ld be anticipated. The
observation is made by passing a measured quantity of water
backwards and forwards for some months through a tube containing
a few grammes of powdered rock. Finally the water is analysed,
and in this manner the amount of dissolved matter it has taken up
is estimated. The rock powder is examined under the microscope in
order to determine the size of the grains, and so to calculate
the total surface exposed to the action of the water. We must be
careful in such experiments to permit free oxidation by the
atmosphere. Results obtained in this way of course take no
account of the chemical effects of organic acids such as exist in
the soils. The quantities obtained in the laboratory will,
therefore, be deficient as compared with the natural results.

In this manner it has been found that fresh basalt exposed to
continually moving water will lose about 0.20 gramme per square
metre of surface per year. The mineral orthoclase, which enters
largely into the constitution of many granites, was found to lose
under the same conditions 0.025 gramme. A glassy lava (obsidian)
rich in silica and in the chemical constituents of an average
granite, was more resistant still; losing but 0.013 gramme per
square metre per year. Hornblende, a mineral

36

abundant in many rocks, lost 0.075 gramme. The mean of the
results showed that 0.08 gramme was washed in a year from each
square metre. Such results give us some indication of the rate at
which the work of solution goes on in the finely divided
soils.[1]

It might be urged that, as the mechanical break up of rocks, and
the production in this way of large surfaces, must be at the
basis of solvent and chemical denudation, these latter activities
should be predominant in the mountains. The answer to this is
that the soils rarely owe their existence to mechanical actions.
The alluvium of the valleys constitutes only narrow margins to
the rivers; the finer _debris_ from the mountains is rapidly
brought into the ocean. The soils which cover the greater part of
continental areas have had a very different origin.

In any quarry where a section of the soil and of the underlying
rock is visible, we may study the mode of formation of soils. Our
observations are, we will suppose, pursued in a granite quarry.
We first note that the material of the soil nearest the surface
is intermixed with the roots of grasses, trees, or shrubs.
Examining a handful of this soil, we see glis