162 
fine siliceous sand, and 11 parts of impalpable 
matter, which was a mixture of ferruginous clay, 
with carbonate of lime. Vegetable or animal 
matters, when finely divided, not only give co- 
herence, but likewise softness and penetrability ; 
but neither they nor any other part of the soil 
must be in too great proportion ; and a soil is 
unproductive if it consist entirely of impalpable 
matter. 
Pure alumina or silica, pure carbonate of lime, 
or carbonate of magnesia, are incapable of sup- 
porting healthy vegetation. 
No soil is fertile that contains as much as 19 
parts out of 20 of any of the constituents that have 
been mentioned. 
It will be asked, are the pure earths in the soil 
merely active as mechanical or indirect chemical 
agents, or do they actually afford food to the 
plant? This is an important question; and not 
difficult of solution. 
The earths consist, as I have before stated, of 
metals, united to oxygene ; and these metals have 
not been decomposed ; there is consequently no 
reason to suppose that the earths are convertible 
into the elements of organized compounds, into 
carbon, hydrogene, and azote. 
Plants have been made to grow in given quan- 
tities of earth. They consume very small portions 
only ; and what is lost may be accounted for by 
the quantities found in their ashes ; that is to say* 
it has not been converted into any new products. 
The carbonic acid united to lime or magnesia^ 
if any stronger acid happens to be formed in the 
