256 HOW CROPS GROW. 



capable of feeding the plant from a dry soil, and hence 

 the soil must be at first profusely watered ; after a time, 

 however, new roots are thrown out, which are adapted to 

 the altered situation of the plant, and then the growth 

 proceeds in the usual manner. 



The reverse experiment would seem to confirm this 

 view. If a seedling that has grown for a short time only 

 in the soil, so that its roots are but twice or thrice branch- 

 ed, have these immersed in water, the roots already form- 

 ed mostly or entirely perish in a short time. They indeed 

 absorb water, and the plant is sustained by them, but im- 

 mediately new roots grow from the crown with great ra- 

 pidity, and take the place of the original roots, which 

 become disorganized and useless. It is, however, only the 

 young and active rootlets, and those covered with hairs, 

 which thus refuse to live in water. The older parts of the 

 roots, which are destitute of fibrils and which have nearly 

 ceased to be active in the work of absorption, are not af- 

 fected by the change of circumstance. These facts, which 

 are due to the researches of Dr. Sachs, ( Vs. St., 2, p. 13,) 

 would naturally lead to the conclusion that the absorbent 

 surface of the root undergoes some structural change, or 

 produces new roots with modified characters, in order to 

 adapt itself to the medium in which it is placed. It 

 would appear that when this adaptation proceeds rapidly, 

 the plant is not permanently retarded in its growth by a 

 gradual change in the character of the medium which 

 surrounds its roots, as may happen in case of rice and 

 marsh-plants, when the saturated soil in which they may 

 be situated at one time, is slowly dried. Sudden changes 

 of medium about the roots of plants slow to adapt them- 

 selves, would be fatal to their existence. 



JSTobbe has, however, carefully compared the roots of 

 buckwheat, as developed in the soil, with those emitted in 

 water, without being able to observe any structural differ- 

 ences. The facts detailed above admit of partial, if not 



