166 



underlying rock belongs to the former series, in others to the latter, but 

 in both cases a peculiarity is noticeable in the composition of these 

 " junction " soils, the nature of which will become evident upon glancing 

 through the next table. 



X. BOKKEVELD NEAR JUNCTION WITH TABLE MOUNTAIN SANDSTONE; OR 



Vice versa. 



i'ivision. 



Ladismith. 

 Robertson. 



Swellendam. 





 Worcester. 



Farm. 



Papenkuilsfontein. 

 Donkerkloof. 



M 



Harmonie I. 

 Het Kruis Pad. 



Het Goed Geloof. 

 Barrydale. 

 De Dooms. 

 Klein Straat. 

 Ezeljacht. 



Unfortunately, not more than eleven soils are available for the pur- 

 pose of illustrating the point involved. It will, however, be seen that, 

 while these eleven soils resemble those of the Table Mountain series in 

 regard to the amounts of lime and phosphoric oxide contained, they 

 approach the Bokkeveld soils in respect of their potash content. This 

 feature is, perhaps, better distinguished on taking the averages, which, 

 for Table X., are as follows : 



Lime 



Potash ... . 



Phosphoric oxide 



042 

 141 

 075 



If these were simply Bokkeveld soils diluted by alluvial sand from 

 the Table Mountain series, one would expect a lowering of all three of 

 the plant food constituents in the same proportion; but, as we see, the 

 reduction of lime and phosphoric oxide is proportionately much greater 

 than that of the potash. It was puzzling at first to account for this, but 

 one explanation seems feasible. Dr. Corstorphine, in his second report 

 on the Geological Survey of the Colony,* remarks on the distinctive char- 

 acter of the lowest beds of the Bokkeveld series, which enables one with 

 little or no difficulty to draw the line of separation between it and the 

 Table Mountain series below. The basement beds of the Bokkeveld 

 series, he says, are felspathic and micaceous, and quite different from the 

 hard compact quartzite of the older series. I am not aware whether it 

 is potash felspar that is here alluded to, and our analyses, as before ob- 

 served, do not furnish data with regard to the total amounts of the plant 

 food constituents in the soil, but it seems just possible that in this fact 

 lies the explanation of the predominance of potash in what appear to be 

 primary soils collected near the common horizon of the two series. It 

 would certainly be of interest to investigate the subject more closely, so 

 as to confirm or disprove the hypothesis here suggested. For this pur- 

 pose samples of soil would have to be collected in such a manner as to 

 render it absolutely certain that they were primary soils representing the 

 vicinity of the line of junction; in the past the collection of soils hag 

 been carried on largely without regard for geological considerations. The 

 time has now arrived, I think, for greater precision in this respect. 



* Ann. Kept. Geol. Commission, 1897, p. 16. 



