516 THE ROYAL SOCIETY OF CANADA 



classes $452,171,396, or approximately a ratio of 20-3 : 8-5 : 4-5. 

 According to this ratio the favorable chances in the 1st conclusion 

 would be 20-3 out of 33-3, and in the 2nd, the volume of fire insurance 

 being there excluded by Part 2 of the evidence, 8-5 out of 13 or 17 

 out of 26. In the preceding years of the semi-decennium 1906-1910 

 the relative amounts of different classes of insurance in Canada were 

 much the same as in 1910. In the 3rd conclusion the probability 

 is fixed approximately by the number of instances where goods in 

 transit are nsured at actual cost, as against those instances where 

 they are insured at other values, especially at the cost of duplication 

 if this, as is occasionally the case, differs materially from the first cost. 

 The ordinary practice is to insure at actual cost, hence a conservative 

 estimate of the probability is 10/11 or 10:1. In the 4th conclusion 

 the probability depends on the number of shipments where goods 

 are insured at the cost of duplication because this differs materially 

 from the first cost, as opposed to those instances where, for any other 

 reasons than such difference, the shipment is insured at other than 

 actual cost. The former instances are comparatively rare, but the 

 latter are much rarer, hence a conservative estimate is 10/11 or 10:1. 



Case 5. 

 Narration. 



(Part 1): "Shortly after midnight a train with two hundred 



passengers left the rails under full speed. Several coaches, in one of 



which was a mother with a child of six months, rolled twenty feet into 



a ditch. It was half an hour before all were rescued from the wreck." 



(No. 1): "Probably the whereabouts of the child would be easily known 

 by its cries." (3/4 or 3:1). 



(Part 2): "The child was lost in the confusion that followed the 



accident, and the grief of the mother was piteous to behold." 



(No. 2): "In view of the child's silence, it seemed indeed hard to believe 

 that the infant could be alive." (2/3 or 2:1). 



(Part 3) : "The infant, however, was found presently beneath a 



seat in the midst of the wreckage, and it was breathing." 



(No. 3); "Very probably it was unconscious from injuries." (100/101 

 or 100:1). 



(Part 4): "On closer examination it proved to be uninjured and 

 sound asleep." 



In this case the narrative of necessary conclusions consists again 

 of the four parts of evidence read uninterruptedly. In the accident 

 in question which occurred near Tottenham, Ontario, June 27, 1908, 



