soLrnn.iTN OK I'liosi'iioKic oxiDK. 205 



conducted to ascertain whether this sliijht increase of the water- 

 soluhle j)hosphates in mixture (a) could be accounted for. A 

 his::h-grade superphosphate (manufactured in Japan) containinji 

 19.37 l^'' ^^ent. of water-.soluhle and 20.39 I'^'i* •-■fiit. of total 

 {)hos])horic oxide, and a total of 28.03 P^^" cent, of combined lime 

 was employed in this and all subse([uent work, and the tests were 

 carried out on similar lines to those already detailed. The sul- 

 j)hate of ammonia emi)kned in this set of analyses was the 

 chemicallv pure salt, not the commercial article as usually sold 

 for fertilising^' purposes, which was employed in the pre\ious 



Here again, as in the case of mixture (a), it will be seen 

 that a similar drop in solubility took place within a few hours, 

 from which there was a recovery after the mixture had stood 

 for one week. The increase in solubilitx noticed in the i)revii)us 

 experiment did not occur. 



When, however, our attention is turned to the (Govern- 

 ment guano mixture (rf). it will be seen that the actual loss of 

 water-soluble phosphates is 21 i)er cent, of the amount originally 

 present. It appeared likely that the large amount of ammonium 

 salt, existing probably as ammonium carbonate in the guano, 

 would account for the high percentage of reversion. St; the 

 experiment detailed hereunder was carried out. 



Ammonium carbonate — extra pure — was freed from ad- 

 herent moisture by extracting tirst with absolute alcohol. The 

 alcohol was allowed to drain otT. and the residue then taken u]) 

 with chemically pure ether. The ether was allowed to eva])()rate 

 oiT in the sun, the residual carbonate was then powdered, and 

 after sifting through a i mm. sieve, it was mixed with an e(|ual 

 weight of superphosphate. This mixture was placed in a glas.^- 

 stoi)pere(l bottle, and the gas (carbon dioxide) was driven oft' 

 rapidly by the action of the acid of the superphosphate. After 

 standing as in i)revious experiment, the following percentages 

 of water-soluble |)hos])horic oxide were obtained: — 



Clearly, therefore, the consi(leral)le decrease of solubility 

 when superphosphates are mixed with Government guano is due 

 to the presence of ammonium carbonate in the guano. 



In the case of kainit (iray remarks that this reversion may 

 be put down mainly to magnesium salt. The most interesting 

 mixture of all is that of superjjhosphate and bonemeal in the 

 above mixtures prepared by me: the reversion had amounted »o 



