WATER CONTENT OP PLANT AND SOIL 29 



saline and alkaline soils, the root hairs, being strongly acid in 

 their nature, must suffer from the increasing alkalinity of 

 the soil. This is a fertile field for investigation and it is 

 hoped that these problems will soon be taken up and solved 

 by the plant physiologist. 



In mesophytes when wilting, the amount of water lost by 

 the soil nearly equals the amount lost by the plant. It might 

 be inferred that a plant ceases entirely to take water from 

 the soil the moment wilting begins, and from this time on 

 it simply undergoes evaporation. However, it can hardly 

 be true that the plant ceases entirely to take water from 

 the soil until it has lost much from its own weight. The 

 facts in the matter may be obtained by a simple epxeriment. 

 If several plants be taken at the time when they begin to 

 lose turgidity and part of them uprooted, and their roots care- 

 fully sealed with wax or cocoa butter, these should lose 

 water more rapidly, if it be true that plants in the soil do 

 not cease to absorb water when they begin to wilt. Both 

 should lose weight equally if the plants in the soil are not 

 obtaining water, and finally, if the sealed plants should lose 

 water more slowly than those in the soil it would be positive 

 proof that the soil is actually taking water from the proto- 

 plasm of the roots. 



The plant absorbs very little water from the soil after 

 it is wilting excessively. From the very nature of the root 

 hairs, it is quite certain that they perish soon after the 

 plant begins to suffer from drought. This is indicated by 

 observations to the effeet that the roots of many annuals die 

 before the stem and that they commence to lose water about 

 the time the leaves die. Should it be found out later that 

 this is certainly the case, the effect will be to increase the 

 estimates of the limit of the non-available water and conse- 

 quently to diminish those of the available water. This being 

 the case the tables at the close of this treatise will need 

 revising, but the general results and conclusions of the work 

 will remain unchanged. 



