s between the 20th of May and the 1st of November.
The rain averages from 12 to 15 inches per annum. In 1852 it
was 12.034 inches; in 1853, 15.473 inches. Although I had no
means of measuring the amount of rain which fell in Londa, I
feel certain that the annual quantity exceeds very much that
which falls on the coast, because for a long time we noticed
that every dawn was marked by a deluging shower, which began
without warning-drops or thunder. I observed that the rain
ceased suddenly on the 28th of April, and the lesser rains
commenced about a fortnight before the beginning of November.

From information derived from Arabs of Zanzibar, whom I met at Naliele
in the middle of the country, the region to the east of the parts of
Londa over which we have traveled resembles them in its conformation.
They report swampy steppes, some of which have no trees, where the
inhabitants use grass, and stalks of native corn, for fuel. A large
shallow lake is also pointed out in that direction, named Tanganyenka,
which requires three days for crossing in canoes. It is connected with
another named Kalagwe (Garague?), farther north, and may be the Nyanja
of the Maravim. From this lake is derived, by numerous small streams,
the River Loapula, the eastern branch of the Zambesi, which, coming from
the N.E., flows past the town of Cazembe.

The southern end of this lake is ten days northeast of the town of
Cazembe; and as that is probably more than five days from Shinte, we
can not have been nearer to it than 150 miles. Probably this lake is
the watershed between the Zambesi and the Nile, as Lake Dilolo is that
between the Leeba and Kasai. But, however this may be, the phenomena of
the rainy season show that it is not necessary to assume the existence
of high snowy mountains until we get reliable information. This, it is
to be hoped, will be one of the results of the researches of Captain
Burton in his present journey.

The original valley formation of the continent determined the northern
and southern course of the Zambesi in the centre, and also of the
ancient river which once flowed from the Linyanti basin to the Orange
River. It also gave direction to the southern and northern flow of the
Kasai and the Nile. We find that between the latitudes, say 6 Deg. and
12 Deg. S., from which, in all probability, the head waters of those
rivers diverge, there is a sort of elevated partition in the great
longitudinal