8 Transactions of the South African Philosophical Society. 
and a coarse, starved-looking grass, with the red sand upon which it 
grows in evidence everywhere. 
Now although the veldt is at best scarcely more than half-covered 
with vegetation, and the climate therefore as much influenced by 
the soil as by the vegetation upon it, yet it seemed d@ priori likely 
that the climate might be affected by the drastic clearance of such 
vegetation as there was, and also by the substitution of blue-ground 
for red over the floors. Such is the problem, and a solution of some 
of its leading aspects is attempted in this paper. 
Relative Thermal Properties of Kimberlite and Red Sand. 
The thermal properties of blue-ground as compared with red sand 
have first to be considered. 
Two small patches, each about 30 inches square, were laid out 
side by side in the most exposed spot available, one of weathered, 
decomposed, virgin blue-ground, the other of red sand. These were 
made as level and smooth as possible. The depth of the deposit of 
blue-ground was about 3 inches, the space beneath and surrounding 
it being red sand. A mercurial Board of Trade thermometer was 
placed in the midst of the blue patch, the centre of its spherical 
bulb being almost exactly 1 inch beneath the surface, and another 
exactly similar and similarly situated in the centre of the red patch. 
These thermometers were both inclined at an angle of about 30° from 
the vertical for convenience in reading, and kept in place by a thin 
wooden prop behind. A thin wooden lath also covered each column 
and scale as a protection against possible hail and missiles. The 
readings were taken at VIII., XIV., and xx. civil time. A spirit radiation 
thermometer of the ordinary pattern was also placed in a horizontal 
position over each patch, and supported by the thinnest wooden forks 
strong enough to carry it, the centre of the spherical bulbs being 
almost exactly three-quarters of an inch above the surface. These 
were also read at VIII., XIV., and xx., and readings of the minimum 
were also taken as well. The experiment lasted from July 4th con- 
tinuously till September 30, 1898, and may be regarded as including 
more or less all types of Kimberley weather. These were the best 
arrangements I was able to make for the purpose. There is no 
doubt, of course, that if larger patches could have been laid out, say 
each of 30 feet square, or better still, if simultaneous observations 
could have been taken, one set on the open veldt and the other in 
the midst of a depositing floor, better results must have been obtained. 
However, the figures given in Table I. will show the general tendency 
of the climatic effects of the two soils. 
