EDITORIAL. 707 



be guided by what has gone before, and his investigation will de- 

 volve into a round of routine. 



In a word, then, research advances by a procession of experiments 

 and observations, based on hypothesis and guided by the foresight 

 which generalization supplies. Science prospers through empirical 

 discoveries, and the theory must not run counter to any positive 

 empirical facts; but science does not stop with their determination. 

 It demands the reason, the connection, the law. These are derived 

 through induction from facts developed by experiment which sup- 

 port the theory or the principle or the conclusion at essential points. 

 Without these steps research lacks the essential features. This is as 

 true in agriculture as in any primary science. 



But research is not the sole source of knowledge and there are 

 many additions to knowledge of a sound, scientific character, whose 

 derivation does not seem to answer the definition of research. 

 Among such are discoveries made perhaps by chance observation, 

 or the recognition of a fact in an experiment, possibly in the course 

 of research but not an outgrowth of it. These may have all the 

 weight of scientific facts, constitute contributions to science, and be 

 strictly original, but still not represent research. This makes them 

 no less valuable or dignified. 



Research carries the idea of a protracted inquiry; it is a con- 

 scious, premeditated effort in the working out of problems usually 

 not solved by a single experiment, but requiring progressive study 

 through a series of experiments from which the answer is gradually 

 evolved. If we are to accept the definition of the sciences, agricul- 

 tural research is to be understood as an excursion into the unknown, 

 involving a combination of various types of effort — hypothesis, ex- 

 periment, generalization. 



It will be recognized that in contrast to research, what in agricul- 

 ture we commonly designate as " experiments " are much simpler in 

 their aim and plan, and lack the finality of research. They are usu- 

 ally satisfied with the empirical result. They may not even be 

 original, for the Hatch Act provides for verification experiments, 

 and many experiments aim merely to test out or adapt the conclu- 

 sions from former work to a new locality or a new set of conditions. 

 They may not even be scientific in the sense that they seek strictly 

 scientific facts, but they should be made with the care and precision 

 of science. Such are many of the commercial experiments, to deter- 

 mine economical methods of practice or relative values. Usually 

 the results are of a more or less temporary nature. 



A large share of the ordinary field and feeding experiments are 

 comparative trials rather than strictly scientific investigations. They 

 are made under presumably quite uniform conditions, and hence the 

 20793°— 14 2 



