78 ADDRESSES 



yield of fruit in this experiment/' In Bulletin 100 of the Pennsylvania State 

 College and Agricultural Experiment Station is given the data in regard to 

 the Massachusetts experiment furnished by the Director of the Massachusetts 

 Experiment Station. It is here stated that "in every respect the treated plots have 

 proven superior to the untreated." A forthcoming report of the Ohio Experiment 

 Station will show that "the increase in fruit production of the mulched and fer- 

 tilized plots, as compared with those receiving no fertilization, has ranged from 

 200 to 1000 per cent, within eighteen months from the time of the first applica- 

 ,tion." One of our agricultural papers recently called attention to the different 

 conclusions reached by the first two experiments stations and raised the question 

 as to what the farmer is to do when authorities disagree. The soil upon which the 

 experiments were conducted are not the same and it seems to me that this fur- 

 nished the reason why such entirely different results were secured. The farmer, 

 therefore, must know which of the soils is like, or most like, that upon his 

 own farm before he can tell which experiments to follow. A soil survey will fur- 

 nish this information. 



WHY CHEMICAL ANALYSIS is UNSATISFACTORY. 



The failure to recognize the importance of the differences in soil is one of 

 the probable reasons why no more satisfactory results have been secured from 

 chemical analyses. Chemists have tried to establish a standard that would apply 

 to all soils. Is this not too much like trying to produce a medicine that would 

 cure all diseases? We know that moisture in the soil is just as essential for the 

 growth of plants as the presence of phosphorus or potassium or nitrogen, but it is 

 not possible to establish a standard or optimum amount which will apply to all soils. 

 An amount which would give the largest growth to a plant on a sand would not be 

 sufficient to prevent its death on a heavy clay. Since this is true in regard to moist- 

 ure, may not be true in regard to other compounds also, although in a less 

 marked degree? The interpretation of the chemical analyses of soils in the past 

 has been undertaken with practically no regard to the differences which are very 

 obvious in the field. The analysis of a sand has been compared with that of a 

 clay, although the two are about as unlike as two soils can be. It is very much to 

 me as if a man were to secure samples of a great many different kinds of grain, 

 analyze them and then endeavor to interpret their composition without first ascer- 

 taining whether he was analyzing corn, wheat, oats, rye or barley. 



THE CART BEFORE THE HORSE. 



A little more than three years ago I was invited to give a talk upon some 

 phase of soil work before the students of one of our largest and most progressive 

 universities. In the course of the address I tried to emphasize the importance of 

 the study of soils in the field. When I had finished the professor who had charge 

 of the soil work in that state said to me "You have taken exactly the opposite 

 view in regard to the study of soils from that along which we have been work- 

 ing. We have thought it best to collect samples from different sections of the 

 state, analyze them and study them in the laboratory, using the information there 

 obtained as an aid in reaching the proper basis on which to separate and classify 

 the soils. You would classify them upon obvious field differences and use the 

 laboratory as a means of explaining those differences which cannot be explained 

 from the field study alone, and the more I think about the matter the more I am 

 inclined to believe that you are right." I might add that the legislature of that 

 state is now appropriating $10,000 a year for the conduct of a soil survey. 



* Bulletin No. 339, page 154, New York Agricultural Experiment Station. 



