Soils of the South-Western Districts of the Cape Colony. 155 
two soils; both are poor in lime, it is true, yet the lime in H is nearly 
double that in E’. The quantity of potash in Ei is also quite normal, 
though it is but moderate in amount in H’. 
Under all the circumstances, to revert again to the Malmesbury 
soils, it seems quite feasible that the process of levelling down the 
hillocks—said to have been attended with success in one instance— 
may in some cases lead to an all-round increase in fertility, notably 
where the hillocks are numerous. 
It appears reasonable to suppose, from what these results reveal 
that an addition of lime would lead to an improvement. In the 
Malmesbury District there are numerous outcrops of lime—for 
instance, on the farms Drooge Vlei, Geelbeksfontein, Springfontein, 
and Lang Riet Vlei. Even if levelling down does not achieve the 
desired result, there should be plenty of lime near at hand on which 
to draw for a supply. 
It may be useful to compare the average percentage composition 
of the hillock soils with those from below. The results, calculated 
on the unsifted soils, are as follows :— 
Phosphoric 
Lime. Potash. Oxide. 
Ehilloeck ground |) 3s..4. 078 073 051 
evel around =. ..c102 010 ‘061 "032 
Calculated in pounds per acre to a depth of 6 inches of soil, this 
would amount to :— 
Phosphoric 
; Lime. Potash. Oxide. 
Hillock ROUT Mees 1,560 1,460 1,020 
Hevelvsroumdly x ssssa: 200 1,220 640 
Hence, generally speaking, to bring the latter soils up to the 
fertility of the former, they would require per acre over half a ton 
of lime, together with about 200 lbs. of potash and 400 lbs. of 
phosphoric oxide; in other words, the equivalent of a ton each of 
Kainit and Thomas’ phosphate per acre. At the same time it would 
not do to rest content with these additions, for so we would only be 
levelling up the general fertility to that of the hillocks, which are 
themselves in want of improvement. In fact, the lime in the latter 
may safely be trebled and the potash and phosphoric oxide doubled. 
Reviewing the entire area covered by these analytical investiga- 
tions, it is difficult, if not impossible, to trace well-defined family 
resemblances between the members of a series of soils overlying the 
same geological formation; similarity between samples is geographical 
rather than geological. For instance, the Caledon soils all present 
a a 
