[bowman] fundamental PROCESSES IN HISTORICAL SCIENCE 517 



forty of the two hundred passengers were injured, fifteen of them 

 seriously, but no life was lost. Old railroaders remarked that they 

 never saw such damage with so few passengers injured. But even in 

 serious accidents, unless they are followed by fire or the coaches are 

 submerged in water, the number killed rarely exceeds one-tenth of the 

 passengers. The child if conscious would be so likely to cry that the 

 probability of the 1st conclusion is fixed approximately by the chance 

 that it had escaped death, plus the further chance that it had escaped 

 injuries which rendered it unconscious. One the above basis in the 

 present case an adult's chance of escaping death would be 9/10 or 

 9:1. A tender infant's would be less. To offset this tenderness and 

 to cover the further chance that the child might be unconscious from 

 injuries, the probability of the 1st conclusion is reduced to 3/4 or 3:1. 

 In the 2nd conclusion, the child being silent, the principal opposing 

 chances are that it was dead or unconscious. Statistics are not 

 available to ascertain exactly these opposing chances, but under the 

 circumstances the hope of finding the infant alive would be compar- 

 atively small, hence the probability that it was dead has been estim- 

 ated at 2/3 or 2:1. In the 3rd conclusion, the child being discovered 

 aliv9 and yet silent, it must have been either (1) unconscious from 

 injuries, or (2) asleep, or else (3) awake and yet quiet. Under the 

 circumstances the chances of (2) and (3) were so slight, and the risk 

 of serious injury was so great, that the probability of unconsciousness 

 from injuries is conservatively estimated at 100/101 or 100:1. 



Case 6. 



A narrative of necessary conclusions may be abbreviated, instead 

 of extended, by additional evidence, not necessarily by way of correc- 

 tion or omission of any point, but because the additional evidence 

 converts a previous conclusion into one of higher value. Thus, in 

 Case 6, "ostensibly" is omitted from the previous narrative in view of 

 Part 3 of the evidence. 



Narration of Case 6. 



(Part 1): "In a newspaper with a daily and weekly edition a 



letter dated May 8 appeared on May 17, ostensibly replying to a 



letter in the paper for May 3." 



(No. 1): "Probably the reply was written at a point far from the place of 

 publication, and both the original letter and the reply appeared in the daily 

 edition, the former on May 3 and the latter on May 17." (3/4 or 3:1). 



(Part 2) : "The reply was dated at the place of publication, and 



the issue of May 17 in which it appeared was the weekly edition." 



(No. 2): "In that case it is highly probable that the issue of May 3 in 



which the original letter appeared was also the weekly edition." (100/101 



or 100:1). 



