V. THE ANALYSIS AND COMPOSITION OF SOILS 99 



tural importance. Sandy soils are sometimes deficient in potash, as 

 are also calcareous soils, whilst clays and loams are often well supplied. 



9. Phosphoric acid. This, the total amount of phosphorus 

 pentoxide extracted from the ignited soil by strong hydrochloric acid, 

 though perhaps not all that is actually present in the soil, certainly 

 includes all that plants are likely to obtain from it. Much of the 

 phosphoric acid of soils is doubtless present in the forms of aluminium 

 and ferric phosphates, which are only slightly assimilable by plants. 

 This is especially the case in soils poor in lime. The phosphorus pen- 

 toxide of a soil is often one of the most important constituents in de- 

 termining its potential fertility. Its amount is generally small, rarely 

 exceeding 0*18 per cent, while in some sandy soils it may fall as low 

 as O02 per cent. Most of the soils of S. Africa are notably deficient 

 in this constituent. 



10. Total. If the analysis of a soil were complete and perfectly 

 accurate, this would, of course, amount to 100 per cent. In actual 

 practice it rarely does so, and this is easily understood. In the first 

 place, certain constituents, e.g., soda, chlorine, sulphuric acid, carbonic 

 acid, manganese and a few others, may be present and are not always 

 estimated, thus tending to make the "total " of those determined less 

 than 100. Then no estimation is absolutely accurate in all experi- 

 mental determinations, errors necessarily occur, and though these 

 errors in the various items may to some extent counteract each other, 

 they do not often exactly do so. Consequently we find that the total 

 of all the various constituents is rarely, and then only by coincidence, 

 exactly 100. At the same time the approximation to this figure is a 

 confirmation of the accuracy of the analysis, and an indication that no 

 important amounts of any constituents have been overlooked. 



11. Nitrogen. This is the total amount of nitrogen, whether 

 existing as nitrates (always small), ammonia, or complex organic com- 

 pounds. It is in the last form that most of the nitrogen is stored in 

 soils, and from these compounds, a gradual supply of nitrates for the 

 plant should be maintained by the action of the nitrifying organ- 

 isms. 



12 and 13. " Available " potash and phosphoric acid. As 

 already stated, no accurate method of actually determining the 

 amounts of potash and phosphoric acid which plants can obtain from 

 a soil is known. Dyer's method, although it is admitted that it gives 

 only empiric results, is believed to furnish the best measure of these 

 amounts, and has been employed in many analyses. 



Many attempts have been made to express the fertility of a soil by 

 some figure, generally derived by giving arbitrary values to each of the 

 important plant food constituents and adding these together. But 

 these attempts have not been successful and have never met with 

 much acceptance. Indeed it is difficult to see how any numerical ex- 

 pression of fertility could be arrived at, even if only the chemical com- 

 position of the soil be considered. Practical fertility, depending as it 

 does upon so many circumstances, would be still harder to express. 

 The physical and mechanical properties of the soil are difficult to 

 measure, and when the size of particles, water retentive power, etc., 

 are measured, it is still almost impossible to interpret them. Much 



