vegetation in low-lying parts, are dangerous all the year round; in such
places the water is deadly at all times unless first boiled.

This malarial poison is distinct from that which produces yellow fever
in America, and is so far unlike it as it is not contagious. The theory
is that the poison is produced below the surface by decaying vegetable
matter in low and dank parts during the more inactive but still warm and
sunny winter season and during the hot months preceding the summer
rainfall. Upon the first rains the malarial poison escapes through the
then softened crust in the shape of vapoury miasms. This happens during
the night, after the surface of the earth has been cooled off. Those
miasms are dissipated or neutralised by the action of the sun. The dewy
grass retains the poison until it is thoroughly dried to the root. All
surface water is liable to that poisonous impregnation. Malarial
manifestations occur all over South Africa, but in progressive degrees
of virulence with the advance to warmer latitudes, and with the descent
from the high table-lands to the coast levels. On the Transvaal high
veldt, for example, a mild form is developed which, in midsummer, to a
small extent, affects and kills sheep. It is called _blaauwtong_, and
does not affect horses. Descending further, this danger to sheep
increases and begins earlier. Below 5,000 feet altitude in the Transvaal
the summer season is dangerous to sheep, and horses and mules are
subject to horse sickness; whilst lower still the same malaria attains
sufficient virulence to attack human beings, and becomes very deadly
upon levels nearing the coast. Komati poort, the frontier railway
station already mentioned, is dreaded as a still worse death-trap than
even Delagoa Bay, where it is very unsafe, say, from December to end of
April. The season of horse sickness terminates upon the appearance of
the first sharp frost in May. The safeguards for human beings consist in
avoidance at night and early morning of low-lying localities, or such