METHODS OF FOREST POLICY. 243 



only by the machinery of the state, which can 

 make inquiries uniform, compel answers, and has 

 no special interests to represent which might 

 influence the reliability of the statements. In 

 forestry statistics especially, the difficulties of as- 

 certaining conditions of supply are beyond the 

 capacity of individual inquiry, owing to the com- 

 plicated nature of the object of inquiry. If there 

 is difficulty in determining quantity and value of 

 standing merchantable timber, which is within the 

 actual vision of the estimater, how much more 

 difficulty must be found in judging the prospec- 

 tive quantity and value of the unperfected crop, 

 the promise of the future ; and this is the essen- 

 tial knowledge upon which is to be based, private 

 as well as state activity with reference to this 

 resource. 



We may only briefly indicate what kind of sta- 

 tistical knowledge would be desirable in order 

 merely to direct public policy. 1 



In the well-ordered state the soils most fit for 

 agriculture should be devoted to systematic food 

 production, but just so should the non-agricultural 

 soils, the absolute forest soils, be devoted to the sys- 

 tematic production of wood-crops ; moreover, as we 

 have seen, the forest in certain situations exercises 

 a potent influence on cultural conditions. Hence 



1 For a fuller discussion see " Considerations in gathering For- 

 estry Statistics," by the author, in Quarterly Publications of the 

 American Statistical Association, 1898. 



