14 Titans actions of the South African Philosophical Society. 



formations must be mentioned. It does not need much discernment 

 to classify as a sandstone soil one taken from the top of Table 

 Mountain, or as a Malmesbnry slate soil one from the upper slopes 

 of Lion's Eump, but when a valley composed of beds of the Bokke- 

 veld series is flanked by sandstone mountains it becomes less easy 

 to predicate to what extent each has influenced the chemical nature 

 of the soil ; still more complicated is the problem when dealing with 

 such districts as Eobertson, where, in parts, quite a large number of 

 rocks contribute to the formation of the soil. 



Had the samples of soils examined been so selected as specially 

 to typify certain definite geological series, it is highly probable that 

 ere now much more would have been learnt regarding the nature — 

 from the farmer's standpoint — of the soils derived from each series ; 

 but circumstances have all along rendered any such system of 

 collection impracticable; indeed, as observed on a former occasion, 

 geographical rather than geological considerations had to guide us in 

 the selection of areas to be investigated. 



In spite of the fact that conditions were not propitious for the 

 selection of typical samples, it has been possible to sort out from 

 the many soils that have been analysed some to typify various 

 geological formations, and the chemical composition of the soils so 

 sorted out on the whole bear out the reasonableness of the classi- 

 fication. 



I propose, first of all, to consider the soils derived from the 

 geologically oldest rocks, and to follow the upward sequence thence, 

 as circumstances permit, ending with the superficial deposits. 



Beginning, then, with the pre-Cape rocks, as the Geological Com- 

 mission has termed them, which underlie the Table Mountain and 

 correlated series, the soils derived from the Malmesbury series 

 require first notice. I regret that by the standard method adopted 

 in our laboratory no analyses have been made of sufficiently typical 

 soils collected within the district which gives its name to this series, 

 but from the Malmesbury beds -of the Paarl and Stellenbosch 

 Divisions the fourteen soils enumerated below were taken, in such 

 places as to be practically representative. The series of rocks from 

 which these are derived consists mostly of hard close clay-slates; 

 the chemical composition of the latter is, for the most part, silicate 

 of alumina, and, therefore, practically void not only of plant food but 

 also of plant-food constituents. 



The percentage proportions of lime, potash, and phosphoric oxide 

 in the soils are what I chiefly desire to direct attention to, but there 

 are also incorporated in the list the percentages of moisture retained 

 in the air-dried soils, as well as the organic matter lost on ignition ; 



