118 



THE ECOLOGICAL RELATIONS OF ROOTS. 



diameter, ran off rather horizontally for a distance of 18 inches or so before 

 turning downward. At 2 feet the root had a diameter of only 2 mm. and was 

 very much twisted and curved, as is characteristic of many plants in pene- 

 trating the hard soil. Practically no branches were given off to the very tip, 

 which was threadhke and unbranched. At a depth of 6 feet, where the soil 

 became looser and more moist, the roots pursued an even course, in striking 

 contrast to the twisted parts in the harder soil. The younger portion of the 

 root is almost white in color and covered with a dense coat of root-hairs. The 

 older parts are dark brown. This description answers for two other plants 

 which were growing rtear it. 



Another plant ga . off two laterals at a depth of 18 and 21 inches respec- 

 tively. These were ea^h a millimeter in diameter and ran rather horizontally 

 to a distance of 2 and 2.5 feet. At a depth of 34 inches another lateral, 2 mm. 

 in diameter, came off and ran almost parallel with the tap. This plant had a 

 woody tap 3 cm. in diameter, which spht at the top and gave rise to three 

 clusters of stems. It tapered rapidly, however, and at a depth of 2 feet was only 

 4 mm. in diameter. The main root outside of branches already mentioned 

 branched but Uttle, as indicated in the drawing. It reached a depth of 13 feet. 



On a half-captured dune in the sandhiUs, 5 or 6 large mature plants were 

 examined, aU of which agreed in having a multicipital crown arising from 

 the 2 to 5 upper forks of the thick, woody tap-root. The surface branching 

 was exceedingly well developed. From the very base of the plant numerous 

 lateral branches arose, varying from 0.5 mm. to 6 or 8 mm. in diameter. All 

 were fairly well rebranched, the smaller laterals and the larger branches of 

 the major ones being densely clothed with root-hairs. In most cases the 

 tap-root, owing to the enormous numbers of large lateral branches, rapidly 

 diminished in size, so that at a depth of 18 inches none were larger than 5 

 nam. in diameter. This stands out in marked contrast to the stout, deep- 

 seated tap of the plains. Indeed, the deepest root examined penetrated to a 

 depth of only 6 feet 11 inches, while most of them reached depths of only 

 4 to 6 feet. The lateral branches, both large and small, formed a dense 

 network of roots filling the soil from 18 to 24 inches on either side of the 

 plant. AU of the larger roots were weU suppUed with small, well-branched 

 absorbing laterals. The divisions of the tap-root were themselves branched 

 and rebranched to such an extent that, as they passed downward, great 

 numbers of roots were found penetrating the soil to depths of 4 or 5 feet. 

 Still other plants were examined which confirm the above descriptions, 

 the taps and larger branches having a strong tendency to run obhquely or 

 even horizontally, rather than vertically downward as in the plains ecads 

 (plate 30, b). 



This affords a very clear case of the effect of the habitat upon root develop- 

 ment. The plains form has a root which is approximately twice as deep- 

 seated as those in the sandhills. While the tap is prominent throughout in 

 the former, it soon loses its dominance in the sandhiU form and often scarcely 

 exceeds in importance some of the stronger laterals. While both forms are 

 supphed with rather abundant surface laterals, in the plains form these are 

 largely confined to the surface foot, while in the sandhills they occur to a 

 much greater depth and are abundant along all of the major branches. The 

 branches in the first foot of the plains soils were mostly less than 1 mm. in 

 diameter and had a horizontal spread seldom exceeding 18 inches. They 

 were practically free from branches. In the sandhiUs the laterals ranged 

 from 0.5 to 8 nun. in diameter and were rather well branched. In number 

 these lateral branches, both large and small, formed a dense network of roots 



