52

soil, be such as to enable it to admit and re-tain as much water as is necessary to vegetation,and no more.

Now we have already seen, that the retentivepowers of moisture are very different in thesimple earths: therefore the proportions inwhich the fertility of a soil requires them to bemixed, must be different in climates and coun-tries that differ considerably in moisture; in thedrier, they must be such as are most retentive;in the moister, such as suffer it to pass or eva-porate more easily.

The same remark extends to situation. Landson a plain should be so constituted as to be lessretentive of water than those situated on a de-clivity ; as is very evident.

So lands that have a retentive or impermeablesub-soil, should be differently constituted fromthose that have one less retentive or more per-meable. The time of the year in which rainmost abundantly falls may also be worthy ofnotice.

These circumstances must, undoubtedly, mo-dify the conclusions that may be drawn fromthe experiments I shall now relate.