706 EXPEEIMENT STATION BECOED, 



In short, the usual field experiments, which form such a prominent 

 feature of station work, teach but little in a definite way as to the 

 life of the plants concerned in them or of the exact conditions of their 

 growth, and they contribute far less than they should to a thorough 

 knowledge and understanding of these plants. 



In too large a proportion of these experiments the main interest 

 and reliance seems to be placed on the weight or bulk of the har- 

 vested crop. This of course is the fimal measure from an economic 

 point of view, to determine whether a practice or treatment is 

 profitable or advisable. But in investigation, and in all experiments 

 except those of the most rudimentary kind, the aim should be to 

 learn not merely the economic result but something of the way in 

 which the result has been brought about, and the effect upon the 

 plant to the imposed conditions. The living plant must be studied 

 quite as much as the final yield. 



These things are highly essential if field experimentation is to de- 

 velop along scientific lines and make progress in laying the founda- 

 tion for an intelligent agriculture. They require far more attention 

 to a certain kind of details than is now given. TJiey require con- 

 tinuous study, with attempts at correlations of growth and specific 

 conditions throughout the season. Otherwise the final result is a 

 composite result of all the conditions, and there is no means for de- 

 termining how much is to be attributed to particular artificial con- 

 ditions imposed, or how far the effect of these has been masked by 

 other unfavorable conditions. 



To make these continuous studies of the plants and their sur- 

 roundings implies a smaller number and less diversity of field ex- 

 periments, and greater attention to those which are undertaken. It 

 will mean more study in planning the work, to provide such checks 

 under more refined conditions of experiment as are necessary to 

 furnish reliable comparisons. It will call for some system of ac- 

 curate measurements at regular intervals, and the recording of un- 

 controlled phenomena wliich may influence growth or physiological 

 function. It will require living with the experimental material — a 

 closer association with the plants and concentration on their study. 



Under such close observation there will be no lack of problems. 

 The investigator with Wsion and imagination will be confronted 

 with them on all sides, and instead of allowing them to bewilder 

 him with their confusion he will devise means to regulate them and 

 take account of them in his plan of experiment. This will make the 

 methods of experimental agronomy more exact and more searching 

 in character, and will make interpretation more sure. It might re- 

 lieve the necessity for some of the repetition that has been going on 

 f<o manv vears. 



