THE PHYSICO-CHEMICAL INVESTIGATION OF SOILS. 317 



virgin soils were easily obtainable and their native growth 

 verifiable; and were actually in part made the subject of chemi- 

 cal investigation. 



In the course of their work, Owen and Peter always care- 

 fully recorded the native vegetation of the soils collected; but 

 neither seems to have formulated definitely the idea that such 

 vegetation might be made the basis of direct correlation of 

 soil-composition with cultural experience. Owen repeatedly 

 expressed to the writer his conviction that such a correlation 

 could be definitely established by close study ; but early death 

 prevented his personal elaboration of the results of his work. 

 Peter likewise stoutly maintained to the last his conviction that 

 soil analysis was the key to the forecasting of cultural possi- 

 bilities; but not being a botanist he did not see his way to put 

 such forecasts into definite form. 



Change of Views. 



In the United States as well, the ancient prejudices have 

 now gradually given way before the urgent call for more de- 

 finite information than could otherwise possibly be given to 

 farmers by the experiment stations, most of whose cultural 

 experiments, made without any definite knowledge of the na- 

 ture of the soil under trial, were found to be of little value 

 outside of their own experimental fields. Even the multipli- 

 cation of culture stations in several states, unaccompanied by 

 soil research, is found to be a delusive repetition of the same 

 inconclusive, random experimenting, since it takes into con- 

 sideration only the climatic differences, but leaves out of con- 

 sideration the potent factors of soil quality and soil variations. 

 At most these were usually mentioned by them in such inde- 

 finite terms as " a clay loam," " a coarse sandy soil," " gray 

 sediment land," and the like; frequently not even with a state- 

 ment of the depth and character of the subsoil and substrata, 

 much less of their geological derivation or correlations. Thus 

 any one not happening to he personally acquainted with the 

 land in question would be wholly without definite data to cor- 

 relate the results with his own case. It is quite obvious that 

 even if only to make possible the identification of new lands 

 with others that have already fallen under cultural experience, 



