On the Food of Plants. 
540 
which is lastly re-dissolved in acid, precipitated by ammonia, collected, 
washed, ignited, and weighed. 
The alkaline aluminous solution is mixed with a solution of hydro- 
chlorate of ammonia, the alumina washed, dried, and after ignition 
weighed. 
The ammoniacal solution from which the two last bodies have been 
separated is reduced to a small bulk, the lime estimated by oxalate of 
ammonia, and the magnesia by phosphate of soda and ammonia in the 
usual manner. A little oxide of manganese sometimes occurs, which 
requires, when its quantity is at all notable, an additional process for its 
extraction. 
Operating thus and using great care, I was surprised to observe that 
sometimes a deficiency was to be found in the result of the analysis very 
far greater than could be explained by any error of experiment. This 
happened with three strong clay soils from Worcestershire belonging to 
the new red sandstone formation, and as more than one repetition of the 
experiment gave the same result, there could be no other conclusion 
drawn but that something had been overlooked ; this could only be alkali 
in an insoluble condition. 
To settle the point, a proper quantity of finely divided matter was 
mixed with five times its weight of pure precipitated carbonate of lime, 
and ignited to whiteness in a platinum crucible for half an hour : the 
strongly aggregated mass was then treated with dilute hydrochloric acid, 
evaporated to dryness, and the silica separated, washed, dried, ignited, 
and found to agree in quantity with that obtained by the other process. 
The solution was mixed with excess of carbonate of ammonia, filtered, 
the soluble portion evaporated to dryness and ignited to expel the am- 
moniacal salt. The residue in the crucible, freed from a little lime and 
magnesia which yet remained, consisted of a white saline substance 
having all the characters of chloride of potassium, and whose quantity, 
when referred to potash, was just sufficient to make up the above-named 
deficiency. 
This experiment, which was repeated several times with the same 
success, suffices to demonstrate the existence, in the soils examined, of 
potash in an insoluble condition to an extraordinary extent, doubtless as 
silicate, and which in the cases referred to amounts to nearly 2 per cent, 
of the whole. It is much to be desired that this should be further 
investigated. 
It would be exceedingly difficult to determine with precision the 
quantity of phosphoric acid in a soil* — a question of great importance. 
The following process answers exceedingly well for the detection of this 
substance, and also for getting some idea of its scarcity or abundance : — 
A suitable quantity of the soil, about 500 grains, is boiled for some 
hours in a few ounces of water containing 100 grains of pure carbonate 
of potash, by which a deep brown solution is obtained and ammonia 
disengaged. This solution, after being filtered, is evaporated to dryness 
and ignited in a platinum crucible to char and destroy the organic 
* This opinion has not been altered by a careful perusal of Sprengel's 
elaborate treatise on the subject of the analysis of soils, Oct. 1843. 
