to those which are here found
on trees are found on the plains of the Kalahari, growing on mere
herbaceous plants. There are several other examples of a similar nature.
Shrubs, well known as such in the south, assume the rank of trees as
we go to the north; and the change is quite gradual as our latitude
decreases, the gradations being herbaceous plants, shrubs, bushes,
small, then large trees. But it is questionable if, in the cases of
mamosho, mobola, and mawa, the tree and shrub are identical, though the
fruits so closely resemble each other; for I found both the dwarf and
tree in the same latitude. There is also a difference in the leaves, and
they bear at different seasons.

The banks of the river were at this time appearing to greater advantage
than before. Many trees were putting on their fresh green leaves, though
they had got no rain, their lighter green contrasting beautifully with
the dark motsouri, or moyela, now covered with pink plums as large
as cherries. The rapids, having comparatively little water in them,
rendered our passage difficult. The canoes must never be allowed to come
broadside on to the stream, for, being flat-bottomed, they would, in
that case, be at once capsized, and every thing in them be lost. The men
work admirably, and are always in good humor; they leap into the water
without the least hesitation, to save the canoe from being caught by
eddies or dashed against the rocks. Many parts were now quite shallow,
and it required great address and power in balancing themselves to keep
the vessel free from rocks, which lay just beneath the surface. We might
have got deeper water in the middle, but the boatmen always keep near
the banks, on account of danger from the hippopotami. But, though we
might have had deeper water farther out, I believe that no part of the
rapids is very deep. The river is spread out more than a mile, and
the water flows rapidly over the rocky bottom. The portions only three
hundred yards wide are very deep, and contain large volumes of flowing
water in narrow compass, which, when spread over the much larger surface
at the rapids, must be shallow. Still, remembering that this was the end
of the dry season, when such rivers as the Orange do not even contain a
fifth part of the water of the Chobe, the difference between the rivers
of the north and south must be sufficiently obvious.

The rapids are caused by rocks of dark brown trap, or of hardened
sandstone, stretching