ii2 THE POPULAR SCIENCE MONTHLY 



Invention of Hypotheses. — The search for explanation of observed 

 facts may be made in some cases by the memory, which may recall an 

 explanation previously learned. In physics and chemistry, the search 

 for explanation is largely aided by experiment; but in the study of 

 land forms experiment serves chiefly to illustrate explanations already 

 reached, rather than to lead to new ones. We are here chiefly con- 

 cerned with the kind of search which calls the investigator's own fac- 

 ulty of invention into play — the kind of search which tries to make a 

 new combination of some pertinent facts or principles of previous 

 acquisition with some of the facts of new observation, in the hope of 

 thereby bringing about a clear understanding of all the facts under 

 discussion. 



The faculty of invention is peculiar in working to a large extent 

 subconsciously. Facts to be explained can be intentionally observed; 

 previously gained knowledge may for the most part be consciously 

 reviewed; but the desired explanatory combination of old knowledge 

 and new facts may not be immediately found while the conscious search 

 for it is going on. Invention is, however, much favored by active 

 observation, and spurred on by an eager spirit of inquiry; it is greatly 

 aided by mental ingenuity, but it is seldom immediately accomplished 

 by conscious intention. However, the faculty of invention can be 

 cultivated if many facts, old and new, are frequently brought to con- 

 scious attention, and the wish for explanation and the search for it are 

 often renewed. Then the subconscious mind will continue the search, 

 and after an interval, during which the matter has been apparently out 

 of mind, an explanation may most unexpectedly awaken attention by 

 springing into consciousness. 



The sudden birth of an apparently successful explanation is truly a 

 most delightful experience; indeed so delightful and encouraging that 

 many an investigator has mistaken it for the climax or crown of his 

 work, and accepted it as the whole truth without further question. 

 But, as has already been pointed out, the too-ready acceptance of an 

 untested invention, as if it were true, is dangerous. The investigator 

 must recognize that it is no great recommendation of an invention, 

 that it explains the partial group of facts that it was made to explain. 

 Of course it must do that; it would deserve no consideration at all if 

 it did not. But in order to deserve acceptance as the true counterpart 

 of past facts, it must do much more. It must explain various facts 

 that it was not made to explain ; facts that it did not expect to explain ; 

 facts that were not thought of, or were not even known at the time of 

 invention, as will appear more fully below. 



The investigator of course hopes that his invention, based on some 

 of the observed facts, will prove to be the true counterpart of some 

 past facts, or of some invisible principle or process, by means of which 



