THE CONDUCTION OF RESEARCH 425 



searched for. An example of the first class would be the determination 

 of the melting point of an element, while the development of a machine 

 for making bricks belongs to the second. The first class is the more 

 straightforward, so that the program can be planned in advance with 

 the possibility of only minor changes, and the final result will be un- 

 certain only in its degree of accuracy. A research of the second class 

 is more difficult, as it is a process of trial and error, the most hopeful 

 ideas being tried out until by the knowledge gained from previous 

 failures, a successful result is obtained. But, unfortunately, in this 

 class of research a successful result can not be guaranteed in advance, 

 and consolation must sometimes be obtained by the knowledge that a 

 well tried failure is a step ahead. 



Perhaps the first thing to do when commencing a piece of research 

 is to make a thorough study of the available literature on the subject; 

 first, to get new ideas, and, second, to prevent the repetition of methods 

 that have been found unworkable. On the contrary, there are certain 

 types of original thinkers who claim that they do not wish to know 

 what any one else has done on the subject, as they believe that this 

 knowledge will hamper the freedom and originality of their work. 

 This manner of working certainly leads to much needless duplication 

 and is not to be recommended except for very special cases. It is 

 unfortunately true that many statements appearing in print are in 

 error, due to poor experimentation, wrong conclusions and arithmetical 

 blunders, and it is strange how long these erroneous statements are 

 accepted. There are many instances where results of this type have 

 been quoted several times by other authors before their absurdity has 

 been made evident. It is, therefore, advisable to take all information 

 on trial, and to take no one's statements for granted without giving 

 them a thorough examination. 



After all the important information has been reviewed and care- 

 fully analyzed, a plan, or several alternative plans, can be laid out for 

 the most promising manner of attacking the given problem. Although 

 the plan will undoubtedly have to be revised or even completely 

 changed during the progress of the investigation, it is much better to 

 have a definite program to start with than to work along in a haphazard 

 manner, principally because of the training gained by the experiment- 

 ers in an orderly procedure. It is foresight and anticipation of diffi- 

 culties at this time that saves endless complications later on, and the 

 value of a research director depends on his ability to keep his men 

 from getting into trouble, or, if this is impossible, to get them out as 

 quickly as possible. The more experience one has had in research the 

 more clearly can he visualize the subsequent procedure, and the final 

 results. 



If the research being undertaken is of a simple character the only 



