358 HOW CROPS FEED. 
contact with the soil, the base may be withdrawn by the 
absorbent silicate, and the acid may unite to lime or mag- 
nesia, The phosphates of lime and magnesia thus formed 
are, however, insoluble, and hence the acid as well as the 
base remains fixed. Again, if the alkali-phosphate be 
present in quantity so great that its base cannot all be 
taken up by the absorbent silicate, then the hydrated 
oxide of iron or alumina may react on the phosphate, chemi- 
cally combining with the phosphoric acid, while the alkali 
gradually saturates itself with carbonic acid from the air. 
It is, however, more likely that organic salts of iron (cre- 
nates and apocrenates) transpose with the phosphate. So, 
too, carbonate of lime may decompose with phosphate of 
potash, producing carbonate of potash and phosphate of 
lime (J. Lawrence Smith). Veelcker, in a number of ex- 
periments on the deportment of the soluble superphosphate 
of lime toward various soils, found that the absorption of 
phosphoric acid was more rapid and complete with soils 
containing much carbonate of lime than with clays or 
sands. 
All observers agree that phosphoric acid is but slowly 
fixed by the soil, Veelcker found the process was not 
completed in 26 days. Its absorption is, therefore, mani- 
festly due to a different cause from that which completes 
the fixation of ammonia and potash in 48 hours. 
As to silicic acid, it may also, as solid hydrate, unite 
slowly with the oxides of iron and with alumina (see Kers- 
ten’s observations, p. 352). When occurring in solution, as 
silicate of an alkali, as happens in dung liquor, it would 
be fixed by contact with solid carbonate of lime, silicate 
of lime being formed (Fuchs, Kuhlmann), or by encoun. 
tering an excess of solutions of any salt of lime, magnesia, 
iron or ammonia. In presence of free carbonic acid in 
excess, a carbonate of the alkali would be formed, and the 
silicic acid would be separated as such in a nearly insoluble 
