210 On the Physical Analysis of Soils. (Marcty 
T ought in the first place to explain why I submit to experiment 
the earths such as they occur in nature, rather than the pure earths 
procured by chemical process ; though I have not absolutely neg- 
lected the latter. I soon perceived that earths prepared artificially 
differ considerably in their physical properties from those which we 
meet with in nature. However, it is indispensable, in order to 
determine precisely the influence of the different earths en vegeta- 
tion, to acquire an exact knowledge of their secondary principles, 
(just as is the case in our researches on plants.) The mere know- 
ledge of the primary principles of an arable earth will be of little 
service. ‘Iwo different kinds of soil may contain the same pro- 
portion of calcareous earth, and yet retain very different propor- 
tions of moisture, become dry at very different intervals of time, 
and censequently produce a very different effect on vegetation, 
according as the lime is under the form of sand or of a fine powder. 
Thus 100 parts of calcareous sand only retain 29 per cent. of water, 
while 100 parts of the same earth in a fine powder retain 85 per 
cent. This difference is still more sensible in silica. When in 
the form of sand it retains only 25 per cent. of water; while 100 
parts of silica, as it occurs in clay, that is combined with alumina 
in each arable soil, retains 280 per cent. of water. is 
Calcareous earth and silica produce under the first form, and 
when they predominate in an arable field, a hot and dry’ soil; 
under the second they render the soil moist and colds Mere che- 
mical analysis will never be sufficient to point out these striking 
differences. 
These phenomena induced me to chuse for the subjects of my 
researches the principal species of earth, which form in greater or 
smaller quantity the upper beds of our globe, of which the arable 
soil is composed. 
1 have accordingly examined in the point of view above stated, 
besides the elements which usually constitute arable soil, namely, 
silicious sand, calcareous sand, the different kinds of clay, cal- 
eareous earth, and vegetable earth, or humus; likewise salphate of 
lime and carbonate of magnesia. I paid attention to the latter in 
consequence of the difference of opinion among philosophers 
respecting its influence on vegetation. I have likewise subjected 
some arable compositions to the same experiments by way of in- 
struction and example. 
As to the method which I followed in my experiments I send you 
the following explanation of it. 
I ascertained the specific gravities of the earths by hydrostatics. 
I weighed repeatedly a well corked flask, first filled with water, and 
then with water and earth atonce. By this means I found the com- 
parative weight of the earth to that of water, and consequently its 
specific gravity. The usual method of finding the weight of a 
determinate volume of the earth by multiplying its specific weight 
by that of water, is not applicable in this case, as it is when we 
ascertain the specific gravity of solid and cohering bodies. It was 
necessary to weigh determinate volumes of these earths in their 
—- 
