430 THE SCIENTIFIC MONTHLY 



work consists in ordering materials, having apparatus designed and con- 

 structed and deciding on the methods to be used. The laying out of the 

 work in a manner to promote efficiency depends largely on the fore- 

 sight and experience of the experimenter or research director. For this 

 reason, it is false economy to employ low grade or inexperienced re- 

 search men, as the saving in the pay roll is more than offset by the de- 

 creased value of the results and the increased cost of the investigation. 



As time is usually as important an item in efficiency as cost, the 

 ordering of necessary supplies should be accomplished as early as pos- 

 sible, and when big delays are impossible to avoid because of the lack 

 of some material, it may become necessary to alter the experimental 

 methods in order to be able to proceed within a reasonable time. It 

 is very costly to have everything set for performing an experiment, and 

 then to find that some small but vital thing has been forgotten which 

 will take weeks to procure. All phases of the preliminary work should 

 be constantly checked over and the most delaying items followed up 

 vigorously. 



The previous discussion of efficiency has been confined to a single 

 piece of research, but usually a number of investigations are going on 

 together in one laboratory, perhaps a separate group of men working 

 on each research, or a group alternating between several types of work 

 as conditions permit. In this case it is the duty of the research director 

 to arrange the work not only so that every one will be kept busy, but 

 so that each man will be working to the best advantage. Every one is 

 more or less of a specialist, and it is of considerable advantage to have 

 each person kept as far as possible on one type of work. This can not, 

 of course, always be done, but by carefully laying out the work ahead 

 in this respect, it will be possible to have the men working in their most 

 efficient positions a large part of the time. The same thing applies to 

 pieces of standard apparatus such as balances, testing machines, etc., 

 so that the work should be planned to use the equipment as efficiently as 

 possible. 



In regard to the selection and training of the research personnel, 

 it will be best to first discuss the types of men available, exclusive of 

 their particular training. In the first place men may be divided rather 

 sharply into two classes, the first we will call practical, and the second, 

 theoretical. The first class have mainly gained their knowledge from 

 experience, are mechanically inclined and know how to use their hands, 

 while the second class have obtained their knowledge almost exclusively 

 from books and have very little commonsense in regard to mechanical 

 matters. For example, one of the theoretical class mav be able to 

 make a complex computation of the stresses in a certain small bolt, and 

 yet when screwing in the same bolt he will calmly twist its head off, 

 simply because he has no mechanical sense. The latter are a type that 



