THE CONDUCTION OF RESEARCH 433 



or value be published or in some way be made available to those who 

 are interested. It now happens, especially in the industrial labora- 

 tories, that much work is done which is never known outside, and there 

 is generally no reason why the results should not be published after 

 the particular organization has received its benefit. Because of this 

 practice a great deal of money is spent in duplicating work that has 

 been already accomplished, and, while a certain amount of duplication 

 is valuable as a check, it is in general very uneconomical. In the same 

 way, it sometimes happens that similar investigations are undertaken 

 simultaneously, and, although their results are later published, it means 

 an unnecessary duplication. In this particular it would be of great 

 value to have a research clearing house where all the work in prepara- 

 tion could be gathered together for general information. 



It may be stated in conclusion that the carrying out of a piece of 

 research will comprise in general the following procedure: First, the 

 similar work of others is studied, especially their difficulties and fail- 

 ures, and from this information a plan of operation is laid out. And 

 if there is any doubt, and there usually is, as to the practicability of the 

 proposed methods, preliminary experiments should be conducted, from 

 which data are obtained for use in designing apparatus and for more 

 completely planning the subsequent procedure. In carrying out the 

 actual work, the first consideration should be accuracy and the second, 

 efficiency, both depending on suitable equipment and on an interested 

 and well-organized staff. Lastly, the experimenter must organize his 

 results, and deduce from them conclusions that will be of value in join- 

 ing them with similar work and in advancing the theory and practice 

 of the subject. Briefly, successful conduction of research depends on 

 the foresight and vision of the experimenter in laying out the work, his 

 accuracy, persistence and manipulative skill in carrying it out, and, 

 lastly, his analytical ability in deducing conclusions from the results. 



VOL. XII.— 28. 



