170 The Plant World. 



as a soil "constant" it will be in spite of these errors and not 

 because they have been removed. In the process as at present 

 conducted there are indeed other errors: the failure to take 

 account of the permanent and semi-permanent flocculations; 

 the errors caused by organic matter and by dissolved salts; the 

 variations caused by the diflerent methods of sieving and of 

 elutriation, etc. But these errors are not inherent. They 

 can probably Ije removed if it seems useful to dj so. The 

 three previously discussed are the crucial ones, and if mechan- 

 ical analysis is ever replaced by something else it will be on their 

 account. 



The Siujace Constant'^. Of course these imperfections of 

 the mechanical analysis have not been suddenly and recently 

 discovered. They have long been appreciated and manv sugges- 

 tions have been made looking toward their avoidance or the 

 substitution of another and more satisfactory physical constant. 

 First of these, chronologically, was probably the suggestion 

 to use as such a constant the total amount of surface of all 

 the particles in a certain weight (or volume) of soil. Indeed 

 it was early suggested that the internal surface (as this is called) 

 might be determined from the mechanical analysis itself and 

 that the inconveniences due to the disunity of expression of the 

 latter might thus be climated. If the mechanical anahsis were 

 reduced to, or expressed as, a surface it coidd be expressed in 

 one unit. Unfortunately, however, the lack of knowledge of 

 conditions within the group limits and of the shapes of the par- 

 ticles make it impossible successfully to determine the internal 

 surface of the soil from the mechanical analysis and if this sur- 

 face is to be used as a constant it must be otherwise measured. 

 Direct measurement being obviously out of the question, it is 

 necessary to find something dependent on the surface and 

 to measure that. This something has been found in the 

 heat of wetting. 



When a dry powder is placed in water a considerable quan- 

 tity of heat is evolved and the quantity of this heat vaiies direc- 

 tly with the total internal surface of the powder and probably 

 in close proportionality thereto. ='' This fact, long known to 



* For the purelv phvsical side of this problem and for references to the literature see 

 Patten, — Trans. Amer. Electrochem, Soc. 11; 387-407 (1907) 



