546 THE ROYAL SOCIETY OF CANADA 



hypotheses or probable conclusions. Any increase of evidence re- 

 sulting from such search may, or may not, confirm the hypotheses or 

 conclusions; it will not change essentially the previous body of neces- 

 sary conclusions, but ordinarily it will extend them in some form, and 

 also give rise to a modified or new series of probable conclusions calling 

 for further investigation. This fructifying process ought to be re- 

 peated indefinitely, until the investigator has exhausted, to the best 

 of his ability, all the sources of information that he can reach. On the 

 basis of this accumulated evidence he then formulates his final, i.e., 

 necessary, conclusions for publicity, e.g., an historian his narrative 

 for publication. Under the prevailing historical method, however, 

 the historian, at this final stage, feels at liberty to adopt for publica- 

 tion, along with the necessary results, those conclusions which he 

 regards as only probable or highly probable. There is, nevertheless, 

 no assurance that conclusions, which seem probable or highly prob- 

 able at this stage, would not fall before further additions of evidence, 

 if such were available, just as many conclusions, which seemed prob- 

 able or highly probable at previous stages of the investigation, fell 

 before additions of evidence that subsequently became available. 

 For this reason conclusions that appear to be only probable or highly 

 probable at the final stage ought to be omitted from the published 

 results. The function of probability here changes. The imaginative 

 faculty which the investigator used freely up to this point in devising 

 probable conclusions or theories for further investigation, he should 

 now employ with the same freedom, but as a purely negative force, 

 in testing the actual necessity of those conclusions which up to this 

 point he has classed as necessary. The homely remarks of a 

 well known detective who holds honesty to be the chief requisite 

 of his calling, and includes the imagination among the other necessary 

 characteristics, illustrate the proper use of theory or probability 

 quite as well in scientific as in criminal investigation: "The thing to 

 do is to muster all known facts, consider them carefully, analyse 

 them, and then formulate a theory on those facts. After the detective 

 has thought, and thought hard and long, he goes out to test this 

 theory. Just as often as the facts change, just so often must the 

 theory change, so as to make it always fit the facts. . . But with 

 every other quality, if honesty be lacking, he will fail. A man who 

 is not honest with others cannot be honest with himself either. And 

 unless a man is honest with himself, he can't trail a sandpiper along a 

 wet beach. For he will pretend, even to himself, that he is travelling 

 the right route, though not really sure in his heart of hearts that he 

 is. And why ? Because it is easier. Dishonesty always seeks the 

 easy way." 



