548 THE ROYAL SOCIETY OF CANADA 



deliberative assemblies that they had wished him dead for years; 

 yet the statement was strictly contemporary and originated from men 

 who were really of the highest principles and knew the facts. The 

 case is indeed quite a problem for the prevailing method which tests 

 the trustworthiness of a record by its origin from a contemporary who, 

 in the words of Bernheim, is "intellectually and morally capable of 

 faithfully communicating the facts." It is, however, only a simple 

 example of the proper function of reasoned probability where the 

 trustworthiness of a record is judged, as it should be, by its own 

 exemplification of the 5 requisites. The dates of birth and death, 

 fields of labor, and similar items of such an obituary are intended to 

 inform the reader according to the reader's interest (2nd requisite) 

 and also exemplify ordinarily the other four requisites, and there- 

 fore should be accepted as prima facie correct. On the other hand, 

 the concluding statement which was the subject of remark may indeed 

 have been intended for the information of the reader according to his 

 interest, but not necessarily so; there is a reasonable probability that 

 the purpose of such a statement may be only to gratify surviving 

 relatives and friends, and wherever such a defect in exemplification 

 of any one of the 5 requisites can be reasonably suggested as probable, 

 the statement or portion affected, together with all deductions there- 

 from, ought to be removed by the investigator from his final results, 

 without benefit of doubt. This principle, however, applies only to 

 reasonably probable defects within the exemplification of the requisites, 

 and not to external probable reasons adduced against conclusions 

 resting on unquestionable exemplification of the requisites. For 

 such unquestionable exemplification establishes the conclusions as 

 scientifically necessary; and merely probable reasons cannot undo a 

 necessary conclusion, if they are simply brought from without and 

 weighed as extraneous objections to a conclusion, acceptance of which 

 is still necessary on other grounds. But a reasonably probable defect 

 in a statement's exemplification of the requisites of trustworthiness 

 is not an extraneous objection to such acceptance, which would still 

 be necessary on other grounds, nor a merely negligible pressure against 

 the statement's foundation. It is a defect within the foundation 

 itself and a shortage in the grounds, all of which are necessary in order 

 to justify acceptance; and thus the statement is no longer free of 

 reasonable doubt. 



Reasoned probability in indirect inferences from primary conclusions. 

 These indirect inferences range from such manifest instances as 

 are noted in Cases 3 and 7 of the experimental test, to difificult de- 

 ductions from evidence corresponding to the conclusions reached on 

 circumstantial evidence in courts of law. Circumstantial evidence 



