ward through the loose mass and out through the meshes of the sieve, when, instead of continuing vertically, they bend toward the moisture which comes from the sawdust and keep close to the inclined surface in spite of gravity.

Fig. 5.—Apparatus to illustrate the Mode in which the Influence of Gravity is overcome by the Effect of Greater Moisture on one Side of the Root. (Sachs.) With a view to seeing whether this sensitiveness to moisture was localized like the sensitiveness to gravity, Darwin covered the tips of a number of seedlings with grease, and then subjected them to an excess of moisture on one side. No bending occurred so long as the tips remained covered. This led him to believe that sensitiveness to moisture is confined to the same part which is sensitive to gravity, and later investigators, using improved methods, have confirmed Darwin's conclusion. The lateral branches, being less controlled by gravity than the main axis, are, as might be expected, more responsive to differences in moisture. So delicate is this sensitiveness that the roots oftentimes seem to work almost intelligently in their search for water. Thus elm roots have been found filling up a drain fifty yards from the trunk, and numerous instances of roots growing into wells and choking water-pipes have been reported.

A very common effect of this special sensitiveness is to regulate the distribution of the rootlets in accordance with the water-shed from the leaves. The greater part of our trees shed the rain outward like a dome or spire, so that the region of earth best watered falls directly under what may be called the eaves: it is just here that the tips of the rootlets occur in most profusion. In the case of shrubs and herbs, which are more apt to grow close together, the water-shed is, of course, mostly indefinite, and as a consequence no regularity is apparent in the distribution of the rootlets; but even among herbs quite definite water-shed is not uncommon, and as with trees the effect upon the rootlets is well marked largely in proportion to the isolation of the plants. Certain kinds shed the water outwardly like the trees (Fig. 6, 1), while others have the leaves so disposed as to act like a funnel and carry the water toward the axial root around which the short rootlets are developed (Fig. 6, 2).