240 Bepokt on Crop Production of the 



" 1. That sodium is never totally absent from plants; and that 



" 2. If indispensable, but a minute amount of it is requisite. 



" 3. That the foliage and succulent portions of the plant may 

 include a considerable amount of sodium that is not necessary 

 to the plant; that is, in other words, accidental." 



Concerning the replacement of potassium by sodium the same 

 author concludes '' that the sodium which appears to replace 

 potassium is accidental, and that the replaced potassium is acci- 

 dental also, or in excess above what is really needed by the 

 plant," and leaves us to infer that the quantity of these bodies 

 absorbed depends to some extent on the composition of the soil, 

 and is to the same degree independent of the wants of vegeta- 

 tion. 



EXPERIMENTS AT THIS STATION. 



These experiments were begun in the winter of 189S-1899 and 

 were continued in the winter of 1899-1900. They were carried 

 on in a forcing house devoted entirely to i^lant nutrition and 

 were in the immediate charge of Mr. O. M. Taylor, who gave the 

 plants the most painstaking care under conditions such as to 

 insure the experiments against errors and accidents. 



THE PLAN AND DETAILS OF THE EXPERIMENTS. 



Tlie soi7.— Plant feeding experiments with natural soil are in 

 some r-espects very unsatisfactory. With such a medium for 

 growth, even the most sterile that can be found, it is not pos- 

 sible to control the supply of plant food within desirable limits. 

 On the other hand no perfectly sterile artificial soil has been 

 discovered which presents conditions the most favorable for 

 plant growth. 



In these experiments pure quartz sand has been used. This 

 was ground from quartz rock by the Berkshire Glass Sand Co., 

 Cheshire, Mass. The degree of fineness admitted of the sand 

 nearly all passing through a sieve .025 inch in mesh. Tests of 

 this sand for phosphoric acid and potash gave no trace of the 

 former and an average of onl}^ .00078 per ct. of the latter. 



