202 SOLL'I!IIJT\ OF PI l( jSF'iiORIC OXIDE. 



The sulphuric acid thus set free would dissolve more of the 

 iron compounds, and so react again and again. In this case we 

 have a continuously increasing quantity of insoluble iron ])hos- 

 phate. and a proportionately diminishing amount of .soluble 

 phosphates.* 



On the other hand, the soluble phosphates may be rendered 

 insoluble much more rapidly, thus : 



Fe.O, + H,Ca P,0, == 2Fe;PoO, + Ca,P,( ). 

 Ferric Super- Ferric Calcium 



Oxide. phosphate. phosphate. phosphate. 



The above would apply in the case of alumina as well 

 as for iron compounds. 



But the solubility of the phosphoric oxide in superphosphate 

 i'-" influenced by other factors than merely calcium carbonate, 

 tricalcium phosphate, and compounds of iron and alumina. To 

 test the extent to which this solubility is affected the following 

 mixtures were prepared: 



{a) Superphosphate and sulphate of ammonia. 

 ib) Superphosphate and sulphate of potash. 

 ( c ) Superphos])hate and kainit. 



(d) Superphosphate and guano from the Union Govern- 

 ment Islands off the South African coast. 

 (c) Superphosphate and bone meal. 



In the Cape Agricultural Journal of June. 1910, there was 

 published a diagram — stated to have been compiled on exper- 

 ience and ])ractice in South Africa— shewing what artificial fer- 

 tilisers could be mixed together without any harmful effects, and 

 what were the mixtures best suited for the farmer. 



On consulting this diagram, it will be seen that the mixtures 

 made by me were those there si)ecially recommended to be made 

 by farmers, with the exception of (c), the mixture of sujjer- 

 l)hos])bate with l)one meal, a combination marked as unnecessary, 

 altbougb sometimes made use of by farmers. 



The above mixtures were made to contain ecjual ])arts by 

 weight of each of the two constituents: these were well mixed 

 and at once placed in glass liottles, closed by means of corks, 

 and the portions were weighed out after the mixture had stood 

 for the re(|uired time. Ik'forc each weighing the sample was 

 again well mixed. 



The sample of su])crplu)sphate used throughout the first 

 part of these experiments was a high-grade article contaitnng 

 17.35 ])er cent of water-soluble pliosi)horie oxide and 2(.).()2 per 

 cent, of lime. 



The guano used represented the average article as sold by 

 the Tnion ( lox ernment. It was a mixture from Malagas, Marctis 

 ;.ii(l Jutten Islands, and had the I'ollowing composition: 



* Aikman : "Manures and Manuring." Second Impression, 399. 



