352 HOW CROPS FEED. 
The water of the soil holds silica, lime, magnesia, alka 
lies, and oxide of iron, often alumina, in solution. In- 
stances are numerous in which the evaporation of water 
containing dissolved salts has left a solid residue of sili- 
cates. Thus, Kersten has described (Jour. fiir prakt. 
Chem., 22,1) a hydrous silicate of iron and manganese 
that occurred as a hard incrustation upon the rock, in one 
of the Freiberg mines, and was deposited where the water 
leaked from the pumps. Kersten and Berzelius have no- 
ticed in the evaporation of mineral waters which contain 
carbonates of lime and magnesia, together with silica, that 
carbonates of these bases are first deposited, and finally 
silicates separate. (Bischof’s Chem. Geology, Cav. Ed., 
Vol. 1, p. 5). Bischof (/oc. e/t., p, 6) has found that silica, 
even in its most inactive form of quartz, slowly decom- 
poses carbonate of soda and potash, forming silicate when 
boiled with their aqueous solutions. Undoubtedly, simple 
contact at ordinary temperature has the same effect, 
though more slowly and to a slight extent. 
Such facts make evident that silica, lime, the alkalies, 
oxide of iron and alumina, when dissolved in water, if they 
do not already exist in combination in the water, easily 
combine when adverse affinities do not prevent, and may 
react upon the ingredients of the soil, or upon rock dust, 
with the formation of zeolites. 
The “pan,” which often forms an impervious stratum 
under peat bogs, though consisting largely of oxide of iron 
combined with organic acids, likewise contains consider- 
able quantities of hydrated silicates, as shown by the 
analyses of Warnas and Michielsen (Mulder’s Chem. d. 
Ackerkrume, Bd. 1, p. 566.) 
Mulder found that when Portland cement (silicate of 
lime, alumina, iron, etc.) was treated with strong hydro- 
chloric acid, whereby it was decomposed and in part dis- 
solved, and then with ammonia, (which neutralized and re- 
