1. If earths do not contribute directly to the food of plants, then would be all soils alike productive; or in other words, if air and water exclusively supply this food, then would a soil of pure sand be as productive as one of the richest alluvion.

2. Though plants may be made to grow in pounded glass, or in metallic oxides, yet is the growth, in these, neither healthy nor vigorous; and,

3. All plants, on analysis, yield an earthy product;[12] and this product is found to partake most of the earth that predominates in the soil producing the analyzed plant; if silica be the dominant earth, then is the product obtained from the plant silicious; if lime prevail, then is the product calcareous, &c. &c. This important fact is proved by De Saussure.

1st Experiment.

Two plants (the pinus abies) were selected, the one from a calcareous, the other from a granitic soil, the ashes of which gave the following products;

GraniticCalcareous
soil.soil.
Potash3 6015
Alk. and mu. sul4 2415
Carbonate of lime46 3463
Carbonate of magnesia6 7700
Silica13 4900
Alumina14 8616
Metallic oxides10 5200

2d Experiment.

Two Rhododendrons were taken, one from the calcareous soil of Mount de la Salle, the other from the granitic soils of Mount Bevern. Of a hundred parts, the former gave fifty-seven of carbonate of lime and five of silica; the latter, thirty of carbonate of lime, and fourteen of silica.

3d Experiment.

This was made to determine whether vegetables, the product of a soil having in it no silica, would, notwithstanding, partake of that earth.—Plants were accordingly taken from Reculey de Thoiry, (a soil altogether calcareous) and the result was a very small portion of silica.