ROOTS AND RHIZOIDS 507 



of other plants, though that claim often is made. . In the light of the experimental 

 fact that the optimum development of roots occurs in a comparatively moist soil, 

 it seems unlikely that plants of a given species should have longer roots in dry soil 

 than elsewhere, although it is possible that for certain species the optimum develop- 

 ment may be found where the percentage of water in the soil is comparatively low.' 

 All that may be stated with certainty is that the roots of xerophytes, as compared 

 with those of mesophytes, are reduced much less than are the shoots, a phenomenon 

 that is not especially difficult to understand. In arid regions the aerial organs of 

 plants are subject to excessive transpiration, a process that retards growth to an 

 amazing extent ; the roots of desert plants, on the other hand, are relatively free 

 from the inhibitory influence of transpiration. 



Whatever may be true of roots in dry soils, there is no doubt that roots in 

 swampy soils are short, not only as compared with the shoots, but actually short, as 

 compared with roots in other habitats. Probably the meager development of 

 swamp roots is explained by such factors as insufficient oxygen, soil toxicity, and 

 low temperature. Growth is known to be inhibited by lack of oxygen, a gas in 

 which swamp soils are relatively poor. If root excretions and products of plant 

 decay contain deleterious or toxic substances, the poor drainage and oxidation in 

 swamps would lead to their excessive accumulation. Low temperature, at least in 

 comparison with that of the air, characterizes bog soils, and it-'is known that low 

 temperature retards root development as well as absorption. 



If further study should show that generally the roots of xerophytes are not 

 only relatively but absolutely more extensive than are those of mesophytes, it 

 is not to be concluded, in the face of opposing experimental data, that long 

 roots are a reaction to arid soils. It is possible that such xerophytes are by 

 inheritance long-rooted. If a hemlock and a red cedar are grown side by side in 

 similar conditions, the former has a meager and the latter an extensive root system. 

 In the course of evolution short- and long-rooted species are likely to have originated 

 in deserts ; of these the long-rooted forms are the more likely to have survived the 

 arid conditions. Similarly both short- and long-rooted forms are likely to have 

 originated in mesophytic habitats, where survival is less a matter of the root 

 system than of stem and foliage characteristics. Obviously much further experiment 

 is necessary before we may know to what extent root variations are reactions to 

 environment, and to what extent they are congenital or characteristic of the species. 



Meager development is not the sole characteristic of swamp roots; in 

 the tamarack and in many other plants the prevalent direction of root 

 growth is horizontal rather than downward. In some cases roots grow 

 directly upward, as in certain palms and mangroves and in Jussiaea, or 

 there are erect projections from horizontal roots (known as knees), as in 

 the bald cypress (fig. 726). The exact cause of these reactions is not 

 known, but it is probable that the small percentage of oxygen in 

 swamp waters and the accumulation of deleterious root excretions and 



'> It has been ascertained, for example, that the roots of some species (as the cotton- 

 wood) are longer, slenderer, and nearer the surface in dry sand than in moist clay. 



