518 THE ROYAL SOCIETY OF CANADA 



(Part 3): "Both letters appeared in both editions, the original 

 letter in the daily for May 3 and the weekly for May 10, and the reply 

 in the daily for May 9 and the weekly for May 17." 



In the narrative of necessary conclusions in this case the succes- 

 sive parts of evidence are added by insertions as well as by extensions, 

 and will read finally, omitting "ostensibly," as follows: 



"In a newspaper with a daily and weekly edition a letter dated 

 May 8 at the place of publication appeared in the daily for May 9 

 and the weekly for May 17, replying to a letter in the daily for May 3 

 and the weekly for May 10." 



The 1st probable conclusion depends chiefly on the proportions 

 of letters respectively in the daily and weekly editions of all papers 

 having these editions. In view of the greater number of daily issues 

 and of the difference of interest which in most papers reduces corre- 

 spondence in the weekly to a minimum, a conservative estimate of this 

 proportion would be 10:1, i.e., there are on the whole ten times more 

 letters in the daily issues than in the weeklies of such papers. An 

 editor, whom the author consulted, estimated the proportion at 

 25:1. The probabiHty has been further reduced from 10:1 to 3/4 

 or 3:1 because the exact interval of fourteen days between the dates 

 May 3 and May 17 suggests the possibility that it was the weekly. 

 It should be noted, however, that here it is not yet in evidence that 

 either of the issues was a weekly; and there are six times as many 

 issues of the daily as of the weekly, between which this exact interval 

 of fourteen days might occur. 



In the 2nd conclusion it is in evidence that the answer was in the 

 weekly. In actual experience men constantly answer letters of the 

 weekly in the weekly, and of the daily in the daily, and they very 

 seldom answer letters of the daily in the weekly, or of the weekly in 

 the daily; hence the high probability of this conclusion, estimated at 

 100/101 or 100:1, that the answer in the weekly of May 17 was a reply 

 to a letter in the weekly of May 3. 



Case 7. 



Narration. 



(Part 1): "A.B., dwelling at Berlin, received simultaneously 

 two letters, both stamped July 26, 1913, at Warsaw in Russia." 



(No. 1): "Probably it was Berlin in Germany and the letters were a day 

 or less in transmission." (100/101 or 100:1). 



(Part 2): "A.B. dwelt at Berlin in Canada, where both letters 

 were stamped August 19, 1913." 



(No. 2): "In that case apparently the letters must have been twenty- 

 four days in transmission." (1,000/1,001 or 1,000:1). 



