tabulate bearing in mind the circumstances tinder which the samples 

 were collected a list of soils formed solely from the beds in question; 

 for, to the reasons which operate against this in other cases, there must 

 here be added this, that the Bokkeveld beds occur mostly in valleys, and the 

 resulting soils are, therefore, as a general rule, diluted, or otherwise in- 

 fluenced, by the disintegration and decomposition products of the rocks 

 composing the surrounding hills and mountains. Hence, there are very 

 special difficulties involved in sorting out many Bokkeveld soils that could 

 really be called typical, from the samples that have been dealt with in 

 our laboratories. I have, however, made a choice of eighteen, which 

 may fairly be considered to represent solely Bokkeveld-derived soils; the 

 results of tho analyses of these are given in the following table : 



Division. 



Ladismith. 

 Robertson. 



Swellendara 



Uniondale. 

 Worcester. 



IX. BOKKEVELD SOILS. 



Fa? m. Fine earth. Water. Lime. 



Buff els fontein. 86'8 



Harmonic I. 80'5 



Harmonie II. 83'9 



Het Kruis. 80'8 



95-2 

 Baden. 49'3 



94-7 



62-7 



81-0 



72-8 



80-3 



89-0 



57-9 



93-2 



81-7 



71-7 



79-0 



70-6 



Poortfontein. 



Uitvlugt. 



Doom Rivers Vallei. 







Lemoens Hoek 

 Kamnassie Wagen Drift. 

 Wilgehouts Rivier. 

 Roode Heuvel. 

 Keurbosch Kloof. 



Stink fontein. 



On comparing the above Tables VIII. and IX., it immediately 

 becomes plain how greatly the Bokkeveld soils differ from those of the 

 Table Mountain series. While the proportions of lime, potash and phos- 

 phoric oxide practically never exceed '1 per cent, in the sandstone soils, 

 they rarely fall below that figure in the typical soils of the Bokkeveld 

 series. The average percentages of plant food in these eighteen soils work 

 out as follows: 



Lime '387 



Potash '231 



Phosphoric oxide '118 



It is difficult to say, in default of a previous thorough investigation 

 of soils which are known to bo productive in this Colony, what propor- 

 tions! of plant food may be considered to render a soil either chemically 

 rich or of normal fertility, but, basing our opinions on European and 

 American standards, we may affirm that, while the sandstone soils shew 

 an all-round poverty, our analyses reveal the average Bokkeveld soil to 

 be satisfactorily provided with all three of the essential inorganic 

 elements of plant food. 



Just here we touch upon a somewhat curious feature, As far as can 

 be gathered, the soils in Table VIII. represent undoubted sandstone soils, 

 and those in Table IX. purely Bokkeveld soils, but in various localities a 

 number of soils have been taken from near the line of junction between 

 the Table Mountain and Bokkeveld series. In some of theso instances the 



