taken up soluble matter, may rise again to the surface bringing the soluble matter up and leaving it at the surface on evaporation. Up to a certain point this is favorable to the plant; beyond the critical point, it begins to be harmful, as abundantly shown in the "alkaline" efflorescences of arid regions.
Besides the water that goes through the soil into the subdrainage, and that which runs off on the surface, enough must be held at all times in the soil during the growing season to supply the plants, and yet not enough to water-log the soil.
Here, then, are a series of possible excesses and deficiencies, between which lie the golden means which give best results. The problem of soil-management thus appears to be a problem of proper balancings and adjustments.
The key to the problem lies in due control of the water which falls on each acre. This water is an asset of great possible value. It should be the habit of every acre-owner to compute it as a possible value, saved if turned where it will do good, lost if permitted to run away, doubly lost if it carries also soil values and does destructive work below. Let us repeat the story of its productive paths. A due portion of the rainfall should go through the soil to its bottomto promote soil-formation there; a due portion of this should go on into the under-drainage, carrying away harmful matter; a due portion should go again up to the surface carrying solutions needed by the plants; a due portion should obviously go into the plants to nourish them; while still another portion should run off the surface, carrying away a little of the leached soil matter. There are a multitude of important details in this complex of actions, but they must be passed by; the great features are clear and imperative.
Experimental studies have shown that, on the average within our domain, crops can use to profit all the rainfall during the growing season, and much or all of that which can be carried over from the non-growing seasons. This greatly simplifies the complex problem, for the highest crop-values will usually be secured when the soil is made to absorb as much of the rainfall and snowfall as practicable. There are, of course, many local exceptions. In securing this maximum absorption and internal soil-work, the run-off, and hence the surface wash, will be reduced to a minimum. It has already been seen that the wash of even this inevitable minimum is likely to be still too great to keep the proper slow pace with soil-generation, when the surface has much slope. Except on very level ground and on lodgment surfaces, there need be no solicitude about a sufficient removal of the soil surface. The practical problem then lies almost wholly in retaining and passing into the soil the maximum of the precipitation. Obviously this gives the minimum of wash to foul the streams, to spread over the bottom