VELD ESTIMATION. 223 



and Kokstad, 31-6: 684; Molteno and Mooi River, 33-3: 66G ; 

 Molteno and Rosetta, 32-2: 67-8; Eosetta and Mooi River, 

 361: 63-9; any three areas, 264: 73-6; four areas, 351: 64-9; 

 average, 35-1 : 64-9. 



(b) Covmwn Genera : Common Species: — ^Mooi River 

 and Kokstad, 441: 55-9 • Rovsetta and Kokstad, 44-3: 55-7; 

 Molteno and Kokstod, 44'3 : 55"7 ; Mooi River and Molteno, 

 43-9: 561; Rosetta and Molteno, 444: 55-6; Rosetta and Mooi 

 River, 44-9 : 551 ; any three areas, 44-5 : 55-5 ; four areas, 

 444 : 55-6 ; average, 44-3 : 55-7. 



The fig-ures for " Total Genera : Common Genera " and 

 "Total Species : Common Species" have been omitted, as 

 the results are of the same order as those given above (a), and 

 were not unanticipated, the ratios expressing the variations 

 due to purely local conditions. 



The figures for Common Genera : Common Species are 

 remarkable, giving practically a constant, and seem to sug'gest 

 some biological relation. To te«t this, I compared many non- 

 dunziekte lists with these, and no correlation was found to 

 obtain. I have therefore postulated that: " A farm is a 

 potentially dunziekte area if the systematic list analysis of its 

 flora, when compared with the floral systematic lists of several 

 well-defined dunziekte farms, shows a ratio of common genera 

 to common species of 44-3: 55-7; and, further, this area may 

 become actively dunziekte-carrying by oA'er-stocking and 

 burning for a few seasons." This latter statement is the result 

 of observations and analysis of a great many dunziekte areas. 



The ecological analysis further enabled one to state that 

 dunziekte is associated with the vegetation characteristic of a 

 definite geological formation, namely, the predominance of 

 Karroo shales (carrying fossils of Glossopteris spp.) on the 

 areas in question. 



But both of these methods failed in themelves to supply 

 the information required, that is, where a condition of the veld 

 (proportion of species present, grazing value, botanical growth 

 state as apart from species lists) was as much likely to be a 

 factor in the causation of the disease as well as, or irrespective 

 of, any specific plant in the area. This seems to be so with 

 stijfziekte, and in some degree with most other such diseases. 



In illustration, three adjacent paddocks on the same ridge 

 at Lidgetton, Natal, are here instanced, each owned by 

 different farmers. 



Paddock A. — Forty-five acres. Almost unbur ed and but 

 slightly grazed ^for thirty years. Floral composition and state 

 nearly that of the primitive veld. Stijfziekte free. 



Paddock B. — Forty-three acres. Burned' and over-stocked 

 (cattle) for thirty years. Greatly altered veld. Definitely 

 stifjziekte carrying. 



Paddoch C. — ^Forty acres. Burned seven odd years, mown 

 several years, stocked eight years, then fallow eight years, 

 then over-stocked two years, veld altering, slight stifjziekte. 



To estimate the differences in such paddocks, careful 



