EDITORIAL. 703 



Science, whether for itself or as applied to any particular class of 

 phenomena, is developed through research. In agriculture we deal 

 largely with natural phenomena of endless variety ; while the object 

 sought is ultimately practical knowledge or information, the stand- 

 ards and the methods must conform to those of science in general. 

 The method of science is that of observation and experiment, guided 

 and supported by reason and generalization. Observation and ex- 

 periment teach facts; and in all science investigation consists in 

 passing up from the fact to the law, and in discovering the facts 

 capable of leading to a law or conclusion. Facts and laws are not 

 artificial or accidental, and are not made by the scientist. Man's 

 only originality in connection with them is their discovery. 



Observation and experiment are therefore fundamental to progress 

 in natural science. Because of this and because " it is better not to 

 see than to see wrongly," these observations should be made with 

 care and skill and so planned as to furnish a clear, decisive result 

 which can be given a safe interpretation. This is the great secret 

 of success in agricultural investigation. The subject and the con- 

 ditions are unusually complex, and the results are very likely to 

 be complex, embodying a variety of factors and dependent upon 

 many limiting conditions. For this reason the conditions of the 

 experiment must be as simple as possible, and limiting factors 

 checked or controlled to the utmost degree. In this way the number 

 of variants will be reduced to a point where safe deductions can be 

 made as to their several effects. The attempt to make an experiment 

 answer too many questions, with a result that is fatal to the whole 

 undertaking, is perhaps one of the most common errors of the over- 

 ambitious worker. 



Facts differ widely in their importance. We must be sure in the 

 first place that what we determine is actually a fact of a permanent 

 character, and not merely the result of a particular combination of 

 circumstances that might never occur again. This means a greater 

 refinement of method than is usual in the common field and feeding 

 experiment, a provision of methods and checks which will bring 

 conditions more largely under control. But among established facts 

 some are pregnant and others sterile, some have no reach — they teach 

 us nothing beyond themselves, and the scientist who has ascertained 

 them has not become more capable of foreseeing other new facts. 

 On the other hand, there are facts of great yield; each of them 

 marks a definite contribution or step in advance and foreshadows 

 the way for further progress. Isolated facts in agriculture do not 

 always have the constructive value that such facts may have in a 

 pure science, because the investigator is skeptical of them, knowing 

 the liability to faulty observation. 



