Phvsical Comi'Ositiox of Cafk Colonv Soils. 



07 



Similar analyses have been made of ssventeen soils from the Gu\ erii- 

 ment farm Elsenburg, and of ten from the division of Steynsburg. 



More comprehensive than an}^ other series of mechanical analyses 

 undertaken by the staff of the Government analytical laboratories was 

 the investigation into the physical composition of four different types 



of soil which occur on the Government experiment station at Robert- 

 son. When the surface soils had been removed to a depth of 8 ins. 

 further samples were collected at each spot, representing every succeed^ 

 ing 8 ins. in depth, up to a total depth of 4 ft., thus making six samples 

 as representing the soil-profile at each of the four spots, or twenty-four 

 samples in all. In the tables below, the numbers 21, 22, 23 and 2-t 

 signify the surface soils to 8 ins. deep ; Nos. 2[a, 22a, 23a and 24a mean 

 the second 8 ins. of soil, and so on. 



The lands consist of alluvial valley soil, and comprise broadly 

 three somewhat different types. These three types are represented 

 respectively in the following tables, on pp. 98 and 99, by Nos. 21, 22 

 and 23. No. 21 is a red-coloured light sandy loam, resting upon a 

 denser and more claye}' subsoil, over 20 ft. in depth. No. 22 is 

 another red sandy loam of considerable depth, o\'erlying a subsoil 

 similar to that of No. 21 ; this soil has a tendenc}' towards alkalinity 

 in parts, and No. 24, in fact, represents an alkaline patch in the area 

 covered by the soil whereof No. 22 is a type. Tlie third type of soil 

 at the station is that of which No. 23 is a sample. It is a grey 

 calcareous loam, with a subsoil of almost white appearance. In this 

 soil brack water is met with a short distance below the surface. 



