CHEMICAL COMPOSITION OF KARROO ASH. 1 39 



these two samples, it is plain that the raw article from which 

 No. 23 was derived was the purer, notwithstanding the farmer's 

 expectations. 



Towards the end of 1901 another sample of Karroo ash 

 (No. 25) was received in the laboratory, but I am not sure of 

 the locality whence it came. It contained 13.22 per cent, of 

 sand. 



Another kraal manure from Laingsburg (No. 26) was 

 analysed in 1902. 



For several years after this the subject of kraal manure and 

 Karroo ash remained dormant, but in 1909 another sample of 

 kraal manure from the Fraserburg Division (No. 27) was 

 obtained and analysed. The manure, like some of the others 

 above referred to. contained a considerable amount of moisture 

 (vis., 37.5 per cent.), and it was therefore first of all allowed 

 to dry by exposure to air. The air-dried manure, which still 

 contained 17.92 per cent, of moisture, yielded 29.05 per cent, 

 of ash, and about 18 per cent, of the ash consisted of potash. 

 Of course, this proportion could not be expected in the crude 

 Karroo ash of the farm. 



In 191 1 three samples of kraal manure and four of Karroo 

 ash were analysed in the laboratory, at the request of the Paarl 

 Farmers" Association, in connection with its w^ork as a medium 

 of distribution for the Western Province farmers. The un- 

 burnt samples were : — 



28. From Letjesbosch Siding, Fraserljurg Division. 



29. Very wet sample from Letjesbosch Siding. 



30. From Stein's Siding, Beaufort West Division. 



These samples w^ere allowed to become air-dry, after which 

 the percentages of moisture, ash. and plant-food constituents 

 in each were determined, and the results given in Table I repre- 

 sent the manure after air-drying. No. 30 was certainly the 

 worst of these three samples. 



The four Karroo ashes analysed at the same time Avere as 

 follows : — 



31. A dry, well-burnt sample from Tromps Graf Siding, 



Victoria West Division (freshly burnt). 



32. A dry, well-burnt sample from an old heap of ash on 



the same farm as No. 31. 



33. A well-burnt ash from Stein's Siding. Beaufort West 



Division. 



34. From Letjesbosch Siding, Fraserburg Division. 



No. 34 contained no less than 23.36 per cent, of pebbles over 

 I mm. in size, together with orange peel, acorns, oak leaves, and 

 feathers, but had apparently, before receiving all those admix- 

 tures, been a w^ell-burnt ash. On account of the quantity of 

 small stones in this sample, the percentages of plant food after 

 removal of the pebbles were calculated, and worked out as fol- 

 low^s : — 



