judea, Syria, and Tyre were supplied in early
days of Scripture history. The Webbe is only six fathoms broad and five
feet deep in the dry season in Waggadeyn; but in the rainy season the
depth is increased to five fathoms. It is navigated by rafts lower down.
Incense, gum, and coffee, are every where abundant around the Webbe and
its tributary streams. Harrar contains about 14,000 inhabitants, and
Berbera 10,000; Sakka about 12,000.

All the early Arabian writers pointedly state, and so also do the
Portuguese discoverers, that the Webbe entered the sea _near Mukdishu_
or Magadoxo. This was no doubt the fact; but from what cause we know
not, the river, after approaching within a short distance of Magadoxo to
the north, turns south-west, and approaching in several places very near
the sea, from which it is only separated by sandhills, it terminates in
a lake about halfway between Brava and the Jub. This is Christopher's
account; but my opinion is, that this lake communicates with the sea
during the rainy season, and even in a small stream in the dry season
also. Christopher pointedly states, that besides filtrating through the
sandhills, it communicates with the sea in two places, between Merka and
Brava; and that this is correct, is proved from the fact, that while the
river near Merka is 175 feet broad, it is reduced to seventy-five feet
near Brava; while the _Geographical Bull._, No. 98, p. 96, states, that
a small river