508 THE ROYAL SOCIETY OF CANADA 



of probability, an inferior and a superior, and even the superior meas- 

 ure cannot produce results individually correct, but its only significance 

 is as to average results in the long run. According to Bernheim, 

 probable conclusions in history rest on the inferior measure, because 

 he states specifically that a probable conclusion in historical investi- 

 gation is one in which the reports or other reasons in favor of the con- 

 clusion outweigh those against it. This is essentially a process of 

 balancing reasons pro and con, described in the Century definition 

 as the conceptualistic or inferior measure. From these definitions in 

 Bernheim and the Century Dictionary it is, therefore, a necessary 

 deduction that probable conclusions in history, except where they 

 represent averaged results, are without significance or value because 

 they rest ordinarily on the inferior measure, which must be less able 

 than the superior to produce individually correct results, the superior 

 in turn being without significance in producing these. 



PROBABILITY TESTED EXPERIMENTALLY. 



Bernheim's definition of probable conclusions in history is essen- 

 tially correct. They rest ordinarily on a variety of opposing reasons 

 of different values, and the respective groups of favorable and unfav- 

 orable reasons as a whole are weighed or balanced against each other 

 in order to strike what historians call a "balance of probability" for 

 or against the conclusion. The probability of conclusions so drawn 

 cannot rise above the level of the inferior conceptualistic measure. In 

 actual experience, however, and therefore also within the historian's 

 special field, cases may arise where two opposing reasons are so prom- 

 inent, and subsidiary reasons, if not excluded entirely, are relatively 

 so insignificant, that the opposing favorable and unfavorable chances 

 may be determined accurately or be estimated with enough accuracy 

 to raise the resulting conclusion to the level of the superior measure 

 of probability. This superior measure is described in the Century 

 definition above quoted as the ratio of the number of occurrences of 

 the event to the total number of occasions in the course of experience 

 and it is there illustrated by the throwing of a die : the probability 

 that the die will turn up ace is 1 /6, or 1 favorable chance out of 6, 

 because the die has equal chances of falling on any one of its six sides 

 and only one of these sides is ace. By a series of cases located in 

 actual experience where the chances for and against the conclusions 

 can be accurately determined or conservatively estimated, the author 

 proposes to test experimentally the deduction made from the fore- 

 going definitions. The cases will be given in narrative form and in 

 harmony with the conditions actually prevailing in historical investi- 

 gation. Almost invariably in the study and writing of history the 



